Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Is a heatpipe really appropriate for this application? The heatpipe expects that the heat source wants to burn up and so there's lots of heat available to vaporize the liquid in the pipe. It's not clear to me whether that situation exists with these Rb standards. My tests with an FE-5680A showed a maximum temperature of about 62C without a heatsink. That's far lower than a CPU or GPU. Some of them run at that temperature *with* the heatpipe. I think that a heatsink/fan (maybe from a video card) equipped with a PWM controller might be a better fit. Many of those combinations have a ducted fan to provide better control of the airflow. That would reduce the effects of drafts and convection. One nuisance with using a video card heatsink is that the back side typically has a raised area that contacts the GPU. For this application you'd have to have a flat back over the entire heatsink. Ed On 8/18/2014 6:58 PM, Angus wrote: On Fri, 04 Jul 2014 02:35:41 +0100, you wrote: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. I finally got around to playing with a couple of laptop heat pipes, fixed to a 25x50x75mm block of aluminium which is fixed to the 12mm thick baseplate. On a quick test of it, a sensor near the end of the baseplate showed 1.5-2x greater variation with temperature compared with just having a fan blow directly onto the baseplate. The oscillator also had to be allowed to run a few degrees C hotter for the heatpipe coolers to work to the same max ambient temp. One cooler had two heatpipes with about 12cm between the aluminium block and the heatsink fins (cast in this case) The other had a single, wider heat pipe with about 5cm between the block and the heatsink (this time with a lot more fine fins) The second cooler was rather more efficient, allowing a extra degreee of cooling at the top end, but more problematic was that it entered 'bang-bang' mode with the analogue temperature controller even sooner, and the temperature fluctuations there were greater. Both were rather worse than with the fan just blowing onto the baseplate. Using a PWM fan controller would help a good bit, but getting more creative with a microcontroller would be better. That way you can give the fan a minimum of a small kick every so often, and vary the repetition rate as well as the duty cycle as more cooling is needed. With feedback from the fan and even air temperature monitoring, you could get a good idea of exactly how much cooling was being applied. Another problem is that the overall temp control range is lower with the coolers - barely 8-9 DegC compared with 12+ DegC with the fan blowing directly on the baseplate. That's mainly the result of the poorer cooling at the top end of the range. The Rb osc fitted during this test was a SA.22c which takes a good bit less power than a 5680A, and the fan blowing onto the baseplate was normally a 60mm one fitted about 50mm away from it. The baseplate was horizontal with the fan blowing onto it from below. Maybe fitting a heatsink directly onto the base would help further with the maximum temp, but it would increase the convection cooling at the minimum temp, reducing the overall benefit. It could also be more susceptible to drafts, and would make the fan control much more delicate. Anyway, that's the results I got with my setup. Other setups and more fine tuning could change things a good bit, but I just wanted to get an idea of how the two cooling methods compared on the same setup. Angus On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider.
[time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
There is no question that direct fan control in combination with a heat sink is the best solution and we use it on FRK and M 100 with proper thermal insulation we get 0.01 C on the back plate and better than 0.1 C on the front. For us the FE 5680 A is not in that class so we looked for a solution that gives us 0.1 C. The shape of the FE 5680 does not lend itself easily for fan cooling if you want to mount it in a chassis horizontally, I did using two L shaped plates with a back plate heat sink and fan. How ever few have access to metal work and it gets quickly expensive. A picture is attached. I did extensive test with heat pipes first with a power resistor on a Alu plate followed by tests with a FRS, FE 5650 and FE 5680. You have to take in to consideration the function of the heat pipe in other words set the temperature of the base plate above the boiling point of the liquid. In my case 46 C was a good tradeoff between fan speed and operating range of the fan. To much heat pipe can also be a problem. No question a uprocessor controlled temperature control would be better, but till now typical time nuts, all talk while we have working analog circuits and boards. If worried about temperature change across the unit it can not be totally be eliminated but if important enclose the unit totally in foam. Easy when you use a heat pipe. I use a an 1/8 Alu base plate between the Rb and the heat pipe so I can also tap threads in to it to hold the heat pipe and I did away with the bottom plate of the FE 5680. Many options. Bert Kehren In a message dated 8/19/2014 3:02:49 A.M. Eastern Daylight Time, ed_pal...@sasktel.net writes: Is a heatpipe really appropriate for this application? The heatpipe expects that the heat source wants to burn up and so there's lots of heat available to vaporize the liquid in the pipe. It's not clear to me whether that situation exists with these Rb standards. My tests with an FE-5680A showed a maximum temperature of about 62C without a heatsink. That's far lower than a CPU or GPU. Some of them run at that temperature *with* the heatpipe. I think that a heatsink/fan (maybe from a video card) equipped with a PWM controller might be a better fit. Many of those combinations have a ducted fan to provide better control of the airflow. That would reduce the effects of drafts and convection. One nuisance with using a video card heatsink is that the back side typically has a raised area that contacts the GPU. For this application you'd have to have a flat back over the entire heatsink. Ed On 8/18/2014 6:58 PM, Angus wrote: On Fri, 04 Jul 2014 02:35:41 +0100, you wrote: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. I finally got around to playing with a couple of laptop heat pipes, fixed to a 25x50x75mm block of aluminium which is fixed to the 12mm thick baseplate. On a quick test of it, a sensor near the end of the baseplate showed 1.5-2x greater variation with temperature compared with just having a fan blow directly onto the baseplate. The oscillator also had to be allowed to run a few degrees C hotter for the heatpipe coolers to work to the same max ambient temp. One cooler had two heatpipes with about 12cm between the aluminium block and the heatsink fins (cast in this case) The other had a single, wider heat pipe with about 5cm between the block and the heatsink (this time with a lot more fine fins) The second cooler was rather more efficient, allowing a extra degreee of cooling at the top end, but more problematic was that it entered 'bang-bang' mode with the analogue temperature controller even sooner, and the temperature fluctuations there were greater. Both were rather worse than with the fan just blowing onto the baseplate. Using a PWM fan controller would help a good bit, but getting more creative with a microcontroller would be better. That way you can give the fan a minimum of a small kick every so often, and vary the repetition rate as well as the duty cycle as more cooling is needed. With feedback from the fan and even air temperature monitoring, you could get a good idea of exactly how much cooling was being applied. Another problem is that the
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
On Fri, 04 Jul 2014 02:35:41 +0100, you wrote: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. I finally got around to playing with a couple of laptop heat pipes, fixed to a 25x50x75mm block of aluminium which is fixed to the 12mm thick baseplate. On a quick test of it, a sensor near the end of the baseplate showed 1.5-2x greater variation with temperature compared with just having a fan blow directly onto the baseplate. The oscillator also had to be allowed to run a few degrees C hotter for the heatpipe coolers to work to the same max ambient temp. One cooler had two heatpipes with about 12cm between the aluminium block and the heatsink fins (cast in this case) The other had a single, wider heat pipe with about 5cm between the block and the heatsink (this time with a lot more fine fins) The second cooler was rather more efficient, allowing a extra degreee of cooling at the top end, but more problematic was that it entered 'bang-bang' mode with the analogue temperature controller even sooner, and the temperature fluctuations there were greater. Both were rather worse than with the fan just blowing onto the baseplate. Using a PWM fan controller would help a good bit, but getting more creative with a microcontroller would be better. That way you can give the fan a minimum of a small kick every so often, and vary the repetition rate as well as the duty cycle as more cooling is needed. With feedback from the fan and even air temperature monitoring, you could get a good idea of exactly how much cooling was being applied. Another problem is that the overall temp control range is lower with the coolers - barely 8-9 DegC compared with 12+ DegC with the fan blowing directly on the baseplate. That's mainly the result of the poorer cooling at the top end of the range. The Rb osc fitted during this test was a SA.22c which takes a good bit less power than a 5680A, and the fan blowing onto the baseplate was normally a 60mm one fitted about 50mm away from it. The baseplate was horizontal with the fan blowing onto it from below. Maybe fitting a heatsink directly onto the base would help further with the maximum temp, but it would increase the convection cooling at the minimum temp, reducing the overall benefit. It could also be more susceptible to drafts, and would make the fan control much more delicate. Anyway, that's the results I got with my setup. Other setups and more fine tuning could change things a good bit, but I just wanted to get an idea of how the two cooling methods compared on the same setup. Angus On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Good Morning Angus Let me start out by saying the FE5680A project was intended for time nuts and the unit will not be part of our bench, so the work on it is limited. As far as Rb's are concerned our focus is on FRK and HP5065. Looking back I would most likely not suggested it, had I known how much time it took up. How ever it also made us take an other look at Rb control, switching from analog to digital and resulted in looking at the FE 405 B. So it is worth it. Using heat pipe for the FE 5680 was the outcome of many test with heat sinks and other heat pipes. The big advantage of the heat pipe is the ease of removing the heat but it also has its limitation. Its biggest limitation is the fact that it works on the principle of converting a liquid to a vapor a process that takes a lot of heat. so you can not expect it to bring the temperature down to 20 degrees. And to much heat pipe is not good either. In the one you see in the picture I may have to deactivate one of the pipes. The ALU p;ate is 3/16 and the LM335 is on the back. I took out the bottom plate of the 5680 and between the two mounting holes of the ALU spine is plenty of room to install a SMD version of the 335. Temperature is 40C but not that critical but range is limited by the function of a heat pipe. You may want to experiment and if you are going to do some serious work contact me off list, not secrets, but the pictures can not be attached on this list because of file size limitation. When I first got the 5680 I did a crude temperature control using a fan a heat sink from an ATT Rb a LM335 along with a bang bang op amp circuit. I previously published picture and schematic. Temperature stability was better than 0.1 C. We have done extensive work on controlling the back plate temperature of a FRK and now get 0.01 C. We learned a lot and I like to share it since it will also apply to the heat pipe. The biggest problem is the fan. We do not have speed control but vary the speed by changing the fan voltage. We use a linear controller also previously published but have to be concerned about fan starting voltage. We want to be able to hold constant temperature over a 10C ambient change. Found that 12 V fan is best and at the same time looked for low noise. Learned also that more heat from the back plane helped with the operating range. Now all voltage regulators and the fan power transistor are along with heat sinks are mounted on the back plate. Side benefit is more stable voltage out of the regulators. This would be an ideal project for time nuts, since the results will apply to other Rb's not only the 5680 and benefit the community. I will stay out of the discussion since we have our hands full. Bert Kehren In a message dated 7/3/2014 9:37:27 P.M. Eastern Daylight Time, not.ag...@btinternet.com writes: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi There is a bunch of stuff on the Wentzel site about their double wall mini-test chambers. It’s probably the best approach to the gradient issue. You still have a gradient around the inner chamber / box, but it’s consistent. You eliminate the coupling of the gradient to the ambient temperature. The downside is the higher thermal resistance of the structure. That may force you right back to thermo electric cooling. I’d try a conventional heat pipe designed for high power desktop CPU cooling first. The things are cheap, come with way more fan than you need and the plate that mates to the CPU is about the right size for this application. Bob On Jul 3, 2014, at 9:35 PM, Angus not.ag...@btinternet.com wrote: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
At 07:23 AM 7/4/2014, Bob Camp wrote: There is a bunch of stuff on the Wentzel site about their double wall mini-test chambers. It's probably the best approach to the gradient issue. You still have a gradient around the inner chamber / box, but it's consistent. You eliminate the coupling of the gradient to the ambient temperature. I dug around a bit and found this: http://www.wenzel.com/wp-content/uploads/Sub-pico-Multiplier.pdf Relevant? -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi Yes, that’s one of the papers that shows their little chamber. I believe it’s mentioned in a couple of other places. Die cast boxes work pretty well for the enclosures. If you want to zero the gradient in the inner enclosure, put a small fan in there. That adds it’s own issues so it’s probably not really something you want to do. Bob On Jul 4, 2014, at 12:46 PM, Scott Newell newell+timen...@n5tnl.com wrote: At 07:23 AM 7/4/2014, Bob Camp wrote: There is a bunch of stuff on the Wentzel site about their double wall mini-test chambers. It's probably the best approach to the gradient issue. You still have a gradient around the inner chamber / box, but it's consistent. You eliminate the coupling of the gradient to the ambient temperature. I dug around a bit and found this: http://www.wenzel.com/wp-content/uploads/Sub-pico-Multiplier.pdf Relevant? -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
keep the heat sensing very close to the heating to avoid thermal control oscillation ! a thermal time constant is large and no to easy to deal with it 73 Alex On 7/3/2014 6:35 PM, Angus wrote: Hi Bert, I am thinking about testing a heat pipe on a fan cooled setup I use. The first temp controlled chassis I did used a peltier and works very well, but was a lot more work to do and is much more power hungry. The main problem I find is not the temp controller itself, but rather the change in the temperature across the chassis as the ambient changes. However good the temp controller is, it only controls a single point, but other points further away from the sensing thermistor can vary a lot. I noticed you posted a picture of a heat pipe cooler a couple of weeks ago - did you happen to compare the temperature across the unit with direct fan cooling and the heat pipe cooler, or with different heat pipes? Angus. On Sat, 28 Jun 2014 12:37:37 -0400 (EDT), you wrote: Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Scott, Could I ask you to post that C crap someplace? Would save me and others some time. This temperature thing is complex enough that I'd like to try some experiments. I want to run an FE5680 in an unheated attic where the temperature swings 40F and then plot some data. On Fri, Jun 27, 2014 at 8:40 PM, Scott Newell newell+timen...@n5tnl.com wrote: What software are you using to communicate with the FE5680. I'd like Just some C crap I tossed together. It's currently running under Cygwin. to try doe experiments. 1) see if we can take advantage some how of the existence of the temp sensor and 2) possibly use the analog input (a faked temp sensor) to control the FE5680. Servo controlling a fan based on the temp signal or VCXO sounds fun. I'm not sure you'd be happy with the fake temp signal--the unit is probably using a custom lookup table to correct the frequency vs temp. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
At 11:01 AM 6/29/2014, Chris Albertson wrote: Scott, Could I ask you to post that C crap someplace? Would save me and others some time. It was posted back in 2012: http://n5tnl.com/time/fe-5680a/control/fe5680_logger.c Beware that the comments may be inaccurate. We now know more about the commands than we did then. There's also a win32 .exe or two in that directory. I'll post something newer. This version spits out the time, the 4 ADC values, the two internal DDS tuning words, the diff, and the user tuning word, such as: 11.141 2163216321774 1140963061 1140662413 0 214 Maybe you can use it as a starting point for your tests: http://n5tnl.com/time/fe-5680a/control/fe5680_logger_4.c I wonder if commands 0x57 and 0x59 might be the tempco lookup table. I've made a rough graph of the transfer function, but it's not good enough to compare. Really need a slow 0-2.5 VDC ramp generator. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi, I fail to see what the benefit is of removing this unless a better temp compensation scheme is used. It is not likely to interfer with the external loop as it reduces the midterm noise that is systematic. It does add some higher rate noise but that is quantization errors of the systematics it reduces. I like to see measurement that support the claim and I am skeptic. As I see it you give the external loop more systematic noise to dampen and the tighter loop you make the more you will expose. Cheers, Magnus div Originalmeddelande /divdivFrån: Scott Newell newell+timen...@n5tnl.com /divdivDatum:28-06-2014 03:50 (GMT+01:00) /divdivTill: time-nuts@febo.com /divdivRubrik: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation) /divdiv /divBert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Because the 96K attachment needs moderator approval I am sending it again without attachment. It is the one I posted before and shows clearly the frequency jumps. So here is the rest of it. Magnus Sorry but I disagree with your statement. First and foremost we have to accept that these devices are not intended for time nuts (metrology). Already my first 1985 FRK monitors the cell current and adjusts the C field accordingly. The HP 5065 does not since it is not intended for the same market, it relies on its temperature control of the total A12 assembly. A closer look at FE specs you will notice the following statement including frequency over or undershoot at any fast or slow temperature slew rate. How do you think they do it. In the case of units that have a DDS in the control loop they do it with the DDS. How else do you explain the attached plot. Again my apologies for not remembering who posted it. I noticed the jumps and mentioned them when I first took a look at the 5680 I did use temperature control in a crude way using what I call the Bang Bang fan control on an ATT FRS heatsink. My YSI did not register any changes but the 5680 noticed and was visible using my Tracor 527 E. The 5680 is not a unit that we will use it turned in to a distraction but we did the controller for time nuts. Our focus when it comes to temperature control is on the FRK that is why Juerg who focused on the 5680 only uses a heat sink. Working out of a basement that is next to a garage data in the winter looks a lot more stable than now. We monitor the tuning word which has a resolution of 6.8 E-13 and when you see no change over long time periods in January and changes now exceeding E-12 now to start asking why. The differences are garage door is open more often and the sun shines part of the time directly in to his basement lab. When the loop is in the long time constant mode to take advantage of the GPS accuracy resolution which increases with time, jumps by the 5680 DDS are no help. It all depends what ultimate accuracy time nuts want out of the 5680. A good fan control along with a fixed setting the temperature ADC input used for frequency control will yield best results. Work by time nuts will help we are not going to do it. The reason is we are totally tied up with work on using the FE 405 B. I stumbled by accident on to it and bought some for testing. Initial tests show for me unbelievable performance specially when it comes to ADEV. It is perfect for GPSDO applications since it is all digital control with a step resolution of 6 e-15. How ever it is also not all perfect since I did detected jumps that I could not explain. Since our testing capabilities are limited I did make Tom aware of the unit and he has now caught the bug. http://leapsecond.com/pages/fe405/ . We think we have traced the problem to again frequency control using oven current have disabled it and hope to have better data to prove that the 3 E-12 jumps are due to the current monitor. Stay tuned. Hope it does not change the ADEV. Not many of you have OCXO's with that low ADEV. We are also using the controller for this device and results look very promising. That is why we call it the Universal Controller. In the future it will control many other devices. But for now please those of you that have the equpment and the expertize focus on the FE 5680 A to make it a viable low cost house reference for a large # of time nuts. Bert Kehren In a message dated 6/28/2014 3:21:48 A.M. Eastern Daylight Time, mag...@rubidium.dyndns.org writes: Hi, I fail to see what the benefit is of removing this unless a better temp compensation scheme is used. It is not likely to interfer with the external loop as it reduces the midterm noise that is systematic. It does add some higher rate noise but that is quantization errors of the systematics it reduces. I like to see measurement that support the claim and I am skeptic. As I see it you give the external loop more systematic noise to dampen and the tighter loop you make the more you will expose. Cheers, Magnus div Originalmeddelande /divdivFrån: Scott Newell newell+timen...@n5tnl.com /divdivDatum:28-06-2014 03:50 (GMT+01:00) /divdivTill: time-nuts@febo.com /divdivRubrik: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation) /divdiv /divBert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi Scott, I reviewed the material at the N5TNL site and it leaves me wondering. There are at least three different and separate temperature controlling areas inside the FEI-5680. One of which has cannot be messed with due to the fact that it is inherent in its design. That would be the posistor (i.e., a PTC resistor) that is attached to the 60 MHz crystal that acts as a general heater. The other two locations are on the physics package itself. One for the lamp area and the other is for the cavity chamber. SO. The lamp area, I surmise is rather simple as it does not require tight temperature control like the cavity chamber area would need. The lamp area just needs a temperature to change and maintain the Rb into a gas form. The cavity chamber area is way more sensitive to temperature as it affects the pressure, frequency and stability of the Rb in the cavity. I suspect that the control mechanism for this area to be more complicated. The block diagram, while pretty general in nature, does show that the system (i.e., internal computer) has an A/D monitoring 4 inputs. Three of the A/D channels are observing system data and the fourth is temperature. What it does not tell us is what temperature or system data it is monitoring. SO... Here is the rub, the only temperature that is truly critical is the cavity temperature. BUT, the system computer does not really control that temperature, except possibly monitor it, as indicated on the block diagram. Thus the real question is, is that the temperature being referred to in the monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? I would think that the temperature reading would have a steadily climbing curve from the application of power to some steady state (relative) value. However, the curve that you selected as temperature seems to rise and become steady (relative) some period of time after application of power and seems associated with the unit going into a lock condition. OR, so you indicate on some of your graphs. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I realize Bert is trying to take FEI to a tighter level, but I wonder if it would be better to add thermal mass to buffer external ambient changes rather than screw with the internal control mechanisms. Particularly seeing as how we have no knowledge of the what the internal firmware is doing. By thermal mass I mean on all sides of the unit. The only way to really achieve that would be make a stirred oil bath container with the FEI suspended in the center of said bath. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: http://www.worldkitchen.com/en/snapware-food-storage/1098437.html It is made out of polypropylene and can handle at least 130 degrees C and it holds just under two and half gallons of oil. Light mineral oil runs around $15 a gallon, so two gallons would be the right amount leaving a little room at the top. Wicking is a problem with wires and cables but using connectors attached to the lid solves that problem by breaking the wicking surface. The final question is how much, if any, external oil cooling would be necessary. That would have to be experimentally determined. The mineral oil, by the way, has a higher flash point then the container and is electrically NON-conductive. Food for thought, BillWB6BNQ Scott Newell wrote: Bert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Hi Most of these lightweight Rb’s do the same thing. They watch the oven current on one or the other section and try to guess the external temperature. Based on that guess they do a simple temperature correction on the unit. The older analog units feed a DC signal into the EFC. The newer digital units feed a signal into the DDS. In both cases (analog and digital), the ADEV of the units can be improved by disabling this “feature”. That of course assumes you are at a constant (as in very constant) abient temperature. In the case of the analog part, it’s the noise on the heater current that gets you. In the case of the digital approach, it’s the tuning granularity of the DDS that messes things up (and possibly heater current noise as well). How constant is “very constant”? That depends on the Rb you have. A good bet is that your device runs better than 2 to 4 ppb over a 100C range without the compensation turned on. That gives you 20 to 40 ppt per degree C. To hit 1 ppt you would need to control the device to better than 0.05 C. If you simply want to hit the 0.1 ppb temperature spec, then you only need a two degree control. If you look at the temperature compensation data words (ddd steps), some Rb’s in a batch are much better than others, so there is no easy way to be sure of the results ahead of time. Bob On Jun 28, 2014, at 3:21 AM, Magnus Danielson mag...@rubidium.dyndns.org wrote: Hi, I fail to see what the benefit is of removing this unless a better temp compensation scheme is used. It is not likely to interfer with the external loop as it reduces the midterm noise that is systematic. It does add some higher rate noise but that is quantization errors of the systematics it reduces. I like to see measurement that support the claim and I am skeptic. As I see it you give the external loop more systematic noise to dampen and the tighter loop you make the more you will expose. Cheers, Magnus div Originalmeddelande /divdivFrån: Scott Newell newell+timen...@n5tnl.com /divdivDatum:28-06-2014 03:50 (GMT+01:00) /divdivTill: time-nuts@febo.com /divdivRubrik: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation) /divdiv /divBert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
On Sat, Jun 28, 2014 at 4:32 AM, wb6bnq wb6...@cox.net wrote: wicking surface. The final question is how much, if any, external oil cooling would be necessary. That would have to be experimentally determined. The mineral oil, by the way, has a higher flash point then the container and is electrically NON-conductive. But it does have a dielectric constant different from air. How much effect would it have at microwave frequencies? Also, what would its effect be on the optical path from the Rb lamp and through the Rb cell? -- Brian Lloyd Lloyd Aviation 706 Flightline Drive Spring Branch, TX 78070 br...@lloyd.com +1.916.877.5067 ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Thanks Bob for putting it in perspective, 0.05 C is very doable, looks like it is also used in some OCXO's. Bert Kehren PS on a related subject I just pulled PROCEEDINGS OF THE ANNUAL SYMPOSIUM ON FREQUENCY CONLTROL ATLANTIC CITY in the late 70's did attend a few because of my involvement in GPS but then did not understand half of it and now want to check what I have learned since. Very interesting reading. In a message dated 6/28/2014 9:36:32 A.M. Eastern Daylight Time, kb...@n1k.org writes: Hi Most of these lightweight Rb’s do the same thing. They watch the oven current on one or the other section and try to guess the external temperature. Based on that guess they do a simple temperature correction on the unit. The older analog units feed a DC signal into the EFC. The newer digital units feed a signal into the DDS. In both cases (analog and digital), the ADEV of the units can be improved by disabling this “feature”. That of course assumes you are at a constant (as in very constant) abient temperature. In the case of the analog part, it ’s the noise on the heater current that gets you. In the case of the digital approach, it’s the tuning granularity of the DDS that messes things up (and possibly heater current noise as well). How constant is “very constant”? That depends on the Rb you have. A good bet is that your device runs better than 2 to 4 ppb over a 100C range without the compensation turned on. That gives you 20 to 40 ppt per degree C. To hit 1 ppt you would need to control the device to better than 0.05 C. If you simply want to hit the 0.1 ppb temperature spec, then you only need a two degree control. If you look at the temperature compensation data words (ddd steps), some Rb’s in a batch are much better than others, so there is no easy way to be sure of the results ahead of time. Bob On Jun 28, 2014, at 3:21 AM, Magnus Danielson mag...@rubidium.dyndns.org wrote: Hi, I fail to see what the benefit is of removing this unless a better temp compensation scheme is used. It is not likely to interfer with the external loop as it reduces the midterm noise that is systematic. It does add some higher rate noise but that is quantization errors of the systematics it reduces. I like to see measurement that support the claim and I am skeptic. As I see it you give the external loop more systematic noise to dampen and the tighter loop you make the more you will expose. Cheers, Magnus div Originalmeddelande /divdivFrån: Scott Newell newell+timen...@n5tnl.com /divdivDatum:28-06-2014 03:50 (GMT+01:00) /divdivTill: time-nuts@febo.com /divdivRubrik: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation) /divdiv /divBert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
I can only answer some of your questions. The tuning word is modified by the temperature but if it came back to the same oven current it will again have the same tuning word to the DDS. Since we have not fount a temperature sensor they most likely use oven current that Bob mentioned. If you do not plan to use it with a controller I would not recommend disconnecting it but if you have good temperature control you should not see a difference. The controller can not selectively enable or disable the temperature sensor and it would not make sense since the controller has no info as to temperature and enabling the frequency control do to temperature would most likely cause a frequency step.. The answer is use a laptop heat pipe it will do a great job when portable and takes little power, Just received some more recharable lithium 12 V 9.8 A batteries. Light weight and low cost Bert Kehren In a message dated 6/27/2014 11:04:31 P.M. Eastern Daylight Time, albertson.ch...@gmail.com writes: So the FE5680A will actually change the DDS tuning word based on an internal temperature sensor? I could see why you might ant to disable this or maybe not depending on how it works. Does the FE5680 first read the user programmed word, apply a delta then write it back or does it ignore user settings. I can't believe it would overwrite a user programmed running word. But this also means there is a sensitive temperature sensor inside the FE5680. Is there any way to read this sensor via the serial port? What software are you using to communicate with the FE5680. I'd like to try doe experiments. 1) see if we can take advantage some how of the existence of the temp sensor and 2) possibly use the analog input (a faked temp sensor) to control the FE5680. The problem with disconnecting the temperature sensor is much worse performance when the GPS is not available in hold over mode. GPS rarely fails but I'm sure some people disconnect the GPS to use the Rb as a portable frequency reference. It would be good if it were temperature compensated while in holdover. One experiment comes to mind: If the resister is removed, can there GPSDO controller selectively enable and disable temperature compensation? What software are you using to support your testing? On Fri, Jun 27, 2014 at 6:50 PM, Scott Newell newell+timen...@n5tnl.com wrote: Bert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Will someone beside us use heat pipe. Would love to have an impendent input. What does it take to get a test going. Scott has done a lot of work, how about some one else step up to the plate. There are a lot of time nuts out there with the 5680A,many for the first time will have a very good reference and some of our experts with proper equipment can make a big difference. Bert Kehren In a message dated 6/28/2014 12:20:12 P.M. Eastern Daylight Time, newell+timen...@n5tnl.com writes: At 04:32 AM 6/28/2014, wb6bnq wrote: monitoring process ? In other words have you traced out the connections to see what is driving the pin you think is the temperature input ? No. I've only traced back from the ADC input to the voltage divider. The next big question is have you monitored the frequency and its stability, externally, to observe what effects are taking place when you disable this input to the A/D ? I have not. That sounds complicated and messy but may be easier than it appears. An appropriate container would be: It does sound messy. I don't think I'm willing to dunk one of my units. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
Bert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
So the FE5680A will actually change the DDS tuning word based on an internal temperature sensor? I could see why you might ant to disable this or maybe not depending on how it works. Does the FE5680 first read the user programmed word, apply a delta then write it back or does it ignore user settings. I can't believe it would overwrite a user programmed running word. But this also means there is a sensitive temperature sensor inside the FE5680. Is there any way to read this sensor via the serial port? What software are you using to communicate with the FE5680. I'd like to try doe experiments. 1) see if we can take advantage some how of the existence of the temp sensor and 2) possibly use the analog input (a faked temp sensor) to control the FE5680. The problem with disconnecting the temperature sensor is much worse performance when the GPS is not available in hold over mode. GPS rarely fails but I'm sure some people disconnect the GPS to use the Rb as a portable frequency reference. It would be good if it were temperature compensated while in holdover. One experiment comes to mind: If the resister is removed, can there GPSDO controller selectively enable and disable temperature compensation? What software are you using to support your testing? On Fri, Jun 27, 2014 at 6:50 PM, Scott Newell newell+timen...@n5tnl.com wrote: Bert asked me to send an update on the FE-5680 tempco mod progress. It appears that the FE-5680A temperature signal (or maybe it's really a current sense signal?) can be disabled by removing a single 10k 0805 surface mount resistor. Using Elio Corbolante's terrific high-res scans, I've noted the resistor location: http://www.n5tnl.com/time/fe-5680a/lobotomy.png Why would you want to disable temperature compensation? As we've seen, the unit's firmware will adjust the DDS frequency as the temperature signal changes. If you're using the '5680 inside a control loop, it's likely to conflict. By removing the resistor, that channel of the 12 bit ADC will be tied to ground through an existing 2.21k resistor. The unit will see a constant 0 counts from the ADC and assume it's really cold. I modified one unit and monitored it for a few hours over a range of temps, running it nice and hot with no heatsink, then blasting it with a fan and placing it on an ice-cold heatsink. I observed no change in the DDS tuning words. It's a really easy mod--remove four screws, set aside the insulator sheet, and apply your hot leucotome/soldering iron. I've also found a simple mod to replace the temperature signal with the output of the unused trimpot. This allows you to simulate any temperature you want. If there's any interest, I'll set up a test and monitor the DDS tuning words as the unit's firmware tries to adjust to the fake temp signal. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
At 10:03 PM 6/27/2014, Chris Albertson wrote: So the FE5680A will actually change the DDS tuning word based on an internal temperature sensor? It appears so. (Again, assuming it's actually a temperature sensor. I guess it could be measuring heater current?) on how it works. Does the FE5680 first read the user programmed word, apply a delta then write it back or does it ignore user settings. I I've not done extensive testing with a user programmed frequency, but it does appear to apply the tempco correction to the user frequency. But this also means there is a sensitive temperature sensor inside the FE5680. Is there any way to read this sensor via the serial port? Yes. The temp signal goes to a 10k:2.21k divider, filtered by a cap, and then to pin 4 of the 4 channel 12 bit ADC. Command 0x22 returns the ADC readings as four 16 bit values. (There's also a VCXO related value, the trimpot, and what I thought was a rail or reference, but now I'm unsure on that one.) What software are you using to communicate with the FE5680. I'd like Just some C crap I tossed together. It's currently running under Cygwin. to try doe experiments. 1) see if we can take advantage some how of the existence of the temp sensor and 2) possibly use the analog input (a faked temp sensor) to control the FE5680. Servo controlling a fan based on the temp signal or VCXO sounds fun. I'm not sure you'd be happy with the fake temp signal--the unit is probably using a custom lookup table to correct the frequency vs temp. For example, I just spun the trimpot down from fullscale to 0. Here's the fake temp signal, the VCXO signal, and the two DDS tuning words as the unit tries to apply tempco correction: 40941870 1140962691 1140662043 40941859 1140962691 1140662043 21853 1140962974 1140662326 01845 1140962974 1140662326 -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DIY FE-5680A lobotomy (disable temp compensation)
At 10:40 PM 6/27/2014, Scott Newell wrote: to pin 4 of the 4 channel 12 bit ADC. Command 0x22 returns the ADC readings Doh! Correction: command 0x5A reads the ADC, command 0x22 reads the DDS tuning words. Sorry for the confusion. -- newell N5TNL ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
