Re: [time-nuts] What's the best way to double 10 MHz to 20 MHz ?
On 05/25/11 07:39 PM, Chuck Forsberg WA7KGX N2469R wrote: If the radio's clock can be trimmed with a voltage, why not divide the radio's 20 MHz clock by two and feed the result into the GPSDO in place of the GPSDO's 10 MHz oscillator. It can not be trimmed via a voltage. Also, I'm not keen to require an external reference. -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? ___ 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] What's the best way to double 10 MHz to 20 MHz ?
Hi You also may want to avoid an oscillator with much 10 MHz content in it. All sorts of odd things can happen with spurs when you have unplanned stuff on the main reference. Another thing to look closely at is - how much of the radio tracks the reference? On some radios, they don't really lock everything up. You get better performance, but not quite what you would expect. Bob On May 27, 2011, at 3:55 AM, Dr. David Kirkby wrote: On 05/25/11 07:39 PM, Chuck Forsberg WA7KGX N2469R wrote: If the radio's clock can be trimmed with a voltage, why not divide the radio's 20 MHz clock by two and feed the result into the GPSDO in place of the GPSDO's 10 MHz oscillator. It can not be trimmed via a voltage. Also, I'm not keen to require an external reference. -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? ___ 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] What's the best way to double 10 MHz to 20 MHz ?
On 27 May 2011 15:21, Bob Camp li...@rtty.us wrote: Hi You also may want to avoid an oscillator with much 10 MHz content in it. All sorts of odd things can happen with spurs when you have unplanned stuff on the main reference. Another thing to look closely at is - how much of the radio tracks the reference? On some radios, they don't really lock everything up. You get better performance, but not quite what you would expect. Bob Yes, a band-pass filter would seem sensible here. To the best of my knowledge, everything is locked to this 20 MHz, so there should be a significant improvement. However, I must admit as to wondering whether its worth the bother. I have a Yaesu FT-ONE, which is a pretty poor design, despite it was the top of the line Yaesu transceiver in its day (~1982) costing $3000. It has a synthesizer for 100 Hz steps and then uses a varicap diode to get the 10 Hz steps! This is not very stable, but in practice is stable enough. There's nothing much one can do about that - perhaps keeping the temperature constant in the vicinity of the varicap and other critical components would help. But it also suffers from the use of more than one reference, so hard to really stabilise. The Kenwood TS-940SAT should be a lot more stable anyway. The Kenwood actually seems a lot better technically to me, despite it costs only $2000 and was released about the same time as the Yause FT-ONE. The Kenwood TS-940S has a built in ATU (not even an option on the Yaesu FT-ONE), FM (optional on the Yaesu FT-ONE), TCXO (not even optional on the Yaesu). I might however look at using a TCXO o OCXO in the Kenwood. Dave ___ 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] What's the best way to double 10 MHz to 20 MHz ?
Imho phase noise is probably as important as long term stability in this application. On Fri May 27th, 2011 10:51 AM EDT David Kirkby wrote: On 27 May 2011 15:21, Bob Camp li...@rtty.us wrote: Hi You also may want to avoid an oscillator with much 10 MHz content in it. All sorts of odd things can happen with spurs when you have unplanned stuff on the main reference. Another thing to look closely at is - how much of the radio tracks the reference? On some radios, they don't really lock everything up. You get better performance, but not quite what you would expect. Bob Yes, a band-pass filter would seem sensible here. To the best of my knowledge, everything is locked to this 20 MHz, so there should be a significant improvement. However, I must admit as to wondering whether its worth the bother. I have a Yaesu FT-ONE, which is a pretty poor design, despite it was the top of the line Yaesu transceiver in its day (~1982) costing $3000. It has a synthesizer for 100 Hz steps and then uses a varicap diode to get the 10 Hz steps! This is not very stable, but in practice is stable enough. There's nothing much one can do about that - perhaps keeping the temperature constant in the vicinity of the varicap and other critical components would help. But it also suffers from the use of more than one reference, so hard to really stabilise. The Kenwood TS-940SAT should be a lot more stable anyway. The Kenwood actually seems a lot better technically to me, despite it costs only $2000 and was released about the same time as the Yause FT-ONE. The Kenwood TS-940S has a built in ATU (not even an option on the Yaesu FT-ONE), FM (optional on the Yaesu FT-ONE), TCXO (not even optional on the Yaesu). I might however look at using a TCXO o OCXO in the Kenwood. Dave ___ 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] What's the best way to double 10 MHz to 20 MHz ?
Hi Looking at this in terms of time nut type stability - you really want to lock the 20 MHz up to something like a TBolt. At that point you have a wire out the back of the radio and all that implies. Keeping the ground loops and RF issues at bay in a transmitter is not trivial. For a timing receiver something like a R1051 or an R6790 might be a better starting point. If it's an internal oscillator (because of the various issues) - spend the time to dig up a 20 MHz part. You loose the Time Nut / WWV is off by 0.0001 Hz capability. You gain a spur free radio that doesn't do something odd when you transmit at frequency using mode . A lot of outfits had a hard time with the sub 100 Hz synthesizer steps back before modern PLL and DDS chips came out. Many of the big boys in the military radio business put out radios that had very predictable spurs from the fine tuning synth. They can be a real pain if the idea is to track something like WWV and extract carrier phase or timing from it. Bob On May 27, 2011, at 10:51 AM, David Kirkby wrote: On 27 May 2011 15:21, Bob Camp li...@rtty.us wrote: Hi You also may want to avoid an oscillator with much 10 MHz content in it. All sorts of odd things can happen with spurs when you have unplanned stuff on the main reference. Another thing to look closely at is - how much of the radio tracks the reference? On some radios, they don't really lock everything up. You get better performance, but not quite what you would expect. Bob Yes, a band-pass filter would seem sensible here. To the best of my knowledge, everything is locked to this 20 MHz, so there should be a significant improvement. However, I must admit as to wondering whether its worth the bother. I have a Yaesu FT-ONE, which is a pretty poor design, despite it was the top of the line Yaesu transceiver in its day (~1982) costing $3000. It has a synthesizer for 100 Hz steps and then uses a varicap diode to get the 10 Hz steps! This is not very stable, but in practice is stable enough. There's nothing much one can do about that - perhaps keeping the temperature constant in the vicinity of the varicap and other critical components would help. But it also suffers from the use of more than one reference, so hard to really stabilise. The Kenwood TS-940SAT should be a lot more stable anyway. The Kenwood actually seems a lot better technically to me, despite it costs only $2000 and was released about the same time as the Yause FT-ONE. The Kenwood TS-940S has a built in ATU (not even an option on the Yaesu FT-ONE), FM (optional on the Yaesu FT-ONE), TCXO (not even optional on the Yaesu). I might however look at using a TCXO o OCXO in the Kenwood. Dave ___ 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] What's the best way to double 10 MHz to 20 MHz ?
On Fri, May 27, 2011 at 6:10 PM, Mark Spencer mspencer12...@yahoo.ca wrote: Imho phase noise is probably as important as long term stability in this application. for real and serious amateur radio dxing it's much more important the phase noise and IMD3 performance of the RX rather than stability. Not that stability doesn't matter, but I'd never trade not excellent PN for stability. Best regards Frank IZ8DWF ___ 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] What's the best way to double 10 MHz to 20 MHz?
List, How about the LM1496 IC DBM? Simple and cheap. Regards, Perrier ___ 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] What's the best way to double 10 MHz to 20 MHz?
Hi As mentioned earlier, phase noise is critical. That would be the main issue. The original suggestion of a diode rectifier doubler has a much lower noise floor. Done from scratch they don't cost much at all. Bob On May 27, 2011, at 12:35 PM, Perry Sandeen wrote: List, How about the LM1496 IC DBM? Simple and cheap. Regards, Perrier ___ 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] What's the best way to double 10 MHz to 20 MHz ?
Yep. But in this case the original poster was looking to improve stability. Perhaps a better way to have phrased my comment would have been to ensure that the phase noise is not degraded vs the original xtal oscilator. I do agree though that phase noise is very important. As a side note I became interested in time nut topics after I purchased a gpsdo to compare the tcxo in my icom 706mkiig to. I then needed to convince my self the the gpsdo was accurate etc (: I did conclude that the tcxo in my icom 706mkiig was jumping by a few tenths of a ppm from time to time (: On Fri May 27th, 2011 12:17 PM EDT francesco messineo wrote: On Fri, May 27, 2011 at 6:10 PM, Mark Spencer mspencer12...@yahoo.ca wrote: Imho phase noise is probably as important as long term stability in this application. for real and serious amateur radio dxing it's much more important the phase noise and IMD3 performance of the RX rather than stability. Not that stability doesn't matter, but I'd never trade not excellent PN for stability. Best regards Frank IZ8DWF ___ 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] Strange temperature peak
I left my Thunderbolt running with Lady Heather started. Returning after a few hours in the room, which is at a constant temperature (underground, no heating, no air conditioning), I found the following plot on the Lady Heather screen : http://www.sdradio.eu/images/ladyheather.gif which shows a narrow and pronounced peak in the temperature. Does anybody have a possible explanation for this ? TNX 73 Alberto I2PHD ___ 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] Strange temperature peak
Hi Alberto, On Fri, May 27, 2011 at 8:45 PM, Alberto di Bene dib...@usa.net wrote: I left my Thunderbolt running with Lady Heather started. Returning after a few hours in the room, which is at a constant temperature (underground, no heating, no air conditioning), I found the following plot on the Lady Heather screen : http://www.sdradio.eu/images/ladyheather.gif since the plot has a step jump, seems only a few tens mC, and then it comes slowly back to the normal track, I'd rule out at least an external temperature change: the thunderbolt can easily detect an hand in its proximity even for a few seconds, but I don't remember ever seen a step change. Could be a firmware/sensor/communication error, but I'm not able to explain why it comes back to normal slowly (but I'm sure there's a firmware reason for that). Frank IZ8DWF ___ 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] Strange temperature peak
On 5/27/2011 8:59 PM, francesco messineo wrote: since the plot has a step jump, seems only a few tens mC, and then it comes slowly back to the normal track, I'd rule out at least an external temperature change: the thunderbolt can easily detect an hand in its proximity even for a few seconds, but I don't remember ever seen a step change. Could be a firmware/sensor/communication error, but I'm not able to explain why it comes back to normal slowly (but I'm sure there's a firmware reason for that). Frank, thanks. I can obviously rule out a hand or a body in proximity of the GPSDO, as when that happened I was alone in the house and moreover the door was locked. I would tend to consider it some sort of glitch either in the firmware or in the temperature sensor. I would be curious to know if somebody else has ever observed in his Thunderbolt a similar phenomenon. 73 Alberto I2PHD ___ 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] Strange temperature peak
Does anybody have a possible explanation for this ? Ghosts, of course! Likely an issue with the temp sensor, their conversion isn't necessarily flat. ___ 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] Strange temperature peak
Hi If you watch the thermometer on the TBolt for long enough, you will indeed see narrow temperature spikes. The gif you posted is a very typical spike. They are fairly rare and they don't repeat. I believe LH averages the readings it gets, so they may simply be a noise burst. The initial jump is the noise pop, and the decay is the averaging taking it out. Bob On May 27, 2011, at 3:32 PM, David VanHorn wrote: Does anybody have a possible explanation for this ? Ghosts, of course! Likely an issue with the temp sensor, their conversion isn't necessarily flat. ___ 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] Strange temperature peak
I have seen it, as a matter of fact two days ago I did see a straight jump in excess of 1 C. I have seen it before. I do not monitor constantly, but will keep an eye out. Bert Kehren In a message dated 5/27/2011 3:19:46 P.M. Eastern Daylight Time, dib...@usa.net writes: On 5/27/2011 8:59 PM, francesco messineo wrote: since the plot has a step jump, seems only a few tens mC, and then it comes slowly back to the normal track, I'd rule out at least an external temperature change: the thunderbolt can easily detect an hand in its proximity even for a few seconds, but I don't remember ever seen a step change. Could be a firmware/sensor/communication error, but I'm not able to explain why it comes back to normal slowly (but I'm sure there's a firmware reason for that). Frank, thanks. I can obviously rule out a hand or a body in proximity of the GPSDO, as when that happened I was alone in the house and moreover the door was locked. I would tend to consider it some sort of glitch either in the firmware or in the temperature sensor. I would be curious to know if somebody else has ever observed in his Thunderbolt a similar phenomenon. 73 Alberto I2PHD ___ 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] Strange temperature peak
I confirm Bob's statement. Such spikes seem to be normal. I am running T-Bolts already since the begin of the time nut action initiated by Tom (TvB) and with a temp resolution of 20m°C. I observed always such spikes up to a few times a day. I do not see a real practical problem. Arnold Am 27.05.2011 21:41, schrieb Bob Camp: Hi If you watch the thermometer on the TBolt for long enough, you will indeed see narrow temperature spikes. The gif you posted is a very typical spike. They are fairly rare and they don't repeat. I believe LH averages the readings it gets, so they may simply be a noise burst. The initial jump is the noise pop, and the decay is the averaging taking it out. Bob On May 27, 2011, at 3:32 PM, David VanHorn wrote: Does anybody have a possible explanation for this ? Ghosts, of course! Likely an issue with the temp sensor, their conversion isn't necessarily flat. ___ 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] Strange temperature peak
Does anybody have a possible explanation for this ? Short answer is the spike is caused by the temperature sensor. This is very typical on all Tbolts at certain temperatures, which depend on the temperature's rate of change and direction. You can change LH's temperature display resolution to hide the small spikes by using G T S 1000 enter. More details: The Tbolt uses a special method to get high resolution data from its temperature sensor by combining three things. The first is the sensor's standard data output which is used for the most significant bits. This gives about 1 deg resolution. 2nd an internal register is used for the middle bits which gives resolution of 0.009 deg and then the Tbolt's firmware averages the internal register data to get the LS bits, with resolution (not accuracy) below 0.001 deg. The return Time constant you see after the jump is the Tbolt's averager working on the Middle bits. All and all it works pretty good MOST of the time, except at major transfer points. The high resolution data interpretation does not work on the e revision sensor parts, so some of the later Tbolts show temp jumps of about 1 deg As far as I know, Lady Heather does special averaging of the temperature data at two different times, neither which apply to your case, so what you see is what the Tbolt is sending out. LH will average the temperature data IF the display filter is turned on with F D 100 Enter, which yours is not, and the LH's temperature controller has a special routine that cuts those spikes out (does not average them in) so that they do not upset LH's S/W temperature control PID loop . ws * [time-nuts] Strange temperature peak I left my Thunderbolt running with Lady Heather started. Returning after a few hours in the room, which is at a constant temperature (underground, no heating, no air conditioning), I found the following plot on the Lady Heather screen : http://www.sdradio.eu/images/ladyheather.gif which shows a narrow and pronounced peak in the temperature. Does anybody have a possible explanation for this ? TNX 73 Alberto I2PHD ___ 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] Strange temperature peak
Maybe a nearby cell phone calling home? Bob Camp wrote: Hi If you watch the thermometer on the TBolt for long enough, you will indeed see narrow temperature spikes. The gif you posted is a very typical spike. They are fairly rare and they don't repeat. I believe LH averages the readings it gets, so they may simply be a noise burst. The initial jump is the noise pop, and the decay is the averaging taking it out. Bob ___ 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] Strange temperature peak
Hi Could easily be. The spikes seem to be random and that would be a source of random RF. Bob On May 27, 2011, at 7:41 PM, Chuck Harris wrote: Maybe a nearby cell phone calling home? Bob Camp wrote: Hi If you watch the thermometer on the TBolt for long enough, you will indeed see narrow temperature spikes. The gif you posted is a very typical spike. They are fairly rare and they don't repeat. I believe LH averages the readings it gets, so they may simply be a noise burst. The initial jump is the noise pop, and the decay is the averaging taking it out. Bob ___ 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] Strange temperature peak
The spikes are due to an artifact in the way the Tbolt firmware reads the temperature sensor chip. It reads two registers and combine the values to get a high-res temp reading. But if the firmware accesses the registers in-between the time that the temp sensor chip updates them it can produce a bogus value. Typically this seems to happen on average of once every couple of hours. Lady Heather flags each spike with a red line at the top of the plot window. Other error events such as holdovers, bad serial data, etc are flagged in the same way... ___ 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] Strange temperature peak
Hello In my case I think it's not a rf problem. I am far away from rf or cell phone stations and the shack is well protected against rf from outside and no cell phone nor other transmitters were used in this room at this time. If it would be an emc problem - why is only the temperature output affected? I rather believe it is an internal digitizing problem, this chip is specified to work with 0.5 deg. C steps, if I am not wrong, and the higher resolution is achieved by a programming trick. Arnold Am 28.05.2011 03:59, schrieb Bob Camp: Hi Could easily be. The spikes seem to be random and that would be a source of random RF. Bob On May 27, 2011, at 7:41 PM, Chuck Harris wrote: Maybe a nearby cell phone calling home? Bob Camp wrote: Hi If you watch the thermometer on the TBolt for long enough, you will indeed see narrow temperature spikes. The gif you posted is a very typical spike. They are fairly rare and they don't repeat. I believe LH averages the readings it gets, so they may simply be a noise burst. The initial jump is the noise pop, and the decay is the averaging taking it out. Bob ___ 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.
