[time-nuts] HP 5334 oscillator performance
I recently took one of my 5334A’s off of the house standard to lend to a friend. I thought it a good idea to check the onboard oscillator before lending it out. Both of my 5334A’s lack Opt 10 oscillators. This particular unit was almost 40Hz high in frequency. I tried to set it to frequency and had trouble getting it closer than 2Hz. The set-ability was very poor. I checked the service manual and found that HP only specified a final frequency within 8Hz. I can now see why. I then turned the instrument off for 12 hours. The next day I powered the unit back up. The oscillator was 20Hz higher than where I had measured it at shutdown the night before. It took well over an hour for it to re-stabilize and overshot the previous set point by 2Hz. At this point I decided to lend a Heathkit IM-2420 to the friend and continued experiments with the 5334A. I now took the second 5334A down and ran the same tests on it. This instrument was somewhat better but still shared many of the problems the first unit. 1) 60+ minutes to stabilize after power-up. Oscillators would start out 9 – 19Hz high and settle in at +- 1Hz. 2) Poor retrace at turn on. Final settling frequency can vary by 1Hz. 3) Poor set-ability due to the ceramic trimmer. 4) Oscillator pulling of 2 – 3Hz when I connected another counter to the rear panel oscillator jack. The oscillator buffer is on the same ECL chip as the oscillator. After looking at the circuit I decided that a new TCXO would be a better solution than trying to modify the existing circuit for better performance. Since these units are used at room temp most of the time, I need not worry about a wide temp range specification. I set the following goals for the replacement oscillator. 1) Improvement of stability by an order of magnitude and set-ability of two orders of magnitude. 2) Replacement to be built on a circuit board that would mount in the oven oscillator connector on the main circuit board. 3) Use of the full time 24vdc at this connector to power the oscillator at all times. Use of an on-card battery supply for oscillator backup. 4) Cheap 5) Use of readily available TCXO oscillator module. After checking Mouser and Digi-Key I settled on a Mouser supplied FOX801BE 10mHz unit @ $13.46. to start my experiments. This oscillator draws only 2ma. and frequency is set by an external trim pot. I would be interested in the experiences of others with respect to the performance of their non-oven 5334’s (or 5328’s) and ideas that support the design goals. Thanks Colin Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] HP 5334 oscillator performance
Hi Colin: A year or so ago I purchased a number of unit oscillators form a number of sources like Digi-Key, Jameco, eBay, etc. These were all XOs, not TCXOs. Some had no markings some were from well known makers. Almost all of them showed high aging rates indicating leaking packages, i.e. they were slowing down and the rate was not getting smaller. You might want to run the unit in a temperature controlled environment for a few days to be sure it's not defective. I did this using a lunch bucket which is an insulated cylinder with an opening of about 5 and about 6 deep. I used a home brewed heater set to the turnover temperature of the XO. Does the 5334 have an external reference input? Have Fun, Brooke Clarke w/Java http://www.PRC68.com w/o Java http://www.pacificsites.com/~brooke/PRC68COM.shtml http://www.precisionclock.com Colin Bradley wrote: I recently took one of my 5334A’s off of the house standard to lend to a friend. I thought it a good idea to check the onboard oscillator before lending it out. Both of my 5334A’s lack Opt 10 oscillators. This particular unit was almost 40Hz high in frequency. I tried to set it to frequency and had trouble getting it closer than 2Hz. The set-ability was very poor. I checked the service manual and found that HP only specified a final frequency within 8Hz. I can now see why. I then turned the instrument off for 12 hours. The next day I powered the unit back up. The oscillator was 20Hz higher than where I had measured it at shutdown the night before. It took well over an hour for it to re-stabilize and overshot the previous set point by 2Hz. At this point I decided to lend a Heathkit IM-2420 to the friend and continued experiments with the 5334A. I now took the second 5334A down and ran the same tests on it. This instrument was somewhat better but still shared many of the problems the first unit. 1) 60+ minutes to stabilize after power-up. Oscillators would start out 9 – 19Hz high and settle in at +- 1Hz. 2) Poor retrace at turn on. Final settling frequency can vary by 1Hz. 3) Poor set-ability due to the ceramic trimmer. 4) Oscillator pulling of 2 – 3Hz when I connected another counter to the rear panel oscillator jack. The oscillator buffer is on the same ECL chip as the oscillator. After looking at the circuit I decided that a new TCXO would be a better solution than trying to modify the existing circuit for better performance. Since these units are used at room temp most of the time, I need not worry about a wide temp range specification. I set the following goals for the replacement oscillator. 1) Improvement of stability by an order of magnitude and set-ability of two orders of magnitude. 2) Replacement to be built on a circuit board that would mount in the oven oscillator connector on the main circuit board. 3) Use of the full time 24vdc at this connector to power the oscillator at all times. Use of an on-card battery supply for oscillator backup. 4) Cheap 5) Use of readily available TCXO oscillator module. After checking Mouser and Digi-Key I settled on a Mouser supplied FOX801BE 10mHz unit @ $13.46. to start my experiments. This oscillator draws only 2ma. and frequency is set by an external trim pot. I would be interested in the experiences of others with respect to the performance of their non-oven 5334’s (or 5328’s) and ideas that support the design goals. Thanks Colin Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Re: [time-nuts] HP 5334 oscillator performance
I was the project manager of the 5334B, which was a reduced factory cost replacement for the 5334A. Of course I thoroughly evaluated the 5334A before deciding what changes to make. I will apologize on behalf of the company for that oscillator in the 34A. It is a really poor design. Giving the 5334A project team the benefit of the doubt, I guess the philosophy was that the built in oscillator is for very non-critical applications and they didn't want to increase the cost to those users but putting a decent oscillator in the standard counter. Many users were like us and had a house frequency standard, so we never used a built in oscillator in a counter. Due to time and cost pressures, we unfortunately inherited the same oscillator in the 5334B. So in both products, please plan on getting a 10811 or using an external frequency reference. Rick Karlquist N6RK Project manager, 5334B, circa 1986 Colin Bradley wrote: I recently took one of my 5334As off of the house standard to lend to a friend. I thought it a good idea to check the onboard oscillator before lending it out. Both of my 5334As lack Opt 10 oscillators. This particular unit was almost 40Hz high in frequency. I tried to set it to frequency and had trouble getting it closer than 2Hz. The set-ability was very poor. I checked the service manual and found that HP only specified a final frequency within 8Hz. I can now see why. I then turned the instrument off for 12 hours. The next day I powered the unit back up. The oscillator was 20Hz higher than where I had measured it at shutdown the night before. It took well over an hour for it to re-stabilize and overshot the previous set point by 2Hz. At this point I decided to lend a Heathkit IM-2420 to the friend and continued experiments with the 5334A. I now took the second 5334A down and ran the same tests on it. This instrument was somewhat better but still shared many of the problems the first unit. 1) 60+ minutes to stabilize after power-up. Oscillators would start out 9 19Hz high and settle in at +- 1Hz. 2) Poor retrace at turn on. Final settling frequency can vary by 1Hz. 3) Poor set-ability due to the ceramic trimmer. 4) Oscillator pulling of 2 3Hz when I connected another counter to the rear panel oscillator jack. The oscillator buffer is on the same ECL chip as the oscillator. After looking at the circuit I decided that a new TCXO would be a better solution than trying to modify the existing circuit for better performance. Since these units are used at room temp most of the time, I need not worry about a wide temp range specification. I set the following goals for the replacement oscillator. 1) Improvement of stability by an order of magnitude and set-ability of two orders of magnitude. 2) Replacement to be built on a circuit board that would mount in the oven oscillator connector on the main circuit board. 3) Use of the full time 24vdc at this connector to power the oscillator at all times. Use of an on-card battery supply for oscillator backup. 4) Cheap 5) Use of readily available TCXO oscillator module. After checking Mouser and Digi-Key I settled on a Mouser supplied FOX801BE 10mHz unit @ $13.46. to start my experiments. This oscillator draws only 2ma. and frequency is set by an external trim pot. I would be interested in the experiences of others with respect to the performance of their non-oven 5334s (or 5328s) and ideas that support the design goals. Thanks Colin Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
