Hi All, This is my first time on the list so apologies if I am going over an old topic. I am repairing a HP 8662A Synthesised Signal Generator that I purchased from a government auction. The unit was purchased without any information about its working state and I have fixing various issues with it. Firstly there was a digital logic problem with the keyboard card that was causing the generator to periodically lock up. When this was fixed the unit came to life but the output signal level was down 5dB and this was traced to a bad step attenuator. I attempted a fix but ended up needing to replace this to get the output steps correct. I have now turned my attention to calibrating the reference timebase and unfortunately this seems to be well out of spec. The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the oscillator seems to be ~40Hz low. I have attempted to adjust the unit via the multi-turn capacitor adjustment but this seems to be at the end stop and the frequency is still low. So my questions are: 1) I assume this module is faulty as it is so far out of spec? The signal generator has been powered off for multiple months but even after having the unit powered on for 24 hours the frequency has not changed significantly. 2) How repairable are these modules? I ended up blowing $200 on the replacement step attenuator so will need to save up for any new parts. I can always use an external 10MHz reference but it is always nice to have equipment complete. 3) From searching the list it seems that the oven could be what is faulty. Would a cold over account for this much error?I have confirmed that the edge connector is supplying 25V on the over supply connections but I don't think the module is warming up. 4) Is it possible  the OCXO has just worn out and has aged beyond the point it can be adjusted back into spec? Regards, Denis

Hi Dennis, The HP 10811 are serviceable. Just download the service manual and study it. I would expect the oven to be faulty with this much error. Most likely the thermal fuse has gone open circuit. The fuse can go open circuit without there actually being any fault with the oven, but there might also be a fault in the oven circuit. The fuse is on the first PCB you get to after opening up the 10811. All described in the manual. If its the thermal fuse, replace with correct thermal fuse. When replacing make sure to not over heat it while soldering. You can of course bridge the fuse with a wire for testing, but then you need to monitor it closely, to be sure it does not over heat. I was lucky to find a HP 8663A in a electronic recycling bin. It was missing PCB's in the PSU and the 10811 was totally burnt to a crisp. Guess the fuse had been bridged. So I needed to get a replacement off ebay. Regards, Askild On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < time-nuts@lists.febo.com> wrote: > Hi All, > > This is my first time on the list so apologies if I am going over an old > topic. > > I am repairing a HP 8662A Synthesised Signal Generator that I purchased > from a government auction. The unit was purchased without any > information about its working state and I have fixing various issues > with it. Firstly there was a digital logic problem with the keyboard > card that was causing the generator to periodically lock up. When this > was fixed the unit came to life but the output signal level was down 5dB > and this was traced to a bad step attenuator. I attempted a fix but > ended up needing to replace this to get the output steps correct. > > I have now turned my attention to calibrating the reference timebase and > unfortunately this seems to be well out of spec. > > The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the > oscillator seems to be ~40Hz low. I have attempted to adjust the unit > via the multi-turn capacitor adjustment but this seems to be at the end > stop and the frequency is still low. So my questions are: > > 1) I assume this module is faulty as it is so far out of spec? The > signal generator has been powered off for multiple months but even after > having the unit powered on for 24 hours the frequency has not changed > significantly. > > 2) How repairable are these modules? I ended up blowing $200 on the > replacement step attenuator so will need to save up for any new parts. I > can always use an external 10MHz reference but it is always nice to have > equipment complete. > > 3) From searching the list it seems that the oven could be what is > faulty. Would a cold over account for this much error?I have confirmed > that the edge connector is supplying 25V on the over supply connections > but I don't think the module is warming up. > > 4) Is it possible the OCXO has just worn out and has aged beyond the > point it can be adjusted back into spec? > > > Regards, > Denis > > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

I had a 10811 that didn't warm up. I'd never taken one apart before but, although a bunch of parts had failed, they were all common (if ancient) op-amps and the like. I didn't have much trouble fixing it once I'd got started. I'd encourage you to go ahead. On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < time-nuts@lists.febo.com> wrote: > Hi Dennis, > > The HP 10811 are serviceable. Just download the service manual and study > it. > I would expect the oven to be faulty with this much error. > Most likely the thermal fuse has gone open circuit. > The fuse can go open circuit without there actually being any fault with > the oven, but there might also be a fault in the oven circuit. > The fuse is on the first PCB you get to after opening up the 10811. All > described in the manual. > If its the thermal fuse, replace with correct thermal fuse. When replacing > make sure to not over heat it while soldering. > You can of course bridge the fuse with a wire for testing, but then you > need to monitor it closely, to be sure it does not over heat. > > I was lucky to find a HP 8663A in a electronic recycling bin. It was > missing PCB's in the PSU and the 10811 was totally burnt to a crisp. > Guess the fuse had been bridged. So I needed to get a replacement off ebay. > > Regards, > Askild > > > > On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < > time-nuts@lists.febo.com> wrote: > > > Hi All, > > > > This is my first time on the list so apologies if I am going over an old > > topic. > > > > I am repairing a HP 8662A Synthesised Signal Generator that I purchased > > from a government auction. The unit was purchased without any > > information about its working state and I have fixing various issues > > with it. Firstly there was a digital logic problem with the keyboard > > card that was causing the generator to periodically lock up. When this > > was fixed the unit came to life but the output signal level was down 5dB > > and this was traced to a bad step attenuator. I attempted a fix but > > ended up needing to replace this to get the output steps correct. > > > > I have now turned my attention to calibrating the reference timebase and > > unfortunately this seems to be well out of spec. > > > > The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the > > oscillator seems to be ~40Hz low. I have attempted to adjust the unit > > via the multi-turn capacitor adjustment but this seems to be at the end > > stop and the frequency is still low. So my questions are: > > > > 1) I assume this module is faulty as it is so far out of spec? The > > signal generator has been powered off for multiple months but even after > > having the unit powered on for 24 hours the frequency has not changed > > significantly. > > > > 2) How repairable are these modules? I ended up blowing $200 on the > > replacement step attenuator so will need to save up for any new parts. I > > can always use an external 10MHz reference but it is always nice to have > > equipment complete. > > > > 3) From searching the list it seems that the oven could be what is > > faulty. Would a cold over account for this much error?I have confirmed > > that the edge connector is supplying 25V on the over supply connections > > but I don't think the module is warming up. > > > > 4) Is it possible the OCXO has just worn out and has aged beyond the > > point it can be adjusted back into spec? > > > > > > Regards, > > Denis > > > > > > _______________________________________________ > > time-nuts mailing list -- time-nuts@lists.febo.com > > To unsubscribe send an email to time-nuts-leave@lists.febo.com > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

Thanks everyone for all of the advice. As predicted the fault was an open circuit thermal fuse. I have bypassed this (temporarily) and I can now trim the oscillator to be 10MHz to the precision of my PM 6672 frequency counter. I am plotting both 10MHz waveforms on a scope and there is minimal slip. I will order a new fuse and then reassemble and then attempt a more accurate calibration when everything has  been allowed to burn in for a while. On a related note, how good is the Hewlett Packet March 1981 Journal article on the HP 10811D? http://hparchive.com/Journals/HPJ-1981-03.pdf There is so much information in this article including how they used a flex PCB to make the oscillator easier to assemble. They even discuss the issues they had with stray capacitance due to the thinner PCB and tearing on sharp edges. Unfortunately today I doubt any manufacturer would be as open about their design process. Regards, Denis On 28/03/2023 6:44 pm, Adrian Godwin via time-nuts wrote: > I had a 10811 that didn't warm up. I'd never taken one apart before but, > although a bunch of parts had failed, they were all common (if ancient) > op-amps and the like. I didn't have much trouble fixing it once I'd got > started. I'd encourage you to go ahead. > > On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < > time-nuts@lists.febo.com> wrote: > >> Hi Dennis, >> >> The HP 10811 are serviceable. Just download the service manual and study >> it. >> I would expect the oven to be faulty with this much error. >> Most likely the thermal fuse has gone open circuit. >> The fuse can go open circuit without there actually being any fault with >> the oven, but there might also be a fault in the oven circuit. >> The fuse is on the first PCB you get to after opening up the 10811. All >> described in the manual. >> If its the thermal fuse, replace with correct thermal fuse. When replacing >> make sure to not over heat it while soldering. >> You can of course bridge the fuse with a wire for testing, but then you >> need to monitor it closely, to be sure it does not over heat. >> >> I was lucky to find a HP 8663A in a electronic recycling bin. It was >> missing PCB's in the PSU and the 10811 was totally burnt to a crisp. >> Guess the fuse had been bridged. So I needed to get a replacement off ebay. >> >> Regards, >> Askild >> >> >> >> On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < >> time-nuts@lists.febo.com> wrote: >> >>> Hi All, >>> >>> This is my first time on the list so apologies if I am going over an old >>> topic. >>> >>> I am repairing a HP 8662A Synthesised Signal Generator that I purchased >>> from a government auction. The unit was purchased without any >>> information about its working state and I have fixing various issues >>> with it. Firstly there was a digital logic problem with the keyboard >>> card that was causing the generator to periodically lock up. When this >>> was fixed the unit came to life but the output signal level was down 5dB >>> and this was traced to a bad step attenuator. I attempted a fix but >>> ended up needing to replace this to get the output steps correct. >>> >>> I have now turned my attention to calibrating the reference timebase and >>> unfortunately this seems to be well out of spec. >>> >>> The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the >>> oscillator seems to be ~40Hz low. I have attempted to adjust the unit >>> via the multi-turn capacitor adjustment but this seems to be at the end >>> stop and the frequency is still low. So my questions are: >>> >>> 1) I assume this module is faulty as it is so far out of spec? The >>> signal generator has been powered off for multiple months but even after >>> having the unit powered on for 24 hours the frequency has not changed >>> significantly. >>> >>> 2) How repairable are these modules? I ended up blowing $200 on the >>> replacement step attenuator so will need to save up for any new parts. I >>> can always use an external 10MHz reference but it is always nice to have >>> equipment complete. >>> >>> 3) From searching the list it seems that the oven could be what is >>> faulty. Would a cold over account for this much error?I have confirmed >>> that the edge connector is supplying 25V on the over supply connections >>> but I don't think the module is warming up. >>> >>> 4) Is it possible the OCXO has just worn out and has aged beyond the >>> point it can be adjusted back into spec? >>> >>> >>> Regards, >>> Denis >>> >>> >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@lists.febo.com >>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >> _______________________________________________ >> time-nuts mailing list -- time-nuts@lists.febo.com >> To unsubscribe send an email to time-nuts-leave@lists.febo.com > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

The thermal fuse on the top board often fail after many years, w/o overtemperature event, just by ageing. You might replace it with a short. Check the oven current, if it’s very low or zero, then it would be the thermal fuse in first place. If it’s a few mA, and maybe the oven won’t get warm, then it might be the NTC which is glued inside the oven mass. I once had a 10811 with intermittently open NTC. I drilled it out, used a standard precision NTC, I think 100k, placed it into the bore hole with some thermal paste, and trimmed the oven set point (written on the XTAL, or the oven mass) according to the manual, but taking the different NTC characteristic into account. The oven still works like a charm. Frank

Great article.  I remember reading it with great interest when it came out.  Bought an early 10811A as a house reference and poured over the extensive manual. Have had either and both the thermal fuse and the thermistor fail on various copies. 73, David N1HAC On 3/29/23 6:16 AM, Denis Dowling via time-nuts wrote: > Thanks everyone for all of the advice. As predicted the fault was an > open circuit thermal fuse. I have bypassed this (temporarily) and I > can now trim the oscillator to be 10MHz to the precision of my PM 6672 > frequency counter. I am plotting both 10MHz waveforms on a scope and > there is minimal slip. I will order a new fuse and then reassemble and > then attempt a more accurate calibration when everything has  been > allowed to burn in for a while. > > On a related note, how good is the Hewlett Packet March 1981 Journal > article on the HP 10811D? > https://nam12.safelinks.protection.outlook.com/?url=http%3A%2F%2Fhparchive.com%2FJournals%2FHPJ-1981-03.pdf&data=05%7C01%7Cdavid.g.mcgaw%40dartmouth.edu%7C0d22f32fcd854208d05808db3077c596%7C995b093648d640e5a31ebf689ec9446f%7C0%7C0%7C638157063203629937%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=%2BPtZs%2FBPdPjBCjRTmf9z3cQ4UAjLAaemtMJlh2n8x%2BQ%3D&reserved=0 > > > There is so much information in this article including how they used a > flex PCB to make the oscillator easier to assemble. They even discuss > the issues they had with stray capacitance due to the thinner PCB and > tearing on sharp edges. Unfortunately today I doubt any manufacturer > would be as open about their design process. > > Regards, > Denis > > > On 28/03/2023 6:44 pm, Adrian Godwin via time-nuts wrote: >> I had a 10811 that didn't warm up. I'd never taken one apart before but, >> although a bunch of parts had failed, they were all common (if ancient) >> op-amps and the like. I didn't have much trouble fixing it once I'd got >> started. I'd encourage you to go ahead. >> >> On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < >> time-nuts@lists.febo.com> wrote: >> >>> Hi Dennis, >>> >>> The HP 10811 are serviceable. Just download the service manual and >>> study >>> it. >>> I would expect the oven to be faulty with this much error. >>> Most likely the thermal fuse has gone open circuit. >>> The fuse can go open circuit without there actually being any fault >>> with >>> the oven, but there might also be a fault in the oven circuit. >>> The fuse is on the first PCB you get to after opening up the 10811. All >>> described in the manual. >>> If its the thermal fuse, replace with correct thermal fuse. When >>> replacing >>> make sure to not over heat it while soldering. >>> You can of course bridge the fuse with a wire for testing, but then you >>> need to monitor it closely, to be sure it does not over heat. >>> >>> I was lucky to find a HP 8663A in a electronic recycling bin. It was >>> missing PCB's in the PSU and the 10811 was totally burnt to a crisp. >>> Guess the fuse had been bridged. So I needed to get a replacement >>> off ebay. >>> >>> Regards, >>> Askild >>> >>> >>> >>> On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < >>> time-nuts@lists.febo.com> wrote: >>> >>>> Hi All, >>>> >>>> This is my first time on the list so apologies if I am going over >>>> an old >>>> topic. >>>> >>>> I am repairing a HP 8662A Synthesised Signal Generator that I >>>> purchased >>>> from a government auction. The unit was purchased without any >>>> information about its working state and I have fixing various issues >>>> with it. Firstly there was a digital logic problem with the keyboard >>>> card that was causing the generator to periodically lock up. When this >>>> was fixed the unit came to life but the output signal level was >>>> down 5dB >>>> and this was traced to a bad step attenuator. I attempted a fix but >>>> ended up needing to replace this to get the output steps correct. >>>> >>>> I have now turned my attention to calibrating the reference >>>> timebase and >>>> unfortunately this seems to be well out of spec. >>>> >>>> The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the >>>> oscillator seems to be ~40Hz low. I have attempted to adjust the unit >>>> via the multi-turn capacitor adjustment but this seems to be at the >>>> end >>>> stop and the frequency is still low. So my questions are: >>>> >>>> 1) I assume this module is faulty as it is so far out of spec? The >>>> signal generator has been powered off for multiple months but even >>>> after >>>> having the unit powered on for 24 hours the frequency has not changed >>>> significantly. >>>> >>>> 2) How repairable are these modules? I ended up blowing $200 on the >>>> replacement step attenuator so will need to save up for any new >>>> parts. I >>>> can always use an external 10MHz reference but it is always nice to >>>> have >>>> equipment complete. >>>> >>>> 3) From searching the list it seems that the oven could be what is >>>> faulty. Would a cold over account for this much error?I have confirmed >>>> that the edge connector is supplying 25V on the over supply >>>> connections >>>> but I don't think the module is warming up. >>>> >>>> 4) Is it possible  the OCXO has just worn out and has aged beyond the >>>> point it can be adjusted back into spec? >>>> >>>> >>>> Regards, >>>> Denis >>>> >>>> >>>> _______________________________________________ >>>> time-nuts mailing list -- time-nuts@lists.febo.com >>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@lists.febo.com >>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >> _______________________________________________ >> time-nuts mailing list -- time-nuts@lists.febo.com >> To unsubscribe send an email to time-nuts-leave@lists.febo.com > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi The use of the thermal fuse was very unique to HP’s specific needs. There is little or no evidence of it actually doing much good ( = protecting things). There is a *lot* of evidence of it failing for no good reason. Save your money for another $40 OCXO and replace it with a short. Bob > On Mar 29, 2023, at 6:16 AM, Denis Dowling via time-nuts <time-nuts@lists.febo.com> wrote: > > Thanks everyone for all of the advice. As predicted the fault was an open circuit thermal fuse. I have bypassed this (temporarily) and I can now trim the oscillator to be 10MHz to the precision of my PM 6672 frequency counter. I am plotting both 10MHz waveforms on a scope and there is minimal slip. I will order a new fuse and then reassemble and then attempt a more accurate calibration when everything has been allowed to burn in for a while. > > On a related note, how good is the Hewlett Packet March 1981 Journal article on the HP 10811D? > http://hparchive.com/Journals/HPJ-1981-03.pdf > > There is so much information in this article including how they used a flex PCB to make the oscillator easier to assemble. They even discuss the issues they had with stray capacitance due to the thinner PCB and tearing on sharp edges. Unfortunately today I doubt any manufacturer would be as open about their design process. > > Regards, > Denis > > > On 28/03/2023 6:44 pm, Adrian Godwin via time-nuts wrote: >> I had a 10811 that didn't warm up. I'd never taken one apart before but, >> although a bunch of parts had failed, they were all common (if ancient) >> op-amps and the like. I didn't have much trouble fixing it once I'd got >> started. I'd encourage you to go ahead. >> >> On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < >> time-nuts@lists.febo.com> wrote: >> >>> Hi Dennis, >>> >>> The HP 10811 are serviceable. Just download the service manual and study >>> it. >>> I would expect the oven to be faulty with this much error. >>> Most likely the thermal fuse has gone open circuit. >>> The fuse can go open circuit without there actually being any fault with >>> the oven, but there might also be a fault in the oven circuit. >>> The fuse is on the first PCB you get to after opening up the 10811. All >>> described in the manual. >>> If its the thermal fuse, replace with correct thermal fuse. When replacing >>> make sure to not over heat it while soldering. >>> You can of course bridge the fuse with a wire for testing, but then you >>> need to monitor it closely, to be sure it does not over heat. >>> >>> I was lucky to find a HP 8663A in a electronic recycling bin. It was >>> missing PCB's in the PSU and the 10811 was totally burnt to a crisp. >>> Guess the fuse had been bridged. So I needed to get a replacement off ebay. >>> >>> Regards, >>> Askild >>> >>> >>> >>> On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < >>> time-nuts@lists.febo.com> wrote: >>> >>>> Hi All, >>>> >>>> This is my first time on the list so apologies if I am going over an old >>>> topic. >>>> >>>> I am repairing a HP 8662A Synthesised Signal Generator that I purchased >>>> from a government auction. The unit was purchased without any >>>> information about its working state and I have fixing various issues >>>> with it. Firstly there was a digital logic problem with the keyboard >>>> card that was causing the generator to periodically lock up. When this >>>> was fixed the unit came to life but the output signal level was down 5dB >>>> and this was traced to a bad step attenuator. I attempted a fix but >>>> ended up needing to replace this to get the output steps correct. >>>> >>>> I have now turned my attention to calibrating the reference timebase and >>>> unfortunately this seems to be well out of spec. >>>> >>>> The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the >>>> oscillator seems to be ~40Hz low. I have attempted to adjust the unit >>>> via the multi-turn capacitor adjustment but this seems to be at the end >>>> stop and the frequency is still low. So my questions are: >>>> >>>> 1) I assume this module is faulty as it is so far out of spec? The >>>> signal generator has been powered off for multiple months but even after >>>> having the unit powered on for 24 hours the frequency has not changed >>>> significantly. >>>> >>>> 2) How repairable are these modules? I ended up blowing $200 on the >>>> replacement step attenuator so will need to save up for any new parts. I >>>> can always use an external 10MHz reference but it is always nice to have >>>> equipment complete. >>>> >>>> 3) From searching the list it seems that the oven could be what is >>>> faulty. Would a cold over account for this much error?I have confirmed >>>> that the edge connector is supplying 25V on the over supply connections >>>> but I don't think the module is warming up. >>>> >>>> 4) Is it possible the OCXO has just worn out and has aged beyond the >>>> point it can be adjusted back into spec? >>>> >>>> >>>> Regards, >>>> Denis >>>> >>>> >>>> _______________________________________________ >>>> time-nuts mailing list -- time-nuts@lists.febo.com >>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@lists.febo.com >>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >> _______________________________________________ >> time-nuts mailing list -- time-nuts@lists.febo.com >> To unsubscribe send an email to time-nuts-leave@lists.febo.com > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi, Generally I would agree that probably the thermal fuse has gone open circuit when there has been no over temperature, more often than it has protected the ocxo from a fault, but the HP 8663A i found, seems to have had the thermal fuse replaced by a wire. I hope my, not to good, pictures makes it through, so you can see the result. [image: P1010305_s.jpg] [image: P1010306_s.jpg] [image: P1010309_s.jpg] Regards, Askild On Wed, Mar 29, 2023 at 8:37 PM Bob Camp via time-nuts < time-nuts@lists.febo.com> wrote: > Hi > > The use of the thermal fuse was very unique to HP’s specific needs. There > is little or no > evidence of it actually doing much good ( = protecting things). There is a > *lot* of evidence > of it failing for no good reason. > > Save your money for another $40 OCXO and replace it with a short. > > Bob > > > On Mar 29, 2023, at 6:16 AM, Denis Dowling via time-nuts < > time-nuts@lists.febo.com> wrote: > > > > Thanks everyone for all of the advice. As predicted the fault was an > open circuit thermal fuse. I have bypassed this (temporarily) and I can now > trim the oscillator to be 10MHz to the precision of my PM 6672 frequency > counter. I am plotting both 10MHz waveforms on a scope and there is minimal > slip. I will order a new fuse and then reassemble and then attempt a more > accurate calibration when everything has been allowed to burn in for a > while. > > > > On a related note, how good is the Hewlett Packet March 1981 Journal > article on the HP 10811D? > > http://hparchive.com/Journals/HPJ-1981-03.pdf > > > > There is so much information in this article including how they used a > flex PCB to make the oscillator easier to assemble. They even discuss the > issues they had with stray capacitance due to the thinner PCB and tearing > on sharp edges. Unfortunately today I doubt any manufacturer would be as > open about their design process. > > > > Regards, > > Denis > > > > > > On 28/03/2023 6:44 pm, Adrian Godwin via time-nuts wrote: > >> I had a 10811 that didn't warm up. I'd never taken one apart before but, > >> although a bunch of parts had failed, they were all common (if ancient) > >> op-amps and the like. I didn't have much trouble fixing it once I'd got > >> started. I'd encourage you to go ahead. > >> > >> On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < > >> time-nuts@lists.febo.com> wrote: > >> > >>> Hi Dennis, > >>> > >>> The HP 10811 are serviceable. Just download the service manual and > study > >>> it. > >>> I would expect the oven to be faulty with this much error. > >>> Most likely the thermal fuse has gone open circuit. > >>> The fuse can go open circuit without there actually being any fault > with > >>> the oven, but there might also be a fault in the oven circuit. > >>> The fuse is on the first PCB you get to after opening up the 10811. All > >>> described in the manual. > >>> If its the thermal fuse, replace with correct thermal fuse. When > replacing > >>> make sure to not over heat it while soldering. > >>> You can of course bridge the fuse with a wire for testing, but then you > >>> need to monitor it closely, to be sure it does not over heat. > >>> > >>> I was lucky to find a HP 8663A in a electronic recycling bin. It was > >>> missing PCB's in the PSU and the 10811 was totally burnt to a crisp. > >>> Guess the fuse had been bridged. So I needed to get a replacement off > ebay. > >>> > >>> Regards, > >>> Askild > >>> > >>> > >>> > >>> On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < > >>> time-nuts@lists.febo.com> wrote: > >>> > >>>> Hi All, > >>>> > >>>> This is my first time on the list so apologies if I am going over an > old > >>>> topic. > >>>> > >>>> I am repairing a HP 8662A Synthesised Signal Generator that I > purchased > >>>> from a government auction. The unit was purchased without any > >>>> information about its working state and I have fixing various issues > >>>> with it. Firstly there was a digital logic problem with the keyboard > >>>> card that was causing the generator to periodically lock up. When this > >>>> was fixed the unit came to life but the output signal level was down > 5dB > >>>> and this was traced to a bad step attenuator. I attempted a fix but > >>>> ended up needing to replace this to get the output steps correct. > >>>> > >>>> I have now turned my attention to calibrating the reference timebase > and > >>>> unfortunately this seems to be well out of spec. > >>>> > >>>> The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the > >>>> oscillator seems to be ~40Hz low. I have attempted to adjust the unit > >>>> via the multi-turn capacitor adjustment but this seems to be at the > end > >>>> stop and the frequency is still low. So my questions are: > >>>> > >>>> 1) I assume this module is faulty as it is so far out of spec? The > >>>> signal generator has been powered off for multiple months but even > after > >>>> having the unit powered on for 24 hours the frequency has not changed > >>>> significantly. > >>>> > >>>> 2) How repairable are these modules? I ended up blowing $200 on the > >>>> replacement step attenuator so will need to save up for any new > parts. I > >>>> can always use an external 10MHz reference but it is always nice to > have > >>>> equipment complete. > >>>> > >>>> 3) From searching the list it seems that the oven could be what is > >>>> faulty. Would a cold over account for this much error?I have confirmed > >>>> that the edge connector is supplying 25V on the over supply > connections > >>>> but I don't think the module is warming up. > >>>> > >>>> 4) Is it possible the OCXO has just worn out and has aged beyond the > >>>> point it can be adjusted back into spec? > >>>> > >>>> > >>>> Regards, > >>>> Denis > >>>> > >>>> > >>>> _______________________________________________ > >>>> time-nuts mailing list -- time-nuts@lists.febo.com > >>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com > >>> _______________________________________________ > >>> time-nuts mailing list -- time-nuts@lists.febo.com > >>> To unsubscribe send an email to time-nuts-leave@lists.febo.com > >> _______________________________________________ > >> time-nuts mailing list -- time-nuts@lists.febo.com > >> To unsubscribe send an email to time-nuts-leave@lists.febo.com > > > > _______________________________________________ > > time-nuts mailing list -- time-nuts@lists.febo.com > > To unsubscribe send an email to time-nuts-leave@lists.febo.com > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

Some of you may have missed [1] the amazing 10811 photos from Askild in the previous posting [1]. In this posting/reply we have -attached- the images, which may show up better for some readers. If you have any comments or questions about list mechanics and infrastructure contact: time-nuts-owner@lists.febo.com -- moderators [1] The time-nuts list accepts attachments without problem, including photos, images, plots, PDF, zip, etc. But it has trouble with html postings or "inline" images. [2] https://febo.com/pipermail/time-nuts_lists.febo.com/2023-March/107631.html https://febo.com/pipermail/time-nuts_lists.febo.com/2023-March/date.html#end On 3/29/2023 11:03 PM, Askild via time-nuts wrote: > Hi, > > Generally I would agree that probably the thermal fuse has gone open > circuit when there has been no over temperature, more often than it has > protected the ocxo from a fault, > but the HP 8663A i found, seems to have had the thermal fuse replaced by a > wire. I hope my, not to good, pictures makes it through, so you can see the > result. > [image: P1010305_s.jpg] > [image: P1010306_s.jpg] > [image: P1010309_s.jpg] > > Regards, > Askild >

Hi Looks like it’s time to head off to eBay and find a $40 replacement. Bob > On Mar 30, 2023, at 2:03 AM, Askild via time-nuts <time-nuts@lists.febo.com> wrote: > > Hi, > > Generally I would agree that probably the thermal fuse has gone open > circuit when there has been no over temperature, more often than it has > protected the ocxo from a fault, > but the HP 8663A i found, seems to have had the thermal fuse replaced by a > wire. I hope my, not to good, pictures makes it through, so you can see the > result. > [image: P1010305_s.jpg] > [image: P1010306_s.jpg] > [image: P1010309_s.jpg] > > Regards, > Askild > > > On Wed, Mar 29, 2023 at 8:37 PM Bob Camp via time-nuts < > time-nuts@lists.febo.com> wrote: > >> Hi >> >> The use of the thermal fuse was very unique to HP’s specific needs. There >> is little or no >> evidence of it actually doing much good ( = protecting things). There is a >> *lot* of evidence >> of it failing for no good reason. >> >> Save your money for another $40 OCXO and replace it with a short. >> >> Bob >> >>> On Mar 29, 2023, at 6:16 AM, Denis Dowling via time-nuts < >> time-nuts@lists.febo.com> wrote: >>> >>> Thanks everyone for all of the advice. As predicted the fault was an >> open circuit thermal fuse. I have bypassed this (temporarily) and I can now >> trim the oscillator to be 10MHz to the precision of my PM 6672 frequency >> counter. I am plotting both 10MHz waveforms on a scope and there is minimal >> slip. I will order a new fuse and then reassemble and then attempt a more >> accurate calibration when everything has been allowed to burn in for a >> while. >>> >>> On a related note, how good is the Hewlett Packet March 1981 Journal >> article on the HP 10811D? >>> http://hparchive.com/Journals/HPJ-1981-03.pdf >>> >>> There is so much information in this article including how they used a >> flex PCB to make the oscillator easier to assemble. They even discuss the >> issues they had with stray capacitance due to the thinner PCB and tearing >> on sharp edges. Unfortunately today I doubt any manufacturer would be as >> open about their design process. >>> >>> Regards, >>> Denis >>> >>> >>> On 28/03/2023 6:44 pm, Adrian Godwin via time-nuts wrote: >>>> I had a 10811 that didn't warm up. I'd never taken one apart before but, >>>> although a bunch of parts had failed, they were all common (if ancient) >>>> op-amps and the like. I didn't have much trouble fixing it once I'd got >>>> started. I'd encourage you to go ahead. >>>> >>>> On Tue, Mar 28, 2023 at 7:28 AM Askild via time-nuts < >>>> time-nuts@lists.febo.com> wrote: >>>> >>>>> Hi Dennis, >>>>> >>>>> The HP 10811 are serviceable. Just download the service manual and >> study >>>>> it. >>>>> I would expect the oven to be faulty with this much error. >>>>> Most likely the thermal fuse has gone open circuit. >>>>> The fuse can go open circuit without there actually being any fault >> with >>>>> the oven, but there might also be a fault in the oven circuit. >>>>> The fuse is on the first PCB you get to after opening up the 10811. All >>>>> described in the manual. >>>>> If its the thermal fuse, replace with correct thermal fuse. When >> replacing >>>>> make sure to not over heat it while soldering. >>>>> You can of course bridge the fuse with a wire for testing, but then you >>>>> need to monitor it closely, to be sure it does not over heat. >>>>> >>>>> I was lucky to find a HP 8663A in a electronic recycling bin. It was >>>>> missing PCB's in the PSU and the 10811 was totally burnt to a crisp. >>>>> Guess the fuse had been bridged. So I needed to get a replacement off >> ebay. >>>>> >>>>> Regards, >>>>> Askild >>>>> >>>>> >>>>> >>>>> On Tue, Mar 28, 2023 at 7:49 AM Denis Dowling via time-nuts < >>>>> time-nuts@lists.febo.com> wrote: >>>>> >>>>>> Hi All, >>>>>> >>>>>> This is my first time on the list so apologies if I am going over an >> old >>>>>> topic. >>>>>> >>>>>> I am repairing a HP 8662A Synthesised Signal Generator that I >> purchased >>>>>> from a government auction. The unit was purchased without any >>>>>> information about its working state and I have fixing various issues >>>>>> with it. Firstly there was a digital logic problem with the keyboard >>>>>> card that was causing the generator to periodically lock up. When this >>>>>> was fixed the unit came to life but the output signal level was down >> 5dB >>>>>> and this was traced to a bad step attenuator. I attempted a fix but >>>>>> ended up needing to replace this to get the output steps correct. >>>>>> >>>>>> I have now turned my attention to calibrating the reference timebase >> and >>>>>> unfortunately this seems to be well out of spec. >>>>>> >>>>>> The timebase is a HP 10811D 10MHz OCXO. The 10MHz output of the >>>>>> oscillator seems to be ~40Hz low. I have attempted to adjust the unit >>>>>> via the multi-turn capacitor adjustment but this seems to be at the >> end >>>>>> stop and the frequency is still low. So my questions are: >>>>>> >>>>>> 1) I assume this module is faulty as it is so far out of spec? The >>>>>> signal generator has been powered off for multiple months but even >> after >>>>>> having the unit powered on for 24 hours the frequency has not changed >>>>>> significantly. >>>>>> >>>>>> 2) How repairable are these modules? I ended up blowing $200 on the >>>>>> replacement step attenuator so will need to save up for any new >> parts. I >>>>>> can always use an external 10MHz reference but it is always nice to >> have >>>>>> equipment complete. >>>>>> >>>>>> 3) From searching the list it seems that the oven could be what is >>>>>> faulty. Would a cold over account for this much error?I have confirmed >>>>>> that the edge connector is supplying 25V on the over supply >> connections >>>>>> but I don't think the module is warming up. >>>>>> >>>>>> 4) Is it possible the OCXO has just worn out and has aged beyond the >>>>>> point it can be adjusted back into spec? >>>>>> >>>>>> >>>>>> Regards, >>>>>> Denis >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> time-nuts mailing list -- time-nuts@lists.febo.com >>>>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >>>>> _______________________________________________ >>>>> time-nuts mailing list -- time-nuts@lists.febo.com >>>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >>>> _______________________________________________ >>>> time-nuts mailing list -- time-nuts@lists.febo.com >>>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >>> >>> _______________________________________________ >>> time-nuts mailing list -- time-nuts@lists.febo.com >>> To unsubscribe send an email to time-nuts-leave@lists.febo.com >> _______________________________________________ >> time-nuts mailing list -- time-nuts@lists.febo.com >> To unsubscribe send an email to time-nuts-leave@lists.febo.com > <P1010305_s.jpg><P1010306_s.jpg><P1010309_s.jpg>_______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com