Hi In your “June 10” stuff: You pull the FET’s out. You connect the heaters. The current in the heaters drops as they warm up. Are the FET’s still out of the circuit or did you put them back in? Bob > On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: > > Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. > Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. > https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ > > cheers, > Wilko > > -----Original Message----- > From: Bob Camp via time-nuts <time-nuts@lists.febo.com> > Sent: 6 June, 2025 15:57 > To: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> > Cc: Bob Camp <kb8tq@n1k.org> > Subject: [time-nuts] Re: repair of a HP E1938A OCXO > > Hi > > If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? > > Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. > > To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. > > Bob > >> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: >> >> Thank you Morris! And thanks to Bob for noting the different temperature labels. >> >> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. >> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. >> >> Anyway, have a look in case you are interested in this project: >> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >> >> Wilko >> >> -----Original Message----- >> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> >> Sent: 6 June, 2025 12:07 >> To: time-nuts@lists.febo.com >> Cc: vilgotch1@gmail.com >> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >> >> Hi all, >> >> Wilko thanks very much for the update - keep them coming! >> >> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. >> >> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. >> >> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. >> >> I hope that helps, >> >> Morris >> _______________________________________________ >> 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 FETs are now back in the circuit. They were out of circuit during the test of the heaters. best, Wilko > On 11 Jun 2025, at 00:00, Bob Camp <kb8tq@n1k.org> wrote: > > Hi > > In your “June 10” stuff: > > You pull the FET’s out. > > You connect the heaters. > > The current in the heaters drops as they warm up. > > Are the FET’s still out of the circuit or did you put them back in? > > Bob > >> On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: >> >> Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. >> Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. >> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >> >> cheers, >> Wilko >> >> -----Original Message----- >> From: Bob Camp via time-nuts <time-nuts@lists.febo.com> >> Sent: 6 June, 2025 15:57 >> To: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> >> Cc: Bob Camp <kb8tq@n1k.org> >> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >> >> Hi >> >> If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? >> >> Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. >> >> To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. >> >> Bob >> >>>> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: >>> >>> Thank you Morris! And thanks to Bob for noting the different temperature labels. >>> >>> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. >>> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. >>> >>> Anyway, have a look in case you are interested in this project: >>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>> >>> Wilko >>> >>> -----Original Message----- >>> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> >>> Sent: 6 June, 2025 12:07 >>> To: time-nuts@lists.febo.com >>> Cc: vilgotch1@gmail.com >>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>> >>> Hi all, >>> >>> Wilko thanks very much for the update - keep them coming! >>> >>> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. >>> >>> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. >>> >>> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. >>> >>> I hope that helps, >>> >>> Morris >>> _______________________________________________ >>> 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 Ok ….. What is causing the current to drop as the oven warms up if the FET’s are out of the circuit during that part of the test? Bob > On Jun 10, 2025, at 6:34 PM, Wilko Bulte <wkb@xs4all.nl> wrote: > > The FETs are now back in the circuit. They were out of circuit during the test of the heaters. > > best, > Wilko > >> On 11 Jun 2025, at 00:00, Bob Camp <kb8tq@n1k.org> wrote: >> >> Hi >> >> In your “June 10” stuff: >> >> You pull the FET’s out. >> >> You connect the heaters. >> >> The current in the heaters drops as they warm up. >> >> Are the FET’s still out of the circuit or did you put them back in? >> >> Bob >> >>> On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: >>> >>> Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. >>> Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. >>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>> >>> cheers, >>> Wilko >>> >>> -----Original Message----- >>> From: Bob Camp via time-nuts <time-nuts@lists.febo.com> >>> Sent: 6 June, 2025 15:57 >>> To: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> >>> Cc: Bob Camp <kb8tq@n1k.org> >>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>> >>> Hi >>> >>> If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? >>> >>> Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. >>> >>> To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. >>> >>> Bob >>> >>>>> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: >>>> >>>> Thank you Morris! And thanks to Bob for noting the different temperature labels. >>>> >>>> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. >>>> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. >>>> >>>> Anyway, have a look in case you are interested in this project: >>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>>> >>>> Wilko >>>> >>>> -----Original Message----- >>>> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> >>>> Sent: 6 June, 2025 12:07 >>>> To: time-nuts@lists.febo.com >>>> Cc: vilgotch1@gmail.com >>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>>> >>>> Hi all, >>>> >>>> Wilko thanks very much for the update - keep them coming! >>>> >>>> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. >>>> >>>> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. >>>> >>>> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. >>>> >>>> I hope that helps, >>>> >>>> Morris >>>> _______________________________________________ >>>> 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 Bob, I re-measured things today. Heater current is really dropping once the oven gets hot. Details are documented on the web page. Apart from discovering a missing resistor that had me puzzled for a bit no real progress unfortunately. 73, Wilko -----Original Message----- From: Bob Camp <kb8tq@n1k.org> Sent: 11 June, 2025 1:02 To: Wilko Bulte <wkb@xs4all.nl> Cc: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> Subject: Re: [time-nuts] repair of a HP E1938A OCXO Hi Ok ….. What is causing the current to drop as the oven warms up if the FET’s are out of the circuit during that part of the test? Bob > On Jun 10, 2025, at 6:34 PM, Wilko Bulte <wkb@xs4all.nl> wrote: > > The FETs are now back in the circuit. They were out of circuit during the test of the heaters. > > best, > Wilko > >> On 11 Jun 2025, at 00:00, Bob Camp <kb8tq@n1k.org> wrote: >> >> Hi >> >> In your “June 10” stuff: >> >> You pull the FET’s out. >> >> You connect the heaters. >> >> The current in the heaters drops as they warm up. >> >> Are the FET’s still out of the circuit or did you put them back in? >> >> Bob >> >>> On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: >>> >>> Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. >>> Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. >>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>> >>> cheers, >>> Wilko >>> >>> -----Original Message----- >>> From: Bob Camp via time-nuts <time-nuts@lists.febo.com> >>> Sent: 6 June, 2025 15:57 >>> To: Discussion of precise time and frequency measurement >>> <time-nuts@lists.febo.com> >>> Cc: Bob Camp <kb8tq@n1k.org> >>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>> >>> Hi >>> >>> If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? >>> >>> Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. >>> >>> To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. >>> >>> Bob >>> >>>>> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: >>>> >>>> Thank you Morris! And thanks to Bob for noting the different temperature labels. >>>> >>>> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. >>>> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. >>>> >>>> Anyway, have a look in case you are interested in this project: >>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>>> >>>> Wilko >>>> >>>> -----Original Message----- >>>> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> >>>> Sent: 6 June, 2025 12:07 >>>> To: time-nuts@lists.febo.com >>>> Cc: vilgotch1@gmail.com >>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>>> >>>> Hi all, >>>> >>>> Wilko thanks very much for the update - keep them coming! >>>> >>>> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. >>>> >>>> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. >>>> >>>> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. >>>> >>>> I hope that helps, >>>> >>>> Morris >>>> _______________________________________________ >>>> 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 That suggests a couple of things: 1) Those FET’s are not what the control loop drives. (if so what do they do?) 2) Something else is what the control loop is driving (see below …) —— OR —— 3) The heaters are something very unusual (if so what?) —— OR — 4) The heaters are damaged Without a full schematic to look at, those are just guesses. In most modern OCXO’s the “heaters” are some sort of solid state device. It might be a FET. It could be a darlington. Bob > On Jun 11, 2025, at 11:01 AM, Wilko Bulte <wkb@xs4all.nl> wrote: > > Hi Bob, > > I re-measured things today. Heater current is really dropping once the oven gets hot. Details are documented on the web page. > Apart from discovering a missing resistor that had me puzzled for a bit no real progress unfortunately. > > 73, Wilko > > -----Original Message----- > From: Bob Camp <kb8tq@n1k.org> > Sent: 11 June, 2025 1:02 > To: Wilko Bulte <wkb@xs4all.nl> > Cc: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> > Subject: Re: [time-nuts] repair of a HP E1938A OCXO > > Hi > > Ok ….. > > What is causing the current to drop as the oven warms up if the FET’s are out of the circuit during that part of the test? > > Bob > > >> On Jun 10, 2025, at 6:34 PM, Wilko Bulte <wkb@xs4all.nl> wrote: >> >> The FETs are now back in the circuit. They were out of circuit during the test of the heaters. >> >> best, >> Wilko >> >>> On 11 Jun 2025, at 00:00, Bob Camp <kb8tq@n1k.org> wrote: >>> >>> Hi >>> >>> In your “June 10” stuff: >>> >>> You pull the FET’s out. >>> >>> You connect the heaters. >>> >>> The current in the heaters drops as they warm up. >>> >>> Are the FET’s still out of the circuit or did you put them back in? >>> >>> Bob >>> >>>> On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: >>>> >>>> Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. >>>> Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. >>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>>> >>>> cheers, >>>> Wilko >>>> >>>> -----Original Message----- >>>> From: Bob Camp via time-nuts <time-nuts@lists.febo.com> >>>> Sent: 6 June, 2025 15:57 >>>> To: Discussion of precise time and frequency measurement >>>> <time-nuts@lists.febo.com> >>>> Cc: Bob Camp <kb8tq@n1k.org> >>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>>> >>>> Hi >>>> >>>> If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? >>>> >>>> Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. >>>> >>>> To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. >>>> >>>> Bob >>>> >>>>>> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: >>>>> >>>>> Thank you Morris! And thanks to Bob for noting the different temperature labels. >>>>> >>>>> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. >>>>> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. >>>>> >>>>> Anyway, have a look in case you are interested in this project: >>>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ >>>>> >>>>> Wilko >>>>> >>>>> -----Original Message----- >>>>> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> >>>>> Sent: 6 June, 2025 12:07 >>>>> To: time-nuts@lists.febo.com >>>>> Cc: vilgotch1@gmail.com >>>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO >>>>> >>>>> Hi all, >>>>> >>>>> Wilko thanks very much for the update - keep them coming! >>>>> >>>>> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. >>>>> >>>>> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. >>>>> >>>>> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. >>>>> >>>>> I hope that helps, >>>>> >>>>> Morris >>>>> _______________________________________________ >>>>> 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 papers on the E1938A indicate that both the rim and lid heaters are resistive and the control loop uses a square root calculation to linearise the temperature control loop. A 12 bit DAC is dithered to provide greater heater drive resolution and the temperature control loop sampling rate is 1Hz. Bruce > On 12/06/2025 04:46 NZST Bob Camp via time-nuts <time-nuts@lists.febo.com> wrote: > > > Hi > > That suggests a couple of things: > > 1) Those FET’s are not what the control loop drives. (if so what do they do?) > > 2) Something else is what the control loop is driving (see below …) > > —— OR —— > > 3) The heaters are something very unusual (if so what?) > > —— OR — > > 4) The heaters are damaged > > Without a full schematic to look at, those are just guesses. > > In most modern OCXO’s the “heaters” are some sort of solid state device. It might be a FET. It could be a darlington. > > Bob > > > On Jun 11, 2025, at 11:01 AM, Wilko Bulte <wkb@xs4all.nl> wrote: > > > > Hi Bob, > > > > I re-measured things today. Heater current is really dropping once the oven gets hot. Details are documented on the web page. > > Apart from discovering a missing resistor that had me puzzled for a bit no real progress unfortunately. > > > > 73, Wilko > > > > -----Original Message----- > > From: Bob Camp <kb8tq@n1k.org> > > Sent: 11 June, 2025 1:02 > > To: Wilko Bulte <wkb@xs4all.nl> > > Cc: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com> > > Subject: Re: [time-nuts] repair of a HP E1938A OCXO > > > > Hi > > > > Ok ….. > > > > What is causing the current to drop as the oven warms up if the FET’s are out of the circuit during that part of the test? > > > > Bob > > > > > >> On Jun 10, 2025, at 6:34 PM, Wilko Bulte <wkb@xs4all.nl> wrote: > >> > >> The FETs are now back in the circuit. They were out of circuit during the test of the heaters. > >> > >> best, > >> Wilko > >> > >>> On 11 Jun 2025, at 00:00, Bob Camp <kb8tq@n1k.org> wrote: > >>> > >>> Hi > >>> > >>> In your “June 10” stuff: > >>> > >>> You pull the FET’s out. > >>> > >>> You connect the heaters. > >>> > >>> The current in the heaters drops as they warm up. > >>> > >>> Are the FET’s still out of the circuit or did you put them back in? > >>> > >>> Bob > >>> > >>>> On Jun 10, 2025, at 2:25 PM, Wilko Bulte <wkb@xs4all.nl> wrote: > >>>> > >>>> Yes, that was not going anywhere, you are right Bob. I somehow decided I had freed the thermistors from the circuit. But I had not. > >>>> Meanwhile I have done some more investigations, more like eliminating things than finding the culprit unfortunately. What I did today is up on the webpage. > >>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ > >>>> > >>>> cheers, > >>>> Wilko > >>>> > >>>> -----Original Message----- > >>>> From: Bob Camp via time-nuts <time-nuts@lists.febo.com> > >>>> Sent: 6 June, 2025 15:57 > >>>> To: Discussion of precise time and frequency measurement > >>>> <time-nuts@lists.febo.com> > >>>> Cc: Bob Camp <kb8tq@n1k.org> > >>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO > >>>> > >>>> Hi > >>>> > >>>> If all the thermistors measure almost the same, why suspect one is out of the circuit / broke? > >>>> > >>>> Since there are a variety of things hooked to the string of thermistors (including the 6.19K R35, it is reasonable to believe that the header pins will show less resistance than the sum of the thermistors. > >>>> > >>>> To double check the connections, measure from the header pins to each of the thermistors. See if the net resistances make sense. If they do, the thermistors are not the issue. > >>>> > >>>> Bob > >>>> > >>>>>> On Jun 6, 2025, at 9:31 AM, Wilko Bulte via time-nuts <time-nuts@lists.febo.com> wrote: > >>>>> > >>>>> Thank you Morris! And thanks to Bob for noting the different temperature labels. > >>>>> > >>>>> In the meantime I have dug a bit deeper (into the rabbit hole maybe?) and added some more pictures & measurements for the thermistor section. Some interesting things were found. > >>>>> My current gut feeling is that one or more of the flex wires and/or their soldering have issues. > >>>>> > >>>>> Anyway, have a look in case you are interested in this project: > >>>>> https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ > >>>>> > >>>>> Wilko > >>>>> > >>>>> -----Original Message----- > >>>>> From: Morris Odell via time-nuts <time-nuts@lists.febo.com> > >>>>> Sent: 6 June, 2025 12:07 > >>>>> To: time-nuts@lists.febo.com > >>>>> Cc: vilgotch1@gmail.com > >>>>> Subject: [time-nuts] Re: repair of a HP E1938A OCXO > >>>>> > >>>>> Hi all, > >>>>> > >>>>> Wilko thanks very much for the update - keep them coming! > >>>>> > >>>>> The hockey puck I sent you is the one from my Z3815A which had been running for about 20 years until the oven stopped heating. I eventually was able to obtain a replacement unit and substituted the hockey puck from it into my original mother board. I did that because I thought there may have been some customization in the original board as the replacement unit warmed up but did not lock in the Z3815A. The mother board I sent is the replacement one I received with the faulty hockey puck on it. That explains the difference in labeling. The replacement hockey puck works perfectly in the Z3815A with its original motherboard which suggests that it’s running at the correct temperature. > >>>>> > >>>>> The higher current draw behaviour for the first minute or so falling to 50 mA afterwards that Wilko has described what what I observed as well on both motherboards. That makes me think the problem is with the thermistors, or their connections, in the hockey puck. The heating elements themselves all show proper continuity when tested with an ohmmeter. > >>>>> > >>>>> I have read somewhere that the series capacitor for the crystal was individually selected for each unit which probably explains why it’s specified on the oven’s label. > >>>>> > >>>>> I hope that helps, > >>>>> > >>>>> Morris > >>>>> _______________________________________________ > >>>>> 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 all, Here are my thoughts about this problem. 1. The problem is is in the hockey puck oven assembly, not the motherboard, evidenced by the fact that the unit works OK with a different oven on the motherboard. 2. There are three resistive heaters, described on the schematic as the rim, the mass lid and the warm-up. Each is fed from the 5 volt supply by a P channel MOSFET labeled Q10, Q9 and Q8 respectively. 3. The three thermistors are most likely NTC units and are in series. The outputs of the thermistor resistive bridge and its reference voltage are fed from the oven connector to the AD7714 ADC chip on the motherboard. The digital output of the ADC goes into the PIC 16C74 microcontroller which controls all the heater MOSFETs (see below). 4. The “warm up enable” gate drive to the Q8 warm up MOSFET is from one of the outputs of the PIC fed through a LMC660 op amp. This suggests that the warm up heater is just switched on and off by the micro. There’s probably no PWM modulation of that drive and the heater is switched off once the oven gets close to operating temp. 5. The gates of the other two heater MOSFETS are fed from LMC600 op amps summing several inputs. One is from a AD7243 DAC (some versions had two DACs) controlled from the PIC. Other inputs are “heater sense” lines from the oven and a current sensing input from the heater return line that passes through R99 - 0.2 ohms and an op amp integrator op amp. 6. The fault is preventing any of the heaters from being switched on. The PIC is therefore entering a state where its inputs are telling it not to run any of the heaters. It’s switching off the DACs and the warm up MOSFET. The inputs to the PIC are digital signals from the ADC which in turn is receiving several analog signals from the thermistor bridge and various reference voltages. That’s why I think the fault lies in that part of the circuit. 7. Therefore you need to carefully check the analog signals going into the ADC. They come from pins on J1 which is the connector to the oven assembly. The relevant pin numbers are 2 (Vref common - grounded in the oven unit), 15 (thermistor bridge ref to ADC ), 8 (therm bridge out to ADC), 7 (2.5 volt ref). I’ll be interested to hear what you find! Regards, Morris

On 6/15/2025 6:12 AM, Morris Odell via time-nuts wrote: > Hi all, > > Here are my thoughts about this problem. > > 2. There are three resistive heaters, described on the schematic as the rim, the mass lid and the warm-up. Each is fed from the 5 volt supply by a P channel MOSFET labeled Q10, Q9 and Q8 respectively. The ratio of heat delivered to the rim vs the lid is crucial for the thermal gain of the oven. This ratio is the same on every unit, and the number was chosen to approximate the average of the values that give infinite thermal gain in typical units. In production, the achieved thermal gain is always at least a few 100's of thousand. Some units could achieve thermal gain in the low millions. Some may be in the low millions but with negative thermal gain. For the really high thermal gains (1,000,000+), the thermal gain becomes a function of ambient temperature. There will be a particular temperature where the thermal gain goes to infinity, and the sign of thermal gain switches polarity on either side of infinity. > > 3. The three thermistors are most likely NTC units and are in series. The outputs of the thermistor resistive bridge and its reference voltage are fed from the oven connector to the The three resistors are nominally identical and the only reason for using 3 is to preserve circular symmetry. Rick N6RK

>> > On 16 Jun 2025, at 02:07, Richard (Rick) Karlquist via time-nuts <time-nuts@lists.febo.com> wrote: > >  > >> On 6/15/2025 6:12 AM, Morris Odell via time-nuts wrote: >> Hi all, >> Here are my thoughts about this problem. >> 2. There are three resistive heaters, described on the schematic as the rim, the mass lid and the warm-up. Each is fed from the 5 volt supply by a P channel MOSFET labeled Q10, Q9 and Q8 respectively. > > The ratio of heat delivered to the rim vs the lid is crucial for the thermal gain of the oven. This ratio is the same on every unit, and the number was chosen to approximate the average of the values that give infinite thermal gain in typical units. In production, the achieved thermal gain is always at least a few 100's of thousand. Some units could achieve thermal gain in the low millions. Some may be in the low millions but with negative thermal gain. For the really high thermal gains (1,000,000+), the thermal gain becomes a function of ambient temperature. There will be a particular temperature where the thermal gain goes to infinity, and the sign of thermal gain switches polarity on either side of infinity. Am I correct in assuming that the warmup heater is an on/off heater which is normally not used? I think I read that somewhere. My guess is that "normally" means it is only used in a cold start situation at very low ambient. But that could be a wrong assumption. > >> 3. The three thermistors are most likely NTC units and are in series. The outputs of the thermistor resistive bridge and its reference voltage are fed from the oven connector to the > > The three resistors are nominally identical and the only reason for using 3 is to preserve circular symmetry. From my measurements, the 3 thermistors are all at roughly the same resistance value, at least at ambient temperature. So that appears ok. What I fail to 'get' is why randomly, and only in 1 in 10 cases (guesstimate) the heater starts on powerup. For about 3 seconds a 3.5A current is drawn from 5V. In all other cases there is no in-rush current, so apparantly no heating start. My assumption is some kind of failsafe stops the PIC from starting the heaters. Wilko > Rick N6RK > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

Today I got NGOCOMM to work, controlling the E1938A heating system. Some interesting observations made, and more open ?? as a byproduct. I am curious what people think of the observations. https://people.freebsd.org/~wilko/HP-OCXO-E1938A/ Wilko