Re: [time-nuts] "The Penultimate HP5065 A15"
I noticed that HP made a particular choice about connecting the ground side lamp power supply. HP use a coaxial cable between the physical assembly and a connector mounted on the chassis, then a single + 20V wire is connected to the voltage regulator board, while the chassis acts as a negative connection. This solution reflects the technology of a "few years ago", now it would be unthinkable. My fear is that other currents circulating on the chassis affect, even if minimally, the power supply voltrage of the lamp. In one of my 5065As I kept the smc connector isolated and with a female smc connectorized coaxial cable I got directly to the 20Volts regulator. I think it is useful in an update of our HP5065A. Luciano Luciano P. S. Paramithiotti tim...@timeok.it www.timeok.it ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
[time-nuts] "The Penultimate HP5065 A15"
Luciano, I agree entirely! I like the motto "do no harm"! When I try a mod such as replacing the dead op amp on the A9 board. I make before and after tests to make sure no degradation has occurred. Since this means parts in 10 -14th AD at 128 Seconds most don't have the capability. However a lot of Time-nuts can test in mid parts in 10-13th from 1 to 100 Seconds so can at least compare that part of the performance. Mods like HP made to move the bridge rectifier from A15 to the chassis can be made without such testing. That said a replacement A15 card to replace a 'Crispy crittered" or badly damaged one would be welcome. Cheers, Corby ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 5, 2020, at 2:55 AM, tim...@timeok.it wrote: > > > I would like to share with you my thoughts on two important points that > concern a little all attempts to modify and upgrade the equipment. > > In general I am very reluctant to modify both the electrically and > mechanically, in this case, the HP5065A. Modifying, even completely a board, > I agree if there is a significant improvement, trying to use the existing > pinout and, in the need for additional connection points, use additional > connectors on the board. As regards the space on the pcb, smd components can > be used whenever possible or doughter boards. Indeed this has been the debate on a *lot* of equipment as it gets old enough to be considered an antique. Do you preserve it in original condition or do you keep it running ( or flying ….). Needless to say, there is a lot of debate about this. > > Remove boards, transformers or anything else I find reluctant, I would > prefer to have the possibility of being able to go back to the original > configuration if necessary. > The area occupied by the optional batteries, which I think almost nobody > uses, can be used for new electronics. > > Using a switching power supply with better performance I agree and since > there is already a dedicated DC input on the back of the 5065A I would prefer > to use the one to connect to an external box that contains the new power > supply and the management of a backup DC input. This eliminates the need to > dismantle anything inside the HP5065A. If you look at the heat generated ( = wasted) running modern <= 3.3V supply parts off of the 20V bus, multiple switchers start to look very attractive. Bringing all that in from “outside” probably isn’t the best / safest way to do it. Using some of the “open space” for a set of supplies is probably a better idea. > > Another important point is that of the certainty of the results of a > change. I mean that most hobbyists who have a 5065A, including me, do not > have the opportunity to measure the proposed improvement effects, first of > all because they do not have a reference such as an HMaser available, nor > even such a refined measuring system to appreciate the improvements made. I > want to remember that between zero and 8kseconds the GPS system (e.g. HP > GPSDO) in our laboratories has an Adev higher than that of the HP5065A and > therefore the measure we do in that range is that of GPS, not that of > rubidium. The most common solution to this is to use multiple reference sources. Finding OCXO’s that are better than a 5065 at 0.1 to 10 seconds is possible in a hobby environment. As mods are done, this is a pretty critical region to examine. A DMTD isn’t a super exotic piece of gear anymore. There are a lot of them in basements. A single mixer setup (which is even easier to build) is quite adequate for a 5065 <-> OCXO comparison. A triple channel setup (still in range for a basement project) would let you do a three corner hat. (with some range to it). If you have a three corner hat setup, the OCXO’s likely can be useful to 100 seconds. Past that a couple of fairly good telecom Rb’s could get you a bit further. There is also the possibility of running more than one 5065 … Not all parameters on the 5065 are at “near noise floor” levels. The close in ( < 20 Hz offset) phase noise is not all that great. One does not need a -170 dbc / Hz floor sort of setup to check it out. As tweaks are done, noise in this region *is* of interest. Comparison against a pretty normal OCXO would be “good enough” in this case as well. Bob > > For this reason I invite all those who dedicate themselves to these very > interesting changes proposed to test the results obtained in depth and to > share them with us with numerical and graphic elements. I take this > opportunity to thank them in advance for their scientific help. > > I want to specify that this is my point of view, it is not a rule and not > necessarily shared by other people. > > thankyou , Luciano > > Luciano P. S. Paramithiotti > tim...@timeok.it > www.timeok.it > > Da "time-nuts" time-nuts-boun...@lists.febo.com > A "Discussion of precise time and frequency measurement" > time-nuts@lists.febo.com > Cc > Data Tue, 4 Aug 2020 13:32:31 -0400 > Oggetto Re: [time-nuts] "The Penultimate HP5065 A15" > Hi > > If a more extensive rebuild is in the works ….. > > +/-20V is (as has been observed) not an ideal voltage for “modern” > electronics. > If you dig into each of the boards, there is a lot of “drop it down right > now” regulation > done on a board by board basis. More or less what might be there: > > A1 Synth &
Re: [time-nuts] "The Penultimate HP5065 A15"
10 years ago I started a project to try to improve on the 5065A by buying an Optical Unit in Europe and only use the temperature control boards. Temperature control with a fan and aging with a DAC based on 20 year experience with my FRK. Age, move and Macular hole in the left eye ended that project. Maybe do a coordinated redesign that is done by multiple members with as the final step being taking the Optical Unit and transfer it to the new chassis is an option. On the present unit if noise is a goal there are multiple HP 10811 with varying performance out there and why not bring out the 10 MHz with an amp. Has been done. Or an other OCXO? Our answer is a clean up loop with an external OCXO. Have good results. With the outstanding performance of my HP5065A the policy is hands off, do not even touch the C field dial in order to get aging data. Past tests have shown that after changing C field it takes time to settle. Bert Kehren Palm City Florida In a message dated 8/5/2020 5:56:17 AM Eastern Standard Time, tim...@timeok.it writes: I would like to share with you my thoughts on two important points that concern a little all attempts to modify and upgrade the equipment. In general I am very reluctant to modify both the electrically and mechanically, in this case, the HP5065A. Modifying, even completely a board, I agree if there is a significant improvement, trying to use the existing pinout and, in the need for additional connection points, use additional connectors on the board. As regards the space on the pcb, smd components can be used whenever possible or doughter boards. Remove boards, transformers or anything else I find reluctant, I would prefer to have the possibility of being able to go back to the original configuration if necessary. The area occupied by the optional batteries, which I think almost nobody uses, can be used for new electronics. Using a switching power supply with better performance I agree and since there is already a dedicated DC input on the back of the 5065A I would prefer to use the one to connect to an external box that contains the new power supply and the management of a backup DC input. This eliminates the need to dismantle anything inside the HP5065A. Another important point is that of the certainty of the results of a change. I mean that most hobbyists who have a 5065A, including me, do not have the opportunity to measure the proposed improvement effects, first of all because they do not have a reference such as an HMaser available, nor even such a refined measuring system to appreciate the improvements made. I want to remember that between zero and 8kseconds the GPS system (e.g. HP GPSDO) in our laboratories has an Adev higher than that of the HP5065A and therefore the measure we do in that range is that of GPS, not that of rubidium. For this reason I invite all those who dedicate themselves to these very interesting changes proposed to test the results obtained in depth and to share them with us with numerical and graphic elements. I take this opportunity to thank them in advance for their scientific help. I want to specify that this is my point of view, it is not a rule and not necessarily shared by other people. thankyou , Luciano Luciano P. S. Paramithiotti tim...@timeok.it www.timeok.it Da "time-nuts" time-nuts-boun...@lists.febo.com A "Discussion of precise time and frequency measurement" time-nuts@lists.febo.com Cc Data Tue, 4 Aug 2020 13:32:31 -0400 Oggetto Re: [time-nuts] "The Penultimate HP5065 A15" Hi If a more extensive rebuild is in the works ….. +/-20V is (as has been observed) not an ideal voltage for “modern” electronics. If you dig into each of the boards, there is a lot of “drop it down right now” regulation done on a board by board basis. More or less what might be there: A1 Synth A2 Battery Charger A3 60 MHz multiplier A4 100 KHz divider A5 Digital Divider (= clock) A6 1 MHz divider A7 AC amplifier A8 Phase detector A9 Integrator A10 OCXO A11 Rb temp control A12 Rb assembly A13 5 MHz buffer A14 Logic A15 Power supply A16 Power for clock A17 Terminal board A18 Jumpers ( = alt for A2) A19 Led Clock board For most uses, A4,A5,A6,A16,and A19 are not required. A2 is just a pair of diodes (A18) rather than a battery charger. A17 is more part of the wiring harness than anything else. Looking at what’s left: A1 synth, this seems to be a target for various replacement schemes. Right now, it has a bunch of positive voltage rails with some circuits running on 20V. Replacement likely would run on <= +12. A3 Multiplier. Again a target for replacement in some schemes. Same supply as A1 for replacement. Existing design runs +20 direct to a lot of circuits. A7 AC amplifier. Now runs +/-20V. Pretty much begs for a modern op-amp based replacement board. +/-12 probably is fine for that board. A *good
Re: [time-nuts] "The Penultimate HP5065 A15"
tim...@timeok.it writes: > Another important point is that of the certainty of the results > of a change. I mean that most hobbyists who have a 5065A, including > me, do not have the opportunity to measure the proposed improvement > effects, first of all because they do not have a reference such as > an HMaser available, nor even such a refined measuring system to > appreciate the improvements made. Well, this depends exactly what you mean by "measure". There is a big difference in being able to measure, with proper uncertainty intervals, that "A is 7.2% better than B", and merely measuring that "A is better than B (by some unknown amount)" Let me give you a concrete example: I am not kitted out for measuring close-in phase-noise. But if I feed two signals of *very* close frequency to my HP5370, hit "TI", "SAMP=100", and "STDDEV", then it shows me a number which correlates strongly with close-in phase-noise. If after modifying one of the two signal sources, that number drops, I have solid evidence of an improvement in close-in phase-noise. Mind you, I could have improved the phase-noise a LOT, and still see the same number, if the phase-noise of the other source dominated. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
I would like to share with you my thoughts on two important points that concern a little all attempts to modify and upgrade the equipment. In general I am very reluctant to modify both the electrically and mechanically, in this case, the HP5065A. Modifying, even completely a board, I agree if there is a significant improvement, trying to use the existing pinout and, in the need for additional connection points, use additional connectors on the board. As regards the space on the pcb, smd components can be used whenever possible or doughter boards. Remove boards, transformers or anything else I find reluctant, I would prefer to have the possibility of being able to go back to the original configuration if necessary. The area occupied by the optional batteries, which I think almost nobody uses, can be used for new electronics. Using a switching power supply with better performance I agree and since there is already a dedicated DC input on the back of the 5065A I would prefer to use the one to connect to an external box that contains the new power supply and the management of a backup DC input. This eliminates the need to dismantle anything inside the HP5065A. Another important point is that of the certainty of the results of a change. I mean that most hobbyists who have a 5065A, including me, do not have the opportunity to measure the proposed improvement effects, first of all because they do not have a reference such as an HMaser available, nor even such a refined measuring system to appreciate the improvements made. I want to remember that between zero and 8kseconds the GPS system (e.g. HP GPSDO) in our laboratories has an Adev higher than that of the HP5065A and therefore the measure we do in that range is that of GPS, not that of rubidium. For this reason I invite all those who dedicate themselves to these very interesting changes proposed to test the results obtained in depth and to share them with us with numerical and graphic elements. I take this opportunity to thank them in advance for their scientific help. I want to specify that this is my point of view, it is not a rule and not necessarily shared by other people. thankyou , Luciano Luciano P. S. Paramithiotti tim...@timeok.it www.timeok.it Da "time-nuts" time-nuts-boun...@lists.febo.com A "Discussion of precise time and frequency measurement" time-nuts@lists.febo.com Cc Data Tue, 4 Aug 2020 13:32:31 -0400 Oggetto Re: [time-nuts] "The Penultimate HP5065 A15" Hi If a more extensive rebuild is in the works ….. +/-20V is (as has been observed) not an ideal voltage for “modern” electronics. If you dig into each of the boards, there is a lot of “drop it down right now” regulation done on a board by board basis. More or less what might be there: A1 Synth A2 Battery Charger A3 60 MHz multiplier A4 100 KHz divider A5 Digital Divider (= clock) A6 1 MHz divider A7 AC amplifier A8 Phase detector A9 Integrator A10 OCXO A11 Rb temp control A12 Rb assembly A13 5 MHz buffer A14 Logic A15 Power supply A16 Power for clock A17 Terminal board A18 Jumpers ( = alt for A2) A19 Led Clock board For most uses, A4,A5,A6,A16,and A19 are not required. A2 is just a pair of diodes (A18) rather than a battery charger. A17 is more part of the wiring harness than anything else. Looking at what’s left: A1 synth, this seems to be a target for various replacement schemes. Right now, it has a bunch of positive voltage rails with some circuits running on 20V. Replacement likely would run on <= +12. A3 Multiplier. Again a target for replacement in some schemes. Same supply as A1 for replacement. Existing design runs +20 direct to a lot of circuits. A7 AC amplifier. Now runs +/-20V. Pretty much begs for a modern op-amp based replacement board. +/-12 probably is fine for that board. A *good* -15 would work for the existing board with minor mods. A8 Phase detector. Replacement probably is all digital. Now runs +20V A9 Later version of the board runs +/-15. Probably would work fine a good +/-12 with minor mods A10 If it’s a 10811, it’s going to need > +18 for the heater and +12 for the OCXO. There are other “at least as good” parts that work fine on +12. A11 Unless you want to redo the heater windings on A12, you are stuck with +20 to +30V. Rest of the board sort of begs for a modern op-amp approach. A12 Lamp assembly is the only load (other than heater windings and C field). It does run on +20V. A13 +20V taken to +9 for everything on the board. Simple mod to run on +12 (or +15 or +10 …). Replacement likely runs on +12 A14 If the upstream boards get changed, this likely does as well. Sort of begs for a $1 MCU and a handful of resistors as a replacement. A15 ( the topic of discussion) So, there are two
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 4, 2020, at 5:23 PM, Poul-Henning Kamp wrote: > > > Bob kb8tq writes: > >>> The "modern" approach to that is to modulate or dither with a >>> good long PRNG to whiten the noise, and while good in theory, >>> it is not _that_ easy to get right in practice. >> >> Of course one could simply cave in to the fact that for the sort of >> gain required here, a pure DC controller works quite well. >> >> I know - that takes all the fun out of it . > > It's not like PI(D) controllers were unknown when the 5065 was > designed, if anything people were better at "servo technology" > back then, than they are today, so I cannot imagine they would > not have considered them, and for some reason ruled them out. > > One problem with PI(D) that if your DUT has impedance (thermal in > our case) on the timescale of the main oscillatory external > disturbances (air-cons in our case), then at best a PI(D) will > introduce a delay in the oscillation, at worst it will amplify it. > > Today you can buy PID controllers where then processor will attempt > to phase-lock the primary disturbance and average it out. That > works great until somebody props the door open for ten minutes > (=half the air-con period) to fix the hinges. Dont ask me how I know. > > This is why I suspect the wien-bridge approach may not be a just a > homage to the HP200, I think they deliberately wanted to shift the > frequencies well north of the 137 Hz. > > Dithering a PI(D) with a PRNG gets you the same effect, but with > wideband noise instead of a single spectral peak. In the same era (1960’s) you could still find OCXO’s done with AC bridge controllers. Their main advantage is getting rid of DC offset voltage issues. Back in the day this was pretty tough to deal with. The world then discovered op-amps and they progressively got better and better. Yes indeed we see chopper based op-amps which are (to a great degree) as good as any AC loop. AC loop OCXO’s pretty much died out by the mid 70’s. Op-amp based controllers have dominated things since the late 70’s. The same is true of a lot of servo systems. You simply don’t need the audio drive anymore …. Bottom line still is: how much effective thermal gain can you reasonably expect to see on something built like the physics package? The structure is large (not a good thing). The two heated zones are coupled (also not a good thing). There is no reason to expect it to have a better thermal gain than your typical OCXO. If your controller would do ok for an OCXO, it should do fine for the physics package. Compare the HP 105 controller to the 5065 setup to see the similarities …. Bob > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 4, 2020, at 3:50 PM, Poul-Henning Kamp wrote: > > > Bob kb8tq writes: > >>> I have not given much thought to A11 yet, and I do find a certain >>> elegance in the wien-bridge approach. It is worth noting that the >>> actual heaters run of the unregulated supply, when I stabilized >>> that, A11 worked less hard. >> >> The AC bridge approach certainly worked well 'back in the day'. > > Actually, what I like is that it makes it possible to decide the > frequency spectrum of the inevitable noise, and in this case put > it up around a kilohertz where it is a lot less likely to cause > trouble. > > The "modern" approach to that is to modulate or dither with a > good long PRNG to whiten the noise, and while good in theory, > it is not _that_ easy to get right in practice. Of course one could simply cave in to the fact that for the sort of gain required here, a pure DC controller works quite well. I know - that takes all the fun out of it ….. Bob > > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Bob kb8tq writes: > > The "modern" approach to that is to modulate or dither with a > > good long PRNG to whiten the noise, and while good in theory, > > it is not _that_ easy to get right in practice. > > Of course one could simply cave in to the fact that for the sort of > gain required here, a pure DC controller works quite well. > > I know - that takes all the fun out of it . It's not like PI(D) controllers were unknown when the 5065 was designed, if anything people were better at "servo technology" back then, than they are today, so I cannot imagine they would not have considered them, and for some reason ruled them out. One problem with PI(D) that if your DUT has impedance (thermal in our case) on the timescale of the main oscillatory external disturbances (air-cons in our case), then at best a PI(D) will introduce a delay in the oscillation, at worst it will amplify it. Today you can buy PID controllers where then processor will attempt to phase-lock the primary disturbance and average it out. That works great until somebody props the door open for ten minutes (=half the air-con period) to fix the hinges. Dont ask me how I know. This is why I suspect the wien-bridge approach may not be a just a homage to the HP200, I think they deliberately wanted to shift the frequencies well north of the 137 Hz. Dithering a PI(D) with a PRNG gets you the same effect, but with wideband noise instead of a single spectral peak. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Bob kb8tq writes: > > I have not given much thought to A11 yet, and I do find a certain > > elegance in the wien-bridge approach. It is worth noting that the > > actual heaters run of the unregulated supply, when I stabilized > > that, A11 worked less hard. > > The AC bridge approach certainly worked well 'back in the day'. Actually, what I like is that it makes it possible to decide the frequency spectrum of the inevitable noise, and in this case put it up around a kilohertz where it is a lot less likely to cause trouble. The "modern" approach to that is to modulate or dither with a good long PRNG to whiten the noise, and while good in theory, it is not _that_ easy to get right in practice. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 4, 2020, at 3:07 PM, Poul-Henning Kamp wrote: > > > Bob kb8tq writes: > >> If a more extensive rebuild is in the works ….. > > Yes, that's where we're headed eventually, but I prefer to do it > incrementally. > >> +/-20V is (as has been observed) not an ideal voltage for “modern” >> electronics. > Right, and being able to drop that down later is in the design. > > However, from a noise point of view, there is a lot to be said for > having local LDO's on individual subassemblies. > >> A11 Unless you want to redo the heater windings on A12, you are >> stuck with +20 to +30V. Rest of the board sort of begs for a modern >> op-amp approach. > > I have not given much thought to A11 yet, and I do find a certain > elegance in the wien-bridge approach. It is worth noting that the > actual heaters run of the unregulated supply, when I stabilized > that, A11 worked less hard. The AC bridge approach certainly worked well “back in the day”. The thermistors that are in there are well adapted to that approach and not quite the value that one might want for a more modern controller. One could deal with that. The biggest issue I see with the design is that there are a few points of failure on the board that could take out the whole device. One could improve the MTBF of the control circuit and upgrade the ratings on the pass transistors at the same time. You also would get rid of the residual audio flying around in the heater winding. There may not be much, but there likely is some (espically if the caps have started to die off ….). Watching how various examples of the 5065 warm up (and ring), I suspect the A11 boards may not be in ideal shape …. > > Unless you plan to run the HP5065 at the freezing point, I expect > the heaters could run of a lower voltage, though +15V may be too low. The real question is - how long do you want to wait for it to warm up at normal temperatures? I suspect it would take a pretty long time to get going in your frozen garage in the middle of the winter, even with full voltage on the windings. If you really want to go nutty: Do an AC controller with an MCU in the middle of it ….. Bob > >> A14 If the upstream boards get changed, this likely does as well. >> Sort of begs for a $1 MCU and a handful of resistors as a replacement. > > I was actually pondering doing that on A17. > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Bob kb8tq writes: >If a more extensive rebuild is in the works ….. Yes, that's where we're headed eventually, but I prefer to do it incrementally. >+/-20V is (as has been observed) not an ideal voltage for “modern” >electronics. Right, and being able to drop that down later is in the design. However, from a noise point of view, there is a lot to be said for having local LDO's on individual subassemblies. >A11 Unless you want to redo the heater windings on A12, you are >stuck with +20 to +30V. Rest of the board sort of begs for a modern >op-amp approach. I have not given much thought to A11 yet, and I do find a certain elegance in the wien-bridge approach. It is worth noting that the actual heaters run of the unregulated supply, when I stabilized that, A11 worked less hard. Unless you plan to run the HP5065 at the freezing point, I expect the heaters could run of a lower voltage, though +15V may be too low. >A14 If the upstream boards get changed, this likely does as well. >Sort of begs for a $1 MCU and a handful of resistors as a replacement. I was actually pondering doing that on A17. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 4, 2020, at 2:49 PM, Poul-Henning Kamp wrote: > > > Bob kb8tq writes: > I would go for the LTZ1000 if you can. The LM399 has gone out of fashion for a reason: [...] >>> >>> Yes, I know, being also a volt-nut :-) >>> >>> Problem is, the A15 PCB isn't all that big, and if you want >>> cooling fins on two TO220's, a lot if it just got occupied. > > Just to follow up on the LTZ1000, this seems to be the smallest > non-SMD implementation of it: > > https://github.com/pepaslabs/px-ref/tree/master/kicad/releases/v2.4.1 > > That's a quarter of the A15 acreage, and more than I think is > warranted. You have already looked at the lamp supply sensitivity and the C-field. I don’t see any reason to go overly nutty with super references. You have a whole bunch of interesting ground issues and have various modules fed *through* the TED. > >> Unless you are running internal batteries, the battery charger board >> (as you have noted) is pretty much empty. For most of us, the 100 KHz >> and 1 MHz boards also are not very useful. They *could* be converted >> to other tasks. > > One of the premises for my current mucking about with kicad was to > make a plug-compatible A15, so that A/B testing would be easy, > both for me and for anybody else who wants to tag along. > >> There is a lot of space on the chassis for the batteries, the 1 pps >> divider and the clock. 1 pps is much easier to do today than it once was. >> Clock may or may not be a useful option. Again, a lot of space to >> expand this or that into. > > Indeed. > > I've kept & even renovated the LED clock, because it was there, but > I used the chance to get rid of the entire divider-chain: > > http://phk.freebsd.dk/hacks/HP5065A/20160112_working_clock/ > >>> BTW: I found a simple solution to this: I use the existing >>> chassis-mount NPN transistor as a pre-regulator, so the LDO's will >>> never see more voltage than they need to for regulation. > >> Pre-reg might also feed the heater circuits. One *could* do a pre-reg on the >> battery charger board. There pretty much has to be some impact to having >> the voltage to the heaters un-regulated …. > > As I said above, I'd like the A15 to be plug compatible if possible. > > Stabilizing the voltage for the heaters helped when I tried it, > but it is not obvious if this is due to misalignment or intrinsic > in the thermostatic regulators. Using OCXO’s as a reference, the supply voltage does indeed get in and mess with the set point. Less so on an external pass transistor design (like the 5065) than on an “all on the oven” approach. My though would be to put the pre-regulator in the A2 / A16 slot. That way it’s not soaking up any of the A15 board space. It also would make it fairly easy to swap things one at a time. In a design where you already have a switcher feeding the A15 (and the heaters), I doubt there would be a significant advantage to a pre-regulator. Bob > > I don't think stabilization on the LM399 level is required, but more > stable than the grid voltage would be good. > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Bob kb8tq writes: >>> I would go for the LTZ1000 if you can. The LM399 has gone out of >>> fashion for a reason: [...] >> >> Yes, I know, being also a volt-nut :-) >> >> Problem is, the A15 PCB isn't all that big, and if you want >> cooling fins on two TO220's, a lot if it just got occupied. Just to follow up on the LTZ1000, this seems to be the smallest non-SMD implementation of it: https://github.com/pepaslabs/px-ref/tree/master/kicad/releases/v2.4.1 That's a quarter of the A15 acreage, and more than I think is warranted. >Unless you are running internal batteries, the battery charger board >(as you have noted) is pretty much empty. For most of us, the 100 KHz >and 1 MHz boards also are not very useful. They *could* be converted >to other tasks. One of the premises for my current mucking about with kicad was to make a plug-compatible A15, so that A/B testing would be easy, both for me and for anybody else who wants to tag along. > There is a lot of space on the chassis for the batteries, the 1 pps > divider and the clock. 1 pps is much easier to do today than it once was. > Clock may or may not be a useful option. Again, a lot of space to > expand this or that into. Indeed. I've kept & even renovated the LED clock, because it was there, but I used the chance to get rid of the entire divider-chain: http://phk.freebsd.dk/hacks/HP5065A/20160112_working_clock/ >> BTW: I found a simple solution to this: I use the existing >> chassis-mount NPN transistor as a pre-regulator, so the LDO's will >> never see more voltage than they need to for regulation. >Pre-reg might also feed the heater circuits. One *could* do a pre-reg on the >battery charger board. There pretty much has to be some impact to having >the voltage to the heaters un-regulated …. As I said above, I'd like the A15 to be plug compatible if possible. Stabilizing the voltage for the heaters helped when I tried it, but it is not obvious if this is due to misalignment or intrinsic in the thermostatic regulators. I don't think stabilization on the LM399 level is required, but more stable than the grid voltage would be good. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
On 8/4/2020 10:21 AM, cdel...@juno.com wrote: > > Back in Nov 2017 I detailed my A15 mods: > > Set Zener current to minimum TC (I achieved < 1PPM/C at the +20V output) > This involved disconnecting the C-field circuit and replacing it with a > dual (in series ) LM299AH and 1PPM resistors. Jumperable settings and a > low TC pot for adjustment. I would have liked to use 3 LM299AH to give me > 21 Volts but would need to switch the input to the unregulated supply to > give enough overhead. The higher the voltage into the C-field string the > lower the TC contribution from the C-field coil itself. > > The posts have the details and schematic. > > Cheers, > Corby To save the rest of you time, here are some of Corby's 5065A posts: "HP5065A C-field mods and optical unit mods" http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090056.html http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090067.html http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090107.html http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090138.html http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090189.html http://lists.febo.com/pipermail/time-nuts_lists.febo.com/2017-November/090255.html /tvb ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi If a more extensive rebuild is in the works ….. +/-20V is (as has been observed) not an ideal voltage for “modern” electronics. If you dig into each of the boards, there is a lot of “drop it down right now” regulation done on a board by board basis. More or less what might be there: A1 Synth A2 Battery Charger A3 60 MHz multiplier A4 100 KHz divider A5 Digital Divider (= clock) A6 1 MHz divider A7 AC amplifier A8 Phase detector A9 Integrator A10 OCXO A11 Rb temp control A12 Rb assembly A13 5 MHz buffer A14 Logic A15 Power supply A16 Power for clock A17 Terminal board A18 Jumpers ( = alt for A2) A19 Led Clock board For most uses, A4,A5,A6,A16,and A19 are not required. A2 is just a pair of diodes (A18) rather than a battery charger. A17 is more part of the wiring harness than anything else. Looking at what’s left: A1 synth, this seems to be a target for various replacement schemes. Right now, it has a bunch of positive voltage rails with some circuits running on 20V. Replacement likely would run on <= +12. A3 Multiplier. Again a target for replacement in some schemes. Same supply as A1 for replacement. Existing design runs +20 direct to a lot of circuits. A7 AC amplifier. Now runs on +/-20V. Pretty much begs for a modern op-amp based replacement board. +/-12 probably is fine for that board. A *good* -15 would work for the existing board with minor mods. A8 Phase detector. Replacement probably is all digital. Now runs on +20V only. A9 Later version of the board runs on +/-15. Probably would work fine on a good +/-12 with minor mods A10 If it’s a 10811, it’s going to need > +18 for the heater and +12 for the OCXO. There are other “at least as good” parts that work fine on +12. A11 Unless you want to redo the heater windings on A12, you are stuck with +20 to +30V. Rest of the board sort of begs for a modern op-amp approach. A12 Lamp assembly is the only load (other than heater windings and C field). It does run on +20V. A13 +20V taken to +9 for everything on the board. Simple mod to run on +12 (or +15 or +10 …). Replacement likely runs on +12 A14 If the upstream boards get changed, this likely does as well. Sort of begs for a $1 MCU and a handful of resistors as a replacement. A15 ( the topic of discussion) So, there are two “customers” for -20V. Both would be happy with a fairly good -15V instead. If the replacement OCXO for A10 tunes 0 to 5V, the need for a negative supply becomes a bit unclear. There are a couple of places you can’t easily get around 20 to 30V. The heater windings on the physics package are the biggie. The Rb lamp driver is probably not worth messing with. If you keep A1, A3, A7, or A8, they will need the existing +20V. +12 or +15 makes more sense for their replacements. A13 could easily be modified and run on +12 or +15. So why all this long winded yack? The ultimate need for +20 is really pretty small. The lamp still needs it. The heaters need something in that vicinity. They don’t need the super regulation or low noise that the lamp needs. The -20 probably does at least as well as a -15 V supply, even as the device sits right now. Long term -15 makes more sense. A bulk +12 (and maybe +5) likely take up the heavy lifting for most of the boards once they are modified. Indeed they would be useful even on the existing boards. Does this change anything you do on the A15 right now? Maybe not. It is worth thinking a bit on though. Bob > On Aug 4, 2020, at 6:35 AM, Poul-Henning Kamp wrote: > > > Magnus Danielson writes: > >> If that is the range you aim to improve, then I strongly recommend you >> do re-read the postings I did about phase noise in detector range, as I >> have showed that phase-noise get's mixed down to DC and increase the >> noise there. With very ugly and hand-wavey hack I was able to >> significantly reduce that noise essentially by cleaning up the >> phase-noise. > > Ohh, absolutely. > > My plan is to put a modulated DDS synth in to get a much more modern > detector system. > > But right now the zener on my a15 is better at measuring temperature > than holding voltage stable, so I'm addressing that first. > > See the 3rd graph here: > > http://phk.freebsd.dk/hacks/HP5065A/20150908_a15/ > I would go for the LTZ1000 if you can. The LM399 has gone out of fashion for a reason: [...] >>> Yes, I know, being also a volt-nut :-) >> >> I have not yet jumped down that rabbit-hole as I am by far not done with >> the current one. :) >> >> I've been looking at the LTZ1000 and one day I may get some. > > You *really* want the version of it which is packaged in the attractive > and *incredibly* useful "HP3458A" cabinet :-) > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incomp
[time-nuts] "The Penultimate HP5065 A15"
>> C-field polarity switch. An experiment. If nothing else >> I will be able to measure the residual magnetic field. There is no requirement to do any thing other than switch the C-field polarity. No optical stuff needed. The PRS10 switches the C-field polarity at a 5HZ rate to reduce magnetic effects and does not do any optical tricks! Back in Nov 2017 I detailed my A15 mods: Set Zener current to minimum TC (I achieved < 1PPM/C at the +20V output) This involved disconnecting the C-field circuit and replacing it with a dual (in series ) LM299AH and 1PPM resistors. Jumperable settings and a low TC pot for adjustment. I would have liked to use 3 LM299AH to give me 21 Volts but would need to switch the input to the unregulated supply to give enough overhead. The higher the voltage into the C-field string the lower the TC contribution from the C-field coil itself. The posts have the details and schematic. Cheers, Corby ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi Poul-Henning, On 2020-08-04 12:35, Poul-Henning Kamp wrote: > > Magnus Danielson writes: > >> If that is the range you aim to improve, then I strongly recommend you >> do re-read the postings I did about phase noise in detector range, as I >> have showed that phase-noise get's mixed down to DC and increase the >> noise there. With very ugly and hand-wavey hack I was able to >> significantly reduce that noise essentially by cleaning up the >> phase-noise. > Ohh, absolutely. > > My plan is to put a modulated DDS synth in to get a much more modern > detector system. That would help naturally, given that OCXO phase-noise is well taken care off. I'm considering swapping out the 00105 oscillators, way too much noise to be useful. > But right now the zener on my a15 is better at measuring temperature > than holding voltage stable, so I'm addressing that first. > > See the 3rd graph here: > > http://phk.freebsd.dk/hacks/HP5065A/20150908_a15/ > Temperature measurements is nice. :) Yeah, I've seen that too. I would go for the LTZ1000 if you can. The LM399 has gone out of fashion for a reason: [...] >>> Yes, I know, being also a volt-nut :-) >> I have not yet jumped down that rabbit-hole as I am by far not done with >> the current one. :) >> >> I've been looking at the LTZ1000 and one day I may get some. > You *really* want the version of it which is packaged in the attractive > and *incredibly* useful "HP3458A" cabinet :-) > How nice of HP ehm Agilent to supply it in a pre-packaged format with some auxillary functionality to go with it. :) Yeah, I got a pair of "HP3457A" cabinets here, but so far no "HP3458A" cabinet. Hopefully soon enough. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Magnus Danielson writes: > If that is the range you aim to improve, then I strongly recommend you > do re-read the postings I did about phase noise in detector range, as I > have showed that phase-noise get's mixed down to DC and increase the > noise there. With very ugly and hand-wavey hack I was able to > significantly reduce that noise essentially by cleaning up the > phase-noise. Ohh, absolutely. My plan is to put a modulated DDS synth in to get a much more modern detector system. But right now the zener on my a15 is better at measuring temperature than holding voltage stable, so I'm addressing that first. See the 3rd graph here: http://phk.freebsd.dk/hacks/HP5065A/20150908_a15/ > > > I would go for the LTZ1000 if you can. The LM399 has gone out of > >> fashion for a reason: [...] > > Yes, I know, being also a volt-nut :-) > > I have not yet jumped down that rabbit-hole as I am by far not done with > the current one. :) > > I've been looking at the LTZ1000 and one day I may get some. You *really* want the version of it which is packaged in the attractive and *incredibly* useful "HP3458A" cabinet :-) -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi Poul-Henning, On 2020-08-04 00:32, Poul-Henning Kamp wrote: > > Attila Kinali writes: > >>> The main reason I dont just repair/replace the A15 is that >>> I want to find out how much instability the PSU contributes. >> There are very few points where supply voltage (in)stability >> contributes to long term frequency drift in a Rb vapor cell standard. > Obviously. > > But it does contribute quite a lot on the sub-hour time-scale, some > of it via temperature-dependencies. If that is the range you aim to improve, then I strongly recommend you do re-read the postings I did about phase noise in detector range, as I have showed that phase-noise get's mixed down to DC and increase the noise there. With very ugly and hand-wavey hack I was able to significantly reduce that noise essentially by cleaning up the phase-noise. Just as Corby's filter trick, which also aims to reduce noise, one need to figure out the dominant noise sources, then reduce them from the top candidate and then down. I have not seen your efforts include the oscillator phase-noise consideration. With that I'm not saying that you attempts is wrong, rather, maybe not the first thing that I would start with. I will pursue my phase-noise approach until I see diminishing returns there, and then look to see what could be the next dominant noise contribution. In this regard I do not see Corbys or your approaches as lost, but rather very interesting, as I come to that point. > > I would go for the LTZ1000 if you can. The LM399 has gone out of >> fashion for a reason: [...] > Yes, I know, being also a volt-nut :-) I have not yet jumped down that rabbit-hole as I am by far not done with the current one. :) I've been looking at the LTZ1000 and one day I may get some. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hi > On Aug 3, 2020, at 6:32 PM, Poul-Henning Kamp wrote: > > > Attila Kinali writes: > >>> The main reason I dont just repair/replace the A15 is that >>> I want to find out how much instability the PSU contributes. >> >> There are very few points where supply voltage (in)stability >> contributes to long term frequency drift in a Rb vapor cell standard. > > Obviously. > > But it does contribute quite a lot on the sub-hour time-scale, some > of it via temperature-dependencies. > >> I do not have studied the 5065 design in all detail (i.e. I might >> be wrong), but the only points where I think it really matters >> is the C-field and the Rb lamb supply. > > Indeed, and those get special attention by me. > >> I would go for the LTZ1000 if you can. The LM399 has gone out of >> fashion for a reason: [...] > > Yes, I know, being also a volt-nut :-) > > Problem is, the A15 PCB isn't all that big, and if you want > cooling fins on two TO220's, a lot if it just got occupied. Unless you are running internal batteries, the battery charger board (as you have noted) is pretty much empty. For most of us, the 100 KHz and 1 MHz boards also are not very useful. They *could* be converted to other tasks. There is a lot of space on the chassis for the batteries, the 1 pps divider and the clock. 1 pps is much easier to do today than it once was. Clock may or may not be a useful option. Again, a lot of space to expand this or that into. > >>> C-field polarity switch. An experiment. If nothing else >>> I will be able to measure the residual magnetic field. >> >> For this to be a valid test, you also need to switch the >> polarity of the light, otherwise you excite different states >> with slightly different magnetic dependence. > > "polarity of the light" - You mean the direction ? > >>> Two +20V on-board LM317-style linear regulators, one for >>> the lamp, one for the rest. Split for noise reasons and >>> to be able to play with the lamp voltage/power. >> >> I would go for a lower noise regulator. > > As would I, but it seems all the candidates I have found either > stall out at 20V abs. max or come in unsolderable packages. > > I'm not convinced the LM317 noise contribution matters anyway, > as long as they are slaved to the LM399 with an OP-AMP, their > long term stability is out of the picture. > >> I would go for one of the inverting µModule regulators from ADI. >> They offer quite low noise/EMI while being easy to use. Only >> downside: They all come in BGA cases, which makes them hard to >> solder by hand. > > Electronics is increasingly leaving the good old soldering iron behind :-/ > >>> The downside of the two linear +20V regulators is that even with a >>> heatsink, they will probably get hot-ish if the internal DC bus is >>> too much over 24VDC. > > BTW: I found a simple solution to this: I use the existing > chassis-mount NPN transistor as a pre-regulator, so the LDO's will > never see more voltage than they need to for regulation. Pre-reg might also feed the heater circuits. One *could* do a pre-reg on the battery charger board. There pretty much has to be some impact to having the voltage to the heaters un-regulated …. Bob > >> My recommendation would be to replace the whole power input >> by a modern switched power supply. > > I have already done that, in a pretty hackish way: > > http://phk.freebsd.dk/hacks/HP5065A/20150930_dcdc/ > > I need to revisit tha and add more filtering, some common mode > noise gets through. > > My plan was to repurpose A2 for a DC/DC brick and a couple of > "ideal diode" power controllers. > >> I recommend using regulators in a TO-220 case >> and something like an Aavid 7021, 7022, 7023, or 5510 heatsink. > > TO-220 (and being indestructible) is why LM317 is my primary candidate. > > Thanks for the feedback. > > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > p...@freebsd.org | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
> On 04 August 2020 at 10:32 Poul-Henning Kamp wrote: > > > > Attila Kinali writes: > > >> The main reason I dont just repair/replace the A15 is that > >> I want to find out how much instability the PSU contributes. > > > >There are very few points where supply voltage (in)stability > >contributes to long term frequency drift in a Rb vapor cell standard. > > Obviously. > > But it does contribute quite a lot on the sub-hour time-scale, some > of it via temperature-dependencies. > > > > > >For this to be a valid test, you also need to switch the > >polarity of the light, otherwise you excite different states > >with slightly different magnetic dependence. > > "polarity of the light" - You mean the direction ? More likely polarisation switching was intended. eg between LH and RH circularly polarised light. A linear polariser and a quarter waveplate are required to produce circularly polarised light from the unpolarised light emitted by a rubidium lamp. Bruce > > ___ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Attila Kinali writes: >> The main reason I dont just repair/replace the A15 is that >> I want to find out how much instability the PSU contributes. > >There are very few points where supply voltage (in)stability >contributes to long term frequency drift in a Rb vapor cell standard. Obviously. But it does contribute quite a lot on the sub-hour time-scale, some of it via temperature-dependencies. >I do not have studied the 5065 design in all detail (i.e. I might >be wrong), but the only points where I think it really matters >is the C-field and the Rb lamb supply. Indeed, and those get special attention by me. >I would go for the LTZ1000 if you can. The LM399 has gone out of >fashion for a reason: [...] Yes, I know, being also a volt-nut :-) Problem is, the A15 PCB isn't all that big, and if you want cooling fins on two TO220's, a lot if it just got occupied. >> C-field polarity switch. An experiment. If nothing else >> I will be able to measure the residual magnetic field. > >For this to be a valid test, you also need to switch the >polarity of the light, otherwise you excite different states >with slightly different magnetic dependence. "polarity of the light" - You mean the direction ? >> Two +20V on-board LM317-style linear regulators, one for >> the lamp, one for the rest. Split for noise reasons and >> to be able to play with the lamp voltage/power. > >I would go for a lower noise regulator. As would I, but it seems all the candidates I have found either stall out at 20V abs. max or come in unsolderable packages. I'm not convinced the LM317 noise contribution matters anyway, as long as they are slaved to the LM399 with an OP-AMP, their long term stability is out of the picture. >I would go for one of the inverting µModule regulators from ADI. >They offer quite low noise/EMI while being easy to use. Only >downside: They all come in BGA cases, which makes them hard to >solder by hand. Electronics is increasingly leaving the good old soldering iron behind :-/ >> The downside of the two linear +20V regulators is that even with a >> heatsink, they will probably get hot-ish if the internal DC bus is >> too much over 24VDC. BTW: I found a simple solution to this: I use the existing chassis-mount NPN transistor as a pre-regulator, so the LDO's will never see more voltage than they need to for regulation. >My recommendation would be to replace the whole power input >by a modern switched power supply. I have already done that, in a pretty hackish way: http://phk.freebsd.dk/hacks/HP5065A/20150930_dcdc/ I need to revisit tha and add more filtering, some common mode noise gets through. My plan was to repurpose A2 for a DC/DC brick and a couple of "ideal diode" power controllers. >I recommend using regulators in a TO-220 case >and something like an Aavid 7021, 7022, 7023, or 5510 heatsink. TO-220 (and being indestructible) is why LM317 is my primary candidate. Thanks for the feedback. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Hoi Poul-Henning, On Fri, 31 Jul 2020 21:31:13 + "Poul-Henning Kamp" wrote: > The main reason I dont just repair/replace the A15 is that > I want to find out how much instability the PSU contributes. There are very few points where supply voltage (in)stability contributes to long term frequency drift in a Rb vapor cell standard. Noise is much more important than stability as it directly contributes to the noise in the FLL. Especially in a circuit like the 5065, where a lot of components don't have the high PSRR that modern components give us. I do not have studied the 5065 design in all detail (i.e. I might be wrong), but the only points where I think it really matters is the C-field and the Rb lamb supply. > The outline idea currently is: > > LM399 self-biased voltage reference > (See: Linear app-note 42, fig 72) > > The ultimate board would use LTZ1000, but I have not quite > convinced myself yet, even if it would be cute to have a > HP5065 which delivered both precise frequency and voltage :-) I would go for the LTZ1000 if you can. The LM399 has gone out of fashion for a reason: Modern bandgap references have similar noise at a fraction of the cost and if you need stability the LTZ1000 is the better option. E.g., the LTC6655 is only about a factor two worse in long term performance (8ppm/sqrt(k hour) vs 20ppm/sqrt(k hour)) and about the same factor in noise voltage. The LTZ1000 offers both better aging (including temp stability) and noise performance. Of course an LTZ1000 costs 4 times as much and that doesn't include the support circuit that you need for temperature control. > C-field polarity switch. An experiment. If nothing else > I will be able to measure the residual magnetic field. For this to be a valid test, you also need to switch the polarity of the light, otherwise you excite different states with slightly different magnetic dependence. > Two +20V on-board LM317-style linear regulators, one for > the lamp, one for the rest. Split for noise reasons and > to be able to play with the lamp voltage/power. I would go for a lower noise regulator. I would recommend the TPS7A45xx.. but unfortunately, they are 20V abs max. A good choice might be the LT3081. Like the other members of the LT3042 family, it offers quite low noise while still being able to deliver 1.5A. It's quite a bit more expensive, though. > One or two pre-regulator current measurement shunts. > > DC/DC-brick switchmode -20V supply with brutal filtering. I would go for one of the inverting µModule regulators from ADI. They offer quite low noise/EMI while being easy to use. Only downside: They all come in BGA cases, which makes them hard to solder by hand. https://www.analog.com/en/products/power-management/umodule-regulators/umodule-inverting-regulators.html > The downside of the two linear +20V regulators is that even with a > heatsink, they will probably get hot-ish if the internal DC bus is > too much over 24VDC. For this reason, and because I may simply run > out of PCB space, I may leave the bridge rectifier out, so it will > only works with EXT-DC. My recommendation would be to replace the whole power input by a modern switched power supply. These are quite low noise and, if used correctly, also low EMI. It also allows you to reduce the voltage, due to better regulation, and thus save on power. Just having an LT8640S as pre-regulator would reduce the dissipated power quite considerably with having low to negligible impact on EMI (Silent-Switcher 2 architecture). But yes, power dissipation will be a problem, you have approximately 0.8W dissipated per difference in voltage. Which means it's already 3W for 24V. While not exporbitantly high, it's already quite considerable if you have no forced air cooling. If you want to go that way, I recommend using regulators in a TO-220 case and something like an Aavid 7021, 7022, 7023, or 5510 heatsink. The larger ones should allow you to even go up to 30V. If you have the space to use a switched power supply and to implement a feedback system such that the output of the switched supply is kept at 2V above the output of the LDOs, you'd reduce power dissipation quite a bit and should be able to get away with D2PAK LDOs and a small heatsink. Attila Kinali -- The bad part of Zurich is where the degenerates throw DARK chocolate at you. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
Having one of the best HP5065A Rb's that Corby has seen right at 4 E-14 at 200 and 400 sec. I have been hands off. My age and macular hole on my right eye has impacted my hardware work. A recent test also shows very low aging over a 9 month period. But last September my environmental monitor was with Tom. The last month I have spend strictly on temperature to better understand what lab temperature effects the Rb. This moment a 48 hour frequency test is in the works because there is a tropical storm that was a hurricane 20 miles east from me. Hope there are no power failures because the Rb is not on a UPS but like all my frequency equipment is on a Power Line Conditioner including the Rb. No power failure I will be able to share frequency versus pressure and temperature. C field control Corby included by my request a fine control using a Vishay trim pot, knowing what I know today I would make the trim pot the coarse and the original Hp the fine. There is little need for coarse adjustment. On the A 15 subject I keep hands off but Juerg is making some changes. In the AC mode there is up to 20 W heat generated between the transformer and the 24 V DC input. Our answer is a Buck Converter on the output of the AC supply down to 16 V for the Li Ion module that in turn feeds a Mean Well SD-50A-24 DC?DC converter. It 24 V output is not adjustable but that is a small price to pay. We have done tests feeding an OSA 8600 directly and have not seen any degradation. But in this case the A 15 board still does its part. The parts are not in the Rb but behind on the back. A pressure, temperature and two DAC board is in the works. Next hurricane the tests will be on a FRK/M100 because that will be my future HP5065A replacement. Bert Kehren Palm City Florida In a message dated 8/2/2020 4:38:31 AM Eastern Standard Time, time-nuts@lists.febo.com writes: Regarding 5065 "A15" When I was about to repair my 5065, the first thing I noticed was thequite high "+22-30V" and its impact on the dissipated heat from theseries regulator Q1. Therefore, I replaced the power transformer T1and the rectifiers with a linear regulated power supply providingfiltered 24V DC. The original circuitry incorporated a battery charger A2but since the accus was (very) dead, it was replaced with a diodejumper board A18 from a second scrapped unit. The original diodesgot hot. They were replaced with modern types. I will follow your progress with the new regulator board with great interest. Best Regards Ulf KylenfallSM6GXV___time-nuts mailing list -- time-n...@lists.febo.comTo unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.comand follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
tim...@timeok.it writes: >A better C field generator is necessary and the LM399 remains a good > solution. It would be interesting to add > barometric pressure compensation. I don't think I can fit that also :-) > Two types of setting could be used to adjust the C field, one with fixed high > stability resistors resistors that can be switched for the coarse and a > potentiometer for fine tuning in order > to mitigate its low temperature stability. already got that. >Excellent solution of two separate power supplies for the lamp and the > rest of the electronics but I have > doubts about the LM317. Since we already have the LM399 available for the > CField, why not use this reference > voltage already present on the board to drive two low noise regulators? That's precisely the plan, but I like the LM317 style short-circuit and thermal protection, so I will probably do it by setting the LM317 voltage with a voltage divider, and then drive the midpoint with an op-amp relative to the LM399 -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
[time-nuts] "The Penultimate HP5065 A15"
Regarding 5065 "A15" When I was about to repair my 5065, the first thing I noticed was the quite high "+22-30V" and its impact on the dissipated heat from the series regulator Q1. Therefore, I replaced the power transformer T1 and the rectifiers with a linear regulated power supply providing filtered 24V DC. The original circuitry incorporated a battery charger A2 but since the accus was (very) dead, it was replaced with a diode jumper board A18 from a second scrapped unit. The original diodesgot hot. They were replaced with modern types. I will follow your progress with the new regulator board with great interest. Best Regards Ulf Kylenfall SM6GXV ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
Re: [time-nuts] "The Penultimate HP5065 A15"
It has been a while since there has been talk of a redesign of the A15 and it is certainly the right way to technologically update the HP5065A that deserves better electronics. A better C field generator is necessary and the LM399 remains a good solution. It would be interesting to add barometric pressure compensation. Two types of setting could be used to adjust the C field, one with fixed high stability resistors resistors that can be switched for the coarse and a potentiometer for fine tuning in order to mitigate its low temperature stability. Excellent solution of two separate power supplies for the lamp and the rest of the electronics but I have doubts about the LM317. Since we already have the LM399 available for the CField, why not use this reference voltage already present on the board to drive two low noise regulators? I hope to be of contribution to this project. Luciano Luciano P. S. Paramithiotti tim...@timeok.it www.timeok.it Da "time-nuts" time-nuts-boun...@lists.febo.com A time-nuts@lists.febo.com Cc Data Fri, 31 Jul 2020 21:31:13 + Oggetto [time-nuts] "The Penultimate HP5065 A15" The A15 board in one of my 5065s is in bad shape, and I have started to look at designing a plug-compatible replacement board. The main reason I dont just repair/replace the A15 is that I want to find out how much instability the PSU contributes. The outline idea currently is: LM399 self-biased voltage reference (See: Linear app-note 42, fig 72) The ultimate board would use LTZ1000, but I have not quite convinced myself yet, even if it would be cute to have a HP5065 which delivered both precise frequency and voltage :-) Op-amp based C-field driver Vishay SMR1DZ resistors for VREF/C-field stability C-field polarity switch. An experiment. If nothing else I will be able to measure the residual magnetic field. Optional adjustable C-field. Optional because the pot may degrade the C-field stability. Two +20V on-board LM317-style linear regulators, one for the lamp, one for the rest. Split for noise reasons and to be able to play with the lamp voltage/power. One or two pre-regulator current measurement shunts. DC/DC-brick switchmode -20V supply with brutal filtering. The downside of the two linear +20V regulators is that even with a heatsink, they will probably get hot-ish if the internal DC bus is too much over 24VDC. For this reason, and because I may simply run out of PCB space, I may leave the bridge rectifier out, so it will only works with EXT-DC. I'll post kicad schematic&pcb once I get further. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
[time-nuts] "The Penultimate HP5065 A15"
The A15 board in one of my 5065s is in bad shape, and I have started to look at designing a plug-compatible replacement board. The main reason I dont just repair/replace the A15 is that I want to find out how much instability the PSU contributes. The outline idea currently is: LM399 self-biased voltage reference (See: Linear app-note 42, fig 72) The ultimate board would use LTZ1000, but I have not quite convinced myself yet, even if it would be cute to have a HP5065 which delivered both precise frequency and voltage :-) Op-amp based C-field driver Vishay SMR1DZ resistors for VREF/C-field stability C-field polarity switch. An experiment. If nothing else I will be able to measure the residual magnetic field. Optional adjustable C-field. Optional because the pot may degrade the C-field stability. Two +20V on-board LM317-style linear regulators, one for the lamp, one for the rest. Split for noise reasons and to be able to play with the lamp voltage/power. One or two pre-regulator current measurement shunts. DC/DC-brick switchmode -20V supply with brutal filtering. The downside of the two linear +20V regulators is that even with a heatsink, they will probably get hot-ish if the internal DC bus is too much over 24VDC. For this reason, and because I may simply run out of PCB space, I may leave the bridge rectifier out, so it will only works with EXT-DC. I'll post kicad schematic&pcb once I get further. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.