Power supply design (2)

[David Hess]

What does the output of U2 do at startup?

What does the non-inverting input of U2 do at startup?

With the output set at 10 volts, pin 3 sits at 3.74 volts.  When I measure during power on at pin 3 the voltage rises to 6.3 volts before falling back to 3.74 volts.  The same thing happens at the inverting input.

Pin 6 outputs 10.65 volts.  On power on I get 17.65 volts.

That answers that then.  The output from U1 is actually the problem.

The inverting input should follow the non-inverting input if the operational amplifier is acting as an operational amplifier.

My WAG (wild ass guess) is that U1 operating open loop during startup before D8 conducts results in overshoot caused by windup of its internal compensation capacitor.  If U1 had a negative offset voltage, then I think it could latch off but we need not worry about that since it is not happening at the moment.  In a production design I would want to test for that.

Try adding a capacitor across D8.  I am not sure what the value should be but a .1 microfarad ceramic or film would be a good place to start.  This will force U1 to operate closed loop during startup and lower noise from the reference as a side effect.

[liquibyte]

That answers that then.  The output from U1 is actually the problem.

The inverting input should follow the non-inverting input if the operational amplifier is acting as an operational amplifier.

My WAG (wild ass guess) is that U1 operating open loop during startup before D8 conducts results in overshoot caused by windup of its internal compensation capacitor.  If U1 had a negative offset voltage, then I think it could latch off but we need not worry about that since it is not happening at the moment.  In a production design I would want to test for that.

Try adding a capacitor across D8.  I am not sure what the value should be but a .1 microfarad ceramic or film would be a good place to start.  This will force U1 to operate closed loop during startup and lower noise from the reference as a side effect.

I've added a 0.1uF ceramic across D8 but I'm still getting about 16.7 volts on startup before it settles back to 10 volts.  I have a jar of many values of ceramics from 1pF to 100nF that I got off Ebay for doing just this kind of thing.

The funny thing is that the version that I have that has a vreg in place of U1 has the exact same issue and it doesn't have a D8 either.

I've even tried loading the output from 2.5K to 3K (6 & 5 15K parallel resistors) to see if it helped but it didn't.  It does help load enough so that RV1 can get the offset of U2 down to around 15mV though.  I'm not sure if the difference between 35mV and 15mV is worth the trouble considering the transient voltage problem.  If it helped in that regard, I'd definitely keep it.

[David Hess]

You could try a larger capacitor across D8.

Try removing C4.

[liquibyte]

You could try a larger capacitor across D8.

Try removing C4.

I tried several from around 0.1uF up to 2.2uF with no luck.  Removed C4 but nothing changed.

I've tried measuring across C1 to see if something is going on from there and I get 61 volts initially settling to 43 volts.  Could this be what's causing this?  Would there be a way to limit this initial inrush to 44 volts tops by using a TVS or something?

[David Hess]

You could try a larger capacitor across D8.

Try removing C4.

I tried several from around 0.1uF up to 2.2uF with no luck.  Removed C4 but nothing changed.

I've tried measuring across C1 to see if something is going on from there and I get 61 volts initially settling to 43 volts.  Could this be what's causing this?  Would there be a way to limit this initial inrush to 44 volts tops by using a TVS or something?

This is a problem and not only because of what it might cause the operational amplifiers to do.  The supply voltage surge could cause this problem although I have no specific experience with TI's Excalibur series of operational amplifiers to know for sure.  Their lower voltage CMOS operational amplifiers would just die with that kind of surge.

The absolute maximum supply voltage for the TLE2141 is 44 volts.  That rating is not for continuous operation and in practice the supply voltage needs to be significantly lower for reliable operation.

You do not want to use a TVS or shunt to try and stop that kind of surge.  A better option would be to use a zener shunt regulator to control and limit the operational amplifier supply voltage to about 36 volts.

[liquibyte]

I was afraid this would eventually come up.  I had concerns back when I started this project that the op amps were being kept too close to their upper limit and that it might have an effect on performance.  What kinds of bad effects do keeping op amps towards their upper supply limit cause?  I was told that everything was within tolerance and not to worry.  Something still kept bugging me about it but since many people had built these and they supposedly worked, I built mine.  I guess not many people have really tested the design until now.

What if I were to use an op amp that had a bit higher limits, say ±48 volt such as this one or this one?  The characteristics are similar and it would put me a few more volts above my input voltage giving a tad more headroom if I can contain the initial surge from the transformer.  Pinout is identical.  The first one has closer specs I believe but the latter has a higher supply voltage range.  I'd really like to solve this but I'm not entirely sure it can be done without a significant redesign.

I have a feeling that I'm going to be looking for something different to build soon.  I wanted to build this so that I could get the ±15 volts to start learning more about op amps and I find it ironic that I don't have a supply that can to help solve this issue.  Can you recommend something to be on the lookout for on Ebay, dual tracking perhaps or at least cheap enough to get two and use in series?

[David Hess]

What kinds of bad effects do keeping op amps towards their upper supply limit cause?

They become generally unreliable and more likely to fail.  Often they immediately fail if you exceed the maximum rating by just a couple volts.

I was told that everything was within tolerance and not to worry.  Something still kept bugging me about it but since many people had built these and they supposedly worked, I built mine.  I guess not many people have really tested the design until now.

The design assumes that the unregulated supply is not spiking above 44 (or 36) volts when power is applied.

What if I were to use an op amp that had a bit higher limits, say ±48 volt such as this one or this one?  The characteristics are similar and it would put me a few more volts above my input voltage giving a tad more headroom if I can contain the initial surge from the transformer.  Pinout is identical.  The first one has closer specs I believe but the latter has a higher supply voltage range.  I'd really like to solve this but I'm not entirely sure it can be done without a significant redesign.

Neither of those are single supply amplifiers so they will require about a -5 volt supply to operate correctly in this circuit.  The opa445 is fast and would probably require frequency compensation.

I have a feeling that I'm going to be looking for something different to build soon.  I wanted to build this so that I could get the ±15 volts to start learning more about op amps and I find it ironic that I don't have a supply that can to help solve this issue.  Can you recommend something to be on the lookout for on Ebay, dual tracking perhaps or at least cheap enough to get two and use in series?

The input voltage surge problem would be easy enough to fix with a 33 or 36 volt zener shunt regulator on the positive supply to the operational amplifiers.  This would be similar to the shunt regulator used to generate the negative supply.

Old Power Designs, HP, and Tektronix power supplies are pretty good.  Full documentation makes them relatively easy to repair and maintain.

[liquibyte]

What kinds of bad effects do keeping op amps towards their upper supply limit cause?

They become generally unreliable and more likely to fail.  Often they immediately fail if you exceed the maximum rating by just a couple volts.

I did notice that even though it's rated at a 44 or ±22 volt max, datasheets often suggest much lower.  I couldn't find anything in this one about that and had assumed with the assurances that things would be ok even towards the upper limit

I was told that everything was within tolerance and not to worry.  Something still kept bugging me about it but since many people had built these and they supposedly worked, I built mine.  I guess not many people have really tested the design until now.

The design assumes that the unregulated supply is not spiking above 44 (or 36) volts when power is applied.

That's what has me so worried about suggesting this supply to anyone to build and use.  Theory differs from reality most days that I'm aware of.

What if I were to use an op amp that had a bit higher limits, say ±48 volt such as this one or this one?  The characteristics are similar and it would put me a few more volts above my input voltage giving a tad more headroom if I can contain the initial surge from the transformer.  Pinout is identical.  The first one has closer specs I believe but the latter has a higher supply voltage range.  I'd really like to solve this but I'm not entirely sure it can be done without a significant redesign.

Neither of those are single supply amplifiers so they will require about a -5 volt supply to operate correctly in this circuit.  The opa445 is fast and would probably require frequency compensation.

Noted.  I didn't think about the single supply feature on the 2141.

I have a feeling that I'm going to be looking for something different to build soon.  I wanted to build this so that I could get the ±15 volts to start learning more about op amps and I find it ironic that I don't have a supply that can to help solve this issue.  Can you recommend something to be on the lookout for on Ebay, dual tracking perhaps or at least cheap enough to get two and use in series?

The input voltage surge problem would be easy enough to fix with a 33 or 36 volt zener shunt regulator on the positive supply to the operational amplifiers.  This would be similar to the shunt regulator used to generate the negative supply.

In your opinion, what sort of wattage should I be looking out for?  D13 is rated at 1W but I'm thinking I'd need 3W or better but I could be wrong.

Old Power Designs, HP, and Tektronix power supplies are pretty good.  Full documentation makes them relatively easy to repair and maintain.

I had looked for some Power Designs stuff but Ebay got eevblog'ed and everything went through the roof when folks here started buying them.  I'll set up a few watches and see if I can come up with a deal or two.

I do appreciate all your help with this.  I was starting to get a little lost and was pulling my hair out.  I think I'll try and put in a zener as you suggest and see if things can be stabilized.

[David Hess]

What kinds of bad effects do keeping op amps towards their upper supply limit cause?

They become generally unreliable and more likely to fail.  Often they immediately fail if you exceed the maximum rating by just a couple volts.

I did notice that even though it's rated at a 44 or ±22 volt max, datasheets often suggest much lower.  I couldn't find anything in this one about that and had assumed with the assurances that things would be ok even towards the upper limit.

Absolute maximum ratings on datasheets are just that.  They do not represent operating conditions conducive to reliability.

Neither of those are single supply amplifiers so they will require about a -5 volt supply to operate correctly in this circuit.  The opa445 is fast and would probably require frequency compensation.

Noted.  I didn't think about the single supply feature on the 2141.

Single supply operation is not necessary if a negative bias supply is used.  As a general rule, single supply operational amplifiers compromise other characteristics so it is better to use dual supply operational amplifiers if feasible.

Nothing in a general purpose power supply requires anything better than a 301 or 741 type of operational amplifier.  I like using OP-27/LT1007 style amplifiers in precision low noise power supplies.

The input voltage surge problem would be easy enough to fix with a 33 or 36 volt zener shunt regulator on the positive supply to the operational amplifiers.  This would be similar to the shunt regulator used to generate the negative supply.

In your opinion, what sort of wattage should I be looking out for?  D13 is rated at 1W but I'm thinking I'd need 3W or better but I could be wrong.[/quote]

I would use a 5 watt 36 volt 1N5365 and a 1 watt 180 ohm series resistor.

[liquibyte]

Single supply operation is not necessary if a negative bias supply is used.  As a general rule, single supply operational amplifiers compromise other characteristics so it is better to use dual supply operational amplifiers if feasible.

Nothing in a general purpose power supply requires anything better than a 301 or 741 type of operational amplifier.  I like using OP-27/LT1007 style amplifiers in precision low noise power supplies.

I think they went with the 2141 as a better fit with the redesign considering it's supposed to do 0-30V @ 0-3A.  One of the things that I've heard is that a 30V supply isn't really necessary but I went with this one to have some overhead.  After struggling with this and having learned a few things, I'm going to get some reading material and try to design my own and see how it works out.  I figure 15V should be enough and the two you mentioned above should be a good starting point.

I would use a 5 watt 36 volt 1N5365 and a 1 watt 180 ohm series resistor.

I was actually looking at the 1N5365 @ Digikey earlier but wasn't sure of the wattage rating so I didn't add it to a cart.  One thing I don't understand, what would the resistor be for?  D13 doesn't seem to need one and it doesn't even get warm.

[David Hess]

I think they went with the 2141 as a better fit with the redesign considering it's supposed to do 0-30V @ 0-3A.  One of the things that I've heard is that a 30V supply isn't really necessary but I went with this one to have some overhead.  After struggling with this and having learned a few things, I'm going to get some reading material and try to design my own and see how it works out.  I figure 15V should be enough and the two you mentioned above should be a good starting point.

I get by with my PS-503 which is effectively two 20 volt supplies in series.  It provides either 0 to 20 volts plus and minus which is just right for powering analog circuits or 0 to 40 volts.

I would use a 5 watt 36 volt 1N5365 and a 1 watt 180 ohm series resistor.

I was actually looking at the 1N5365 @ Digikey earlier but wasn't sure of the wattage rating so I didn't add it to a cart.  One thing I don't understand, what would the resistor be for?  D13 doesn't seem to need one and it doesn't even get warm.

I am not sure why they included D13 but it is only dropping the voltage by 10 volts and is not regulating the supply to U3.

The zener shunt regulator would be configured the way R3 and D7 are with the series resistor dropping the voltage.  44 volts - 36 volts = 8 volts across the resistor.  8 volts / 180 ohms = 44.5 milliamps.  8 volts * 44.5 milliamps = 356 milliwatts in the resistor and 36 volts * 44.5 milliamps = 1.6 watts in the zener diode.

A 3 watt 36 volt 1N5938 could be used as well.

The current in the zener diode is actually lower than 44.5 milliamps by any current that U1 and U2 are drawing.

[liquibyte]

I am not sure why they included D13 but it is only dropping the voltage by 10 volts and is not regulating the supply to U3.

The way it was explained to me is that since the positive rail is sitting so close to the limit of the op amp and that because they pull the negative rail 1.3V below ground that a voltage drop was required to keep from overloading the op amp.

The zener shunt regulator would be configured the way R3 and D7 are with the series resistor dropping the voltage.  44 volts - 36 volts = 8 volts across the resistor.  8 volts / 180 ohms = 44.5 milliamps.  8 volts * 44.5 milliamps = 356 milliwatts in the resistor and 36 volts * 44.5 milliamps = 1.6 watts in the zener diode.

A 3 watt 36 volt 1N5938 could be used as well.

The current in the zener diode is actually lower than 44.5 milliamps by any current that U1 and U2 are drawing.

Makes sense.  I've got a few other things I'm putting together for another order so I'll add those to the cart.  I'm ordering some OP-27's too, I'd like to try my hand at a dual supply from scratch.  I've got a question about that by the way.  When you build two separate supplies and hook them together in series to get positive/negative, do you have to rate the components in the design higher than if you just used it as a single supply?  If so, is there a general rule such as 2x or 2.5x.  For instance I build a supply that uses 50V caps, would I have to use 100V caps or even higher?  I know there are instances where diodes across the output in a certain configuration is required.

[David Hess]

The zener shunt regulator would be configured the way R3 and D7 are with the series resistor dropping the voltage.  44 volts - 36 volts = 8 volts across the resistor.  8 volts / 180 ohms = 44.5 milliamps.  8 volts * 44.5 milliamps = 356 milliwatts in the resistor and 36 volts * 44.5 milliamps = 1.6 watts in the zener diode.

A 3 watt 36 volt 1N5938 could be used as well.

The current in the zener diode is actually lower than 44.5 milliamps by any current that U1 and U2 are drawing.

Makes sense.  I've got a few other things I'm putting together for another order so I'll add those to the cart.  I'm ordering some OP-27's too, I'd like to try my hand at a dual supply from scratch.

There is nothing wrong with TL081s (The TL071 is or used to be a TL081 graded for low noise.), 741s, and 301As.  The lowest cost precision amplifier is probably the OP-07.  OP-27s are more expensive and their extra speed may require a more complex design so I would get something working with a cheaper amplifier first.  Their added performance is not going to make a difference in a general purpose power supply.

I've got a question about that by the way.  When you build two separate supplies and hook them together in series to get positive/negative, do you have to rate the components in the design higher than if you just used it as a single supply?  If so, is there a general rule such as 2x or 2.5x.  For instance I build a supply that uses 50V caps, would I have to use 100V caps or even higher?  I know there are instances where diodes across the output in a certain configuration is required.

This just depends on the implementation.  All of the dual output designs I am familiar with are either separate supplies or a pair of supplies that share a common ground.  In both cases each supply or side just sees its own voltage.

[liquibyte]

There is nothing wrong with TL081s (The TL071 is or used to be a TL081 graded for low noise.), 741s, and 301As.  The lowest cost precision amplifier is probably the OP-07.  OP-27s are more expensive and their extra speed may require a more complex design so I would get something working with a cheaper amplifier first.  Their added performance is not going to make a difference in a general purpose power supply.

I get lost looking at the damned things and the datasheets don't help.  I did compare the pricing though and for 10 OP-27's you can get 10 each of the others.  I did that instead.  I already have a few LM358's that I got from somewhere as well.

This just depends on the implementation.  All of the dual output designs I am familiar with are either separate supplies or a pair of supplies that share a common ground.  In both cases each supply or side just sees its own voltage.

What I was looking to do when I started this project was play around with op amp circuits and wanted a positive/negative supply to use.  When I built these, I used quality components and made sure I didn't skimp on anything just in case.  I was worried about the caps so I upped the recommended voltage from 50 to 63 on all of them for margin.

If I wire the resistor and zener between the rectifier and filter caps that would put it before the 10V zener as well.  There's 33.5 volts into pin 7 on U3 right now with 35.1 volts across pin 7 and pin 4 which would be reduced to around 26 volts and 27.6 volts respectively.  Should I cut the trace to that and run a wire to before the 36V zener or would the voltage drop matter all that much being that it's the current limiting part of the circuit?

I plan to eventually redo the boards because I've learned a few things since I did these, they were my first attempt at designing one and they are a pain in the ass to get in and out.

[liquibyte]

Parts will be arriving on Monday.  I'll post here with the results when I'm done testing.  I'm still wondering if a soft start circuit might not be a bad idea though.

I think I need to find a better alternative to the terminal blocks I used for power to board connections.  I used these because I don't have a crimper yet and I thought it would make things easier to get in and out.  Turns out I was wrong there.

[paul18fr]

my question is probably quite naive, but I'm wondering if numerical simulations can help of course I've no experience in electonics ? ... (in mechanical engineering I commonly use specific sovers)

I found some free ones :
- Qucs which has a GUI : http://qucs.sourceforge.net ...
- SPICE http://bwrcs.eecs.berkeley.edu/Classes/IcBook/SPICE/
- CircuitLab https://www.circuitlab.com
- and probably many others ....

Paul

[liquibyte]

my question is probably quite naive, but I'm wondering if numerical simulations can help of course I've no experience in electonics ? ... (in mechanical engineering I commonly use specific sovers)

I found some free ones :
- Qucs which has a GUI : http://qucs.sourceforge.net ...
- SPICE http://bwrcs.eecs.berkeley.edu/Classes/IcBook/SPICE/
- CircuitLab https://www.circuitlab.com
- and probably many others ....

Paul

Don't forget LTspice IV and TINA-TI.

I suck at spice simulation.  I'm not sure if it's even possible to model transformer inrush let alone what David Hess said about op amps.  I guess you could have a go at it.  I'm sure with your background you'd get much closer than this 8th grade graduate could.  Not that I can't do math, I just haven't gotten that far into things yet.

Simulating startup of an operational amplifier strikes me as pretty difficult to do.  Some operational amplifier models will definitely not handle it correctly if only because they do not always model common mode input range violations.

[David Hess]

If I wire the resistor and zener between the rectifier and filter caps that would put it before the 10V zener as well.  There's 33.5 volts into pin 7 on U3 right now with 35.1 volts across pin 7 and pin 4 which would be reduced to around 26 volts and 27.6 volts respectively.  Should I cut the trace to that and run a wire to before the 36V zener or would the voltage drop matter all that much being that it's the current limiting part of the circuit?

Cut the trace and use the resistor and zener as a shunt regulator for at least U1 and U2.  I would alter the design to use it for U3 as well and get rid of D13.  The bottom of the zener can go to the negative supply so the voltage across the operational amplifiers is restricted to 36 volts total assuming that the negative supply can handle the current.

There are better ways to do all of this but adding the 36 volt zener is the most simple.

Note that some operational amplifiers will run at 36 volts with a 44 volt absolute maximum voltage and some will run at 30 volts with a 36 volt absolute maximum voltage.  Do not use the ones with a 36 volt absolute maximum voltage at 36 volts!

I plan to eventually redo the boards because I've learned a few things since I did these, they were my first attempt at designing one and they are a pain in the ass to get in and out.

I have done that before.

[David Hess]

my question is probably quite naive, but I'm wondering if numerical simulations can help of course I've no experience in electonics ? ... (in mechanical engineering I commonly use specific sovers)

SPICE tends to have problems in situations like these because the models are not usually designed to operate outside of narrow bounds.

[liquibyte]

If I wire the resistor and zener between the rectifier and filter caps that would put it before the 10V zener as well.  There's 33.5 volts into pin 7 on U3 right now with 35.1 volts across pin 7 and pin 4 which would be reduced to around 26 volts and 27.6 volts respectively.  Should I cut the trace to that and run a wire to before the 36V zener or would the voltage drop matter all that much being that it's the current limiting part of the circuit?

Cut the trace and use the resistor and zener as a shunt regulator for at least U1 and U2.  I would alter the design to use it for U3 as well and get rid of D13.  The bottom of the zener can go to the negative supply so the voltage across the operational amplifiers is restricted to 36 volts total assuming that the negative supply can handle the current.

There are better ways to do all of this but adding the 36 volt zener is the most simple.

Just to clarify because I was thinking in terms of the 10V zener configuration as in circuit A in the picture I did.  You mean for me to do it like circuit B right?  If so, I was thinking B1 but then you said that B2 is a possiblility if I remove the 10V zener.

Note that some operational amplifiers will run at 36 volts with a 44 volt absolute maximum voltage and some will run at 30 volts with a 36 volt absolute maximum voltage.  Do not use the ones with a 36 volt absolute maximum voltage at 36 volts!

I always follow the datasheet which is why I questioned the 44 volt level on the other board.  I thought that it was too close to the limits of the op amps and given any untoward occurance, failure would be inevitable.  I ordered several of each of the ones you had mentioned and the first thing I did was save the datasheet to my op amp datasheet folder.  I also went and got some AN's and other reading material on the subject.

I plan to eventually redo the boards because I've learned a few things since I did these, they were my first attempt at designing one and they are a pain in the ass to get in and out.

I have done that before.

I have a better idea of the constraints of the case and plan on using a different scheme for the terminal blocks.  I haven't decided the best way to do it yet but I'll come up with something.

[David Hess]

If I wire the resistor and zener between the rectifier and filter caps that would put it before the 10V zener as well.  There's 33.5 volts into pin 7 on U3 right now with 35.1 volts across pin 7 and pin 4 which would be reduced to around 26 volts and 27.6 volts respectively.  Should I cut the trace to that and run a wire to before the 36V zener or would the voltage drop matter all that much being that it's the current limiting part of the circuit?

Cut the trace and use the resistor and zener as a shunt regulator for at least U1 and U2.  I would alter the design to use it for U3 as well and get rid of D13.  The bottom of the zener can go to the negative supply so the voltage across the operational amplifiers is restricted to 36 volts total assuming that the negative supply can handle the current.

There are better ways to do all of this but adding the 36 volt zener is the most simple.

Just to clarify because I was thinking in terms of the 10V zener configuration as in circuit A in the picture I did.  You mean for me to do it like circuit B right?  If so, I was thinking B1 but then you said that B2 is a possiblility if I remove the 10V zener.

B2 may be better in general but if you use the zener regulated 36 volt supply to power U3 (and maybe the Q3 circuit) as well, then the 10 volt zener circuit will not be needed.  They are separate issues.

I would use the B2 connection and increase the negative supply to -2.5 to -5.0 volts but if you use 44 volt operational amplifiers instead of 36 volt ones, then you can do this and use the B1 connection instead so the maximum output voltage will be a little higher.

What would be clever is to change the Q3 circuit by using the 10 volt zener to level shift the output of U3 so it can drive Q3 even with Q3 supplied by the higher voltage.  Then the current limit LED will light during current limiting or high input voltages.  Or that could just be confusing.  /shrug

[liquibyte]

B2 may be better in general but if you use the zener regulated 36 volt supply to power U3 (and maybe the Q3 circuit) as well, then the 10 volt zener circuit will not be needed.  They are separate issues.

I would use the B2 connection and increase the negative supply to -2.5 to -5.0 volts but if you use 44 volt operational amplifiers instead of 36 volt ones, then you can do this and use the B1 connection instead so the maximum output voltage will be a little higher.

I'd have to do too much with the boards as they are to change the negative supply so I'm going to go with desoldering the cathode of D13 and running a wire from it to before the 180R resistor.  The rest of it is just a matter of soldering the zener across the bottom of R1 and soldering the 180R to the positive side of the filter cap where it comes into the terminal block.  C1 and the bridge rectifier are off board, R1 is on the board.

What would be clever is to change the Q3 circuit by using the 10 volt zener to level shift the output of U3 so it can drive Q3 even with Q3 supplied by the higher voltage.  Then the current limit LED will light during current limiting or high input voltages.  Or that could just be confusing.  /shrug

I looked up what you were talking about and didn't even know that was a thing or at least what is going on is what it's called.  I'm not seeing how to do it as it applies to U3 and Q3, so your last statement applies for the moment.

I think what I need to do is visit more service manuals on BAMA and the like to see how certain problems have been solved amongst various power supplies over time.  Given that these are proven designs, or at least were at the time, I think reverse engineering would be a more logical approach to understanding than magic smoke experiments.  I've been rather lucky in that regard.  I've only burnt a few transistors and resistors so far, and once an exploding cap.  To be honest, I grabbed the wrong voltage but it did scare the crap out of me when I powered on.  I used to work for a place that built motor braking solutions and it was nothing compared to when the labs there blew stuff up, it was always fun when an IGBT let go.

[liquibyte]

My goodies came in this morning and I've managed to get everything in there like I posted in the pic in my above post.

The good news is that 61 volt spike is gone.  The bad news is that I'm still getting a fairly decent spike around 36 volts with the PS recalibrated at 30V 3A which is tp be expected given a 36V zener I suppose.  At this point I don't know if this is inherent to linear supplies and inrush or if the circuit just needs some rework to catch this and shunt this based on P1's setting somehow.  I'm too new to this to know off the top of my head what needs done but I have a feeling that a solution might be in the PS-503 service manual schematic.  I might have to build one of these just to probe around in it.  Some of the parts are unknown such as some the transistors and all of the op amps.  The manual doesn't say much about these other than their part designation so I looked up the part number and get back C741C so I'm assuming some sort of oldish 741 op amp.

Either way, it's working much better than it was and even though it's still spiking, at least it's not as bad as it was.  One of these days I'll manage enough money to get a real scope and perhaps help tame this thing if I haven't already.  I have another meter in mind and I think that I probably need to get an analog or two just for when things like this are going on.  I do appreciate your help with all of this.  I'll keep updating for posterity if I figure anything out.

[liquibyte]

The really bad news is that this is a 3A power supply so the modifications are not going to work.  If I load this down at 3 amps, the resistor is going to smoke.

[David Hess]

The good news is that 61 volt spike is gone.  The bad news is that I'm still getting a fairly decent spike around 36 volts with the PS recalibrated at 30V 3A which is tp be expected given a 36V zener I suppose.

The output should never rise above the set point when power is applied or removed.  I am inclined to believe that there is more than one problem but the lack of any regulation or protection on the supply to the operational amplifiers I consider a design flaw.

At this point I don't know if this is inherent to linear supplies and inrush or if the circuit just needs some rework to catch this and shunt this based on P1's setting somehow.  I'm too new to this to know off the top of my head what needs done but I have a feeling that a solution might be in the PS-503 service manual schematic.  I might have to build one of these just to probe around in it.  Some of the parts are unknown such as some the transistors and all of the op amps.  The manual doesn't say much about these other than their part designation so I looked up the part number and get back C741C so I'm assuming some sort of oldish 741 op amp.

The PS-503 design is pretty dense and unconventional in some respects even if you ignore the complexity added by being a dual output design with tracking.  There are changes I would make like having the output amplifiers operate with a fixed gain and varying the output from the reference because it would make the frequency response more predictable.

The PS-503 operational amplifiers are just 36 volt 741 types and not critical.  Besides qualifying their sources, Tektronix graded incoming 741s and the ones with high bias current, high offset current, high offset voltage, and high offset voltage drift are the -00 ones used in non-critical designs.  They actually graded them for 40 volt operation as well instead of buying the 741/A/M version.

One of my PS-503s has a pair of 301A operational amplifiers for U55 and U155 but they and one of the 741s are in sockets so I think someone who knew what they were doing repaired it in the past.  They added the compensation capacitors needed by the 301As to the back of the printed circuit board.

Either way, it's working much better than it was and even though it's still spiking, at least it's not as bad as it was.  One of these days I'll manage enough money to get a real scope and perhaps help tame this thing if I haven't already.  I have another meter in mind and I think that I probably need to get an analog or two just for when things like this are going on.  I do appreciate your help with all of this.  I'll keep updating for posterity if I figure anything out.

Some multimeters have a pretty fast peak detect mode.

Any storage oscilloscope is handy in cases like these.  DSOs are particularly useful for diagnosing startup problems and even an old and low performance unit can handle something like this.  I very rarely need to use anything better than a Tektronix 2230 which is the oldest DSO that I can honestly recommend.

I would never recommend a USB DSO unless it had a unique and useful feature like the ones from CircuitGear which support low frequency vector network analysis.  That is particularly useful for designing high performance power supplies.