Linear power supply project

[Therm Mr.]

The parts list shows the Front panel pots as a 50k 10-turn ww for volt adjust. That one I found a Bourns for at Mouser. The other for the current side it shows a 10k 10-turn CERMET. That one I cannot find.  Is there something special about the CERMET that makes it better for current? Any idea of where I may find one?  My net search only finds 1-turn panel mount CERMETs and the 10-turn are all ww.

Thanks

Therm

[pqass]

So now I am ready to run the footprint assignment tool. I still need to add the fan power and control circuit and any mods that may be needed for the display. Should I wait till this is done before starting to assign footprints? I am going to read back through the display discussion likely several times so I can understand it better before my next step forward.

I have been creating some parts list carts at DigiKey and Mouser. I may even need to add in Newark and Texas Instuments to get it all. I am ordering multiples of most everything figuring It won't hurt to have some spares for future experiments and projects.

@Kleinstein you mentioned some TL062 chips. I would like to order some of those as well as the LM358's I already have in my cart. I would like to be able to interchange like pqass suggested by using sockets. My question is there are many different suffixes on them any particular suffix I should choose? Also TL431LP's also have many suffixes I am guessing the TO-92 package does the suffix matter?

WRT S2A and B (on HPs schematic):  The switch to change transformer taps is ganged to to the switch in the CREF generation section. Why? Because if you double the output voltage possible you are only allowed that at half the current; as you're limited by the VA of the transformer.  The "S1A" label on a contact of the S2B switch is just nuts.      The (momentary) switch called CC SET is just to choose your set vs. actual current shown on the display.  I don't know why they called it S1A when there is no B in sight.

You can assign footprints anytime (once you annotate everything the first time).  Just save it in the Assign Footprints popup.  If you don't delete the symbol, your choice will remain there.  You can then add more symbols, then annotate the new ones in the Assign Footprints popup again (previous assignments should remain).  Just use generic footprints DIP (double and single op amp layouts) and you'll have lots of op amps to choose from.

The parts list shows the Front panel pots as a 50k 10-turn ww for volt adjust. That one I found a Bourns for at Mouser. The other for the current side it shows a 10k 10-turn CERMET. That one I cannot find.  Is there something special about the CERMET that makes it better for current? Any idea of where I may find one?  My net search only finds 1-turn panel mount CERMETs and the 10-turn are all ww.

I wouldn't worry too much about exotic pots or 1% odd-ball value resistors just yet. I think you should prototype the bare minimum circuit on a breadboard or perfboard, with a single darlington, jelly-bean single-turn pots, and a collection of op amps to try would be enough to get basic voltage regulation working.  You could then add the Current Error Amp,  CV/CC Indicators, and then the meters last.  At each checkpoint, you'll be testing and gaining knowledge.  Then finally update the parts piecemeal to their final form (eg. better transformer, parallel darlingtons, 10T pots, posts, etc.).  Your implementation won't be exactly like the version in the service manual.

Once the prototype validates your specific design, you can update the schematic and produce a PCB layout that you know will work. PCB layout without prototyping guarantees that there will be multiple errors that bodge wires can't fix.  It's a recipe for frustration.

[enut11]

I am not aware of a 10T panel mount cermet pot. Both of the pots on my E3612A feel the same in use and appear to be WW. The advantage of cermet is usually over carbon, ie lower TC.
enut11

"Cermet – Ceramic based materials exhibit a high degree of stability over a wide temperature range.  Resistive elements that incorporate Cermet inks will typically have very low temperature coefficients as compared to other resistive materials.  This characteristic is important for any application that has to operate in a wide temperature range.  Cermet potentiometers are very stable and typically have a TC that is expressed PPM. The TC for a typical Cermet potentiometer is 150 ppm/°C (0.015%) which is significantly below Conductive Plastic or Carbon"

[pqass]

See attached for a photo of the 10T pots from my E3611A.  They are Bourns 3590S-A17-503 (-103).
I can't find the A17 variant anywhere.  It might be the code for metal bushing and metal D shaft.

[Therm Mr.]

WRT S2A and B (on HPs schematic):  The switch to change transformer taps is ganged to to the switch in the CREF generation section. Why? Because if you double the output voltage possible you are only allowed that at half the current; as you're limited by the VA of the transformer.  The "S1A" label on a contact of the S2B switch is just nuts.      The (momentary) switch called CC SET is just to choose your set vs. actual current shown on the display.  I don't know why they called it S1A when there is no B in sight.

You can assign footprints anytime (once you annotate everything the first time).  Just save it in the Assign Footprints popup.  If you don't delete the symbol, your choice will remain there.  You can then add more symbols, then annotate the new ones in the Assign Footprints popup again (previous assignments should remain).  Just use generic footprints DIP (double and single op amp layouts) and you'll have lots of op amps to choose from.

Thanks.
That S2B was messing with me.


Its good to know I can start assigning footprints while waiting on parts.
 

I wouldn't worry too much about exotic pots or 1% odd-ball value resistors just yet. I think you should prototype the bare minimum circuit on a breadboard or perfboard, with a single darlington, jelly-bean single-turn pots, and a collection of op amps to try would be enough to get basic voltage regulation working.  You could then add the Current Error Amp,  CV/CC Indicators, and then the meters last.  At each checkpoint, you'll be testing and gaining knowledge.  Then finally update the parts piecemeal to their final form (eg. better transformer, parallel darlingtons, 10T pots, posts, etc.).  Your implementation won't be exactly like the version in the service manual.

Once the prototype validates your specific design, you can update the schematic and produce a PCB layout that you know will work. PCB layout without prototyping guarantees that there will be multiple errors that bodge wires can't fix.  It's a recipe for frustration.

I finally concluded after thinking about it I would order parts and start testing (excited to start) before I committed to a PCB design. It makes perfect sense to test first. I was putting the horse before the cart in my excitement.


Thanks for all the advice and help. I am grateful and thankful for you and all the others here who have helped me with this. I am having fun and learning a bunch.

See attached for a photo of the 10T pots from my E3611A.  They are Bourns 3590S-A17-503 (-103).
I can't find the A17 variant anywhere.  It might be the code for metal bushing and metal D shaft.

Thanks that's great to have that number. I just checked my cart at Mouser and just yesterday I put the 503 50k and 103 10 k in my cart (you must have been putting out vibes ). I was doubting them but not since you posted that pic of them. I searched Bourns for the A-17 and it said it was a non standard item. Maybe HP special ordered it. I will also add a couple of inexpensive 1 turns for testing.

Therm

[Therm Mr.]

I am not aware of a 10T panel mount cermet pot. Both of the pots on my E3612A feel the same in use and appear to be WW. The advantage of cermet is usually over carbon, ie lower TC.
enut11

"Cermet – Ceramic based materials exhibit a high degree of stability over a wide temperature range.  Resistive elements that incorporate Cermet inks will typically have very low temperature coefficients as compared to other resistive materials.  This characteristic is important for any application that has to operate in a wide temperature range.  Cermet potentiometers are very stable and typically have a TC that is expressed PPM. The TC for a typical Cermet potentiometer is 150 ppm/°C (0.015%) which is significantly below Conductive Plastic or Carbon"

Interesting, Thanks for the info. I am learning the same value components can be very different depending on materials and construction. So much to know.

Therm

[Therm Mr.]

Would this meter work? You can even program a custom setting if you can't use one of the presets. They Cost $72.00 but look pretty nice. They are 12-24 ac/dc supply power. But should not be hard to provide. Or are these something different entirely?

[rolycat]

For metering I would go for something more in keeping with the LED display on the original E3610A.
You can buy 4 digit panel meters showing both voltage and current for just a few dollars on eBay and similar sites, and they are usually fairly accurate.

For example:
https://www.ebay.com/itm/335327568961

[Therm Mr.]

For metering I would go for something more in keeping with the LED display on the original E3610A.
You can buy 4 digit panel meters showing both voltage and current for just a few dollars on eBay and similar sites, and they are usually fairly accurate.
(Attachment Link)
For example:
https://www.ebay.com/itm/335327568961

Thanks, I saw those but could find no wiring diagram or specifications on them and with no info its kind of hard to know what I am buying. I like the style though.

Therm

[rolycat]

Most such meters work with a supply voltage anywhere between 4 and 28 volts. Power consumption is typically around 20mA and accuracy is within a percent or better - there are usually two trimmer potentiometers on the back for fine adjustment.
They are cheap and versatile enough that I generally keep one or two in my parts stock.
There are a number of different styles, but they generally use the same wiring scheme. There are two plugs on the back:

• Three thin wires, usually black, red and yellow. Black and red are the power for the meter, and yellow is the voltage sense wire.
• Two thick wires, usually black and red. These are the current sense wires and go to a shunt resistor on the meter PCB.

There's a description and diagrams here of how to wire a similar meter:
https://protosupplies.com/product/dual-display-0-100v-0-10a-panel-meter/

[enut11]

Would this meter work? You can even program a custom setting if you can't use one of the presets. They Cost $72.00 but look pretty nice. They are 12-24 ac/dc supply power. But should not be hard to provide. Or are these something different entirely?

Not my style but more importantly I find it very handy to be able to resolve mV and mA with a power supply. Great for testing zener diode breakdown and optimum LED operating current and more. This means 5 full digits (99999) for the voltmeter and 4 digits (9999) for the ammeter. These are a little more expensive but tend to be more accurate.
This voltmeter is about $14AUD delivered. eBay item number:335203244430

enut11

[xavier60]

Most such meters work with a supply voltage anywhere between 4 and 28 volts. Power consumption is typically around 20mA and accuracy is within a percent or better - there are usually two trimmer potentiometers on the back for fine adjustment.
They are cheap and versatile enough that I generally keep one or two in my parts stock.
There are a number of different styles, but they generally use the same wiring scheme. There are two plugs on the back:

• Three thin wires, usually black, red and yellow. Black and red are the power for the meter, and yellow is the voltage sense wire.
• Two thick wires, usually black and red. These are the current sense wires and go to a shunt resistor on the meter PCB.

There's a description and diagrams here of how to wire a similar meter:
https://protosupplies.com/product/dual-display-0-100v-0-10a-panel-meter/

From my experience, the 4 digit types have very accurate current measurement because they have chopper auto zeroing.
The big annoyance is the refresh being slower than the usually advertised 3 times per second. It can actually be once per 2 seconds.

[Kleinstein]

The very cheap 3 and 4 digit modules may use a µC internal ADC and oversampling to get just enough resolution. This can result in a rather noisy or slow update. I have seen less of the 5 digit modules lately - a few years ago there was a teardown shwoing a microchip 18 bit SD ADC inside, suggesting 3-4 samples per second if no extra oversampling is used.

The instructions are often a bit vague on how to connect. At least they usually show if separate wires for the supply are used. Given the low price it may be worth getting a unit upfront and experiment with the interface.

For the current reading the shunt on the module can complicate things a bit. It may be easier to use votlage reading modules and amplify the shunt voltage if needed.

[Therm Mr.]

Those meters are pretty cheap I will order some just to experiment with. At minimum I will learn something. I can wait a couple of weeks for them.

The Trumeter has my interest because of the True RMS or peak feature. I was wondering if maybe they would solve a lot of the configuration issues caused by the cheaper meters?

I should be placing an order with both Digi and Mouser this week. They had most of what I need. I will still need to go back and read this thread again before sending the order. To make sure I have all the alternatives covered. I have over ordered to be sure I have some extras in case some of the smoke escapes on the cheaper stuff. Plus to have some different op-amps to try.

In the meantime I have some components to start to experiment with. Transistors, regulators, op-amps misc stuff I have laying around. I need to get back in the groove after 40-50 years on the bench. I have been using a 12 volt 2 amp supply from my old DSL modem i converted. I run it through a ltm8067 DC to DC converter. It works for now.

Thanks

Therm

[Kleinstein]

The output of the supply is DC. So no need for true RMS and hardly any need peak detection.

[Therm Mr.]

The output of the supply is DC. So no need for true RMS and hardly any need peak detection.

I did not know True RMS was only for AC power. I had to look it up makes perfect sense.

Is there any advantage to using toroidal T-former I saw a build where one was used? Pros, cons?

[xavier60]

I had to switch to a toroidal transformer because the leakage flux from the EI type I originally used caused the display on my DSO to wobble.
It's classed as medical grade, I think due to it having electrostatic shielding.
https://www.eevblog.com/forum/projects/linear-lab-power-supply/msg2388873/#msg2388873

[Kleinstein]

Toroidial transformers have pros and cons: on the pluse side they are more compact / lower weight, lower magnetic leakage and usually better efficiency. As down sides they may not have as good an isolation (not 2 really separate chambers), a bit more tricky to mount and have a larger turn on current spike (may need extra circuit beyond some 150 VA). The lower output resistance has pros and cons (less voltage sag under load, but larger current spikes for the rectifier).

[Therm Mr.]

I had to switch to a toroidal transformer because the leakage flux from the EI type I originally used caused the display on my DSO to wobble.
It's classed as medical grade, I think due to it having electrostatic shielding.
https://www.eevblog.com/forum/projects/linear-lab-power-supply/msg2388873/#msg2388873

Nice work on that.

[Therm Mr.]

@andrewtaylor recommended I use a dual 12v 4a secondary T-former for the primary power. I found the one attached.

Is this the correct dual secondary transformer config?

I assume I will be switching between these windings for High and Low range is that correct?

The one thing that puzzles me is the 117/234 VAC primary? Does it matter that its not 115/230?

I have not yet found a dual 15v 200 ma for the Reference/Bias supply yet. Any advice on finding one?

[Kleinstein]

For the regulator supply one should get away with less than +-12 V DC as in the HP schematics. Some +-8 V (so that a 5 V ref is OK) should be enough.
So one could get away with less than 15 V AC for the transformer, maybe only 2 x 12 V.

For the main transformer rating one has to keep in mind that 1 A DC from rectifier and filter cap needs about 1.6 A of RMS AC current from the transformer. So the 2 A rated transformer would be good for some 1.2 A of DC output with the higher voltage.

The exact voltage for the primary is not that relevant. The 117V winding would be perfectly fine with 110 V or 100 V, just the secondary scales down accordingly too. So not 24 V but more like 22.5 V at nominal load. A transformer is fine with lower voltage - just with a lower power rating as the maximum current still stay the same.

[Therm Mr.]

I am seeing T2 as a center tap transformer. So, I think I would need a 14v or 16v secondary winding with each side of center being 7 or 8 v? Or is my thinking flawed?

And with T1 I am seeing a 2x?v winding wired in parallel with 3 common, 4 high range winding and 5 low range winding. Am thinking correct or am I wrong and misunderstanding something?

I have placed my orders and I should have all the components I need early next week to begin testing on the breadboard. Though the display will be another week. I guess its on a slow boat from China.


Thanks

Therm

[Kleinstein]

To get a +-12 V auxiliary supply for the regulator one would need a transformer with something like 2x14 V - 16 V.
Just 16 V with a center tap or 2x8 V would not be enough for the +-12 V like in the HP plan. It could still work with some +-6 V for the regulator part. This should still be enough though it could be tight with a 5 V reference. There is not real need to have +-12 V for the OP-amps. They are perfectly OK with less.
A lower voltage could be a thing if the voltmeters are power from the same supply.


As shown T1 also has a center taped winding and is using only part of the winding for the high current - so no parallel connection. This way the permissible current at the lower voltage setting is usually not all the way 2 x , but a little less (e.g. 1.6 x). There is not easy, unviersal ratio as there is loss from the primay and secondary and the local vs global loss question.
Ideally there are  2 completely separate windings in parallel for higher current or in series for the higher voltage. However this would need a less common 3 (or 4) pole switch to also have poles to switch the max voltage/current.

[Therm Mr.]

To get a +-12 V auxiliary supply for the regulator one would need a transformer with something like 2x14 V - 16 V.
Just 16 V with a center tap or 2x8 V would not be enough for the +-12 V like in the HP plan. It could still work with some +-6 V for the regulator part. This should still be enough though it could be tight with a 5 V reference. There is not real need to have +-12 V for the OP-amps. They are perfectly OK with less.
A lower voltage could be a thing if the voltmeters are power from the same supply.

So would a 2x14v be producing 28v since the center tap is grounded and the bridge is being fed by the outsides?

As shown T1 also has a center taped winding and is using only part of the winding for the high current - so no parallel connection. This way the permissible current at the lower voltage setting is usually not all the way 2 x , but a little less (e.g. 1.6 x). There is not easy, unviersal ratio as there is loss from the primay and secondary and the local vs global loss question.
Ideally there are  2 completely separate windings in parallel for higher current or in series for the higher voltage. However this would need a less common 3 (or 4) pole switch to also have poles to switch the max voltage/current.

Regarding T1, so when the switch is set to high it has one 12v higher amp output? Then when switched to Low it now has a 24v lower amp output? Is this correct?

Sorry if these are dumb questions.

Thanks

Therm

[Kleinstein]

A 2x14 V transformer with the center tap at ground would give about +-18 V at the filter capacitors. With a small transformer and low load the voltage can be some 20 % higher. So about OK (with plenty of reserve for a lower mains) for the +-12 V regulators.

For T1 one one would have to choice of something like 24 V / 2 A or 12 V up to about 3.2 A  DC (assuming a transformer rated for some 3.3 A AC).