Show your favorite and most used benchtop PSU

[David Hess]

@ David Hess
Thank you for showing your favorite set up.

Could you elaborate the interesting things(at least to me) in the pictures?

1) Why two wire connection instead of one from the input cap? Is it factory default or you modified it?

The original aluminum electrolytic capacitors, one of which is shown in that photo for comparison purposes, were wired the same way.  One side of each wire goes to the rectifier and the other side of each wire goes to the input to the regulator so the resistance in series with the capacitor itself is minimized.

When I replaced the big can style capacitors, I selected replacements based first on diameter so they would snap into the gray mounting brackets which held the original capacitors.

Also seen in that photo is an internal LED I added to show that the power supply is active when the cover is off.

2) What those two resistors do between Lo and Hi terminal of DM502? Why two Rs instead of one?

1.10 Mohms is just a more convenient test resistance for the high resistance range and I did not have any 1% resistors higher than 1.00 Mohm handy.  1.00 Mohms would leave the most significant digit potentially changing between 0 and 1.  That DM502 was measuring temperature using a probe at the time for no particular reason.

3) What is the purpose of the caps on the second DM502?

I was replacing solid tantalum capacitors which had shorted in a Tektronix PG506 pulse generator.  The NOS (new old stock) solid tantalum capacitors shown in the photograph were being burned in and tested for leakage before installation.

The leakage shown is 3.1 microamps for 10 capacitors in parallel but 2.48 microamps of that is caused by the 10 Mohm shunt resistance of the voltmeter so the actual average capacitor leakage was 0.062 microamps which is considerably better than the maximum rated leakage of about 4 microamps which might be specified for a typical 25 volt 15 microfarad solid tantalum capacitor.

[nadona]

Thank you very much again for the detailed explanation.

Could you draw how all these are connected for leakage test?
Why leakage test for tantalum? Are you doing it because they are NOS or you do all the time for tantalum for a certain applications?

The picture of tantalum caps drew my attention because recently two of my Datron 1062 were stopped working. Same reason. A tantalum cap  in the power supply section shorted while bringing down the analog board's input resistance to less than 1ohm. First time I smell the tantalum. I have two of 1082 and thinking to replace the same tantalum to other type of cap.

Maybe I start a new thread on how to test leakage and why?

Regards,

[David Hess]

Could you draw how all these are connected for leakage test?

There is not all that much to it.  The power supply is configured as a current limited 0 to 40 volt source with one of the multimeters measuring low values of current in series.  The other multimeter is measuring the voltage across the capacitors which were all tied in parallel.  If the leakage current had been high, I would have disconnected them in sets to find the bad one or ones.

I just hacked the test together in a few minutes and let it run overnight.

Why leakage test for tantalum? Are you doing it because they are NOS or you do all the time for tantalum for a certain applications?

These were NOS solid tantalum capacitors of unknown heritage; there is a local surplus electronics store here which has bins of them and they are inexpensive compared to new ones.

There *are* sometimes reasons to grade capacitors for low leakage but in this case it was just to weed out bad or marginal ones before installing them into the circuit.  Why replace a shorted tantalum capacitor with another bad or marginal capacitor?

I had already tested them for capacitance and dissipation factor.  It would be nice if someone made a portable capacitor tester which did capacitance, dissipation, ESR, and leakage but I have never seen such a thing.  When I was picking the capacitors out of the bins, I had my multimeter with me to check each one for capacitance and resistance which is not as good as a real leakage test at the rated voltage but better than nothing.  I did not find any bad ones.

The picture of tantalum caps drew my attention because recently two of my Datron 1062 were stopped working. Same reason. A tantalum cap  in the power supply section shorted while bringing down the analog board's input resistance to less than 1ohm. First time I smell the tantalum. I have two of 1082 and thinking to replace the same tantalum to other type of cap.

When solid tantalum capacitors first became available, they were optimistically advertised as *not* requiring voltage derating for reliable operation.  In this specific case, 20 volt capacitors were used on the plus and minus 16.5 volt supplies although I have run across 16 volt solid tantalum capacitors used on 15 volt supplies and 6.3 volt solid tantalum capacitors used on 5 volt supplies before.  One was shorted and after replacing it, the *other* one shorted a few hours later while I was watching so I stopped by the surplus electronics store and picked up some suitable replacements for all of the similar tantalum capacitors in the circuit; this was cheaper than ordering new ones online.

The replacements were rated at 25 volts which I figured was derating enough.

In general you can replace solid tantalum capacitors with good aluminum electrolytic capacitors of 2 to 4 times the value to get an equivalent ESR but if you have an inexpensive source of solid tantalum capacitors, I would use the same type as a replacement because they have better high frequency characteristics.  No matter what kind of capacitor is used, reliability will be improved if they are voltage derated.

[nadona]

@David Hess;
As I expected, more than perfect answer

Regards,

[BravoV]

Shameless bump .... more please ... 

[PTR_1275]

Hi guys,

Fairly new here, but I have been salivating my way through the pages of power supplies. Good to see some TTi power supplies and I'm surprised that I haven't seen many "Big" power supplies.

I do a little bit of electronics work, small level PIC work mostly, then the main work I do is testing battery chargers and solar regulators for battery charging, so I need a fair bit of current for testing them. The lab isn't where I would like it to be but when compared to a year ago I have added a lot of extra stuff. Currently I have 5500w of DC load bank capability in the rack with a HP 6050 mainframe on the way.

Power supplies that I use are

TTi PL330. 0-32v 0-3a linear supply
Manson 0-32v 0-21a SMPS
The big one (Recently bought) HP 6012A. Really cool auto ranging 0-60v 0-50a (Maximum of 1000w).

[dom0]

70 V (+-0-35 V), 500 mA supply. Transformer is grossly over-dimensioned ; probably has something like 200 VA. Based on HP schematics. Heat sink is dimensioned for ambient temperature of 50 °C.

Adjustable tracking ratio with calibrated 1:1 (actually... 1:-1) position ; adjustable current limit for both channels at once. Two digit meters, very fast -> transients are easily observed. Separate output on/off switch.

Total weight about 4.7 kg. Case is a recycled steel drawer with aluminium top and back panels. Transformer mounted flipped on steel beams.


Very handy, used it all the time even before I completed the meter lately.

[Whales]

I've just obtained a mains->~12Vx20A  toroid.  Once I'm finished my current project I'm going to rewind this and make myself a ~24V variable power supply with current limiting. 

[4cx10000]

First picture, yeah I know real bad, on top right of the instrument shelf is a Thurlby. Had it for years but now I'm planing to order a TTi CPX400D, any comment on that PSU?

[steve30]

Here's a picture of my PSU. It is shown powering a couple of white LEDs. I retrofitted my own variable linear PSU board into the old fixed output PSU.

It gives a fixed +5V (up to a couple of amps) and -5V (a few milliamps), and a variable 2.2V to 24V @ just over an amp.

The red panel meter measures current in mA, and the green panel meter reads the variable voltage. It really could do with some better panel meters as these cheap ebay ones aren't too accurate. The blue 10 turn pot adjusts the voltage and the silver single turn pot adjusts the current limit. Obviously things need mounting in the case properly.

The circuit design isn't excellent but it works reasonably well for my use. Hopefully I'll make a page on my website about it in due course.

[alsetalokin4017]

HP 721A, modded with 10-turn voltage pot and turncounting knob:

 

[Ysjoelfir]

My absolute favorite PSU is the first one, the colourful one (no, it's not sparkling!). My father build it in the 1980s after a plan made by ELV - one of the few products designed by ELV that doesn't have ANY special ELV branded Item in it. Sadly, it broke down shortly after my father died because of my inhability to think clearly at that time. I will repair it in the future, but maybe I need some help from you there.
However, why do/did I love this thing?
- 19" Rackmount formfactor
- Two independant voltage sources, one can go slightly into the negative, the other one can be current limited
- nice feel, rigid build quality
- ground lift switch
- fixed output voltages
- 4 wire output at one channel
- remote controllable
- I love analog instruments!
- Dad build it

No cons at this one, for what it was designed, it seems perfect to me.

We used that thing in the basement lab, so when I was working in my teenagers room I needed something, too. I first build the big black 19" rackmount thing, which is basicaly a high voltage tube regulated power supply on the one channel, combined with a variac (is that the right word for a variable auto transformer?) and a 40V 25 amps power transformer at the second channel. This thing was very usefull when I started working with tubes, but it has some negative aspects...

pros:
- generates nice high voltage for most common tube experiments (up to 360V at 100mA short time)
- Output voltage is very stable because of a big output cap and good regulation of the EL34 tube
- 40V 25A AC output is more then enough for most experiments.
- I still love analog instruments...!

cons:
- DC voltage and current too low to use in advanced tube experiments
- DC path is dangerous and has a slow regulation due to big output capacity
- no low voltage high current DC output
- high voltage transformer is too small to handle 100 mA continuously. It should have been able to deliver up to 450V, but with any significant load attached it drops down to around 360V.
- nobody wanted to carry that beats up to the third floor when I moved into this apartment, so I had to carry it by myself

One of the biggest cons at this one, besides the immense weight, was that it didn't deliver any DC voltage with anything above 100 mA power (actually, it didn't deliver any DC voltage below 50V at all...). So I had to get another one. After some projects of building a variable switch mode power supply (which failed spectacular) I was frustrated and bought the third one, a noname Power Supply which is capable of delivering 2x 0-30V / 3A. This is what I use now on a regular base, mainly because I didn't repair Dads old PSU untill now. That thing is quite okay, it's one of those range switching PSUs that use relais to switch the secondary transformer windings. I found that under different brands all over the world, for example I saw one labled Mastech, one from Conrad Toolcraft and so on...

pros:
- Quite compact. No heavyweight!
- two independant channels with enough voltage/current capability to work with
- Channels can be combined in series or parallel.
- GND is floating but can be connected to earth via a bridge
- stays cold due to the switching-thingy

cons:
- switching-thingy produces some noise which can disturb sensible electronics.
- digital displays... meh. I don't like them. It's nice for fine adjusting the voltage, but for amps I definitely prefer the old fashioned analog meter.
- cheap build quality

[BFX]

Here is my favorite power supplies in order of usage 

1. 2x HP6632B 0-20V/ 0-5A
2. Tesla SZ 3.81 0-30V/0-1A
3. R&S NGMO1 0-15V / 0-5A
4. HP6200B 0-40V/0.75A or 0-20v/1.5A

[dom0]

What is that RF thing with a dozen BNCs on the front panel? Some kind of combined analyzer-generator (like those cell tower survey things)?

[BFX]

What is that RF thing with a dozen BNCs on the front panel? Some kind of combined analyzer-generator (like those cell tower survey things)?

It's Anritsu 3681A signal generator.

[bingo600]

My latest from *bay, bought for opamp fun.
It really looks like that , mint condition.



/Bingo

[MatthewEveritt]

My main power supply, a solartron as1164.2

[Vgkid]

That Solartron is actually a neat looking power supply.

[MatthewEveritt]

Thanks, I like it. The three dials for voltage are really nice for quick setting / tweaking, and have the bonus of making it look far more complex and impressive than it really is.

[Macbeth]

Hmm... I'm thinking this supply is more of the female variety...



  

[Joule Thief]

Hmm... I'm thinking this supply is more of the female variety...



  

with knobs that nice, I would prefer the WOMAN on top. 

[coromonadalix]

I have a kepco ate 75-8, gives 80v at 8 amps,   modded with 2x 4 1/2 digits panel meters, modded the pots with multi-turns ones, heavy as hell, 864 watts of juice ... a very noisy 120vac fan inside.
Had to put the fan on a speed controller/dimmer,  but practical when i short this puppy, it output lots of heat at maximum outputs.

[med6753]

I have several  homebrew supplies but this is my "go to" when requiring high current. 1.2 VDC - 15 VDC at a continuous 10 Amps. The heart of the supply is a LM396K regulator which is the big brother to the LM317. The case was salvaged from a 1950's EICO Kit 6V/12V automotive battery charger/eliminator.



A peak at the inside. It ain't elegant but it works. Hidden under the heatsink is a fan.

[SharpEars]

At up to 4000 volts, I think I've got it covered. Here it is lighting an LED with 10 mA (V setting: 4kV C setting: 10 mA, CC mode triggered):

[julius oe.]

Hi there,

here is my main PSU-Frontend for high current and fixed voltage projects. As you can see it's going on a usual ATX-male plug, so i can use different ATX-supply units (in case of defects).





Some extras: Analog AmpMeter, hour-counter, isolated-ground, power LED, "all ok?" LED, decorative broken Op-Amp.

I could have save me all those plugs by using a nice switch 
that idea appeared just last week... on the other hand side: I'm actually using a modded computer power plug for a +12V,5V,Grd.,-12V rail. comes in handy sometimes.

dodgy but it dose what i need!