Power supply overshoot when disabling output

[electricar]

Hi folks,

so this is my first post after reading a lot in this forum First I have to thank all the people around here for helping and discussing about very intersting topics. I learned a lot!

So this is my problem now:
I got this power supply
http://www.amazon.de/Wentronic-Stabilisiertes-Labornetzger%C3%A4t-Ampere-Display/dp/B001C6FQF0/ref=sr_1_1?ie=UTF8&qid=1452881325&sr=8-1&keywords=Wentronic+Stabilisiertes+Labornetzger%C3%A4t

I bought it in 2012 and it worked like a charm until a few weeks. Now it produces an overshoot everytime I disable the output with the right red output button.
E.g.:
Settings: 3,3V with current limit 0,25A
Load: 100R or without load, doesn't matter.
I got some screenshots in the attachments.
I also tried to reverse engineer the output stage, but I think that the problem is actually in the control loop after looking at the schematic.

It is very troublesome...

Does anybody has the real schematic of this power supply or even own the same supply and has got the same problems?
Do you guys have some ideas at what I should look for? I also checked the electrolytic capacitors and added some in parallel but it didn't help, too.

Thank you very much in advance for your help!

Regards

[uncle_bob]

Hi

We had a bunch of supples back 10 or 20 years ago that put out a "full voltage" spike on shutdown. Our decision was that chucking the poorly designed supplies was cheaper than continuing to blow stuff up. Even if we fixed the first problem, who knew that other fatal design flaws were lurking in the beast.

Being the doubting type, I took one out of the garbage and brought it home. The thought was that "if I *always* disconnect the load first I will never have a problem. That was before it caught fire ....

Bob

[wraper]

check the electrolytic caps on the output.

[electricar]

I already checked them.

While looking for other power supplies I found one that looks similar to mine from the outside and the inside. I noticed other power transistors and additional fine adjustment for the current but I think they are quite similar.

Perhaps this helps solving the problem?! 

[mij59]

Check the relay's for bad or welded contacts.

[electricar]

Check the relay's for bad or welded contacts.

How to do that? Never did this before.
Thanks!

[Kleinstein]

Checking the contacts can be done by just measuring in circuit. Welded contacts would keep the relay stuck to one position.

It is also possible to just have too much delay between the relays. The relay disconnecting the output should not be slower than disconnecting the regulator. It does not not look like a good design in this respect.

[electricar]

After replacing the two 220uF electrolytic capacitor by one 120uF aliminium polymer capacitor the spike only remains when disabling the output. "Only" a little overshoot from 3,5V to 9,6V 

I've checked the relays. Is it possible that when the relay is damaged, the two alternating pins are separated when the coil is not powered and connected when the coil is powered? Actually they should never be connected!?
I'm talking about the two pins 14 and 12 of relay K2 in my schematic and the upper and lower pins of the relay K1a in the schematic of the Basetech BT-305 power supply.

[uncle_bob]

It is also possible to just have too much delay between the relays. The relay disconnecting the output should not be slower than disconnecting the regulator. It does not not look like a good design in this respect.

Hi

.... or .... It does not look like a good design.

Bob

[mij59]

After replacing the two 220uF electrolytic capacitor by one 120uF aliminium polymer capacitor the spike only remains when disabling the output. "Only" a little overshoot from 3,5V to 9,6V 

I've checked the relays. Is it possible that when the relay is damaged, the two alternating pins are separated when the coil is not powered and connected when the coil is powered? Actually they should never be connected!?
I'm talking about the two pins 14 and 12 of relay K2 in my schematic and the upper and lower pins of the relay K1a in the schematic of the Basetech BT-305 power supply.

On the data sheet of the HK3FF relay the  contacts are not numbered, where do the numbers come from ?
Best way is to remove the relay and test the contact resistance in the powered and non powered state.

[electricar]

Those pin numbers are not the same as on the actual relay. I chose this eagle symbol only for making clear the wiring.
I tried to remove the relay but it's really hard to heat up all those pins! 

[mij59]

Those pin numbers are not the same as on the actual relay. I chose this eagle symbol only for making clear the wiring.
I tried to remove the relay but it's really hard to heat up all those pins! 

I guess you don't have a hot air de soldering tool, so leave it.
I can't attach the data sheet, check the relay according to the data sheet, the common contact is in the middle on the left side, in the off powered state it connects to the terminal on the bottom right

[Refrigerator]

If it only started doing this now it could be a cracked solder joint, check around the pins of components, especially the bodged ones ( like the resistor near the multiturn pot ).
Also if you reverse engineer the board a little more you might find the problem.

[electricar]

I somehow managed to desolder the relay and it works normally. After adding the relay onto the pcb the overshoot somehow is "only" about 1,4V.

[mij59]

I somehow managed to desolder the relay and it works normally. After adding the relay onto the pcb the overshoot somehow is "only" about 1,4V.

Sounds like some kind of mechanical problem, could still be the relay's or what Refrigerator pointed out.
Try tapping the pcb and relay's with the ( isolated ) handle of a screwdriver.

[electricar]

It didn't change anything 

Besides that I don't want to spend so much more time in finding the problem, I have other projects which I want to complete. And because this is my only power supply, I urgently need another for the time until I finish my own power supply (draft schematic already finished).
I wanted to ask you what power supply you could recommend as "second source", although it would be my "first source" until I finish my own power supply.
What do you think about the Korad KA3005P. As second source it would suffice I think. Can you recommend other power supplies?

Thank you in advance!

[uncle_bob]

Hi

Used supplies abound on eBay. They can be dirt cheap. Anything with an HP label on it should be fine. The older stuff is a bit better bet than the newer ones (easier to work on). One that does not sell for quite the premium is Power Designs. They made darn good supplies. Since they are no longer around, they can go for some pretty low prices. I have a number of them. They all work very well.

Sitting in front of me are a pair of GW GPS-3030's. They are a nice modern supply. They work well. The pair I have spend 20 years hard at work before somebody decided to upgrade all the benches to digital readouts. They were *very* affordable.

Bob

[electricar]

I know that in the US there is a lot on eBay. Unfortunately here in europe there aren't so much of these power supplies on eBay 
Besides that I only need a decent power supply for testing the seperate parts of my own designed power supply until it works The bought PSU doesn't have to be the best in the world and full of features. Only an output switch would be nice (this feature is btw really rare on the low budget end!).
I really think the KA3005D/P has the the most features for it's money. I'm only a little bit unsecure about some comments about soft buttons not working properly...

[uncle_bob]

Hi

I've heard a lot of the "only shows up in the US" comments before. What really happens is that your search turns up a lot of dealers from the US (and elsewhere) that are selling stuff you don't want and doing it at prices you never ever would pay. I typically spend a few weeks sorting through things to find anything worth biding on. Normally I find that after 40 or 80 hours of searching, the "magic item" is outside the US rather than inside. Is it worth the effort to get a very solid power supply for $20 ... who knows.

Bob

[Refrigerator]

An ATX power supply could be  a reasonable alternative for the time being.

[electricar]

I bought the Korad KA3005P and the first thing I did was to look at the output voltage when enabling and disabling it through the button. Unfortunately it has some serious overshoot problems when enabling the output. I couldn't see anything when disabling it. In the end, first I was satisfied with the look and feel of the PSU, but I will send it back because of the overshoot problem.

Edit: I forgot to look at the revision of the PSU. Do the newer revisions have the same problem? If yes, there is another reason the speed up the development of my own supply

[de.leon]

So just received my Korad KA3005d and out of the box there is an a/c voltage between ground and - terminal as soon as the main power button is pushed.  Regardless of settings, then there is a spike in both a/c and d/c well above the settings when on button is pushed, finally a larger spike when off button is pushed disengaging the output.  I know some of these issues came up in Dave’s review but that was a long time ago.  Figured they had fixed the problem…guess not.  Expect I’ll be returning this unit, still looking for a good dependable linear ps that does not cost an arm and leg.  He only way to use this unit is to plug and unplug your bananas after and before touching the off/on switch defeating its purpose.

[de.leon]

Ok went at this thing again.  Using a Fluke 115 I measured at the output binging posts using min/max function
It appears d/c is ok, no spikes or random voltages. Sadly a/c is a different animal. 
With unit plugged in and power button turned on I get 0.113v ac across + or  - and GND.
When I turn the output on, I get a 4.75 spike of a/c  across + and - to GND.
When changing voltage (rotating the knob) values I get 1.62v a/c  + and - to GND.
When turning off the output I get 2.06v a/c + and - to GND.
A/c reads 0 at all other times when output is turned on.
Don’t know if this is normal operation for a linear supply.
Don’t know if this is enough to damage circuits or oscilloscope.
If anyone can advise I would welcome the advice.  I would be happy to do a simple mod to correct rather then send it back to poor old Amazon. If not, what are the recommendations out there for hobby power supplies. Would prefer linear, don’t need computer interface, just a few memory slots.
Thanks