Rigol DP712 Output Shorted; Fuse? / I Screwed Up

[zanfar]

If you want the backstory, skip to the end.

----

I have a Rigol DP712 50V/3A power supply, and I think I blew something in the output. With nothing connected, when I turn the output on, it immediately switches to CC mode (setpoint is 5V/500mA). The rear fuse appears okay, and I would expect the unit not to turn on at all if that was the problem.

The short: Is this worth diagnosing and repairing myself?

The long: How do I troubleshoot this? I have not opened the case as that will break the warranty seal. I know those are marginal at best, but I'd rather not until absolutely necessary.

----

Backstory:

I am an idiot. I saw the "#919 - How To Charge Li-Ion/LiPo Batteries With A Power Supply" EEVBlog video and had a battery pack I thought I could save. I verified the terminals and polarity, but when I attached the battery to the power supply, something popped. I detached the pack and turned off the PS immediately. It has been off and unplugged from the wall since. The pack is a 3P2S, as far as I can tell. At the time of the failure, the output was set to 12.6V/500mA, OVP at 13V, and OCP at 700mA.

I'm aware I may have just let the smoke out of $300; you're welcome to roast me, but please do so in parallel with advice. Thanks.

[Harry_22]

Dear Zanfar, don't worry this is normal situation if you work with batteries.
Just open the cover find powerful diode near output and put it out.

[zanfar]

Sorry, can you elaborate?

I think you're saying to find a large diode connected to the output and replace it; is that correct?

If so, I'm curious how an overvoltage or overcurrent situation could blow something on a device rated for 50V/3A and how, in a blown state, that would short the output.

Thanks for the suggestion, regardless. I probably won't have time to crack the case until the weekend.

[Harry_22]

Could you send a photo of your battery connection to the power supply. Don't connect completely just show the wire ways and may be we find a reason.

You don't "crack the case" but open it carefully and repair the device like a service engineer.
Will find something like this:

[zanfar]

I had time today to open the power supply. Once open, there was a clear smell of smoke.

The first image is of the battery connection, as requested. The only changes have been to disconnect the positive (red) lead from the power supply and connect the DMM to verify the battery voltage (11.467VDC). The PCB was scavenged from a cheap charger, and the terminals are (from right-to-left in the image): B+/B3+ (Red), B2/B2+ (Green), B1/B1+ (Blue), T1 (White), B-/B1- (Black). If it's not clear, the leads are connected to B+ and B-.
The second image is of the power supply operating in the failed state. Again, the only changes were to disconnect the outputs.

The third image has what I now think is (at least part of) the problem. I assume the goal will be to replace the TO-220 component, which I believe to be a Diodes, Inc. S12M15-600B (https://www.diodes.com/assets/Datasheets/S12M15600B_LS.pdf)

Are there any other places I should double-check? Are there any caveats to desoldering and replacing this part? I am a moderately skilled solderer and have a quality station available.

Thanks again for you time.

[Harry_22]

Many thanks for photos. They make the conversation more meaningful.

So you connected battery correctly.
The PS is working. It provides current 0.5 A to some short circuit "load". And we can calculate it 0.58V / 0.5A = 1.16 Ohm

You see the capacitor at output. If you put PS off this capacitor have to be discharged. Correct?
I guess that it is discharged by this thyristor.
You connected the battery to PS which was switched off and all the possible current passed to this thyristor.
Also your PS has OVP as well as OCP functionality. This somewhat complicates the logic of work, especially with the battery.
I can see other powerful parts at output. Seems they are also involved in PS protection.

What should be done is to eliminate the short circuit. You put out S12M15 thyristor and check the PS.

[macboy]

Nearly all lab supplies will have a reverse-biased rectifier diode across the output terminals. In some "normal" usage cases, this will prevent a reverse voltage from appearing across the terminals, which could damage various components (output transistors, output capacitors, etc.). Such a case is when using two supplies in series (to get higher output voltage), and one goes into current limiting mode, while the other continues to try to push more current.

Anyway, this diode is easily damaged by doing things like connecting a powerful battery or another power supply in reverse, or sometimes even just by using the supply/supplies in "normal" series operation. I've seen 1 A reverse diodes on a 3 A supply! You can imagine what would happen in the above scenario if the second supply is pushing 3 A through a 1 A diode.

Rectifier diodes usually fail short circuit. Of course, not a perfect short, but with some resistance (it is a semiconductor after all). Based on the symptoms you see, I would strongly suggest looking for this diode and testing it. You can check if the diode is the problem simply by removing it (or lifting one leg), then testing the supply.

You could also test it with a multimeter in diode test mode. Turn off and unplug the supply. Short the terminals (to discharge the output capacitor, just in case) then remove the short. With a multimeter in diode mode, connect test leads to the + and - outputs of the supply. Test in both directions. If the diode is good, the meter will show approx 0.6 V forward voltage (a good diode) in one direction, and open/over limit in the other direction (it may start low and slowly climb up, as the output capacitor of the supply is charged by the multimeter test current). If the diode is bad, you will measure 0.6 V or lower in both directions.

[zanfar]

One last question: is there anything special about the compound between the TO-220 package and the heatsink, or is it normal thermal paste? I have verified that the heatsink and package case are electrically connected in the factory state, so I shouldn't need anything insulating.

[tooki]

One last question: is there anything special about the compound between the TO-220 package and the heatsink, or is it normal thermal paste? I have verified that the heatsink and package case are electrically connected in the factory state, so I shouldn't need anything insulating.

Just normal thermal paste. Assemble the replacement however the original was.

[tooki]

You don't "crack the case" but open it carefully and repair the device like a service engineer.

”Cracking the case” (or “cracking open…”) is completely normal English colloquial usage for “opening the case”. It doesn’t imply that care is not taken.

Not that I think you should be trying to lecture people about language usage, when in your own first reply you said this, which is not correct English, neither formal nor colloquial, and in fact led to confusion:

Just open the cover find powerful diode near output and put it out.

(FYI, to “put out” has several meanings in English, but none of them is “replace” or “remove”, or anything else that would apply here.) It’s fine to make language mistakes, we all do, but then don’t admonish others for what in fact aren’t even mistakes!

[Harry_22]

Nothing special. Usual termal compound. Yes heatsink has a connection with thyristor anode.
Could you make a photo from othe side to see the rest TO-220 devices.

You can ask as many question as required until you fix the bug.

PS
Meanwhile I found the disclaimer notes from Rigol:

[Harry_22]

Dear Tooki, thank you for corrections. English is not my native language. One of my aim to be here is to improve my language skills.

[zanfar]

I have this photo, let me know if you want additional ones. Visually, the rest of the board appears fine.

[Harry_22]

please specify the name of this part

[zanfar]

I believe it is an On-Semi MBR10100G Schottky Diode (assuming you're not asking about the Cap  ).

https://www.onsemi.com/download/data-sheet/pdf/mbr1080-d.pdf

[Harry_22]

and back side of PCB also

[Harry_22]

seems Schottky is the protection diode and it is alive because you connected the battery correctly.

[zanfar]

...and a better front-side.

[zanfar]

I ordered the replacement parts, and I've been re-reading the posts here:

You connected the battery to PS which was switched off and all the possible current passed to this thyristor.

By this, do you mean that had I turned the PS on first, then connected the battery, I may not have blown the diode?

I'm not sure which order I operated, nor am I now sure of which order the video showed. I'm not planning on trying again, just curious and trying to understand.

[thm_w]

I don't think you're an idiot. The video is deceptive and says charge "relatively safely". You can see the discussion in the original forum post: https://www.eevblog.com/forum/blog/eevblog-919-how-to-charge-li-ionlipo-batteries-with-a-power-supply/50/

I don't think its safe and I now would only use a proper two quadrant supply or one with specific charging output (RD6006). If I didn't have those then the diode on the output.

The "right" way I guess would be to turn on the PSU first at a voltage just at the battery voltage. But again, I would not do that, its just unnecessary when a diode can be used.

[zanfar]

Thanks; the idiot part isn't me following the video. I was waiting for parts to arrive for other projects and was somewhat bored. I saw the video and jumped straight into "this is something I can do now!" instead of working through the problem. I don't need the battery, and I have a LiPo charger/balancer in storage--I just didn't take my time.

[Harry_22]

Dear Zanfar, unfortunately your power supply does not allow the direct battery connection.
No one could have foreseen this.
The fact is that the capacitor is installed at output. It is discharged through thyristor when the voltage is removed by pressing on/off button. When you connected the battery you voltage obviously was off and current passed through the thyristor. The swelling of thyristor case shows that the current was very large.
This can be corrected by installing an external diode between PS and battery (see scheme below).
I can't understand only the function of p-channel MOSFET.

[IanJ]

I can't understand only the function of p-channel MOSFET.

Hmmm, possibly when the thyristor turns on the mosfet will turn on also so that when the output cap is discharged it will also pull/reference the neg output down to earth via the mosfet, i.e. stop any spikes when the thyristor turns on.......if the other end of the 390ohm is tied to earth.

Just a thought.

Ian.

[shapirus]

It's interesting. What's the rationale behind using a thyristor for something as simple as discharging the output cap? Why don't they use a resistor instead of it?

[zanfar]

I would guess due to the fact that it can be switched and therefore is not shorting the output during operation. But that's just a guess. This assumes that the thyristor is only used to discharge the cap, but the fet/resistor branch means there is still something unknown. I was casually working on reverse engineering a schematic of the output board; maybe I'll step that up and see what I find.