5
\$\begingroup\$

I hope the question is not off-topic: I want to know "why" it failed more than how I should use my equipment, even though the questions are probably linked.

The power supply is a 0-3A, 0-30V lab power supply (LABPS23023). I fried it trying to charge a 20AH 3.2V LiFePo single cell battery. I had set the supply to 3.8V for the initial charging as recommended, and the current limitation pot halfway (~1.5A, way below the max recommended 1C). Since the battery was only 3.3V stock, and that the power supply had both voltage and current limitation, I thought I would be safe.

Result = heavy instant smoke in the power supply. Not even a blown fuse.

Did the battery "feed" power into the PS instead of the opposite, and the latter would be seriously unable to withstand it?

I highly suspect my understanding is both naive and incomplete. Would a simple heavy duty diode have protected the power supply, but would it be charged in this case?

The schematics of the PS is here: enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ Even if connecting a fully charged battery in reverse, the resulting amperage should just blow a frontend fuse – but then again, if there is such a fuse. Most power supplies come with at least a rough schematic. Do you have any insight what exactly broke? \$\endgroup\$ – Marcus Müller Mar 13 '17 at 10:03
  • 1
    \$\begingroup\$ A PSU that limits both voltage and current would fulfill the two most important requirements. \$\endgroup\$ – Dampmaskin Mar 13 '17 at 10:04
  • 1
    \$\begingroup\$ But the voltage limit of a typical 0-30 V lab power supply is set with a single turn potentiometer. Attempting to dial in precisely 4.20 V is quite hard, and the pot may turn by itself if e.g. subjected to vibration (resulting in overcharging). Furthermore, the output voltage of some shoddy power supplies is dependant on temperature and/or current. I have personally blown up a 2S 1 Ah Lithium ion battery through overvoltage, while attempting to charge it like this (thankfully I kept it under constant supervision, so I just tossed it out of the window before the eruption). \$\endgroup\$ – jms Mar 13 '17 at 10:13
  • 1
    \$\begingroup\$ Can you open the PSU without voiding the warranty? It would be interesting to see exactly what blew. That being said, keeping the warranty intact seems like the most important thing right now. \$\endgroup\$ – Dampmaskin Mar 13 '17 at 10:19
  • 2
    \$\begingroup\$ I have once got the smoke out in the same way. I connected the battery first. The voltage and the current limit was carefully preset. The smoke appeared when I switched the PSU on. But that was my own idiotic design. Theoretically it's possible that your Welleman either is not lach up proof. \$\endgroup\$ – user287001 Mar 13 '17 at 10:36
6
\$\begingroup\$

Mea culpa, well, sort of. The flyback diode was dead (V32 on the schematics - no fuse indeed), and it was short-cutting the output, hence the smoke also above in the circuit. I replaced them and it worked again.

Now for "why": I have two of the same 20AH LiFePo battery. And one packaging is plain wrong! Yes, the plastic wrapping with the label is the other way round: inverted positive and negative terminals. I had probably double-checked with one, and assumed that the other was alike. I am going to yell loud at the supplier (a well-known company though). It could have cause fire, with such currents. Amazing.

Conclusions:

  1. One can never check enough the polarity (batteries were alike, I tested the right one, and probably plugged the other)
  2. Do it one at a time, it would not have occurred
  3. Do not even trust the labels and packaging! :(
  4. KISS: when the clamp diodes are dead, a reverse polarity is the obvious thing to check first (so @Bradman was right indeed)
  5. My power supply works well again, warranty void (bah)

Bonus: I replaced the SMT diodes (1N5402) by even more powerful diodes (31DF, 300V,60A peak) + fixed them with silicone so they would not even unsolder themselves on a terrific case (like here I guess).

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.