Can a reverse polarity Lithium Ion Battery connection damage an HP Power Supply?

[SharpEars]

I was using an HP 6626A power supply over the weekend to charge a lithium ion 3.7V battery. I made the stupid  mistake of connecting the battery backwards (i.e., reverse polarity) to one of the 50W outputs which at the time was programmed to supply 4.2 V at 0.5 A. The length of the mistake was somewhere on the order of five seconds and the battery at the time was supplying approximately 3.85 V. Unfortunately, this all happened so fast that I didn't even catch a glimpse at what the supply displaying during this time period.

The lithium ion battery seems not to have been damaged or affected in any way. I am wondering if I could have damaged the power supply doing this or if it has enough protection circuits in it to be able to tolerate a "momentary" (on the order of five seconds) reverse polarity battery connection such as this.

I am not even sure exactly what would happen electrically in a scenario like this. Clearly the supply would be supplying its max programmed current of 0.5 A at +4.2V while battling an unknown current at -3.85 V.

Thanks in advance for any thoughts, ideas (or condolences). I am sure that I am neither the first nor the last person to make this sort of mistake.

- Mr. Will Check Twice Next Time before turning on the power!

[Fungus]

Have you connected up a multimeter and tried it?

[SharpEars]

Have you connected up a multimeter and tried it?

It seems to be measuring and operating normally.

I searched through the manual for this scenario and the closest I found was that it mentions a "reverse diode" characteristic when two of the outputs are connected in series and the supply is in CC mode (See: First paragraph at the top of Page 60 of the Operating Manual), where a reverse voltage can be present at (presented to?) an output. I am not sure what this means and it is not clear to me if the supply has protection from this sort of thing, specifically with the scenario I presented it with per my original post.

[SharpEars]

As I read up on how lithium batteries are constructed, I am learning that most (all?) of them have reverse polarity protection as part of their control circuit. I am guessing that this must have kicked in and probably saved both the battery and the supply, but am still not 100% percent certain.

By the way, the battery in question was from a Nintendo New 3DS XL that I bought used for my son (without a charger). Of course he wanted to play right away and so I had to hack together a charging solution . Fortunately his dad had an HP 6626A lying around in the basement. Unfortunately, his dad lacked attention to detail (attributing this to the stress of condoling my son) resulting in that one fateful moment of noticing that I had connected the red wire to the "-" prong and the black to the "+", followed by the corresponding jerk of the Li-ion battery away from the PS leads when I saw the error of my ways.

[deadlylover]

Aren't you just putting them in series and then shorting the output? The current limiter should kick in so no harm done. I tried it on a SMU, no drama. 

The battery might have over current protection as well.

[HighVoltage]

I have no particular experience with the HP 6626A.
But I have learned my lessons with battery charging and HP/Agilent power supplies.
Some of them have a "crow bar" that kicks in and shortens the output to almost 0 ohms (to protect the DUT against over current and over voltage)
At one time I had a brand new Agilent PSU and the crow bar kicked in, while I had a 12V battery hooked up for charging.
The magic smoke came out the side of the PSU.
I was able to repair it but it was a lot of work to fix that burned PCB.

In your case it seems the PSU is still alright.
Just test voltage and current output
 

 

[SharpEars]

I have no particular experience with the HP 6626A.
But I have learned my lessons with battery charging and HP/Agilent power supplies.
Some of them have a "crow bar" that kicks in and shortens the output to almost 0 ohms (to protect the DUT against over current and over voltage)
At one time I had a brand new Agilent PSU and the crow bar kicked in, while I had a 12V battery hooked up for charging.
The magic smoke came out the side of the PSU.
I was able to repair it but it was a lot of work to fix that burned PCB.

In your case it seems the PSU is still alright.
Just test voltage and current output

Fortunately a 3.7V lithium ion battery is not a 12 V (car?) battery. It has a protection circuit (with hopefully reverse polarity protection and over-current protection) and can put out far fewer amps into the crowbar if it does kick in (which I am not sure that it did). Also, I can envision a significant voltage drop on the side of the lithium ion if 500 mA of reverse current got fed into it (and its protection circuit).

As I reflect on this, I realize that there was a significantly greater chance of the battery being damaged (and I am referring to the protection circuitry and not the Li-ion cells) than the supply, given its (i.e., the battery's) capability (and protection circuits).

By the way, if you remember, which Agilent PSU was it that you burned out, just out of curiosity? Many of the high current ones (i.e., not the 6626A) make excellent programmable battery chargers if you write your own software and use GPIB. You can really top those batteries off like no automated charger can.

[Kilrah]

Doing this should have done nothing more than "a bit more than a short circuit" from the PSU's point of view, and as it was set to 500mA current limit you simply discharged your battery with 500mA for a few seconds. Nothing to worry about for either the PSU or battery.

[nctnico]

When charging batteries it is really necessary to use a series diode between the PSU and the battery. If something goes wrong the battery may short into the power supply causing major damage. Some power supplies have a crow-bar circuit which shorts the output if it detects something is wrong. With a battery attached directly this event will burn away wiring and PCB traces (I've seen the results of such an event).

[HighVoltage]

Fortunately a 3.7V lithium ion battery is not a 12 V (car?) battery.

No, it was a smaller 12V Lab battery but still lead type with good significant initial current

By the way, if you remember, which Agilent PSU was it that you burned out, just out of curiosity? Many of the high current ones (i.e., not the 6626A) make excellent programmable battery chargers if you write your own software and use GPIB. You can really top those batteries off like no automated charger can.

It was the Agilent E3648A, dual output PSU
Luckily, the PCB in this area was only a two layer board and I could repair it.
Before this accident, I was not aware of the Agilent "crowbar".
 

[SharpEars]

Doing this should have done nothing more than "a bit more than a short circuit" from the PSU's point of view, and as it was set to 500mA current limit you simply discharged your battery with 500mA for a few seconds. Nothing to worry about for either the PSU or battery.

I don't think that the current limit setting on the HP PSU applies when the PSU is sinking current rather than sourcing it, which it would have been in this case (depending of course on the aggressiveness of the protection circuitry in the Li-ion battery).

[nctnico]

Many of the high current ones (i.e., not the 6626A) make excellent programmable battery chargers if you write your own software and use GPIB. You can really top those batteries off like no automated charger can.

Why would you need extra software for that? Just use a series diode, set the voltage to the maximum charging voltage and the current to the maximum charging current. The power supply will do CC/CV charging by itself (I charge batteries this way every now and then).

[Performa01]

Doing this should have done nothing more than "a bit more than a short circuit" from the PSU's point of view, and as it was set to 500mA current limit you simply discharged your battery with 500mA for a few seconds. Nothing to worry about for either the PSU or battery.

I don't think that the current limit setting on the HP PSU applies when the PSU is sinking current rather than sourcing it, which it would have been in this case (depending of course on the aggressiveness of the protection circuitry in the Li-ion battery).

How do you think the PSU would sink current?

As the battery is connected in reverse, the output voltage would be e.g. -3.7V, which is <0V, hence 'a little more than a short' from the PSU's perspective.

Ideally, the PSU still sources 500mA, which now acts as a discharge current for the battery (because it's connected the wrong way).

In practice, there will most likely be a beefy reverse polarity protection diode across the output, which limits the output voltage to some -0.8V and pretty much shorts the battery - but the PSU still sources just 500mA.

If there is a crowbar, it would
1) only fire if there is an overvoltage, which is clearly not the case
2) not be effective anyway because of the wrong polarity

[SharpEars]

Doing this should have done nothing more than "a bit more than a short circuit" from the PSU's point of view, and as it was set to 500mA current limit you simply discharged your battery with 500mA for a few seconds. Nothing to worry about for either the PSU or battery.

I don't think that the current limit setting on the HP PSU applies when the PSU is sinking current rather than sourcing it, which it would have been in this case (depending of course on the aggressiveness of the protection circuitry in the Li-ion battery).

How do you think the PSU would sink current?

As the battery is connected in reverse, the output voltage would be e.g. -3.7V, which is <0V, hence 'a little more than a short' from the PSU's perspective.

Ideally, the PSU still sources 500mA, which now acts as a discharge current for the battery (because it's connected the wrong way).

In practice, there will most likely be a beefy reverse polarity protection diode across the output, which limits the output voltage to some -0.8V and pretty much shorts the battery - but the PSU still sources just 500mA.

If there is a crowbar, it would
1) only fire if there is an overvoltage, which is clearly not the case
2) not be effective anyway because of the wrong polarity

I would think that the following happens in a reverse polarity scenario due to the battery's protection circuit:

The battery goes high impedance by way of either an in-line MOSFET or a very fast acting (i.e., <=200 ms) auto-reset fuse that trips due to overcurrent (from the battery) trying to deal with the 0.5 A reverse polarity load imposed by the PSU.

[TheSteve]

When charging batteries it is really necessary to use a series diode between the PSU and the battery. If something goes wrong the battery may short into the power supply causing major damage. Some power supplies have a crow-bar circuit which shorts the output if it detects something is wrong. With a battery attached directly this event will burn away wiring and PCB traces (I've seen the results of such an event).

If you want protection against a short you should use a fuse, not a diode. when charging lithium cells you want precise control over the voltage.

[KL27x]

I would think that the following happens in a reverse polarity scenario due to the battery's protection circuit:

The battery goes high impedance by way of either an in-line MOSFET or a very fast acting (i.e., <=200 ms) auto-reset fuse that trips due to overcurrent (from the battery) trying to deal with the 0.5 A reverse polarity load imposed by the PSU.

I'm with Kilrah, here, from perspective of the battery. I'm not too keen on this "reverse polarity load," idea. It's just a load of 500mA to the battery. Charging a cell properly is what I would call a "reverse polarity load."

The "reverse polarity protection" on the cell is nothing more than a switch that cuts out when the voltage of the cell decreases below a certain cutoff, around 3V. And that doesn't happen at 500mA current-limited load until the battery discharges sufficiently.

The question is whether or not the PSU can limit the current in this scenario. I plugged a li ion battery backwards into a small charging circuit, once, and the chip quickly got hot and fried. The battery was fine, but I unplugged it after only a few seconds. For all I know, the current was still properly limited up until the chip fried, but the voltage drop absorbed by the chip was >4V instead of ~1V. But I wouldn't bet on it. Because in order to limit current, the psu/chip is going to do this by futzing with the output voltage. But if it's a linear regulator, it can probably not reduce voltage below 0... so perhaps it can't limit current, at all, in this scenario? Perhaps if it was a buck output, it could? Not really sure.

But in case the psu can't limit the current, perhaps the battery could save the PSU by virtue of its max discharge protection circuitry/fuse. Not the "reverse polarity protection." IMO. But I have to admit I'm not completely sure I know what I'm even saying, lol.

[vince53]

I have a couple of power supplies that I will use to charge batteries directly when needed.  I would never hook up a battery to a power supply without a series diode.  Even if the polarity is correct, bad things can happen if the line voltage on the power supply goes out and the battery is still connected.

One of my supplies is a Xantrex 60v, 20 amp switching supply.  I believe there was a sentence in the instructions about using a series diode when hooked to a battery. I believe it does have an over voltage crowbar circuit. I have the output from the back of the supply going to a diode on a small heat sink and pair of binding posts.  The sense leads from the supply are hooked on the binding posts so the voltage drop across the wires and diode is negated.

On my Astron linear supply I had enough room to mount the diode inside and add a second pair of binding posts on the front panel.  The voltage on these posts doesn't match the meter but I use it in constant current mode and keep an eye on it.

For my  RC batteries I use chargers with balancing circuits.

Vince

[relic38]

I connected a 5S Li-Ion pack (18V, 2200Ah for RC airplanes) backwards to my bench supply.  The resulting arc scared the living crap out of me.  That channel of the supply went dead immediately.  The battery was fine.  the battery is rated to output 35C (i.e. 80A) continuous, so I have no idea exactly how much current was flowing for the fraction of a second it was connected.

Anyway, I found that the supply had reverse polarity protection in the form of a large (5W) diode across the output.  It had fused into a short, which continued to short the supply output (I'm glad it didn't blow open). I had a similar diode in a junk bin so I replaced it. 
The supply has been fine since.  Lesson learned.

[SharpEars]

I connected a 5S Li-Ion pack (18V, 2200Ah for RC airplanes) backwards to my bench supply.  The resulting arc scared the living crap out of me.  That channel of the supply went dead immediately.  The battery was fine.  the battery is rated to output 35C (i.e. 80A) continuous, so I have no idea exactly how much current was flowing for the fraction of a second it was connected.

Anyway, I found that the supply had reverse polarity protection in the form of a large (5W) diode across the output.  It had fused into a short, which continued to short the supply output (I'm glad it didn't blow open). I had a similar diode in a junk bin so I replaced it. 
The supply has been fine since.  Lesson learned.

I would be surprised if a Nintendo New 3DS XL's battery was rated to even supply 0.5C (1750 mAh capacity). I think the protection circuit would kick in way before that (and most likely within several hundred ms) to prevent the internal cell from overheating. I suppose the proof of this was the fact that I didn't even see so much as a spark when I connected the supply leads to the battery terminals. Also, the battery felt cool to the touch seconds after the fact. I wish I had glanced at the supply display before disconnecting to see what it was up to (or better yet had some sort of external measurement device that recorded what was happening in those precious few seconds).

[relic38]

I would be surprised if a Nintendo New 3DS XL's battery was rated to even supply 0.5C (1750 mAh capacity). I think the protection circuit would kick in way before that (and most likely within several hundred ms) to prevent the internal cell from overheating. I suppose the proof of this was the fact that I didn't even see so much as a spark when I connected the supply leads to the battery terminals. Also, the battery felt cool to the touch seconds after the fact. I wish I had glanced at the supply display before disconnecting to see what it was up to (or better yet had some sort of external measurement device that recorded what was happening in those precious few seconds).

Yeah, a device battery is probably 1C or 2C discharge rated.  Also it iwas probably over0current protected.