Diode across battery pack - does it have a purpose?

[bookaboo]

I'm attempting to optimise battery life on a remote IoT sensor and turned my attention to the battery pack which the customer has sourced. Looking at the suppliers drawing I noted the 1N4007 diodes across the cells. The supplier insists they are for "anti reverse charging" (it's a non rechargeable pack).

I can't see what function the diode performs in this configuration? The pack is built to drawing with the diodes fitted as shown and are 1N4007, not a zener or anything.
Is there a gap in my understanding? They look pointless.

[RoGeorge]

1. When cells are connected in series, the one that discharges first will become reverse polarized by the other stronger cells.  That can damage the reverse polarized cell.  By adding the diode, the maximum negative voltage will be limited at -0.5V, which might not completely ruin the reverse polarized cell.

2. The reverse polarized cell might show a high impedance and cut the current, even if the other cells might still have enough voltage and current.  With the diode, the whole pack will remain low impedance, because an open diode is almost like a short-circuit, so it will let the other good cell(s) to deliver power to the load.

3.  Protect the cells against trying to recharge them with wrong polarity charger (if connected to the charger in reverse, by mistake).

[voltsandjolts]

There is no charger. These ER34615 cells are lithium metal primary cells.
Connecting them in parallel like this makes me nervous, it assumes they are well matched (voltage and capacity) and if one fails with internal short the explosion will be twice as big as you might expect.
And lithium metals do explode violently if you don't treat them nice.

[RoGeorge]

The diodes have nothing to do with internal shorts.  With or without diodes, batteries do not explode as easy as you may think.

It's OK to have the diodes there, they will make the battery pack last longer.  There is nothing to worry about.

[bookaboo]

1. When cells are connected in series, the one that discharges first will become reverse polarized by the other stronger cells.  That can damage the reverse polarized cell.  By adding the diode, the maximum negative voltage will be limited at -0.5V, which might not completely ruin the reverse polarized cell.

I've seen that on lead acid, and come to think of it on NiMH packs. It's probably the reason they have it here.

3.  Protect the cells against trying to recharge them with wrong polarity charger (if connected to the charger in reverse, by mistake).

These are non-rechargeable, but that also would make sense.

There is no charger. These ER34615 cells are lithium metal primary cells.
Connecting them in parallel like this makes me nervous, it assumes they are well matched (voltage and capacity) and if one fails with internal short the explosion will be twice as big as you might expect.
And lithium metals do explode violently if you don't treat them nice.

Yes these are Lithium-thionyl chloride (Li-SOCl2), they have a much higher internal resistance than Li-Ion or Li-Po.
Dead shorting a single cell would give you 120mA if you were lucky, there is a ER34615M which would delivery higher current than ER34615, but still much less than other Lithium types. This is why my attention was drawn to the pack in the first place, it's the old "transmit burst current issue" causing some edge case issues.
The packs are manufactured in China, the customer has been using them for years with zero issue, but yes it's good to be diligent.

[coppercone2]

cool to know, is there any other tricks for lithium thionyl chloride batteries? Like if you wanted to make a backup pack. I think its kind of cool to have a anytime 20 year power source ready to go, with like those D cell xeno's

I had a 4s string of AAA NiMH cells cook off like that I think, the wrapper on one of the cells looked wrinkly.

Interesting how the circuit behaves when you open circuit on one of the batteries lol, especially for HV strings. 

I did make some lab power sources with D cells, but I fuse them, hope its enough, but usually if I am using that its for meticulous work and I keep my eye on them. Also a nicer form factor for carrying (i.e. testing trailer lights, testing a fan), and I can put whatever fuse I need for the load on the front panel, and it has banana jacks. I just put like 10x D cells in series with a fuse, but I usually don't use this thing for more then a few minutes. I made 9, 14 and 5V ones (approx). Also thought to make some smaller ones with altoids tins that use AAA's and maybe 9V cells for special applications, and maybe a coin cell bias supply that is HV. Kind of jealous how they used to have 60V batteries in the store back then, and now we are stuck at low voltages (when you need 500uA or something).

I really like it for trouble shooting bulky equipment that is worked on in a mechanical area, like a welding machine with a fan. You don't need to manage them as much as lead acid batteries too, those are obnoxious, never seem to work when you need em.

I don't put switches on these, I opted to make a banana connector plug in switch (you can even do double throw), because I figure that is going to break on me. Maybe for the small altoids tin ones I would put switches, but for the D cells they are a nusence. Just get banana plastic plugs to keep them isolated when not in use.

[bookaboo]

I imagine the internal resistances would sum in a series pack, so even less current supply capability. You could parallel the series packs but that's straying perhaps into the dangers outlined above.

The capacity/volume ratio is very hard to beat though. You could use them to trickle recharge a more useful battery like Li-Ion.

Here's a very useful appnote forpowering IoT devices using ER34615 or similar:
https://www.quectel.com/wp-content/uploads/2021/03/Quectel_Low_Power_Design_User_Guide_V1.3.pdf

[voltsandjolts]

There is no charger. These ER34615 cells are lithium metal primary cells.
Connecting them in parallel like this makes me nervous, it assumes they are well matched (voltage and capacity) and if one fails with internal short the explosion will be twice as big as you might expect.
And lithium metals do explode violently if you don't treat them nice.

The diodes have nothing to do with internal shorts.  With or without diodes, batteries do not explode as easy as you may think.
It's OK to have the diodes there, they will make the battery pack last longer.  There is nothing to worry about.

Read what I wrote. I didn't even mention diodes. I've worked with lithium metal batteries for >20 years. I've met people that have been involved in the explosions, including one who lost hearing in one ear when a battery pack exploded in a small room, and another whose truck windows were blown out as they were driving when a battery pack on the back seat exploded. So, be careful, treat them with respect.