low voltage(3.7v) high current(135A) short circuit protection

[benyB]

I am making multiple 8 x parallel(lithium iron 18650 3.7v 3400mA) battery packs that should be under 100Wh capacity.

The individual cells are not protected for over current. there for my ebike.

Each x8 cell pack will need some kind of short circuit protection >100A , 135A ( 500w )

The x8 cell pack has no BMS.

I don't need charge voltage protection as the x8 cell packs will be connected together in series to make one large 36v battery that will have a BMS.

I'm trying to design a modular battery that is designed to come apart.

I need some protection against the small packs shorting out and going into thermal runaway ( catching fire )  !

Should i used an existing battery protection IC and a few mosfets to increase the power ?

It seams battery protection IC's monitor mV drop rather than detect current with with a shunt?
I could calculate volt drop of a 500W draw on my 8xbattery pack and set the IC to trigger enough mosfets to switch off for 3.7v 500W maybe ?
But I get a feeling this is way over complicated?

If possible I want to avoid having a full BMS in each battery (they are just 1 large 3.7v cell ) and another BMS to control the 10 batteries hooked up in series !

can anyone point me to a design that could achieve the low voltage high current protection ?

How about a massive thermistor? is that even slightly safe?

Its for an E-bike that can draw up to 500W peek

[f4eru]

strange inefficient design.
Why ?

[Siwastaja]

By Kirchoff laws, same current flows through all cells in series. If you cut the path anywhere, no current can flow. As a consequence, current measurement and overcurrent (including short circuit) protection is done once, on the pack level.

Possibility of current taking alternative routes is prevented by good mechanical design.

If you try to come up with a design where each series element is separately protected "for modularity", the cost and complexity will blow through all roofs. Such protection needs to handle the full pack voltage, after all, so single-cell integrated solutions won't work. Just don't do it.

[f4eru]

and also a lot of connectors at low voltage and high current in the path without reason is wasterful from both financial and ohmic losses points of view