Looking for high dynamic range fuel gauge

[Waseem]

Hi,

I'm working on a design that uses a beefy 40V LiFePO4 battery pack, and can draw current anywhere from a couple of mA when sleeping all the way to 40A when the load is engaged. I want to use one fuel gauge chip but I'm having trouble selecting the current sense resistor. If I place a 1 mohm resistor for example, I don't get enough voltage drop at low current for most chips that I found that can work at 40V. And if I place a 20 mohm resistor, I get an unacceptable amount of power loss during high current 0.02*40*40 = 32W.
I contemplated the use of two fuel gauge chips, one for high current path and one for low current and then do some firmware wizardry to combine the two readings, but I would like to find a more elegant and easier solution if it exists.
Any help is appreciated.

[Terry Bites]

You can still buy an old fossil ic, the LM3915 is a
driver for a Logarithmic Dot/Bar Display. Hack a cheap kit https://www.ebay.co.uk/itm/394643376897

[Kleinstein]

Using a kind of seprate measurement system for the high current part (e.g. motors) and low power part (e.g. µC and sensors, motor controllers) is likely a way to go. The high current part is likely only using power for some time an no power the rest of the time.  To make it work easy with the SW the higher power part should take truely no power when off.

[wasedadoc]

You can still buy an old fossil ic, the LM3915 is a
driver for a Logarithmic Dot/Bar Display. Hack a cheap kit https://www.ebay.co.uk/itm/394643376897

@Terry Bites. Please explain how using an LM3915 addresses the OP's question.

[uer166]

The more elegant solution would be to ignore the sleep current entirely and reset the fuel gage at known points (say 100% or 0% SoC). This of course wouldn't work well if you have it sleeping for years.

A fuel gage/coulomb counter doesn't really tell you how close the pack is to cutoff anyway, the cell voltage does.

[Someone]

A fuel gage/coulomb counter doesn't really tell you how close the pack is to cutoff anyway, the cell voltage does.

Yes, no point trying to measure currents down near the self discharge rate of the cells.