If someone thought there was a good chance to hack these battery meters for more discharge current, I'd send one to you.
I might even pay!

 

10 of them were $6.50 each.
No connectors with them though.

http://www.aliexpress.com/item/5-in...2637987266.html?spm=2114.13010608.0.64.UMbgvn

Is there more documentation available? The specification says the minimum input voltage is 6.5V. Does that mean the power to the unit itself, or the battery under test? My main interest would be to test AAs.

 

Have to google it. User manual, no specs.
I'm reading cells down to 2 volt. Hoping it will go down to 1.5 in lithium mode for my LTOs.
The discharger won't go below 3v.
The balancer runs at 2.4v. (5s)

 

Looking good. Time to scale it up.
After running all night.
12s max = 2.323
min = 2.311
12mv unbalance min to max.

 

Just because I have it working doesn't make it a proper design.

Really nice to see someone say it. This concept seems lost on the new generations. Both in code, and in circuit design.

 

Testing going better than I hoped.
After testing to balance 18 unmatched 12ah cells somewhere in the middle, bulk charging left 2s over charged.
Resetting my balance board to max voltage, I connected it, and it quickly brought down the high cells. Even at 1/2 amp rate.
It seems near the top there is not much energy stored. It takes far less to top balance than expected.

Need to get a second board built and charge with shunts across all 18 cells.
Testing without feedback resistor gets me to a few mv difference in shunt settings.

What I found works best to balance my balance board, is to use a stiff power supply at set voltage.
While monitoring current, I adjust each pot to near cutoff. 100ma works well.
As I'm using a cell for voltage source, it takes a second trip to verify no voltage sag.

If anyone wants to play with this circuit, I have a half dozen extra boards.
A few parts need to be added, so I will be making new boards soon, for faster assembly.
These are fine, just a time killer.

 

Full scale test underway. 18s 90+ah 4.2 volt max.
Parallel Chevy Volt packs.

Set shunts at 4.1v. Charged til all were shunting. Led on.

Now let it sit overnight and see what I get.

Balance was bad before test. Let batteries sit to long and they got way below critical.

 

Thinking of a test using 1157 lamps as load. May be simpler than remote mounting load resistors.

They offer reverse resistance change near 4v, but can't think why that would be bad.

One filament is over an ohm cold the other well under.

Test boards have 4.7ohm 5w on 12s and 3ohm 12w on 6s.

Any lower, need fan or heat sinking. Neither is a deal breaker.

 

View attachment 109922

Think I see now. In the data sheet circuit, there is current limit resistor in the high rail.
Moving load to collector circuit requires base resistor.

This with 1157 worth a shot?

 

Question about heat.

Testing going well. Balances within a few mv. Seems dependent on pot resolution more so than circuit design.

Trying the same circuit using both filaments of 1157 bulb. Draws about 1.5 amps during shunt. Lights slightly.
This may not cost more than 12w load resistors and not need cooling. As more heat will raise resistance.

Checking temperature of pnp when not saturated.
Specs are 150C.
Am I ok with case temperature of 115C?
Thermocouple held on with wood clothespin. Should be about a wash. Heat-sink wise?

 

Shunt circuit update.
Lamps make great load.
12v 50w and 6.3mm pins fit perfectly on board layout.
Sink about 2amps when set to 2.65 volts for my LTO pack.

24s 60ah pack