16 Cell 18650 Charger/Discharger Board Module

[hpeyerl]

Cool. To a coder the current must look pretty amateurish. But hey, it works

Not criticizing. Happy there'll be some way for us to contribute.

 

[nz_lifer]

Rev2 design finalized.

.Removed level shifter for shift registers, not required for charger/float mosfets.

[SOLVED] MOSFET gate voltage higher than voltage at drain/source. Safe? - Page 1

[SOLVED] MOSFET gate voltage higher than voltage at drain/source. Safe? - Page 1

www.eevblog.com

. Added a low side current sense resistor, 0.05 Ohm, giving us 0.05V per Amp.
I put it on the low side to allow easy non differential high resolution measurement by the ADC.
Hopefully the MUX does not introduce too much noise into the readings.

Dissipating 0.05W at 1A. If all 16 are at full power discharge, 0.8W will be dissipated into the PCB.
A lot less than the potential 67Ws the megacell PCB has to deal with.
Will it be enough to affect our cell temperature readings?
Hopefully not if I keep the heat conduction paths limited so it keeps the heat local to resistor.
The resistor is rated to 2W, so 0.05w (1% rated power) should not raise it's temperature much.

The 1115 ADC does allow for up to 0.30V negative voltage to be measured.
Will allow for some cool stats during charging, like capacity put into the cell.
There will need to be some funky maths to accommodate different starting voltages.
Or charge/discharge to get the cells to a reference starting voltage.

.Going with a 4 Layer board.
This doubles the cost but makes routing so much easier.
From us$10 to $20 is still good for a 16 cell system.
Exact pricing after I upload to JLCPCB.

 

[nz_lifer]

Let the routing begin

 

[nz_lifer]

Routing is part science and part art.
241 nets
355 components.
My first 4 layer board, could take a while.
Using a PS4 controller to help with some hotkeys.

 

[Bubba]

Routing is part science and part art.
241 nets
355 components.
My first 4 layer board, could take a while.

Is there auto routing or does it have to be done manually.
I remember using Orcad back in the day on a 386 cpu.

I always started with required locations of the most important items first.
Followed by placing the items by the most connections that could be straight (if that makes sense to you) .i.e. data lines

So I agree. Part science and part art! Hey if it doesn't look good i can't work right?
Have fun.

 

[nz_lifer]

Is there auto routing or does it have to be done manually.
I remember using Orcad back in the day on a 386 cpu.

I always started with required locations of the most important items first.
Followed by placing the items by the most connections that could be straight (if that makes sense to you) .i.e. data lines

So I agree. Part science and part art! Hey if it doesn't look good i can't work right?
Have fun.

Auto routing for something this complex is pretty hard.
Especially with 'custom' components on there.

Now that I have each cell's component placement figured out, it is the home stretch. Another 20 hrs?

 

[Wolf]

Looking good. Very nice project.
Just a side note while working on my cells in parallel tester I did an resistance check on the style of cell holders I believe you are going to use and was shocked at the actual resistance. Almost 7mΩ. Maybe yours are different but if not I'm not sure how this could impact your C/D tester if at all.
\

Wolf

 

[Bubba]

View attachment 23382View attachment 23383

Looks like it is coming along nicely. Please don't take me wrong ... but starting to look like the layout of the MegaCell charger?

Hey could I recommend you to look at 3D printing your own Cell holder based on Watts-on https://secondlifestorage.com/index...ringslot-18650-battery-holder.5849/post-48201

He has several versions. I personally like the lifting version.

 

[Eric44]

Cool project, don't stop

 

[nz_lifer]

Looking good. Very nice project.
Just a side note while working on my cells in parallel tester I did an resistance check on the style of cell holders I believe you are going to use and was shocked at the actual resistance. Almost 7mΩ. Maybe yours are different but if not I'm not sure how this could impact your C/D tester if at all.
View attachment 23389\

Wolf

My two test at 10 and 12mΩ. ±1
At full discharge of 1A the holder could be dissipating 0.025W out of our total 4W, 0.6%
Total of 25mΩ, both ends.
The mosfets themselves are 10-30mΩ across the discharge current range, 0.8 - 1.1A.
Readings are taken after the mosfets
As long as we know we can account for it.
But I think we are talking down in the 0.5% of our accuracy, maybe less.
Starting temperature is going have more of an effect.

Discharging at High and Low Temperatures

Watch for limitations when using a battery at adverse temperatures and how to overcome the effects

batteryuniversity.com

Looks like it is coming along nicely. Please don't take me wrong ... but starting to look like the layout of the MegaCell charger?

Hey could I recommend you to look at 3D printing your own Cell holder based on Watts-on https://secondlifestorage.com/index...ringslot-18650-battery-holder.5849/post-48201

He has several versions. I personally like the lifting version.

At this stage the only thing in common with the megacell is that it has 16 cell capacity as well

Cell holders, they do look good.
Main issue is anything times 64 adds up quick.
The off the shelf ones are good enough for now for me. Anyone else is always welcomed to Gucci up their build.

Pluses of the MegaCell Charger over this

+Minor setup required. Connect power. Network setup.
++Functional software. It is good until you hit one of the undocumented features.

This over the MCC.
+Cost. You will need to add cell holders, tp5100 chargers, power resistors.
+Open design.
?Accuracy. Voltage and thermal accuracy.
+PCB is not being used as a heatsink.
+Separation of cells should limit thermal transfer of a heater heating up the cell next to it.

Potential accuracy as I do not know what is in the MCC.
Or how much electrical noise my layout will create.
I have tried to keep the low voltage adc signal traces from running in parallel with the high current traces.

Screenshots from this video for close ups of MCC pcb.

Order placed.
ETA in my hands is probably Friday 5th Feb?
Came to roughly us$38 per board delivered. 5x boards $189.

A few small changes on the fly during routing.
Not as tidy as I wanted but should be good enough to test.
I have not done a ground pour to fill in the spaces to assist with initial troubleshooting.
Hopefully I have caught any major bugs.

If anyone is keen on a set to help with coding, I am willing to cover some of the PCB production costs.
Shipping any spares from down here (New Zealand) is going cost too much.

 

[hpeyerl]

I'll dive into some coding. I don't need subsidization. I also don't need to be the only person doing the coding in case anyone else also wants to jump in.

 

[nz_lifer]

Files you will need to order from JLCPCB.COM
Attached zip.
I believe if you sign up as a new user you get $20 off?

Step by step order process attached.
Looks like a part went out of stock between when I placed my order and making this walk through.
Extra step substituting part included.

 

[nz_lifer]

Need to give this device name to refer to while developing.
How about the SLS16M?

 

[hpeyerl]

Need to give this device name to refer to while developing.
How about the SLS16M?

Apparently you need help with coding _and_ naming.

I'll think on this very urgent task.

Chargie McCharger?
The Chargurator?

Shall I wait to order until you've built one and put electrons through it?

 

[nz_lifer]

Project creep already. I know right?

I like your line of thinking.
Since does more than charging, full testing.

Tester Mc Testie? TMT16
Testanator 16?


I would wait till I get my set in hand to catch any bugs I missed.

 

[Eric44]

"Ultimate Charger" or "18650 TMT16"

 

[hpeyerl]

Testease!

hmm. perhaps not.

 

[Bubba]

Could call it the CMB16 for Charge, Measure, Balance
Since it's the only unit I have seen that balances holder to holder. Wait, it has that feature..yes?

 

[nz_lifer]

Could call it the CMB16 for Charge, Measure, Balance
Since it's the only unit I have seen that balances holder to holder. Wait, it has that feature..yes?

Balance, technically, yes.
When it switches to float mode each bank will effectively become a 16p pack.

I don't know how accurately we will be able to measure current though.
We will have to watch our float termination currents too if we are cashing that <5% accuracy.

Test of Battery Charge Termination Test

 

[nz_lifer]

Production complete. Waiting on shipping now.