Prototype: charge, discharge (16) 18650 or 26650 or mixed.

[J_Mack58]

It’s been in my mind for awhile charge, discharge, display capacity top balance and tackle internal resistance another day. I got Arduino programming down… Python on a Raspberry pie no problem. Allen Bradley Micologix PLC with panelview would be high end… I got the parts. Automation Direct Micro PLC could be good and cheap. A journey of a thousand miles begin with one step…..Something to hold the freakin cells. Finally I got something.

 

[Oleksii]

Long wires - not good. Iron contacts - not good.
Good luck to you

 

[J_Mack58]

Long wires - not good. Iron contacts - not good.
Good luck to you

Let me see how the circuitry and programming go then I will invest in copper. Wires were meant to be cut.

 

[Oleksii]

More you work with current measurement, more you will love thick COPPER wires
Share details more often, so I could give some advises.
Btw, did you evaluate all other existing projects to consider them as a start point?

 

[J_Mack58]

More you work with current measurement, more you will love thick COPPER wires
Share details more often, so I could give some advises.
Btw, did you evaluate all other existing projects to consider them as a start point?

I did not consider any other projects. the beauty of me learning is me messing up and doing things wrong. Plus my project is based
on the parts I have collected to try to build this 16 cell beast. But your advice is welcome, matter of fact I got a question. I was planning on discharging the cell through a 1 ohm 20 watt, I see we have a drawer full at work, sound like a good ideal? Also I have this 5 volt @7amp power supply if I’m charging and monitoring can I hit the cell with 5 volts and cut charging at 3.6 volts?

 

[J_Mack58]

Most of us here are heavy into Electronics or Electrical and programming. Although I can Python on a Raspberry Pie or program a Arduino my strong skill set is programming a Programmable Logic Controller (PLC). They are very precise and they have to be for robot control of car assemble and gas refinery…two places I have worked. So with that said I started my 16 cell charger/discharger using a PLC. We have many old Allen Bradley Micrologix and SLC 500’s awaiting e-waste because my plant use the Allen Bradley Contrologix 5000. So I took a 7 slot PLC.
Slot 1 is for the CPU. Slot 2-5 is 4 analog cards with 4 Channels each. This is how I read my cell voltages. They can read 0 to 10 volts with .00001 accuracy. Slot 6 is 16 channel relay output module. Through these relays I delivery 5 volts to each individual cell. Yes I have a 5 volt 16 amp power supply ready to deliver the juice because the goal is to really charge all cells at 1000 mA. I’m not sure what raw voltage a commercial charger is putting to a cell but I figured 5 volts would be okay. The last slot is a 16 channel output card that switch relays that put 3 ohm resistor per cell to ground to discharge the cells. To be honest I chose the wrong wattage for my 3 ohm resistor but I think the relay is okay. With the hardware all wired up programming is easy. It’s ladder logic. Now get this, we have human machine Interfaces (HMI) that we getting rid of. I will program one to give all the bells and whistles one can imagine.

 

[J_Mack58]

This power supply delivers what it says. The main wire will be 10 AWG. When it split to 16 cells it will be the size you see in the picture. That wire did handle the 16 amps but it did get warm. I knew better just took the power supply for a quick spin 0 to 16 amps in 1 mS.

 

[Hanssing]

I like it, reusing old discarded AB-PLC's. But Ladder yuck
I did not go into details on the AI-card, but are you sure it is not just a 12bit? Most of the old ones from back then was.
10 volts, resolution 0.00001 => 1000000 => ln2(1000000) = 19.9, so more or less 20bits ADC-resolution. Don't think thats right, but it would correspond to the mantissa of IEEE floating point real32, which you get out of the scale-calculation.

Edit: I just saw your picture, it looks like the scaling is 32767, so that would be 15bits resolution => 10Volt/2^15 = 0.0003 volt/adc-bit.
Thats also what the block shows with ADC-input of 1 <=> 0.0003 Volt


Must admit I've been toying with the same idea, but I would use either Beckhoff or CoDeSys on an RPI for house-control and automation.

 

[J_Mack58]

I like it, reusing old discarded AB-PLC's. But Ladder yuck
I did not go into details on the AI-card, but are you sure it is not just a 12bit? Most of the old ones from back then was.
10 volts, resolution 0.00001 => 1000000 => ln2(1000000) = 19.9, so more or less 20bits ADC-resolution. Don't think thats right, but it would correspond to the mantissa of IEEE floating point real32, which you get out of the scale-calculation.

Edit: I just saw your picture, it looks like the scaling is 32767, so that would be 15bits resolution => 10Volt/2^15 = 0.0003 volt/adc-bit.
Thats also what the block shows with ADC-input of 1 <=> 0.0003 Volt


Must admit I've been toying with the same idea, but I would use either Beckhoff or CoDeSys on an RPI for house-control and automation.

LOL @ “Ladder yuck“ . Man I know but I been doing Ladder since AB PLC-2 on a teach terminal…then DOS 3.0. . We have Beckhoff on our vision systems at work, they not throwing them out anytime soon.