Sid Battery Balancers.

My first experience in computers was my dad's TRS 80 Model 1. When I was 3 years old he got one that was fully decked out for around 700 bucks. I remember that number because my mom said they could make 1 and a half house payments. At the time I think the model 2 or 3 was out but he wanted a fully loaded model 1 for some reason.

I remember loading cassette tapes and running games on it. I was so confused that I put my teddy rupskin music tape in it and couldn't figure out why it wouldn't play. Good memories.  Then around 1986 he sold it because we ended up getting a Nintendo and he seemed to enjoy that more.

We wouldn't get another computer till 1992. We got a lucky hell of a deal because Michigan State University was doing a sell off of computer equipment.  We found out through my uncle who was a student at the time and MSU forgot to put the add in the paper for it. We pretty much had the run of thr place with a few other people who knew. Keep in mind students rarely showed to these things because they couldn't afford the gear.

We picked up a 486dx2-66 which we later found out was actually stolen from a chemist lab that another studen and staff member had worked to buy real cheap. It was labeled as a 386SX33 for 300 dollars. 

The computer had 32MB of ram, an Adaptec (2940 or 1 gen older) card and 4 Seagate drives running in raid 0. It was a super micro motherboard. Need less to say it was loaded. My uncle later found out about thr planned theft and we contacted the university and offered to return it but they never got back to us. So we kept it.

Thing is it didn't come with a power cable, keyboard, and monitor. So over the next month we added parts to get it running. It had a beautiful viewsonic 17 monitor which was cutting edge at the time.

It had a turbo button which my dad spent 2 hours getting the pinouts right to read 66 when the turbo button was pushed and 8 when it wasn't pushed. Thing is the turbo button didn't do anything. But we didn't care. 

Many years we had that computer but computers were getting so fast that it was practically useless after just a few years. Eventually we then upgraded to a pentium 133.

So there is my early story. 

Oy, Mine was useing a commador 64, then a Adam Computer, then a Packer Hell, then we modded the packard hell to its max memory I think it was 32meg had a 1g hard drive, two video cards voodoo, a cpu that was modded to fit in the socket was a amd K6, a evergreen brand 😛  running win 95.

Oh those were the good old days, and not to mention had a system with dual pent 3 slot 1 water cooled from a 55gal tank with two video cards  water cooled as well, 

On 2/16/2021 at 12:50 PM, Sid Genetry Solar said:

Hey, I started out on a pocket RadioShack TRS-80 handheld computer--saving programs on a cassette tape.  Then to an Apple //e with the dual disk drive (and the 64kb expansion card).  Never managed to get the 3.5" floppy drive working, and never bothered with the cheekily named Apple Sider (hard drive).  After the //e went kaput (literally instantly--one moment it was working, then suddenly the monitor went blank as the computer died on the spot), my programming endeavours went to the Windoze systems.

Been through the Basic Stamp (2P40 to be specific), wore the memory out with so many write cycles, and found that it was a bit too slow.  From there found a PIC16F872 amongst multiple boxes of stuff we'd gotten at a yardsale (longtime tinkerer passed away)...so yes, I am very much aware of "tiny" memory systems.  Technically I'm a millennial...so by the time I got the aforementioned stuff to play with, it was 'hand-me-downs' from Grandpa.

Hey, if you want the complete source code to "Brick-Out" written by Bruce Tognazzini in 1981(?) for Applesoft Basic...umm...I have it ;-).  Printed it out with an Apple ImageWriter about 18 years ago.

My comment about 2MB of memory was in reference to the question asking about whether I was using a Linux OS--nope, not enough memory for that.

If you had that, you were a late comer. 🙂 The computer I was taking about was already gone by the time that unit came out. It was a Heathkit H-8. It was replaced by an Osborne 1. We still have it. The next one as a Tandy 1000 SX. It had a big hard drive. All of 20 MB. That drive was replaced with a 40 MB WD unit that I still have. It was replaced by a 340MB HDD. Then I bought my first real one for myself ( already had an Atari 800 ), a US Logic 486DX2 ( Computer City  brand ) with a 340MB HDD and an IBM G-20 CRT monitor.  The HDD soon needed a twin.  I had learned computer programing in 1973 on an IBM360/370. Dad had taken an Army computer course in 1966 for his work in the Navy. Tiny Core Linux is about as small as you can get and it is 11MB. Whereas CP/M ran on 64KB. Funny thing about the Z-80 chip used in the H-8, it also was used in one of my HP printers.

Anyways, that is off the beaten path. Any Idea when I can buy 8 of the balancers?

Sean has all of the remaining prototype balancer units.  If they trip due to overload/overheat, they will not auto reset--you have to discharge the battery below the balance threshold voltage.  There's also a few minor code bugs (i.e. display indicating "BAL" while they're actually still in idle mode), however these do not affect the actual functionality.

I still haven't ordered a production run of balancers...the 1st shipment of GS inverters is hopefully going to be delivered in a few days, and I have to get firmware v1.1 ready...with all the features that customers are going to expect.  Lots of stuff on my plate, and I can only do so much.

19 minutes ago, Sid Genetry Solar said:

Sean has all of the remaining prototype balancer units.  If they trip due to overload/overheat, they will not auto reset--you have to discharge the battery below the balance threshold voltage.  There's also a few minor code bugs (i.e. display indicating "BAL" while they're actually still in idle mode), however these do not affect the actual functionality.

I still haven't ordered a production run of balancers...the 1st shipment of GS inverters is hopefully going to be delivered in a few days, and I have to get firmware v1.1 ready...with all the features that customers are going to expect.  Lots of stuff on my plate, and I can only do so much.

https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcSG6WDT853kRRXm0tvcBPL0TpSlG2iPWW48pg&usqp=CAU

🤣 Just poking at you.

Ask Sean, I don't know if he's sold any.  'Course it's after business hours now...oh, and let's not forget that I forgot to ship him the (required) mica insulators...nor that we do not have source for heatsinks.  Very crude start, but that's what we have.

Well, I've try calling Sean to see about getting at least 7 of them.  Its a no go for me.  I need them to do the 18650 so I can use the two packs in my system, otherwise I'm not useing them until I have them balance correctly.

Ordered a small production run of 400pcs "fancy" balance shunts with the MCU + LCD + serial comm functionality.  Let's say 2 weeks before I get 'em.

Haven't looked to source heatsinks yet, but most people making their own batteries can come up with something 😉.  Will provide some mica insulators with 'em (though no heatsink compound yet...so many things to find suppliers for...!)

So far it appears that that the price will be ~$12/ea (for single-cell unit)...for a 16S system, you'll need 16 units.

Good to hear they are ordered.  Thanks

I want 16 as soon as you get them.  I sent a message to Sean here on the forum.  Guess that was a poor choice as he hasn't read it.  Sent one to Ben (about transformer winding) also. same results.  Notification of messages received must not be turned on here.

Anyway, I'm hoping to remove the heatsinks from the byd battery modules (if they ever get here!).  One or two of those large heatsinks should work well for mounting balancers with fans.  I have some others if that doesn't work out, so one way or another...

Currently not charging above 54.4v (3.4v/cell) but figure after balancers installed and settled in, I should be able to bump the charging setpoint up gradually, letting the balancers smooth cells out.  That's assuming they will stay in balance once I get them there.  They aren't new modules after all...

Might have to stick with 54.4v, but would like to get to 56v.  Guess we'll see.

I'll buzz Ben and let him know.  Oh, and Sean too 😉

Yeah, 3.4vpc is kinda low...barely over the nominal 3.3v voltage, so it will be extremely difficult to get any charge into the cells--not to mention a very low charge current.  With the balancers installed, if you're connected to solar (i.e. >5A), gently bumping the charging setpoint up is a good idea.  You should easily get to 56v; the balancers are pre-configured for 3.55v (+/-0.02v, thanks a lot resistor tolerances!)  I run my MPPT "absorp" setpoint at 56.5v, and the "float" setpoint at 56.0v.  (Much more voltage difference results in the batteries discharging for several minutes to the lower voltage.)

LiFePo4 seem to hold balance quite well once initially balanced.  Its Li-Ion that seems to be a bit more of a difficult creature.

16 hours ago, Sid Genetry Solar said:

Ordered a small production run of 400pcs "fancy" balance shunts with the MCU + LCD + serial comm functionality.  Let's say 2 weeks before I get 'em.

Haven't looked to source heatsinks yet, but most people making their own batteries can come up with something 😉.  Will provide some mica insulators with 'em (though no heatsink compound yet...so many things to find suppliers for...!)

So far it appears that that the price will be ~$12/ea (for single-cell unit)...for a 16S system, you'll need 16 units.

So that means I'll need 30 to start with. 28 to use and 2 to blow up. And I have a giant heatsink to use. Like 24 inches long by 12 inches wide.

So how much surface is from everything for the heat sink to connect?  I thought about buyings some olf fashing iron iron :P.  Luckly I have a drill press now, haha.  once I get the balancers I'll see what I need, but I'll look for a real iron flat bar of some kind, that will absorb the heat, if I can find one that is a grill style that might work better.

I'll have to look back and see what size wire is needed for those, if I could get by eith a thin enough wire I could use a solderless bread board and be able to use a switch to switch between banks  or better yet, use a one way diod per each cell and balancer and can have four cells per one balancer 😛  Just kidding.  that would cause a over amprage problem. lol.

oh well, I'll just buy a bunch of them. since I'm buidling three more 4s 30 p batts right now.  New cells this time.

  Dang, I need 40 of them as it stand right now.  geesh.  Thats$480   I might just buy in spirts then lol.

As long as the cells are in parallel (same voltage, same chemistry, same negative, same voltage packs, etc.) you can easily parallel the balance wires and use a single set of balancers for the whole pack.  I have 44 cells in parallel on a single balancer on my system.  242Ah of capacity for the balancers...no problem.

The thinner the wire...the more voltage drop you will get, and the higher the cells will be able to go while the balancer is working.  I'd advise at least 16AWG, and keep it short as reasonably possible.

Considering that they can easily put out 20W of heat EACH (20W being the size of a small soldering iron), this is nothing small.  Worth noting that both people that I've given prelim balancers to immediately went for the cheapest little dinky heatsinks they could find on Amazon.  Those couldn't keep a fly cool in the rain.  The bigger (and more surface area) the better.

Obviously, once the cells are balanced, the balancers should not get very hot in operation.  For LFP (LiFePo4) it's the first few initial charge cycles that generate a lot of heat.  Seems Li-Ion never really fully balances out, and will run the balancers a good bit harder.

That heatsink should be big enough ;-).

As long as you keep the voltage across the balancer under 6v, they shouldn't fail.  I've "cooked" them without a heatsink, and the thermal protection trips pretty quick.  (Not saying they can't fail, just we haven't had any fail...yet.)

(If the main transistor fails short, the screen will flash "FAIL"; it'll be a constant sink of ~100mA.)

16awg, thats pretty thin.  Well the cells I'm putting together right now are brand new 2650 for 120 of them for $228 isn't bad. and they small enough to fit in a ammo box with enough room for the bms, sweet. 114Ah for $228 at 12v. .  comes out at $165 kw.  I probley won't need a balancer on these if I just build the cell banks and connect them and then charge, they should be ready to just go.

Still waiting on the keyfog Sid. it's not came in yet.