diy solar

diy solar

Best practice for connecting BMS sense leads?

Yes and the lines are separated into busbars and sense wires. The question is where do you want to measure voltage? Given that the first negative and last positive do not have any interconnects to other terminals, where is the best spot to get a consistent voltage measurement for each cell in the pack without introducing another variable?
I put mine somewhat close to the terminal, off to the side. At this distance, I think the internal accuracy of the BMS is the largest limiting factor.
 
I put mine somewhat close to the terminal, off to the side. At this distance, I think the internal accuracy of the BMS is the largest limiting factor.
That may be the answer. It would not be that much more work when I make up busbars for my future 3P16S pack since every 3 cell terminals will be connected by that parallel connection. That way my torque of the nuts will all be through only a Nord Lock washer without introducing a ring terminal into that equation.
 
That may be the answer. It would not be that much more work when I make up busbars for my future 3P16S pack since every 3 cell terminals will be connected by that parallel connection. That way my torque of the nuts will all be through only a Nord Lock washer without introducing a ring terminal into that equation.
Exactly.
 
From a Linear Technologies data sheet - this is typical of all BMS designs that I have studied so far.
In English - it is simply saying that when measureing a cell voltage, it automatically turns off balancing just before to get an accurate reading. It then turns the balancing back on.....all of this in a very small slice of time. Because of this arrangment, the length of the balance leade (different resistances) is really not an important thing. When zero current is flowing, the length of the wire is not a factor.

1610160030165.png
 
Egads, stainless steel bolts? Stainless steel has a resistivity of about 76, more than 40 times that of copper. OK, place all the ring terminals you want on top of that poor connection.

I don't know the conditions under which you measured, but SS studs patently do not affect my battery. In fact, I have an SS bolt, an SS washer, the BMS ring, and another, huge, SS washer, on top. I've run 220 amps through the EVE cells and the bus bars (2 each of 15x2mm) have less than 10 degree F rise after 1 hour. It is my firm belief that the bolt/stud&nut do not significantly affect current-carrying capabilities. The best thing one can do to minimize resistance is properly clean the terminals and bus bars just prior to installation.

I'm using an Orion BMS (very high accuracy) and a Klein 3-decimal-place meter to measure everything and everything is gold.

Edit: To be clear, the Orion and the Meter (measuring at the bus bar) agree with each other. In other words, it's not affecting battery resistance or BMS measurement to any meaningful degree.
 
Last edited:
Because of this arrangment, the length of the balance leade (different resistances) is really not an important thing. When zero current is flowing, the length of the wire is not a factor.
I have noticed that the leads in Commercial packs like Chevy Volt, Nissan Leaf and Tesla modules are of varying lengths so my preference is an artifact of the constant reconfiguring of my packs and never wanting to have to have a lead that is too short.
 
I soldered my BMS wires to the ring terminals. I also had to lengthen some of the BMS wires so they would reach and soldered those connections and put heat shrink over them. Busbar first, then ring terminal then nut. Obviously putting the ring terminal or anything else under the busbar (except the cells terminals) is a no no and conflicts with the busbars purpose in this application. I am using 20mm stainless steel grub screws and stainless steel serrated flange nuts.

I need to reconstruct my pack and when I do I plan to solder the BMS wires directly to the busbars. Once my pack is finalized I do not intend to ever take it apart unless I have a defective cell. That's very unlikely to happen, but I will have access to the terminals in case I feel the need to add an active balancer later on. I have already removed all the shrink wrap from the busbars. I have heat shrink to put around the busbars and BMS wires when I redo my pack. The point made about putting the ring terminal under the nut is a good one. I am not sure if doing so effects torque readings and I don't like that is spins under the nuts when the nuts are torqued down.

EDIT: My pack will be stationary.
 
My reasons for preferring a solder ring to a crimp ring are two-fold:

1/ solder is a reliable connection electrically - no need to ‘ohm them out’ or screwing around if they don’t match.
My experience in high volume electronics manufacturing says poor solder joints are the number one cause of problems.
I have literally seen hundreds of thousands of them. This made by people and machines who only had one job, solder it properly.
Make sure you know how to solder, that includes cleaning off the flux you used. (you did use flux, right?)
Not cleaning (or using) flux is a novice mistake, one that particularly insidious for a sense circuit.
There is reason manufacturers use crimped connections, and it isn't price (crimp connectors are extra cost).
You really don't ohm out a connection you've soldered? Wow.

From my experience, some of the youtube "how to solder this connector" crowd make me cringe. Crimping is easier, and more reliable for a novice.
2/ for a longer-lasting connection when wires will be getting moved around, I agree crimp is superior to solder, but that’s really not the use case for these sense wires.

So if there is any downside to solder other than that the connection won’t last as long if the wires move or get tugged-on (and than many consider crimping easier/more convenient than soldering), I’d go for a crimped terminal, but it it’s only that, I prefer solder...
If you're proficient with a soldering iron, I'd prefer it too. It does however still require the correct tools and the knowledge how to use them. Soldering a tiny wire to a big heatsink like a busbar is not what you should be learning on.
 
Personally I have quite a bit of soldering experience. My first soldering iron was one of those big Weller's. I think it was 60/100 watts. Too big for sure but it got me started playing with electronics. That was back around 1969 or 1970.

It is important to use a high quality solder. I am using Kester 44. It's my understanding removal of flux isn't necessary unless soldering on high impedance PCB's. We are just soldering a tiny tinned wire to a ring terminal or busbar. I don't see any problems with that.

If one is inexperienced soldering and wants to learn, then a little practice will do. Just have to make sure the solder flows and is bright and shiny when cooled off.
 
I don't know the conditions under which you measured, but SS studs patently do not affect my battery. In fact, I have an SS bolt, an SS washer, the BMS ring, and another, huge, SS washer, on top. I've run 220 amps through the EVE cells and the bus bars (2 each of 15x2mm) have less than 10 degree F rise after 1 hour.
Measuring temperature rise is a poor criteria to differentiate good electrical connections. You need to measure voltage drops. Measure battery voltage (directly at the terminals) with no load and then with your 220A load. Now you can calculate battery resistance:

R = V_noLoad - V_load) / 220A

Remove sense wire lugs for this test. Install polished AL bolts (no lock washers). Repeat measurements.
 
Measuring temperature rise is a poor criteria to differentiate good electrical connections. You need to measure voltage drops. Measure battery voltage (directly at the terminals) with no load and then with your 220A load. Now you can calculate battery resistance:

R = V_noLoad - V_load) / 220A

Remove sense wire lugs for this test. Install polished AL bolts (no lock washers). Repeat measurements.

Yes. I had two points -- 1) current flowing through the bolt isn't necessary to safely retrieve 1C from these cells; 2) I've not observed a significant difference between BMS values and my DMM. I have measured voltage drop between many points. It's been so long that I don't remember the values. I'll do some testing (as much as I can - I don't have AL bolts - but I can measure differences between terminal face and various components.
 
I use a 16 ton hydraulic crimper on larger wires and Klein 1005 hand crimper on smaller terminals. I use Marine grade adhesive lined heat shrink tubing on all connections. Never had an issue.
 
Personally I have quite a bit of soldering experience. My first soldering iron was one of those big Weller's. I think it was 60/100 watts. Too big for sure but it got me started playing with electronics. That was back around 1969 or 1970.
I personally have seen a few million boards of various types, and diagnosed their problems. Seriously, I worked in the test engineering lab where (starting in 1981) we manufactured 30,000 boards per week (the original IBM PC) for the 5 years I was there. We were a contract company, i.e. we manufactured for lots of companies, Compaq, HP, Apollo, Adaptec, etc. Everything from power supplies to memory boards, modems, printer boards, etc. Over 5 years time, literally millions of boards. Bad solder joints are the number one problem, we had QC people whose ONLY job was to inspect the soldering. Even people who have done it for years as a living make mistakes. Just trying to impart my experience (since it is very relevant, and very extensive). I've run test labs for military avionics, NASA, Motorola, etc. Last 20 years I've been in software engineering. But soldering hasn't changed much in the past 100 years or so.
It is important to use a high quality solder. I am using Kester 44. It's my understanding removal of flux isn't necessary unless soldering on high impedance PCB's. We are just soldering a tiny tinned wire to a ring terminal or busbar. I don't see any problems with that.
Kester 44 is good, it's the "gold standard" and of course has rosin core, and is NOT lead free. Weller is great, but to solder onto a busbar, you are going to need the correct tool, a 20 watt iron designed and sized for a PCB won't work, you need something larger (like for copper pipes). Certainly soldering guns aren't expensive, but they also might cause problems with 22 gauge wire like BMS sense leads. Flux is NOT a good conductor, impedance and resistance are two different things. You are wanting to sense voltage in the millivolt range, not run high frequency signals through it or carry large currents. A good quality crimp will have just as low a resistance, has less of a learning curve, and stands a much better chance of success for people without extensive experience. Be sure to prep the busbar with a dab of flux (DO NOT use anything but rosin based flux for electrical connections), heat to temperature, add solder, then introduce the (already tinned) wire to the solder pool. Be sure to clean off the flux when done. Or, buy a crimp tool (you'll need one anyway) and the correct crimp connectors and wire strippers (you'll need them anyway as well). If your ohm meter can't tell the difference between a crimp connection and a soldered connection, why bother with solder (again, soldering does take experience). One or two crimp connections on some test wire, and you are good to go. Soldering to bus bars on the other hand is like learning on one of the most difficult solder joints possible. Speaking from experience, for an untrained person, crimping has a much higher success rate. We actually analyzed numbers and defects.

Always be sure and ohm out the connection, end to end. Someone scoffed and said it wasn't necessary on a soldered connector, decidedly an incorrect assumption. Great way to blow out your BMS.

If one is inexperienced soldering and wants to learn, then a little practice will do. Just have to make sure the solder flows and is bright and shiny when cooled off.
Seriously, this is like picking an Indy race car to learn to drive.

Someone (maybe you) was talking about solder ring terminals, that's a different story. Much easier, just make sure to clean the flux off completely. Then again, you won't get better performance from that or more accurate measurements, it'll just be the same as a good crimp.
Since this is a DIY solar forum, I'm speaking to people with little to no experience. A crimp connection is every bit as good, and much more likely to be done correctly if you are inexperienced. I understand the desire to solder, I really do, but the end results when done properly are identical, and one just has a much higher rate of success.

Obviously, it's your system and you are more than welcome to construct it as you wish. For the average DIY, crimping is a much better option and works just as well. The topic however is best practice for connecting your sense leads. The most likely to succeed is crimping using the correct tools, I'm not talking a pair of pliers and a steak knife.
 
I've not observed a significant difference between BMS values and my DMM
I would expect them to be the same if the devices are calibrated. Both are presumably measuring voltage at the top of the busbar. When I have probed the SS stud I think I have seen a subtle difference but that is not a scientific observation.
 
My experience in high volume electronics manufacturing says poor solder joints are the number one cause of problems.
I have literally seen hundreds of thousands of them. This made by people and machines who only had one job, solder it properly.
Make sure you know how to solder, that includes cleaning off the flux you used. (you did use flux, right?)
Not cleaning (or using) flux is a novice mistake, one that particularly insidious for a sense circuit.
There is reason manufacturers use crimped connections, and it isn't price (crimp connectors are extra cost).
You really don't ohm out a connection you've soldered? Wow.

From my experience, some of the youtube "how to solder this connector" crowd make me cringe. Crimping is easier, and more reliable for a novice.

If you're proficient with a soldering iron, I'd prefer it too. It does however still require the correct tools and the knowledge how to use them. Soldering a tiny wire to a big heatsink like a busbar is not what you should be learning on.
No.

I first clean and flux the busbar and melt a nice 1/4” diameter dab of solder onto it (I have a first powerful Iron for this). Then after cleaning and fluxing both the solder dab and the tinned end of the wire, I use my precision iron to melt some solder onto the tinned end of the wire then lightly press the molten solder and the wire tip onto the surface of the solder dab on the busbar until they fuse. It’s worked well for me so far...
 
No.

I first clean and flux the busbar and melt a nice 1/4” diameter dab of solder onto it (I have a first powerful Iron for this). Then after cleaning and fluxing both the solder dab and the tinned end of the wire, I use my precision iron to melt some solder onto the tinned end of the wire then lightly press the molten solder and the wire tip onto the surface of the solder dab on the busbar until they fuse. It’s worked well for me so far...
Great way to get a cold solder joint. Your mileage may vary. Using a precision soldering iron on a busbar (a large heatsink), you sooner or later will get a cold solder joint.
 
Great way to get a cold solder joint. Your mileage may vary. Using a precision soldering iron on a busbar (a large heatsink), you sooner or later will get a cold solder joint.
I do this sort of stuff with my precision JBC station regularly - no issues. I use the medium flat tip, not the ultra-fine needle tip.
Still, it is precision application of thermal energy.

Side note, I also use no-clean tacky flux from ChipQuick to avoid a lot of the cleaning requirements. Works extremely well, even on the extremely sensitive 20Ghz+ PCB work.
 
One of my objectives when soldering is to always have some sort of mechanical connection first. It can be a simple twist or bend sometimes.
If I were going to solder onto a bus bar, i would probably drill a small hold about the size of the wire first ...insert the wire and make an bend it back against the bus bar on the other side .... then solder.
 
I do this sort of stuff with my precision JBC station regularly - no issues. I use the medium flat tip, not the ultra-fine needle tip.
Still, it is precision application of thermal energy.

Side note, I also use no-clean tacky flux from ChipQuick to avoid a lot of the cleaning requirements. Works extremely well, even on the extremely sensitive 20Ghz+ PCB work.
Trust me, I'm just as anal retentive about my connections as the next guy. I fully understand the desire to solder. But I am just pointing out that a crimp connection in this application is every bit as accurate, and much easier to accomplish. I have a very nice Weller soldering station myself, a Fluke 87, a Tektronix oscilloscope, etc. I just think it is wrong to solder as a "best practice", the manufacturer recommends crimping for a reason. Neither type of connection will get you more accurate results when done properly.
 
Back
Top