diy solar

diy solar

Improving Contact Area on Welded Stud Pads

Anyone doing continuous monitoring of resistance across battery pack to provide warning system of terminal conductivity degradation?

for example, measure resistance of pack including all cells and busbar at 100% and compare from cycle to cycle. if pack goes from 20mOhm at 100% to being 46mOhm at 100% then operator can confidently say either cell or busbar resistance has increased.

i write firmware for arduino and it’s fun, wondering if anyone else using AC resistance measurement principle
 
Anyone doing continuous monitoring of resistance across battery pack to provide warning system of terminal conductivity degradation?

for example, measure resistance of pack including all cells and busbar at 100% and compare from cycle to cycle. if pack goes from 20mOhm at 100% to being 46mOhm at 100% then operator can confidently say either cell or busbar resistance has increased.

i write firmware for arduino and it’s fun, wondering if anyone else using AC resistance measurement principle
Interesting idea. My approach when I get my pack up and running is there will be a temperature sensor on each terminal logged through an RPI. I don't know how effective it will be in assessing small changes as the heat may be sinked too quickly, but was implementing it more to monitor for bigger problems.
 
as you say, monitoring temperature of terminal and/or cell is easier. most reliable too i suppose

either way if the terminal is 60C maybe it’s time to stop delivering power lol

thanks for your insight
 
Interesting idea. My approach when I get my pack up and running is there will be a temperature sensor on each terminal logged through an RPI. I don't know how effective it will be in assessing small changes as the heat may be sinked too quickly, but was implementing it more to monitor for bigger problems.
may i ask the model name of temperature sensor you are using?
 
as you say, monitoring temperature of terminal and/or cell is easier. most reliable too i suppose

either way if the terminal is 60C maybe it’s time to stop delivering power lol

thanks for your insight
My only concern is that copper can carry a lot of heat away, but hopefully the delta would stand out against the other terminals. Is a fun add on to the project that I'll fully delve into the raspberry pi monitoring side when winter comes around. The sensors are made though, glued into appropriately sized tube lugs with thermal glue.
 
My only concern is that copper can carry a lot of heat away, but hopefully the delta would stand out against the other terminals. Is a fun add on to the project that I'll fully delve into the raspberry pi monitoring side when winter comes around. The sensors are made though, glued into appropriately sized tube lugs with thermal glue.
sounds like you have a very robust solution already regarding temperature sensor installation ?

cool idea using a 1-wire protocol thermometer!
 
My only concern is that copper can carry a lot of heat away, but hopefully the delta would stand out against the other terminals. Is a fun add on to the project that I'll fully delve into the raspberry pi monitoring side when winter comes around. The sensors are made though, glued into appropriately sized tube lugs with thermal glue.
I'm spoiled.

I bought a flir IR camera/multimeter recently and it's amazing for this sort of thing.
 
I'm spoiled.

I bought a flir IR camera/multimeter recently and it's amazing for this sort of thing.

Now you'll have to do image processing software to automate it.

I figured a series string of PTC fuses - if any one gets hot it trips the relay. Thermistors scanned by analog mux if you want to log.

I have done primitive image processing software (scan sequence of frames for a spot of laser light to detecting timing pulse), but to me that's a PITA. KISS.

I did consider FLIR clamp meter with IR camera, but I think it didn't even do DC amps.
 
Thermistors scanned by analog mux if you want to log.
For future experiments, this is the path that I plan on taking.

thermristor + analog mux + 16-bit ADC

B3950 NTC thermristor [adafruit]​

1630040024124.png

CD74HC4067 analog multiplexer [sparkfun]​

1630040185473.png

ADS1115 16-Bit ADC [adafruit]​

1630040097387.png

cost of materials..

(0.95usd * 4 thermristor) + (5.5usd * 1 mux) + (15usd * 1 ADC) = 24.3 usd 4S pack

(0.95usd * 8 thermristor) + (5.5usd * 1 mux) + (15usd * 1 ADC) = 28.1 usd 8S pack

(0.95usd * 16 thermristor) + (5.5usd * 1 mux) + (15usd * 1 ADC) = 35.7 usd16S pack
 
For a small number of sensors (not exceeding the selectable I2C addresses), you can get a calibrated part for $0.70


I think it allowed 8 addresses. After that you'd need a mux.
I used this one, and one issue was it failed to read address pins reliably on power up, so responded to wrong address.
Workaround was to issue I2C "General Call Reset", which caused it to read address again.
 
I made some washers for Michael to sell in his group buy with cells.


I laser cut them out of 1000 series Annealed Aluminum (absolutely dead soft, very conductive). They made a perfect fit and increased surface area to 3/4". But everybody thought it wasn't that great of an idea, and not too many were sold.
I do have a box of a couple hundred of each size (6mm and 8mm studs). If anyone is interested, shoot me a PM.20210902_154310.jpg
 
trying to evaluate welded studs M6, M8, and threaded holes. I realize the M8 would have the better clamping force. Any thoughts on whether the smaller M6 stud would actually have larger surface area for buss bar to lug contact over the larger M8 stud? Is the stud base the same size on both M6 and M8 with just the studs being different diameters? Is greater clamping force better then a greater surface area? Is it a wash?
 
trying to evaluate welded studs M6, M8, and threaded holes. I realize the M8 would have the better clamping force. Any thoughts on whether the smaller M6 stud would actually have larger surface area for buss bar to lug contact over the larger M8 stud? Is the stud base the same size on both M6 and M8 with just the studs being different diameters? Is greater clamping force better then a greater surface area? Is it a wash?

Regardless of how it's attached, the aluminum holding the stud in place cannot handle the max torque of an M6 stainless steel stud.

The M8 torque capacity is well above what the aluminum can handle, and I believe is overkill. You'll never use the tightening capacity of the M8 in this application - if you need that much pressure, then you need to increase the pad and use two or more studs per contact.
 
trying to evaluate welded studs M6, M8, and threaded holes. I realize the M8 would have the better clamping force. Any thoughts on whether the smaller M6 stud would actually have larger surface area for buss bar to lug contact over the larger M8 stud? Is the stud base the same size on both M6 and M8 with just the studs being different diameters? Is greater clamping force better then a greater surface area? Is it a wash?

I think the pressure which can be achieved with M6 stud and slightly larger contact area is lower than optimum, before the knee of the curve <contact resistance>/<pressure>
I think the larger M8 stud, with more torque and more force over smaller contact area is better.

The resistance to rotation provided by aluminum terminal in cell is a concern, as Stienman mentions. But I think if torque is applied by holding busbar to prevent rotation (or holding sandwiched pair of cells with busbar between them), rather than holding a single cell with busbar and nut, that would take the torque.

I might also consider a drop of Loctite on the thread first - that would lubricate between nut and thread, resulting in higher clamping force for a given torque. When working near the pull-out (or stripping) force of a bolt one must be careful to use correct torque for lubricated vs. dry (e.g. you don't lubricate lug nuts.) But for these cell terminal studs I think the torque used is well below that limit (you could look up torques of stainless bolts to compare.)
 
Regardless of how it's attached, the aluminum holding the stud in place cannot handle the max torque of an M6 stainless steel stud.

The M8 torque capacity is well above what the aluminum can handle, and I believe is overkill. You'll never use the tightening capacity of the M8 in this application - if you need that much pressure, then you need to increase the pad and use two or more studs per contact.
Just so I understand, you vote for the M6 over the M8? Or either would be fine? I didn't mention my application and this was more of general question regarding stud contact area, but i draw ~25-50A continuously with a 600W inverter and a 30 amp MPPT solar system. Higher spikes to 100A for a few seconds when starting the gennie.
 
These cells can easily handle over 200 amps with the M6 holes (torque M6 nuts to <= 40 in-lb). I've done 100% SoC -> 0% -> 100% using 200+ amps for charge & discharge. Cell terminals were same temperature as the cell. Overall temp increase was ~10-15F. So the numbers you're talking about is peanuts for these.
 
These cells can easily handle over 200 amps with the M6 holes (torque M6 nuts to <= 40 in-lb). I've done 100% SoC -> 0% -> 100% using 200+ amps for charge & discharge. Cell terminals were same temperature as the cell. Overall temp increase was ~10-15F. So the numbers you're talking about is peanuts for these.
I might also consider a drop of Loctite on the thread first - that would lubricate between nut and thread, resulting in higher clamping force for a given torque. When working near the pull-out (or stripping) force of a bolt one must be careful to use correct torque for lubricated vs. dry (e.g. you don't lubricate lug nuts.) But for these cell terminal studs I think the torque used is well below that limit (you could look up torques of stainless bolts to compare.)
My research for the M6 threaded holes is to use blue Loctite as it helps with the Al/SS combination and is much cheaper than the product made for that. It is my intent to put the Loctite in the hole first, actually between the bottom and bolt thread. (this is preferred vs. on the threads due to hydraulic action) Then when putting on the busbars and BMS leads, Nordlock washer, and flange nut, there will actually be "pull" on the stud as it is being drawn to the top of the threads vs. midway. An infinitesimal amount, but a lot better than any downward force. This is using the hex head grub bolt such that it isn't attempting to thread into the hole. It is also suggested to put NOALOX on the threads. A member on here is a strong advocate of the stuff.
 
Just so I understand, you vote for the M6 over the M8? Or either would be fine? I didn't mention my application and this was more of general question regarding stud contact area, but i draw ~25-50A continuously with a 600W inverter and a 30 amp MPPT solar system. Higher spikes to 100A for a few seconds when starting the gennie.

If your studs look like the ones [link="[URL]https://diysolarforum.com/threads/improving-contact-area-on-welded-stud-pads.26869/post-319290[/URL]"]Stepandwolf posted[/link], then without washers you have about 64mm^2 of surface area on the aluminum to connect your busbars to.

Using a busbar current calculator, you can run a continuous 51A through that cross section with minimal temperature increase if both sides are aluminum. You can surge much higher than that, of course, but you should average 50A or less in the short terms (minutes). Note that if your busbars are copper, then your margin increases substantially - not just because the connection is lower resistance, but because the copper will soak up the heat and spread it out, increasing the speed it dissipates any heat generated at the connection. This will be somewhat offset by the reduced area of contact due to oval and slightly oversize holes in the busbar, but I expect the copper to more than make up for that.

So I don't believe you need to go to extra effort to increase the contact between the cell and the busbar. Make sure you use copper busbars, clean all oxidation off the terminal and busbar, use an appropriate antioxidant, and tighten to the recommended torque.
 
I bought a 2mm flat bar of 99% pure copper on Amazon $17.99. I cut it to length to cross both terminals and then drilled 10mm holes so it is tight against the 10mm post and sits flat on the 13mm pad. I then used a washer and nut to tighten it down so I get contact with the 13mm pad, the 10mm post and the washer and nut on the post. I do not think you can get anymore contact then with this method and the copper conducts electric at a much higher efficiency.

 
Back
Top