diy solar

diy solar

Fire!! Never cover LiFePO4 with wood!!!

Even if using a thread locker the stud can be pulled out. Here is a photo of one of mine that pulled up with the aluminum threads included. That is red loctite you see on the stud. I didn't use a primer but so far all of the other studs are ok. The aluminum is so soft it's really easy to strip. I am not entirely sure why this stud pulled up but I do recall some studs had more wiggle room than others while screwing them in. I have since bought a beam style torque wrench.
Not that the stud is out, measure the depth of the hole. My stripped terminal was tapped less than 4mm deep.

I’ll be checking the depth of all tapped terminals of new cells from now on (before I secure a grubscrew with Loctite Red).
 
I totally agree.

Sadly noalox and ox-gard aren't available in Thailand, and shipping+ tax will make one tube over $75.
(Not the $4 one, that will be about 22,50, but I doubt it's enough for 160 terminals)

Oxidation of silver isn't a problem.
The conductivity of silver oxidation is close to non oxidized silver.

Copper oxidation isn't a real problem (with copper to copper) sure the oxidation changes the way the electrons behave, 2 oxidated copper (with thight contact) counteract eachother makeing a good contact again.
This is the reason you don't need to clean each strand of wire, but just compressing them is enough.

Aluminium is the bad boy here.
Milliseconds after you cleaned it already oxidises, and after a few minutes that far that it will interfere with the contact.

Covering the terminal contact areas with a little drop of silver paint after cleaning creates a seal, like oxi-guard or noalox.
With the big difference that silver is conducive and noalox/oxi-gard are not. (Di-electric grease)

I agree that different metals is asking for troubles.
Not the trouble I encountered!!!

Best is to use aluminium bus-bar and if you can get, aluminium threaded rods.
With aluminium nuts.

Once you start with different metals..
You are screwed.

Most likely not in the period of installation, but 15-25 years.

Most cells don't have environment that speed up the galvanic corrosion process.
No aviation or salt water.
No huge temperature swings.

304 stainless steel is the most "Neutral" out there, it doesn't get eaten away from galvanic corrosion.

Yes, that makes the aluminium and the copper the sacrifice metal.
Tinning or zinc plated makes that the sacrifice metal.

Inside the aluminium threads?
Zinc plated grub screws get damaged just by being in the box together.
There will be bare metal.
If you don't believe me, place it a few hours in salty water and lay it outside for a few days.
All those rust spots are damaged areas.

At those locations the bare iron have full contact with the aluminium. The zinc surrounding is corroded quickly.

Loctite is a good solution.
Not just because it locks the nut, it also seals the thread.
No oxigen = no galvanic corrosion.

There are many discussions about it.
Most people use the brass bolts they get with the cells.
Or what ever they had available.

I'm yet to see an eaten away thread hole from galvanic corrosion.

That has a real clear signature.
No smooth walls like stripped thread, but uneven.

As long as no one is able to show the galvanic corrosion problem in the thread, it only exist in theory.

Outside the thread, the contact that have air exposure...
Yes, absolutely.
Anyone with a car, that cleaned their lead acid terminals from the white/blue/green powder know what that looks like, lead and copper with probably some acid air., And different temperatures.

That is a whole different story.

304 steel rods with Loctite or Jbweld (or similar Epoxy) are the best.
Applied on the rod, then screwed in.
You can see the Loctite react with the aluminium oxide in the thread.
It turns almost directly black.
Having Loctite on the bottom is useless.
Yes, the interface and chemistry means the aluminum oxidizes in preference or instead of the copper. Unfortunately aluminum oxide is the worst conductor of the bunch. The objective is to just keep oxygen away from the aluminum, and actually that small tube of Ox-Gard would work for about that many connections, you just really only need a thin film, wipe it on with your finger after cleaning, then connect the terminal and busbar. You just want to keep oxygen away from the connection.
 
Not that the stud is out, measure the depth of the hole. My stripped terminal was tapped less than 4mm deep.

I’ll be checking the depth of all tapped terminals of new cells from now on (before I secure a grubscrew with Loctite Red).
No need to measure the depth. You can tell by counting the threads on the stud. I count 5 plus the bottom of the stud where it is tapered. That hole is 6.5 to 7mm deep. Besides I already repaired it with JB Weld.

You made a good point about the depth of the hole but in this case that didn't cause the problem. I believe it was most likely due to sloppy tapping.
 
No need to measure the depth. You can tell by counting the threads on the stud. I count 5 plus the bottom of the stud where it is tapered. That hole is 6.5 to 7mm deep. Besides I already repaired it with JB Weld.

You made a good point about the depth of the hole but in this case that didn't cause the problem. I believe it was most likely due to sloppy tapping.
Threads are easy to damage in production. I'm guessing most of these terminals may be tapped by hand. Sometimes it's worse with a tapping head.

I know one machinist (personally, well actually it was me..) that was using a tapmatic tapping head in a drill press. This uses a clutch to engage CW rotation when entering the hole, and when retracting, it pulls up, engaging the clutch to reverse and back out the tap.

If you try speeding up the process, pulling upwards before the tap stopped, that force, along with the rotational inertia of the tapping head puts a load on the threads. It's tough to see any damage the machinist caused by this. The threads gage out fine. But micro stress cracks at the root of the threads can fail when torquing at assy, or worse in use in the final system. I was lucky enough to have a failure early, with a failure analysis lab to dissect the parts and look at the granular structure.

Rolled threads in aluminum are stronger than cut threads due to hardening by cold working.

It may be best to use epoxy to install the studs. Grind off the bottom of your studs to reduce imperfect threads. File off the first thread into a ramp to allow it to form any poor thread at the bottom of the terminal hole. The MIL spec screws are supposed to allow 3 imperfect threads. I've seen 4-5 imperfect thread on some MIL fasteners. Commercial hardware could have more bad threads near the ends.
 
This stuff sounds like it might be better to use over regular loctite :

 
This stuff sounds like it might be better to use over regular loctite :


Although sold under the name "Loctitie", appears to be an adhesive rather than a thread locker.
It is meant to make electrical connections.

For these studs, electrical connection really isn't important. Due to differences in resistance and path length, I expect virtually all current to flow through interface between busbar and terminal, not through resistive stainless stud.

Only reason anything is being considered (other than perhaps anti-seize grease to prevent galling) is that the aluminum threads are so delicate and get stripped easily. Exacerbated by the fine craftsmanship and quality control provided by back-alley machine shops tapping terminals in grade-B cells that were originally designed to be laser welded.
 
Only reason anything is being considered (other than perhaps anti-seize grease to prevent galling) is that the aluminum threads are so delicate and get stripped easily. Exacerbated by the fine craftsmanship and quality control provided by back-alley machine shops tapping terminals in grade-B cells that were originally designed to be laser welded.

And the tapping is done by the lowest bidder (in China)!
 
@diyernh, thanks for the anecdotal information about carbon and aluminum. A grease or paste containing silver seems the better choice where conductivity is desired. I found this video informative.
 
Last edited:
This stuff sounds like it might be better to use over regular loctite :


Dang! That's some expensive stuff.
Waaah1.gif
 
Installing a fuse in the balance leads will lead to increased resistance in the circuit leading to erroneous V readings for the cells by the BMS. It's the same reason balance leads should be kept the exact same length as differing lengths of wire have different resistance. Some will argue such a short distance would not affect the V at the end, however resistance is resistance and when dealing with 3.65v per cell, even a small resistance has a huge effect.

Have you verified this, or are you just extending common knowledge regarding conductors carrying significant current?

The BMS, like all properly designed voltmeters, is a very high impedance load. This is proper engineering practice because a sensitive voltmeter should not place ANY load on the circuit. This is why the “ohms per volt” specification of your DVM is an important indication of the quality of the DVM. This is also why any “voltage sense” leads in your device can be such small wire gauge.

This means that the BMS lead is, for all practical purposes, carrying no current. Since the lead carries no current, there can be no voltage drop. And; since there is no current, a fuse in line will not alter the reading of cell voltage. If you did see a difference, it means the BMS is poorly engineered. My JBD BMS’s exhibit no such sensitivity to lead length, much less a “huge” effect.

Never-the-less, at this time, I don’t intend to install fuses in BMS sense leads for the other reasons stated above.
 
Last edited:
From http://www.orionbms.com/manuals/pdf/wiring.pdf (bolding mine):

If external cell tap fuses are used, they should be as close to the cell terminals as possible, kept as low resistance as possible within reason, and be rated no larger than 5A. Fuses must always be rated for the maximum possible DC working voltage with a suitable DC interrupting current rating for the application. Certain electrical codes or regulations may dictate the maximum fuse size, and it may be smaller than 5A depending on wire gauge and other considerations. Fuses should be selected to have the lowest possible resistance to prevent degrading the cell voltage measurement accuracy. Series resistors are not permitted as current limiting devices for cell tap wires as they may over-heat or limit necessary current to blow internal fuses on the BMS.

Sounds like a qualified "yes" to me.
 
From http://www.orionbms.com/manuals/pdf/wiring.pdf (bolding mine):



Sounds like a qualified "yes" to me.
The phrase
“kept as low resistance as possible WITHIN REASON“
tells me that it isn’t a serious concern.

That same manual confirms that, as long as cell voltage is less than 5-volts, keeping the protective zener diode intact and non-conducting, no current flows through the sense lead. The manual appears carefully written to limit DIY‘ers from making ill-considered modifications. It is a quite comprehensive, and interesting manual.

We can all emphasize whatever factors we think are important. It is still truth that a lead for a properly designed sensitive voltmeter carries essentially no current, and can drop no voltage. I don’t have experience with other brands of BMS, but mine have a neat voltage calibration feature that can also null out any installation variations. Since a sensitive voltmeter is a static load, to the degree that it is ANY load, this works well to null out installation variables.

A typical BMS input impedance would be 10 million to 100 million ohms. At 3.5 volts, it would draw from 3 to 30 micro-amps. A typical fuse might add 0.2 ohm’s of resistance. The voltage drop across the fuse would be a most-trivial 0.6 to 6 microvolts.
 
Last edited:
I found and ordered oxi-gard $45 for 8oz including transport and tax to Thailand.

Tap set with the sharpest bottom tap I could find for M8, $35,-

I already have press drill so that is easy.

Tapping my M6 in copper went perfectly with electric screwdriver.
It didn't suddenly stop but rattled, each rattle made it go a little deeper.
They don't have to be strong, it's mostly to connect lugs.

Depending on the number of cells I need to do and how difficult it goes I might do the first part like that.
Last bit by hand.

The engineers who have done this a lot:
What lubricant to use?
For tapping/drilling iron I always used old engine oil.
That wasn't easy to clean from the copper.
What is good for aluminium?
 
For aluminum, consider "tap magic", kerosene (probably same thing), isopropyl alcohol.
 
I found and ordered oxi-gard $45 for 8oz including transport and tax to Thailand.

Tap set with the sharpest bottom tap I could find for M8, $35,-

I already have press drill so that is easy.

Tapping my M6 in copper went perfectly with electric screwdriver.
It didn't suddenly stop but rattled, each rattle made it go a little deeper.
They don't have to be strong, it's mostly to connect lugs.

Depending on the number of cells I need to do and how difficult it goes I might do the first part like that.
Last bit by hand.

The engineers who have done this a lot:
What lubricant to use?
For tapping/drilling iron I always used old engine oil.
That wasn't easy to clean from the copper.
What is good for aluminium?
Just ask, I would be happy to mail you some. My wife is Ukrainian, I am used to filling out customs forms. You might have to pay import duties, but that is up to the government, and for an item under $10 should not be much.
 
Even if using a thread locker the stud can be pulled out. Here is a photo of one of mine that pulled up with the aluminum threads included. That is red loctite you see on the stud. I didn't use a primer but so far all of the other studs are ok. The aluminum is so soft it's really easy to strip. I am not entirely sure why this stud pulled up but I do recall some studs had more wiggle room than others while screwing them in. I have since bought a beam style torque wrench.
I mentioned in this thread about imperfect lead in threads. Your photo shows a good example. At the bottom of the stud, there is about 2 threads that were imperfect. By cutting off the stud about 4-5 threads, you can grab a couple more threads near the bottom of the tapped hole. Grind the stud with a 30 degree chamfer and file down the lead of the thread.

It may not seem much, but if the stud has a few bad threads, and the tapped hole has some imperfect threads, there could be very little engagement. Getting 6 threads instead of 4 is a 50% increase.
 
For aluminum, consider "tap magic", kerosene (probably same thing), isopropyl alcohol.
So is it a bad idea to use tapping oil when tapping an aluminum stud? I thought that is what Cinergi used when he tapped his...

[edit: now that I understand ‘Tap Magic’ is a tapping oil (and Kerscene is also a paraffin oil) and not isopropyl alcohol, I’m guessing you were suggesting kerosene or Tap Magic for tapping and isopropyl alcohol to clean up the tapping fluid after the threads are tapped, not to perform the actual tapping, right?]
 
Last edited:
So is it a bad idea to use tapping oil when tapping an aluminum stud? I thought that is what Cinergi used when he tapped his...

[edit: now that I understand ‘Tap Magic’ is a tapping oil and not kerosene, I’m guessing you were suggesting kerosene and isopropyl alcohol to clean up the tapping fluid after the threads are tapped, not to perform the actual tapping, right?]
Any oil is better than no oil. I used 3 in 1 oil Clean it up with alcohol or acetone then put noalox or similar on
 
Back
Top