diy solar

diy solar

The secret to compression

Great breakdown of tradeoffs, thank you for that insight

I think foam and springs are the way to go for me.

Inner tubes are less robust and the issues with gasous mixtures and expansion and moisture and maintentance and noise and energy consumption are pretty absent with foam and springs.
My battery doesn't have any springs or compressible "foam" pads. My compression setup is fixed... Just wood with some threaded rods and lock nuts, and a few very thin plastic sheets between the cells. If I was going to do it again though I'd use the foam sheets.

Pic of the innards during assembly

1632161948443.png
 
My battery doesn't have any springs or compressible "foam" pads. My compression setup is fixed... Just wood with some threaded rods and lock nuts, and a few very thin plastic sheets between the cells. If I was going to do it again though I'd use the foam sheets.

Pic of the innards during assembly

View attachment 65515
Thank you, nice wood work by the way!

I heard that it's only 0.5mm or so per cell, so 4 cells -> 2mm max, 8 cells -> 4mm max

Lots of wood have nice flex characteristic. So 2mm I can imagine not mattering.

In some crazy row of like 32-64 cells I imagine this stuff justifies more and more elaborate approaches.

Cheers, thanks for sharing your build
 
Thanks. I feel it's decent enough for my purpose, though I don't feel like a great carpenter. FWIW I used some leftover 8020 extruded aluminum rails in order to join the box corners, and that really helped provide strength, and ensure everything is square.

0.5mm thickness seems small to me. I thought the neoprene stuff was sold in like 3/16" or 1/4" thickness. (1/8" would be ~5mm I think). I could probably go back and accommodate a few mm by loosening the lock nuts if that's all which is required. My "case" is really tight, but I've got about 1/2" I can play with this winter while the RV is in storage.
 
Without having an opinion one way or another, my research revealed two related takes on the subject. First being, it's better to use smaller cells, 100Ah or less, in parallel to create cell blocks, rather than using large capacity individual cells, due to structural integrity. Second being clamping or binding is recommended only in mobile applications to prevent movement inherent to the application.

Personally I wonder how much the average operator understands the immense pressure the common threaded rod can exert, a simple screw jack can lift a car without much effort. Many cells arrive slightly swollen, is this nessersary a bad sign? Or is it the nature of the design allowing some movement, this is something I'm reluctant to voice an opinion on, but I certainly wouldn't base my actions on a single study. The move to ever larger capacity cells to maintain series only connection may be where the problems arise.

Furthermore using automotive examples of different cell chemistries in conjunction with different cell structure is a weak argument, apples to oranges, they need to be constrained because they're in a mobile application, in a stationary installation, earthquakes not withstanding, the cells are not subject to the same conditions.

The beauty of a personal opinion is that nobody can disagree with it, the truth will however reveal itself one way or another, given enough time.
This thread has provided me with inspiration and knowledge for future builds, better awareness of different types of "foams". Yahoo!

I want to use the Big Cells 280/310Ah but Too Anxious given my engineering skills currently and the complexity of managing the aluminum flexture.

The approach you described of 100Ah or less and parallel blocks is the path I chose after being discouraged by Big Cell Flex and Terminal Engineering Challenges(tm) of other intrepid builders.

Will have individual 500A (350A continuous) shunt on each 100Ah pack to continuously verify balance of charge/discharge between conductors to busbar. And probably individual contactor per pack in addition to FET BMS per pack.

Everyone has different situation and needs etc! Bring on the info! This forum rocks.
 
Thank you, nice wood work by the way!

I heard that it's only 0.5mm or so per cell, so 4 cells -> 2mm max, 8 cells -> 4mm max

Lots of wood have nice flex characteristic. So 2mm I can imagine not mattering.

In some crazy row of like 32-64 cells I imagine this stuff justifies more and more elaborate approaches.

Cheers, thanks for sharing your build

My 16s setup with two 8s rows was clamped down with wingnuts finger tight (that's a fair amount of torque). When at a low SoC, the cells were LOOSE inside the fixture.
 
Say I had a simple DC solenoid air bleed valve that doesn't leak...

And some cheap air pump? Maybe with another DC solenoid valve?

And this 25psi ported pressure sensor, and it reads 24-bit values out (2^24=16777216 max reading value) compensated over temperature range 0-50degC, total error band near ±1.25% so like ~1.5e-6 PSI resolution or one millionth of a PSI granularity of observation..

Do you suppose this is a decent starting point?

Combined with a microcontroller that activates the bleed valve when reading goes over set threshold, and activate pump when under set threshold?
16476-SparkFun_Qwiic_MicroPressure_Sensor-01.jpg



Kinda spitballing implementation ideas, I have some of those ported sensors and silicone tubing around. Seems fun to hack on this.
The electronics end is not in my wheel house but the 10,000' view for the plumbing would be:


High pressure air supply> mechanical regulator to knock air pressure down to something a lot lower but higher than desired. 5 or 10 psi over the maximum pressure required > Needle valve (very close to) > Solenoid > T > Pneumatic wedge (or what ever you are using to apply the force)

From the 3rd port on the T > Needle valve (very close to) >Solenoid > Atmosphere.

When pressure reaches lower limit due to leakage, temperature, atmospheric pressure or SOC the first solenoid opens up and slowly bleeds pressure into the system. When upper limit is reached due to leakage, temperature, atmospheric pressure of SOC, the 2nd solenoid opens and slowly bleeds pressure to atmosphere until set point is reached.

You dont need any temperature compensation. Pressure is the only relevant measurement.

Depending on how good you are at setting up the needle valves you may want to add some volume to the system in the form of a small tank to reduce the rate at which the pressure changes when adding or bleeding off air.

You could replace the high pressure air source and regulator with the low pressure air pump you were talking about in your post.
 
Last edited:
The electronics end is not in my wheel house but the 10,000' view for the plumbing would be:


High pressure air supply> mechanical regulator to knock air pressure down to something a lot lower but higher than desired. 5 or 10 psi over the maximum pressure required > Needle valve (very close to) > Solenoid > T > Pneumatic wedge (or what ever you are using to apply the force)

From the 3rd port on the T > Needle valve (very close to) >Solenoid > Atmosphere.

When pressure reaches lower limit due to leakage, temperature, atmospheric pressure or SOC the first solenoid opens up and slowly bleeds pressure into the system. When upper limit is reached due to leakage, temperature, atmospheric pressure of SOC, the 2nd solenoid opens and slowly bleeds pressure to atmosphere until set point is reached.

You dont need any temperature compensation. Pressure is the only relevant measurement.

Depending on how good you are at setting up the needle valves you may want to add some volume to the system in the form of a small tank to reduce the rate at which the pressure changes when adding or bleeding off air.

You could replace the high pressure air source and regulator with the low pressure air pump you were talking about in your post.
Thank you! These notes will be in my mind if I do find/assemble the parts to try this silly idea.
 
An interesting discussion on achieving and maintain optimum clamping pressure. :)

I just went with plywood and they are clamped with a single threaded rod with a nyloc tightened to "about this tight" because RFT would be too much pressure.
 

Attachments

  • PXL_20210921_113256679.jpg
    PXL_20210921_113256679.jpg
    207.2 KB · Views: 43
An interesting discussion on achieving and maintain optimum clamping pressure. :)

I just went with plywood and they are clamped with a single threaded rod with a nyloc tightened to "about this tight" because RFT would be too much pressure.
I think your results will be just as good as those that use springs, pneumatics, foam, or whatever.
 
Well.

Just waiting on my cells. If they show up undamaged I will order the class T fuse and neoprene I'm going to line the box with and start charging.

In the mean time most of the rest of this crap need to be put in to clean up the hack job I did in my camper previously to get through the summer.

20210921_090928.jpg
 
I'm a little late to the game here, but I would like to share my viewpoint on compression and see what y'all think. Also, forgive me if this has been brought up already.

Compression is about maintaining the shape of the cell, to prevent bulging when at high SOC etc. But wouldn't an airbag ( like mentioned on page 7) just form around that "bulge" and apply pressure to areas that are unnecessary (ie the edges of the cell)?

I see foam doing the same thing, since it is not a FIXED surface, and can move, it will allow the cell to bulge and pressure be applied elsewhere.

A flat piece of sturdy metal (or wood) does not move easily, and I feel will keep the cell in place better, no?
 
I'm a little late to the game here, but I would like to share my viewpoint on compression and see what y'all think. Also, forgive me if this has been brought up already.

Compression is about maintaining the shape of the cell, to prevent bulging when at high SOC etc. But wouldn't an airbag ( like mentioned on page 7) just form around that "bulge" and apply pressure to areas that are unnecessary (ie the edges of the cell)?

I see foam doing the same thing, since it is not a FIXED surface, and can move, it will allow the cell to bulge and pressure be applied elsewhere.

A flat piece of sturdy metal (or wood) does not move easily, and I feel will keep the cell in place better, no?
Yes, regarding the air, but foam does not just "move" away. In fact as the poron compresses in any given local area the pressure it applies will increase at that location.

That said, the flat plate is how their fixture is presumably designed. I believe this was discussed. The compromise on consistency is springs of a known rate while making them as long as possible to reduce variation over the length of compression they will experience.

The poron is a compromise on space efficiency.
 
I'm a little late to the game here, but I would like to share my viewpoint on compression and see what y'all think. Also, forgive me if this has been brought up already.

Compression is about maintaining the shape of the cell, to prevent bulging when at high SOC etc. But wouldn't an airbag ( like mentioned on page 7) just form around that "bulge" and apply pressure to areas that are unnecessary (ie the edges of the cell)?

I see foam doing the same thing, since it is not a FIXED surface, and can move, it will allow the cell to bulge and pressure be applied elsewhere.

A flat piece of sturdy metal (or wood) does not move easily, and I feel will keep the cell in place better, no?
The airbag would need to have a ridged surface between it and the cell in the same way a spring would need to act on a ridged plate.
 
An interesting discussion on achieving and maintain optimum clamping pressure. :)

I just went with plywood and they are clamped with a single threaded rod with a nyloc tightened to "about this tight" because RFT would be too much pressure.
Nice clean build there

I prefer the German method of "gutentight"
 
It seems to me that an airbag would provide more consistent 12PSI pressure than a rigid surface... meaning, the airbag is technically better in that respect.
 
It seems to me that an airbag would provide more consistent 12PSI pressure than a rigid surface... meaning, the airbag is technically better in that respect.
Except that's not how the original test was designed and carried out in the lab setting.

It was done with a rigid plate under a fixed total force, not "12psi".
 
Except that's not how the original test was designed and carried out in the lab setting.

It was done with a rigid plate under a fixed total force, not "12psi".

Oh, that’s news to me. I never saw any documents on the details of the test and I didn’t think anyone here had uncovered those details. PSI vs kgf is irrelevant however (as it’s the same).
 
Oh, that’s news to me. I never saw any documents on the details of the test and I didn’t think anyone here had uncovered those details. PSI vs kgf is irrelevant however (as it’s the same).
Well yeah you can convert one to the other but that assumes perfectly evenly distributed force over the given area and an unyielding plate for both, as well as the entire cell returning the equal and opposite force.

However, it does not directly translate when you consider that the cells bulge from the center.

The force at the middle at full charge will be higher than the force at the outside as it expands into the plate and increases locally.

So as I said before the springs are a compromise on constant pressure while maintaining a similar "flat plate" style of force application while the poron is a compromise on overall consistency of delivery applied across the cell in favor of space savings.

In both cases the force directly in the center will increase as it expands, perhaps even far beyond the "x psi" applied as the total force stays basically the same during the test, so my vote is for space savings.


If I did this in my basement for home use a good long set of springs and plywood backed up by an aluminum plate would be my go-to as space is less critical.
 
Back
Top