That is definitely something I am considering. 
its not just about price but space. sometimes it comes down to what components you can fit in your enclosure.
I would love to use a proper 200a bms for my battery but I cant because this thing is the size of a graphics card.I’d pay $200 for an AS that could do 20s 200a reliabley with some smart features… Just me though
Absolutely!
One of my biggest concerns though is that small is great, universally desired, but small is terrible for reliability. It’s a delicate balancing act.
For me personally, it’s not worth doing this unless I can some up with a good way to keep the unit relatively small without sacrificing reliability. There are some cool ways to spread heat and understanding MOSFET failure mechanisms can really help the unit survive the abuse. Price is a concern but is secondary.
Finding time to do the design work, component searching, etc., is almost the hardest part of all. ![]()
It would be big though and probably a lot more than $200 unless made in very high volume. The 20S requirement means high voltage MOSFETs and they have a high resistance so you need a ton of them to get that 200A rating.
The smart features require a bit of programming time and a lot of testing time and the cost of that all needs to be amortized across the units being sold…really adds up quick unless the sales volume is huge….like an extra $100 per unit in low volume.
IMHO, the low reliability, high volume China units have really skewed our understanding of what is involved, what is really needed, to create small, full-featured, reliable products. I’m just being cranky now but that kind of device, especially if high voltage and high current, costs a lot.
I do appreciate your thoughts though, thanks. Knowing what specs are desired, budgets, etc., really helps me decide what mix of features/specs to use and whether it can be made reliable at an (almost) affordable price.
Like 18 FETs? And a PCB with thick copper? Wait is not an ESC, 4 FETs in parallel isn’t enough?
Just another quick thank you…
While thinking about your desire for 20S and 200A, a really tough combo to design for, I had a bit of a breakthrough regarding something that was really negatively affecting the cost and feature set of the switch.
I have already spent a couple hundred hours thinking about this thing but this new approach never occurred to me until now, thanks to your message. Yea…sure…I might have thought of it eventually but I wanted to give credit where credit is due.
So…thanks.
Hopefully the idea works out! So often it’s the tiniest little detail that derails something but perhaps not this time.
We’re forced to use high-resistance FETs because of the 20S requirement (in general, the higher the voltage rating the higher the resistance of the FET) and that creates a lot of heat. Even a 10oz copper PCB will quickly saturate thermally and rise to high temps at those current levels.
Sure, a heat sink can help but spreading the heat out by using lots of FETs makes the heat sink a lot more efficient and significantly lowers the junction temperatures.
A lot depends on the FETs though. Maybe four $13 FETs could do it but if we used twelve $2 FETs instead we can lower the unit price a lot and spread out the heat too.
So much will have to come out in testing. I can use heavy copper PCB’s, specialized PCB material, certain FETs, heat spreaders, heat sinks, etc., but there are so many tradeoffs with each…price, size, weight, assembly time, etc.
What would think would be the typical wire gauges and numbers in parallel (if applicable) used for a setup like this?
honestly i would pay 2 or 300 just for an AS that will never break. dumb or smart it really doesn’t matter, as long as it works and doesn’t pop even if we do stupid things (within hardware limit) with it.
What voltage and current ratings (continuous and 5 sec burst) would you want it to have?
i don’t run high end spec like others, so pretty much anything u create that fit others’ need would work for me.
right now, my top end board runs at 12s and 60a cont. (80a burst), not that fancy compare to what others are doing
But important nonetheless as there can be multiple models for this switch so owners of moderate power setups aren’t paying for ratings they don’t need.
Thanks!
that’s even better! although if i have the space, i would probably just get the best one out of all, just to futureproof myself
6awg AS150’s or QS10
If it’s bulletproof and can do 20s 200a I’d pay top dollar ![]()
Does anyone parallel up 10AWG or other gauges?
I don’t think so? Got an example… I think I know what you mean but just want to be sure beforehand
No photos, sorry.
Instead of using a single positive 6AWG wire for the positive and a single 6AWG wire for the negative does anyone use something like two 10AWG wires for the positive (both soldered to the same connector pin) and two 10AWG wires for the negative (soldered to the same connector pin)?
The positive and negative would each have two or more wires in parallel for each. Instead of using one large wire, 6AWG, you would use multiple smaller wires.
This makes the wire easier to bend and also divides the current up between the wires, making it easier to get even more current handling.
I ask because this determines whether there are multiple holes or solder pads on the AS switch circuit board for the wiring.
No one really does that, but I’d be open to doing that if it’s safer 
I’ve seen people double up 12AWG, but I haven’t seen anyone need large enough current to warrant 6AWG anyways 
Why not just leave a large enough pad that you could do either method? I can see benefits to either system depending on the packaging of the build.
