DIY spot-welder, FET-switched, processor-timed adj. pulse

Batteries, Chargers, and Battery Management Systems.
okashira   10 kW

10 kW
Posts: 957
Joined: Nov 17 2014 6:45pm
Location: Houston, TX

Re: JP spot welder

Post by okashira » Mar 31 2016 3:44pm

We make the holders, they are great because after welding I epoxy between cells above the holders. After it cures, you can cut off the outsides of the holders to really make the pack envelope small as possible.
And there's still a little air gap for airflow if you so desire.

I use 0.2mm nickel for the sheet and 0.3mm nickel for stacking
"Peace out - I'm done."
-- Bernhard Riemann, 1859

edisonmotors.net
Tesla Model S 18650 cells - like new - and custom packs from Tesla cells for sale. Any shape Any power Any capacity.
PM me or search my threads.

User avatar
fechter   100 GW

100 GW
Posts: 14060
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: JP spot welder

Post by fechter » Mar 31 2016 4:20pm

Offroader wrote:
Thanks, so you understand my concern about the other side connecting to different cells. I drew out a lot of diagrams and also don't think it would matter. Each individual series groups have exactly the same length wire.

Yes, I intend to actually weld a very small nickel strip, maybe 3 mm wide between the parallel groups for balancing. I will then run the balance wires to the bms. The reason for a thin strip is in case of a short it will limit the current draw from the other cells in parallel.
I didn't see the parallel connection so was really scratching my head.
The resistance of your wire is probably much less than that of the nickel strip, so the wire lengths are not going to be very critical. Depending on the copper gauge and the nickel, we can do the math on it, but WAG the copper is like 10x less resistance than the nickel strip for a given length.

Keep in mind there is "ideal" and there is "good enough". With an odd configuration, achieving ideal is going to be very difficult. The basic idea is to have many series connections so the current gets spread out and not concentrated on a single conductor. At the pack ends, you need a single conductor, so the simplest approach is to just use heavy copper so the voltage drop is minimal and the cells all discharge equally. Where you can run into trouble is when using only nickel strips and making every series connection through a single conductor.

Two goals in the layout from an electrical standpoint are
1. make sure the conductors don't get hot.
2. try to equalize the drain between cells in a parallel group.

Equalizing the drain is not so important if your discharge rate is low. When running 2C or higher, it might matter.
Another goal is ease of construction. Large nickel sheets that cover all the cells in a group is good for this.
Durability and safety are also very important. Avoiding potential shorts caused by vibration or bike crashes is high on my list.
"One test is worth a thousand opinions"

User avatar
Offroader   10 MW

10 MW
Posts: 2461
Joined: Sep 08 2013 9:03pm
Location: USA

Re: JP spot welder

Post by Offroader » Apr 01 2016 11:12am

For those interested about current or cell equalization, here is a post by doctorbass.

https://endless-sphere.com/forums/viewt ... 96#p920396

You can see in the image I actually designed my pack like the bottom right in the picture.

Image

User avatar
fechter   100 GW

100 GW
Posts: 14060
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: JP spot welder

Post by fechter » Apr 01 2016 11:59am

In any of those, if the parallel connections are really hefty, the sharing will be OK. The main pack connections on the ends need the hefty parallel conductors in most cases.
"One test is worth a thousand opinions"

okashira   10 kW

10 kW
Posts: 957
Joined: Nov 17 2014 6:45pm
Location: Houston, TX

Re: JP spot welder

Post by okashira » Apr 01 2016 1:18pm

Offroader wrote:For those interested about current or cell equalization, here is a post by doctorbass.

https://endless-sphere.com/forums/viewt ... 96#p920396

You can see in the image I actually designed my pack like the bottom right in the picture.

Image
...

That doesnt matter if you make the connections properly like fetcher said.
Current will always share
"Peace out - I'm done."
-- Bernhard Riemann, 1859

edisonmotors.net
Tesla Model S 18650 cells - like new - and custom packs from Tesla cells for sale. Any shape Any power Any capacity.
PM me or search my threads.

User avatar
mistercrash   10 kW

10 kW
Posts: 947
Joined: Mar 09 2011 11:01am
Location: Guelph, Ontario, Canada eh!

Re: JP spot welder

Post by mistercrash » Apr 05 2016 6:47pm

Offroader, I sent you a pm.
Send $10 at paypal.me/Ray1964 and you will be in my thoughts and prayers.

markz   100 GW

100 GW
Posts: 7414
Joined: Jan 09 2014 11:38pm
Location: Alberta Canada

Re: JP spot welder

Post by markz » Apr 05 2016 7:37pm

Why you flooding the JP spot welder with battery hook up scenerios?

User avatar
riba2233   1 MW

1 MW
Posts: 1832
Joined: Jun 30 2013 10:23am
Location: Croatia

Re: JP spot welder

Post by riba2233 » Apr 06 2016 5:13am

markz wrote:Why you flooding the JP spot welder with battery hook up scenerios?
It's ok :)

User avatar
bigbore   10 kW

10 kW
Posts: 507
Joined: Sep 25 2009 7:43am
Location: Italy

Re: JP spot welder

Post by bigbore » Apr 06 2016 1:43pm

okashira wrote:
eric1565 wrote:BigBore... Couple of issues I see with that battery setup, and .3 nickle.

1. run the copper wire over all the cells, and not just solder a small section to the cells. The nickle strip connecting the cells together does not conduct current well at all!!!!!
2.That battery seems small. We run anywhere from 60ah on A123's to 222ah on tesla cells.
3. The wires look small. 8 gaga is a bit small for dumping 2k amps. we use 4 gage.

Eric
Also needs to connect the diode to the battery terminal not to the middle of the lead. You are missing alot of inductance
So, I have improved the setup:
Image
Image

User avatar
riba2233   1 MW

1 MW
Posts: 1832
Joined: Jun 30 2013 10:23am
Location: Croatia

Re: JP spot welder

Post by riba2233 » Apr 06 2016 5:05pm

Now that's a killer :mrgreen: You've got to love massive buss bars 8)

VirgilBudBrigman   100 µW

100 µW
Posts: 8
Joined: Oct 29 2015 11:46am

Re: JP spot welder

Post by VirgilBudBrigman » Apr 08 2016 7:00pm

Hey guys,

I blew a MOSFET last night. I'm not sure about installing the diode-mod just yet, but I have a question regarding replacing the FETs:

Do you think (6) of these will be compatible?: http://www.digikey.com/product-detail/e ... ND/5958884

The values are roughly the same as the IRF1324 that comes stock, except the Vdss (max voltage?) is 40V vs 24V.

Thanks!
Last edited by VirgilBudBrigman on Apr 08 2016 11:31pm, edited 1 time in total.

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 08 2016 7:33pm

That link doesn't get to a device. Just a generic page.

VirgilBudBrigman   100 µW

100 µW
Posts: 8
Joined: Oct 29 2015 11:46am

Re: JP spot welder

Post by VirgilBudBrigman » Apr 08 2016 11:35pm

Alan B wrote:That link doesn't get to a device. Just a generic page.
My bad, here is the part link: http://www.digikey.com/product-detail/e ... ND/5958884

Also here is a comparison between it and the stock IRF1324: http://www.digikey.com/short/38ndh7

Price is roughly the same($3.86 vs $3.31), so that's a non-issue.

Thanks!

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 09 2016 2:28pm

The question is, what are you trying to accomplish by changing FETs?

The FETs you propose have higher forward resistance at a lower current measurement.

Their higher voltage rating won't protect them from the inductive spikes.

They take more current to drive the gates.

They cost more.

Digikey has relatively few in stock.

The Datasheet link for the IXYS part doesn't work.

And they haven't been tested for this kit's use.

Don't get me wrong, IXYS makes great FETs and in all probability they will work, but it is important to first minimize and manage the voltage transients. Extra cable between the switch and the welding probes stores energy and this feeds back into the FETs when the pulse ends. The voltage this transient reaches is essentially unlimited. So adding a few volts of FET rating doesn't prevent the problem.

I don't have one of these welders yet but I would fabricate it with absolute minimum wire length to the probes (and battery). And incorporate the freewheeling suppression clamp diode, which is a standard part in solid state drivers for relays, as an example.

User avatar
fechter   100 GW

100 GW
Posts: 14060
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: JP spot welder

Post by fechter » Apr 09 2016 3:44pm

There are a lot of part choices out there.

Something with a lower on resistance might be better.
Alan, can you see a better one?
"One test is worth a thousand opinions"

jmz   100 W

100 W
Posts: 105
Joined: Jan 16 2015 8:31pm
Location: Christchurch, New Zealand

Re: JP spot welder

Post by jmz » Apr 09 2016 5:36pm

I would be interested to know how quickly these welders switch.

My own one, which is designed with RCD snubbers (interleaved on the bus between switching FETs) instead of flyback diodes, deliberately has a very slow turnoff (4us, still safely within the device SOA), with 800 Ohm gate resistors, to allow the current to gradually commutate to the snubber capacitors in the presence of their finite inductance (~10nH).

The advantage of this is that it safely captures the energy from all circuit inductance, not just that present in the leads, and the FETs never have to absorb any energy in avalanche operation.

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 09 2016 6:28pm

fechter wrote:There are a lot of part choices out there.

Something with a lower on resistance might be better.
Alan, can you see a better one?
I haven't looked much, but doing some quick parametric sorting I see that the IRF1324 has a fairly low tolerance for single pulse Avalanche Energy compared to some others in this performance regime, about a quarter of a Joule compared to 1.5J or 1.9J for some other parts in the 1.3 milliOhm range.

I suspect it is this Avalanche event that is blowing FETs, so selecting a part with a higher avalanche capacity would be reasonable, however dissipating the energy elsewhere with diodes or slowing down the turnoff may be a better approach.

Slowing down the turnoff, as jmz suggests is probably the most cost effective approach, is there a requirement to turn off so quickly? The tradeoff is a slow turnoff will increase heating in the IR loss of the FET. So one can only go so far with that. Interleaving with snubbers and slowing the turnoff sounds like a good engineering approach. A 4uS turnoff with a 1000A pulse would be about 0.01 or 0.02 Joules so not much heating.

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 09 2016 6:32pm


User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 09 2016 7:51pm


User avatar
fechter   100 GW

100 GW
Posts: 14060
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: JP spot welder

Post by fechter » Apr 09 2016 10:21pm

Alan B wrote:This looks good also.

auirfb8409

http://www.irf.com/product-info/datashe ... fs8409.pdf
$3.28ea from DigiKey. 600+ in stock. Those look pretty good.

They keep coming out with better parts all the time.
"One test is worth a thousand opinions"

rio9210   100 mW

100 mW
Posts: 49
Joined: Sep 29 2015 3:04am

Re: JP spot welder

Post by rio9210 » Apr 10 2016 1:48pm

Alan B wrote:This looks good also.

auirfb8409

http://www.irf.com/product-info/datashe ... fs8409.pdf

For Europe buyers, this should be ok ?

https://www.reichelt.de/IRFB-7430/3/ind ... EARCH=IRFB

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 10 2016 2:00pm

rio9210 wrote:
Alan B wrote:This looks good also.

auirfb8409

http://www.irf.com/product-info/datashe ... fs8409.pdf

For Europe buyers, this should be ok ?

https://www.reichelt.de/IRFB-7430/3/ind ... EARCH=IRFB
The pictured part there looks really sketchy. Lots of counterfeit parts out there.

rio9210   100 mW

100 mW
Posts: 49
Joined: Sep 29 2015 3:04am

Re: JP spot welder

Post by rio9210 » Apr 10 2016 6:14pm

Alan B wrote:
rio9210 wrote:
Alan B wrote:This looks good also.

auirfb8409

http://www.irf.com/product-info/datashe ... fs8409.pdf

For Europe buyers, this should be ok ?

https://www.reichelt.de/IRFB-7430/3/ind ... EARCH=IRFB
The pictured part there looks really sketchy. Lots of counterfeit parts out there.

You think it`s chinese sub-stuff ? They haven`t the right pic for every item they sell.

User avatar
Alan B   100 GW

100 GW
Posts: 7550
Joined: Sep 11 2010 7:43am
Location: San Francisco Bay Area, USA
Contact:

Re: JP spot welder

Post by Alan B » Apr 10 2016 6:17pm

The pic certainly doesn't look legit. That's all I am responding to.

okashira   10 kW

10 kW
Posts: 957
Joined: Nov 17 2014 6:45pm
Location: Houston, TX

Re: JP spot welder

Post by okashira » Apr 10 2016 10:31pm

I think you will find those mosfets will blow very quickly.
Stick with the irf1234s. Their low voltage makes them more tolerant of avalanche.
With my diode mod I still running wonderfully at over 3kA now. I can get full flagging on 0.3mm nickel and I'm not even on full power.
I run 9V with a 0.001 Ohm battery

Further to the diode mod, I suggest looking into ~12V rated TVS diodes (or ~14V if you're using a 12V lead acid) and place them across the source drain. Put a few in parallel, and they will take the avalanche to further protect the fets.
"Peace out - I'm done."
-- Bernhard Riemann, 1859

edisonmotors.net
Tesla Model S 18650 cells - like new - and custom packs from Tesla cells for sale. Any shape Any power Any capacity.
PM me or search my threads.

Post Reply