JP spot-welder, FET-switched, timed adj. pulse

[Offroader]

fechter,

Yes red light is on as normal. Green light blinks when foot pedal is pressed which means the MCP1407 is getting the signal to open the gates.

I'm thinking it could be the MCP 1407 chip.

Anyway to test this chip? I'm thinking I need to send voltage to the gates somehow to see if they open.

But why would the MCP 1407 chip fail? Would blowing the fets do this? I did have the PIC chip short on me already but I fixed that.

 

[Offroader]

OK Something doesn't make any sense at all. When checking continuity between the Source pin and the aluminum bar that attaches to the battery, I get no continuity at all. This is the pin that has the solder that I built up to the aluminum bar.

It doesn't make any sense considering the solder is right up to the bar touching it. This should explain why I'm not getting any voltage at the probes. I never removed this bar when replacing the fets.

Take a look at the picture, no continuity between the points that I circled in red and connected with a red line. If I understand the circuit correctly these two points should be connected.

You can see the solder pushed up against this aluminum bar. Should there be continuity between these two points, and how could there not be? I'm very confused now.

The only thing I can think of is I used a different solder than the one I used previously to build up the traces. Could this solder not be connecting with the previous solder?

I'm thinking my MCP1407 may be OK after all.

 

[markz]

> , I noticed however that there are sparks every time he weld. Any idea why ?

Its all about pressure and tip contour. I think a pointier tip is better, mine are more rounded off.
When I use little pressure it sparks a lot, a lot of pressure and no sparks.
The pressure is important, consistent pressure and tips.

-------------------------------------------------------------------------------------------------------------

Thanks Riba for your time and effort in making this wonderful device. That would be great if you could get the schematic posted with corresponding parts and specifications.

 

[rio9210]

> , I noticed however that there are sparks every time he weld. Any idea why ?

Its all about pressure and tip contour. I think a pointer tip is better, mine are more rounded off.
When I use little pressure it sparks a lot, a lot of pressure and no sparks.
The pressure is important, consistent pressure and tips.

-------------------------------------------------------------------------------------------------------------

Thanks Riba for your time and effort in making this wonderful device. That would be great if you could get the schematic posted with corresponding parts and specifications.

I don't think that guy, while welding, just "gently" lay down those tips. And, like previously mentioned in this 45 page post, pushing down too hard it's counterproductive for having good welds.
So, if we're excluding pressure issues, what else could lead to having a lot of sparks ?

 

[riba2233]

OK Something doesn't make any sense at all. When checking continuity between the Source pin and the aluminum bar that attaches to the battery, I get no continuity at all. This is the pin that has the solder that I built up to the aluminum bar.

It doesn't make any sense considering the solder is right up to the bar touching it. This should explain why I'm not getting any voltage at the probes. I never removed this bar when replacing the fets.

Take a look at the picture, no continuity between the points that I circled in red and connected with a red line. If I understand the circuit correctly these two points should be connected.

You can see the solder pushed up against this aluminum bar. Should there be continuity between these two points, and how could there not be? I'm very confused now.

The only thing I can think of is I used a different solder than the one I used previously to build up the traces. Could this solder not be connecting with the previous solder?

I'm thinking my MCP1407 may be OK after all.

Lift the source bar and check, there is a good chance that pcb source pads broke. Fortunately there is a really nice way to fix that!

 

[Offroader]

Yes it looks like the pads are cracked. One is clearly really cracked and destroyed on the right. The others look to have a hairline crack in them.

This is really odd because I didn't remove the bar or really do any excessive soldering. Why would this happen? I wonder if my fets blew because I wasn't making contact here with all of my fets, unless this happened only when I desoldered my fets. But look how neat I desodered them in the picture below, I didn't even remove all the solder from that area.

How do I fix it?

Thanks

 

[fechter]

Pretty amazing it blew through the copper traces.
One idea we considered before is to use a copper ring terminal on each source leg and bolt it to the aluminum bar. Bolt first, solder later to ensure the holes line up. That should eliminate the possibility of blowing the trace. You must have been running some seriously high amps to do that.

On the sparking/sticking issue, any residue on the surface of the nickel could cause this. You might try cleaning the nickel with alcohol or fine sand paper first and see if it makes a difference. I think the basic problem is high resistance at the contact point. Cratered tips don't help either. After cleaning up the tips with a file, I noticed an improvement.

After about 1,000 welds, my tips developed increasingly large flat spots.
To fix this, I removed the tips from the handles and placed them in my electric drill and spun them against a flat file. This shaped them back to pointy very quickly.

 

[mistercrash]

Pretty amazing it blew through the copper traces.
One idea we considered before is to use a copper ring terminal on each source leg and bolt it to the aluminum bar. Bolt first, solder later to ensure the holes line up. That should eliminate the possibility of blowing the trace. You must have been running some seriously high amps to do that.

Good idea fechter, I think I'm going to try that since I did the same as Offroader :lol:

 

[mauimart]

In regards to the electrodes sticking I have found that High Strength 182 copper works better than some of the other copper alloys. It's specifically designed for resistance welding. I got some 1/8" rod from McMaster. http://www.mcmaster.com/#catalog/122/3802/=11gefmk.

I filed the tips into dome shapes and I have been successful in getting decent to good welds using 0.15mm nickel with 4 to 5ms pulses and a deep cycle marine lead acid battery. I also did some limited testing with 0.30mm nickel and although the welds look clean they are not sticking. This was at up to 28ms pulses.

Disclaimer: I am new to spot welding and just got my welder working. Pictures of my setup and welds below.

 

[macribs]

In regards to the electrodes sticking I have found that High Strength 182 copper works better than some of the other copper alloys. It's specifically designed for resistance welding. I got some 1/8" rod from McMaster. http://www.mcmaster.com/#catalog/122/3802/=11gefmk.

I filed the tips into dome shapes and I have been successful in getting decent to good welds using 0.15mm nickel with 4 to 5ms pulses and a deep cycle marine lead acid battery. I also did some limited testing with 0.30mm nickel and although the welds look clean they are not sticking. This was at up to 28ms pulses.

Disclaimer: I am new to spot welding and just got my welder working. Pictures of my setup and welds below.

Where did you get that spot welder? DIY or ready made? Never seen those cell holders before? 3D print?
Nice info on the little LCD screen

 

[mauimart]

Where did you get that spot welder? DIY or ready made? Nice info on the little LCD screen

It's my own creation based on what others have done here on the forum. Basically the same design as JP with an added user interface that allows me to select and view weld pulse times. I also have a counter that keeps track of the number of welds. I thought of adding a bit more functionality (battery source voltage monitoring, trickle charge circuit for battery, etc) but I keep getting distracted. Would also like to clean up the layout and transfer to double-sided board.

 

[Offroader]

My welder is alive once again. I decided to modify it and solder copper wire from the source pin directly to the aluminum bar. I used some aluminum solder flux I had and special solder then I used regular solder on top of the special solder to solder the wire. I also fixed the trace by soldering a bridge and filing down the aluminum bar to make room, but I wasn't really happy with this so I added copper wire soldered to the aluminum bar.

It probably is just me but it seems the welder is working better than ever, but probably just me. Seems welds are better somehow.

Welding .3 nickel is possible by cutting it in half or grinding a groove in it. But the welds don't always seem to be as strong as with .2 nickel. I was able to put a couple of nice holes in .3 by cutting it in half. The problem with .3 nickel is that it is hard to get consistent good welds like you can with .15 or .2 nickel,

I still don't recommend .3 nickel, and I believe you are better off using a wider .2 nickel instead. Remember .3 x 8 mm = .2 x 12 mm .

For my battery build I decided after trying .15 , .2 and .3 nickel that I will custom cut .2 x 12 mm strips from a 90 x 10000 nickel sheet to get the thickness that I need.

 

[riba2233]

Glad you've fixed it! I developed a solution for that, similar to what fechter described, but hadn't had time to post it.

I'm always against 0.3 mm nickel since it's to thick and you can never be sure if it welded good unless you use too much welding power that can very easily make the cell leak, and then you have a big problem.

 

[Allex]

Here is my picture showing cell end cap to the right, it is about 0,3mm thick So welding with 0,3mm nickel could be risky. A bit to much current and you will be making holes in the cells.

 

[fechter]

My welder is alive once again. I decided to modify it and solder copper wire from the source pin directly to the aluminum bar. I used some aluminum solder flux I had and special solder then I used regular solder on top of the special solder to solder the wire.

Wow, I've never seen solder that actually works on aluminum like that. That should take the current.

 

[Offroader]

My welder is alive once again. I decided to modify it and solder copper wire from the source pin directly to the aluminum bar. I used some aluminum solder flux I had and special solder then I used regular solder on top of the special solder to solder the wire.

Wow, I've never seen solder that actually works on aluminum like that. That should take the current.

Yes it is special flux. It seems this stuff only became available during the last few years and before that there was nothing.

The welder seems so much better since I directly soldered direct into the source pins from the aluminum bar. Seems I can get better welds at lower pulse time and they are better also. I wonder if I may have not been using all of my fets. Maybe some of them lost connection to the aluminum bar.

 

[fechter]

The welder seems so much better since I directly soldered direct into the source pins from the aluminum bar. Seems I can get better welds at lower pulse time and they are better also. I wonder if I may have not been using all of my fets. Maybe some of them lost connection to the aluminum bar.

That's possible. Even if the traces were intact, the resistance may have caused uneven distribution of the current between FETs.

On the thicker welds, the pros alway slot the ribbon so the current is not shunted by the ribbon.
The welder electrodes go on opposite sides of the slot. This is not necessary on .15mm nickel, but probably wouldn't hurt. It is definitely needed on thicker stuff.

 

[grip911]

here is a pack I just made with Jakov`s welder .the okashira diode update seems to help a lot. https://www.facebook.com/photo.php?fbid=555874401239500&set=pcb.555874531239487&type=3
BTW It took 6000 welds in all

 

[riba2233]

Very nice! Can you show us where it goes into

 

[Offroader]

I was checking my gate 82r resistors on my JP Welder and one of them seems to not be stable.

In the video it is the last resistor I check or the 3rd and you can see the number jumping around and the line bar moving. It is the only one like this.

Will this cause any problems?

Thanks.

[youtube]wpgQCoyhG0Y[/youtube]

 

[riba2233]

Well I don't think is should cause any problems, but if you can I would advise to replace it. Maybe it's loosing contact inside itself, maybe it's only drifting.

 

[mistercrash]

My welder is back. A couple weeks ago I blew the traces like Offroader did So I took fechter's idea and got some #6 ring terminals and crimped and soldered them on each FET source leg and screw the terminals on the aluminum bar using the already existing threaded holes in it. I just welded some .3 slotted Nickel on a practice cell at 12 ms and I can't even rip it off. I'll edit this post later to put a pic in.

To answer Offroader's post below, yes I did change the six FETs but the welder still did not work because the trace was gone underneath the aluminum bar. Now the pics of the ring terminals, the dial I made and last the 0.3 Nickel strip I welded on the cathode, ripped off and then welded the other end on the anode and ripped it off again to take the pics. Yes I feel the welder is working better with the source legs screwed to the aluminum bar.

 

[Offroader]

My welder is back. A couple weeks ago I blew the traces like Offroader did So I took fechter's idea and got some #6 ring terminals and crimped and soldered them on each FET source leg and screw the terminals on the aluminum bar using the already existing threaded holes in it. I just welded some .3 slotted Nickel on a practice cell at 12 ms and I can't even rip it off. I'll edit this post later to put a pic in.

Can you post a picture of this modification.

Are you saying you blew just the trace and not the fets?

Are you also saying the welder now welds better with this modification?

I hope I wasn't stupid and replaced the fets when I didn't need to, and all i needed to do was rebuild the trace. I'll have to test my old fets to see if they work.

 

[grip911]

Very nice! Can you show us where it goes into

It`s an extension pack for the 911 https://www.facebook.com/photo.php?fbid=538140583012882&set=a.328898887270387.1073741828.100004508347334&type=3

 

[riba2233]

Hell yeah!

Do you have any videos or internal pictures?