SPOT WELDER WITH PIC16F88 tested - YOUTUBE video

[Ichan]

I use 47000 uF x 12 for a simple battery tab welder, do not know exactly the thickness of the nickel tab - about 0.2 mm i guess.

The wiring resistance and the quality of the electrode will play important factor and the properties of the welded metal too.

-ichan

[kalpakchiev]

I use 47000 uF x 12 for a simple battery tab welder, do not know exactly the thickness of the nickel tab - about 0.2 mm i guess.

The wiring resistance and the quality of the electrode will play important factor and the properties of the welded metal too.

-ichan

That is about 0.5F At what voltage level You charge capacitors?

[kalpakchiev]

ORCAD schematic and PCB ready. To cut it on CNC machine it is single side with bridges on top layer.

[sadman]

ORCAD schematic and PCB ready. To cut it on CNC machine it is single side with bridges on top layer.

Hi Friend

you did good job and nice add on tool for work shop, i have one question can you please specified working voltage of car audio capacitor (1F)

regards
sadman

[Ichan]

That is about 0.5F At what voltage level You charge capacitors?

24V.

-ichan

[kalpakchiev]

Hi Friend

you did good job and nice add on tool for work shop, i have one question can you please specified working voltage of car audio capacitor (1F)

regards
sadman

As far as I know that type of capacitors are from 0.3F up to 12F +/- 15% and all of them are rated for 16VDC/21VDCpeak or 20VDC/24VDCpeak.  A good ESR is less than 0.002 Ohm. Some of them are equipped with on board digital voltmeter.

[sadman]

Hi Friend kalpakchiev

can you please share pcb file in pdf format as i do not have orcad

regards
sadman

[kalpakchiev]

Hi Friend kalpakchiev

can you please share pcb file in pdf format as i do not have orcad

regards
sadman

REQUESTED PDF. LCD and KEYPAD will be ready tonight. Best regards

[kalpakchiev]

Work is in progress - I have the capacitor , proper cable and main PCB is ready. Next is LCD and KEYPAD PCB.

[kalpakchiev]

Display is ready. Tonight I'll assemble all together for first test

[SuperDrive]

Good progress, waiting for it.

[kalpakchiev]

So the purpose was to design a reliable, medium power and at last but not least easy to build SPOT WELDER. First tests in real world are done and I think the experiment was successful .

Next step is to refine the project. Thing to do:

1. Power supply for pic processor MUST be separated

2. Using TC4422 limits the performance up to 18V - I have to change the driver

3. Is You can see in video there are 3 modes SINGLE, DOUBLE and CONTINUOUS - continuous  mode have to be reworked
  
4. I have to think about the charge and discharge circuits - in my opinion there are better ways to do it.

The work is still in progress so suggestion to this will be of great help.

Video 1
http://www.youtube.com/watch?v=PqPzzsDhFFo

Video 2
http://www.youtube.com/watch?v=Bb6R_LCmBOc

Video 3
http://www.youtube.com/watch?v=hG6S0Q5yUoY

Some notes:
For discharge are use 5xIRFZ48 in parallel, without sinking and they are enough (up to 18V tested)

For charging is use 1xIRFZ48 again without sinking, but it needs sink or another one in parallel because gets really hot

[borberk]

In my opinion you should increase charge voltage up to 30V. This can help when welding not so clean metal pieces and capacitor can have smaller capacitance. You know formula Q=C*U.

[kalpakchiev]

You are right but the capacitor I use in these project is rated 20VDC/24VDC Peak so 20V is highest limit. When this project is finished I can make some tests with these capacitors. It will be fun .

[borberk]

Capacitors on picture are 250V. If longer charge time is acceptible you can charge through greater resistance constantly without FET switch. 

[kalpakchiev]

Yes they are. But in this project is used this one:1F/20VDC - attached picture.

 Probably longer charge time will be acceptable and constant charge will make circuit more simple, but in this case I have to take precaution that charging voltage not to exceed 20V and also to control voltage of the capacitor when welding very thin material.

[Ichan]

Have you try it welding things?

-ichan

[kalpakchiev]

only a few beer caps and a screw driver by mistake . My camera battery is flat now. I'll upload video later when recharge it and also I have to buy a tripod

[kalpakchiev]

I want to rebuild charging circuit and Your opinion will be very helpful for me. What do You think about these two:

[titi]

Hi,

In my opinion, i prefer the original design.

The First one (old schematic.jpg) doesn't work, you need charging CAP and then Discharge then on Load, on your design CAP is in parallel with Load, CAP never charge.
The Second One (new schematic.jpg), the charging MosFet is at the best place, but the discharge MosFet is floating, since charging MosFet is open when discharging, you could not turn off discharge MosFet.

something that is important for driving mosfets,  is that more powerful the mosfet is and more capacity Gate Source is important.
So you need have to charge the small capacity between Gate Source very quicky, and when you want it to turn off quicky, discharge it quickly.
Each IRL1004 has 5330 pF between Gate and Source, if you put 10 Mosfets in parallel you have 10*5330pF= 53.30nF, it is for that there is IC Driver (to accelerate turn on and off).
If you want to do a good driver, the best is a push pull but it is very important to put a small resistor beetween push pull and the gate to limite charge and discharge current on the gate (10 or 6.8 Ohms).
In that configuration turn off and turn on time are the same.
If you charge with 470 Ohms and discharge with 10k, you will have somes problems.
In your design, don't forget to put a resistor between Base and Emiter for transistors (for 2 PNP it is very important).

I hope, it will help to improve your design.

Best Regards.

[kalpakchiev]

Indeed You are very helpful.

On bot schematics Resistor 0.01R is only for tests in Proteus and represents resistance of the circuit during weld (when output is shorted). In real PCB it does not exists.

On first schematic (old one) I see a problem with place of the charging transistor. I wanted to change the place of the load (capacitor) and to put the transistor at LOW SIDE. And trying this I come to second schematic (new one).

But the main reason to post my question is just that floating SOURCE of discharge transistor. I can not figure out how to fulfill both requirements : both charge transistor and discharge transistor to be at LOW SIDE.

In your design, don't forget to put a resistor between Base and Emiter for transistors (for 2 PNP it is very important).

Do You mean to add these two resistors R21 and R22

[kalpakchiev]

In this configuration I have to use P channel and N channel MOSFETS.

[borberk]

You are right and firing P FET will not work. Idea removed.

[kalpakchiev]

I am sure there is a way just i can't see it

Again this schematic with floating SOURCE. If during weld I also turn on charge transistor then discharge transistor will have GND. The voltage of the capacitor can't exceed max. allowed because it is discharging at the same time. But with this method  welding current will have to pass trough both the transistors, which I think is no good. There must be some other way

[borberk]

Isn't it so that in welding time weld current flows only through weld FET and main C and through charge FET only a current which is almost the peak charge current because battery is practicaly loaded with 1ohm+Rweld+Rcharg?