Picaxe spot welder controller

U

UAX

New Member
I'm new to Picaxe and would like to start with this project. I don't have much experience in programming anything except a little Visual Basic on the PC. I have already experimented with the Picaxe Editor / simulator and I'm suprised how simple the basics are.

I've built a battery tab spot welder which I'd like to upgrade with a Picaxe controller. The welder uses a 12V car battery, 10x irf1324 mosfets and a MCP1407 mosfet driver which I'd like to control with the Picaxe. The Picaxe should generate a short pulse ( a few ms), a short pause and a welding pulse (5-50ms). The pulses will be adjusrable with a potentiometer but that's pretty simple.

My questions are:

- Will the Picaxe generate these short pulses accurately (enough)?

- The mosfet driver needs a >1.8V logic HIGH and <1.3V logic LOW input so it can be connected directly to an output of the Picaxe? Do I need a pulldown resistor on the Picaxe output?

Thanks!
 
C

Circuit

Senior Member
Welcome to the PICAXE forum - and with such an interesting project!

Yes, you can generate the short pulses; the PULSOUT command is calibrated in 10 microsecond units at the slowest clock-speed. (See http://www.picaxe.com/BASIC-Commands/Digital-InputOutput/pulsout/ )
No, you should not need a pulldown resistor on the PICAXE output; the logic LOW and HIGH specifications are within your limits. (edit: I thought that the intervening driver would hold the circuit whilst the PICAXE pin floated on power-up but I stand corrected).
My only thought is that having such a huge current pulse running near a PICAXE chip could give some problems with interference. The control circuit may need a metal enclosure and be prepared for some unexpected resets.

But you certainly have caught my interest with this project; could you post a schematic of the device as you have it right now? A pen-on-paper sketch can be photographed and the .jpg uploaded to this site if you are not using formal schematic-drawing software. Also, perhaps, a more detailed description of your general arrangement; I would rather like to follow in your footsteps on this one.
 
Last edited:
hippy

hippy

Technical Support
Staff member
A pull-up/down resistor is highly recommended because a PICAXE pin can be a floating input when first powered up, during download or it may accidentally or inadvertently be set as an input as the program runs. You must ensure the MOSFETs do not accidentally trigger unless the PICAXE has control of those.

You may also have to take other precautions to prevent inadvertent MOSFET triggering.

It may be better to use one ( or fewer ) high current MOSFETs rather than more lower current MOSFETs in parallel. Those with more experience with MOSFET devices may be able to advise on that.
 
U

UAX

New Member
I will give you more information about the device + a schematic later today, I'm pretty busy right now. The spot welder is working already and a similar one has been built by other people so I won't change anything about the fets, etc.

I'm a little worried about possible problems with the picaxe too, especially as it would be located in a metal enclosure together with the fets! I think I'll need some serious shielding around the chip...
 
G

Goeytex

Senior Member
- Will the Picaxe generate these short pulses accurately (enough)?
Click to expand...
Yes

- The mosfet driver needs a >1.8V logic HIGH and <1.3V logic LOW input so it can be connected directly to an output of the Picaxe? Do I need a pulldown resistor on the Picaxe output?
Click to expand...
There should be a pulldown resistor either on the Picaxe output pin or pulldowns on the individual Mosfet Gates ( or both). Why? To assure that the gates of the MOSFETS are not floating during the short power up time. If the gates are floating the Mosfets will conduct and there will be a short duration output pulse at power up.

If it were my circuit. I would have a 10k pulldown on the Picaxe output pin as well as a 10K to 47K pulldown on the gate of each of 10 Mosfets. It's good insurance against unwanted glitches at power up, and will keep the Mosfet off in case of a Mosfet Driver open circuit failure.

To help isolate the Picaxe from switching noise, place a series resistor from the Picaxe output pin and the Mosfet Driver inputs(s)
1K might be a good value. The ideal resistor size can be calculated by looking at the MCP 1407 Datasheet. The MCP1407 needs 130uA to switch properly so make sure the resistor allows each mosfet driver to get about 150uA
 
Last edited:
hippy

hippy

Technical Support
Staff member
UAX said:
The spot welder is working already...
Click to expand...
That is handy. You can build the PICAXE circuit, not controlling the welder but just regularly pulsing a LED for longer than it would when controlling the welder. That way you can check it is working and identify any resets or lock-ups. Connect it to the welder battery supply and see how it behaves when the welding arc is triggered. That should help you identify and iron out any major noise or interference issues.

One thing to watch out for is pick-up on the programming cable which may appear to the PICAXE as a download initiation. It is probably advisable to add a DISCONNECT command at the start of the program to avoid that and perhaps unplug the programming cable when using the welder.
 
W

WHITEKNUCKLES

New Member
For those interested in tab welding of batteries and electronic hardware the link is a shining example of photographic documentation.
Low voltage (2.6v) resistance welding can be the cheapest form of tab welding but there are links to definitive information that includes capacitor discharge.
There are convincing demonstration videos of battery welding and steel panel fabrication.
The software is for the Arduino but the circuit appears to be well suited to Picaxe using burst firing as all relevant parameters are explained.

Dave

http://www.avdweb.nl/tech-tips/spot-welder.html#h1-spot-welder-series-configuration
 
Last edited:
J

Jeremy Harris

Senior Member
I built a capacitive discharge battery tab welder a few years ago. It welds nickel tabs to cells pretty well, but as pointed out above, the induced fields around the discharge cables make anything metallic within a 100mm or so of them physically jump, so the potential for induced current interference is high. Just keeping the control circuitry well clear of the electrode cables, and using good decoupling, close to the Picaxe pins should do the job.

My welder just uses a large capacitor bank (around 1F), charged to between 16 and 18V, discharged in a single pulse through two copper rod electrodes into the resistance weld area. Not ideal, as in a perfect weld you should really have a short cleaning pulse before the main welding pulse, but it was cheap (eBay, for both the 1500A thyristor and the bank of large, low ESR capacitors, which are coupled up in parallel with 20mm x 3mm copper bus bars).
 
U

UAX

New Member
I haven't been able to continue this because I spent a whole day trying to program the Picaxe I bought! Now it seems my AXE027 download cable is putting out inverted signal, so I'll probably have to return it...

My spot welder is very similar this one, the design is very straightforward, cheap and simple: http://dgroebe.free.fr/punktschweiss.htm

Things I have changed are:

- I use to-220 irf1324's mounted to a metal enclosure

- I have added decoupling capacitors to the MCP1407 like show in the datasheet

- I use a car battery instead of the A123 pack

- Have made nice electrodes from brass tubes and the cores of an underground cable
 
J

john2051

New Member
Hi, I've been looking for a way to re-attach the tab to a 10Ah NiCd cell.
It looks like either version may work.
Using the capacitor discharge method, what are the chances of the electrolytics sustaining
damage and how would this affect their lifespan?
Thanks for an interesting project.
john
 
U

UAX

New Member
rossko57:267648 said:
UAX said:
I haven't been able to continue this because I spent a whole day trying to program the Picaxe I bought! Now it seems my AXE027 download cable is putting out inverted signal, so I'll probably have to return it...
Click to expand...
AXE027 is supposed to run inverted data, the Picaxe expects inverted data.
Have you worked through these?
http://www.picaxe.com/FAQs/Fault-Finding/
Click to expand...
I started a thread about the problem and the cable seems faulty. The inverted signal of my AXE027 is inverted, ie. it's not inverted. ;)
 
You must log in or register to reply here.
Top