Hillhater
100 TW
this seems to work .. 
:lol:
[youtube]UU7QC5Uby6M[/youtube]
:lol: :lol:
[youtube]UU7QC5Uby6M[/youtube]
Are we overthinking the tab welder ?
- Thread starter Hillhater
- Start date
kdog
10 kW
ha! Offer him $50bucks (with free shipping of course!) you'll be popping cans in no time!
jonescg
100 MW
AvE does some pretty funny videos 
skeetab5780
1 MW
nice and simple, reminds me of doctor bass stick welding with his a123 26650 pack back in the day
How would you do the anode of the cell?
How would you do the anode of the cell?
Hillhater
100 TW
....try the " Mk II"......skeetab5780 said:
[youtube]tlm-C_fiGS8[/youtube]
skeetab5780 said:
Good question. I like the super precise timing circuit too. I'm surprised the relay doesn't weld it's contacts.
Chalo
100 TW
I'm in the process of choosing a design and components for a simple battery tab welder. My inclination for the purposes of low cost and simplicity is to use a relay, momentary button switch, and a 12V starting battery.
It's easy to get a 12V mechanical contactor rated for 500A or more. But I reckon that such a device will have an abbreviated lifespan when used to deliver short pulses at high current, thousands of times. For this reason I'd prefer to use a DC-DC solid state relay, which is cheap and easy to find up to 100A rating.
So my question is this, for whoever's familiar with the issues: how much current do tab welders actually deliver? Sure, the short circuit current could hypothetically be as high as the current source and the cables allow-- but this is resistance welding after all. The spot welds and nickel strip function as power resistors that clamp the maximum current to something less than short circuit current. Is it plausible to use a 100A rated solid state relay to switch a tab welder?
These things can be placed in parallel for higher continuous current capacity, but the inrush current must not exceed the capacity of a single relay. I reckon I'd parallel at least a couple of them to be on the safe side.
Any suggestions? I suppose I can use a rinkydink SLA battery to limit current, but that's no help if it limits current too much to make the welds quickly.
It's easy to get a 12V mechanical contactor rated for 500A or more. But I reckon that such a device will have an abbreviated lifespan when used to deliver short pulses at high current, thousands of times. For this reason I'd prefer to use a DC-DC solid state relay, which is cheap and easy to find up to 100A rating.
So my question is this, for whoever's familiar with the issues: how much current do tab welders actually deliver? Sure, the short circuit current could hypothetically be as high as the current source and the cables allow-- but this is resistance welding after all. The spot welds and nickel strip function as power resistors that clamp the maximum current to something less than short circuit current. Is it plausible to use a 100A rated solid state relay to switch a tab welder?
These things can be placed in parallel for higher continuous current capacity, but the inrush current must not exceed the capacity of a single relay. I reckon I'd parallel at least a couple of them to be on the safe side.
Any suggestions? I suppose I can use a rinkydink SLA battery to limit current, but that's no help if it limits current too much to make the welds quickly.
Alan B
100 GW
These DC-DC SSR's are basically an FET in a plastic case. They should have an opto-isolator which will slow the pulse down as well, which may or may not be desirable.
Since the simple circuits seem to be marginal with six chosen FETs in parallel, the chances are that the DC SSR will not hold up. Currents are generally in the 1000-2000A range.
Paralleling a half dozen DC-SSR's would cost more than the common circuits, and there still needs to be some timing circuitry to make consistent timing.
There's probably too much resistance in them as well, which would protect them a little from the high currents, but make getting a good weld difficult.
The low cost Arduino based circuits would seem to be a better direction to put effort into.
But, in the interest of science, let us know how it worked out.
Since the simple circuits seem to be marginal with six chosen FETs in parallel, the chances are that the DC SSR will not hold up. Currents are generally in the 1000-2000A range.
Paralleling a half dozen DC-SSR's would cost more than the common circuits, and there still needs to be some timing circuitry to make consistent timing.
There's probably too much resistance in them as well, which would protect them a little from the high currents, but make getting a good weld difficult.
The low cost Arduino based circuits would seem to be a better direction to put effort into.
But, in the interest of science, let us know how it worked out.
I measured mine at around 1800A during the pulse. A 100A rated SS relay will be able to handle quite a bit more for a short pulse, but maybe not 1800A. A mechanical contactor will most likely weld the contacts on the first pulse (and not really fast enough for doing 8ms pulses). Lower weld currents with longer time might work but you run into electrode sticking issues. It would be interesting to know what parts are inside the 100A SS relay.
Addy
100 W
If you're using an SSR for an AC mains powered spot welder, it should be on the primary side of your transformer. The currents will be much smaller on the primary side. I use one of those SSRs with my spot welder and a variable timing circuit and it serves me well. I have not noticed the SSR getting hot, though I haven't done any marathon welding sessions.
e-beach
10 MW
I was watching the video below and that guy did a nice job with his battery tab welder. He has a second video that updates his system including the abandoning a digital timer.
As for what can last....? How much can a car solenoid take? The guy in the video below says thousands of pulses.
I also had the notion, for the sake of accuracy, to consider a small bank of capacitors to control the relay.
Something like this: Press a normally open switch to fill the capacitor/s. Let the switch go. Push a normally open switch that discharges the capacitor voltage into the relay for a quick and accurate burst of voltage that is controlled in duration by the chosen capacitance of the capacitor/s.
This way you will get the same amount of time in a repeatable manner. It would take some experimentation, but once the needed capacitance is determined your timing is set and adjustable by adding or removing more or less capacitance.
[youtube]o1NFbchHeM8[/youtube]
As for what can last....? How much can a car solenoid take? The guy in the video below says thousands of pulses.
I also had the notion, for the sake of accuracy, to consider a small bank of capacitors to control the relay.
Something like this: Press a normally open switch to fill the capacitor/s. Let the switch go. Push a normally open switch that discharges the capacitor voltage into the relay for a quick and accurate burst of voltage that is controlled in duration by the chosen capacitance of the capacitor/s.
This way you will get the same amount of time in a repeatable manner. It would take some experimentation, but once the needed capacitance is determined your timing is set and adjustable by adding or removing more or less capacitance.
[youtube]o1NFbchHeM8[/youtube]
Chalo
100 TW
I don't intend to use capacitive discharge, so I don't think I'll be seeing 1000A+ from a lead acid battery alone. It will only be as much current as 12V can shoot across the resistance in the conductive path, up to the limit of the battery if the resistance is that low.
My thinking is that with lower peak current from a battery rather than a capacitor, it's less important to get a super brief pulse duration. I'd like to end up with a tab welder that works by tapping a momentary switch to determine the duration, which can then be operated by feel for different cells and strips. I considered using an interval-on time delay relay to set a uniform and repeatable pulse duration, but I haven't found one with resolution in the range of fractions of a second.
My thinking is that with lower peak current from a battery rather than a capacitor, it's less important to get a super brief pulse duration. I'd like to end up with a tab welder that works by tapping a momentary switch to determine the duration, which can then be operated by feel for different cells and strips. I considered using an interval-on time delay relay to set a uniform and repeatable pulse duration, but I haven't found one with resolution in the range of fractions of a second.
e-beach
10 MW
I was musing about a simple discharge circuit being that no being is a robot, having a simple and inexpensive timer would be a nice assurance against an accidental lead thumb on the push button switch. Burning a hole in the nickle strip at the end of a parallel would be bad. 
Having looked at some simple momentary switch circuits in the past few minutes it looks like one would also need transistor to control the relay.
Having looked at some simple momentary switch circuits in the past few minutes it looks like one would also need transistor to control the relay.
Alan B
100 GW
The Arduino circuits are already pretty simple and cheap. A car battery, a few FETs in parallel, a $3 micro to generate a couple of pulses.
The FETs are lower voltage and lower resistance than the usual FETs used in controllers, and are often cheaper. Shorting a car battery can generate pulses in the 1000 amp region, so it takes a few in parallel.
Consistency of welds is important. Battery pack failures can be serious.
Are people really feeling that's too complicated? Or are they just unaware how simple they can be??
The FETs are lower voltage and lower resistance than the usual FETs used in controllers, and are often cheaper. Shorting a car battery can generate pulses in the 1000 amp region, so it takes a few in parallel.
Consistency of welds is important. Battery pack failures can be serious.
Are people really feeling that's too complicated? Or are they just unaware how simple they can be??
e-beach
10 MW
This one looks like one could use the good power mosfet's from a blown controller. The one he uses are:
[youtube]pQ53sk9ksxQ[/youtube]
Visual learning anyone?
[youtube]pQ53sk9ksxQ[/youtube]
e-beach
10 MW
Alan B said:
In my case I have never messes around with anything arduino anything so I ask you this....
How simple is it? Got Details?
Thanks!
Alan B
100 GW
The code is published, the tools are free. It is about as simple as loading some software, plugging in a USB cable, selecting a file and clicking the "download" button.
Remember, Arduino was designed by and for Artists, not techies.
Remember, Arduino was designed by and for Artists, not techies.
Hillhater
100 TW
remember Alan,..some of the posters on here have trouble figuring out how to solder !Alan B said:
And, i for one, would no more attempt to make a tab welder, than i would a soldering iron...not whilst there are very nice ones available for ~~$100 !
footbaldd
100 mW
e-beach
10 MW
Alan B said:
aahhh Alan B......and here I wanted to be spoon-fed this time...
But now I got to do some research.
Can't you do the research for me?
Chalo
100 TW
I think that programming Arduino is too high a barrier to getting this job done, if it can be just as easily done with straightforward discrete components.
Ideally, I'd use an interval-on delay relay with a dial from 0.1 seconds in 1/10 second increments or smaller. I'd use it to switch a solid state relay or group of them to drive a 12V load from a car or motorcycle battery. I'd trigger these with a momentary foot switch. That's three components and a battery, no programming required. The only wear parts would be the electrode tips.
If I can find a suitable delay timer, I'll do it that way. Otherwise, I'm going to try it even simpler, with me as the delay timer.
Ideally, I'd use an interval-on delay relay with a dial from 0.1 seconds in 1/10 second increments or smaller. I'd use it to switch a solid state relay or group of them to drive a 12V load from a car or motorcycle battery. I'd trigger these with a momentary foot switch. That's three components and a battery, no programming required. The only wear parts would be the electrode tips.
If I can find a suitable delay timer, I'll do it that way. Otherwise, I'm going to try it even simpler, with me as the delay timer.
e-beach
10 MW
I was think the same thing....
eBay has things like this https://www.ebay.com/itm/DC12V-LED-...e=STRK:MEBIDX:IT&_trksid=p2055119.m1438.l2649
The only way to know if it works is by purchasing it.
Anybody game? Unless something out there is better?
eBay has things like this https://www.ebay.com/itm/DC12V-LED-...e=STRK:MEBIDX:IT&_trksid=p2055119.m1438.l2649
The only way to know if it works is by purchasing it.
Anybody game? Unless something out there is better?
footbaldd
100 mW
Using the Arduino IDE on the Atmega chip is not that complicated. I find with most programming and command line, it is confusing for a few minutes, but then is like riding a bike. Especially with the endless online resources. I am using Albert's pcb board and programming, which is a lot more complex than what would be needed for a DC based control circuit. Using the Atmega, a momentary switch or foot pedal, and a SSR with batteries would be pretty easy to do.
Alan B
100 GW
The code is already written and published, loading it into the chip is about as hard as putting a file on a USB stick.
The best results come from dual-pulses so there are three times involved, though primarily the second pulse is the one you adjust. It need to be adjusted to compensate for your setup - the resistance in the battery and system, etc.
There are plenty of threads on ES and elsewhere describing these tools, or you can buy one for well under $100 in many cases.
It is certainly easier than trueing a wheel.
Search for DIY battery welder arduino and quite a few variations pop up.
Messing around trying to make a simpler one will almost certainly cost you more time and money, and you'll probably end up buying or copying a proven design anyway.
The best results come from dual-pulses so there are three times involved, though primarily the second pulse is the one you adjust. It need to be adjusted to compensate for your setup - the resistance in the battery and system, etc.
There are plenty of threads on ES and elsewhere describing these tools, or you can buy one for well under $100 in many cases.
It is certainly easier than trueing a wheel.
Search for DIY battery welder arduino and quite a few variations pop up.
Messing around trying to make a simpler one will almost certainly cost you more time and money, and you'll probably end up buying or copying a proven design anyway.
footbaldd
100 mW
If you just need to upload the code then it is a cakewalk. For a DC dual pulse timer that controls a digital output off of a digital input momentary switch, the code is about as simple as it gets anyways.
