Looking for advice on homemade spot welder design

...

This is the one that I was thinking about. Not in the DropBox after all:

formatting link
Bob

Reply to
Bob Engelhardt
Loading thread data ...

I was thinking of a simple half-wave rectifier - namely, a diode in series with the primary winding. This way you get a half-sine pulse in every cycle, then zero volts. Like i said, i never seen this before and would have to check on how it stresses the transformer.

Another option would be using a reactance in series (most probably a capacitor, to compensate a bit for the inductive load) to limit the primary current. The capacitor needed would probably need to be quite large (and mains rated) though.

Reply to
plisandro

is not going to stick. Is there a special rod that would not stick in use like this?

The tips won't stick as long as their combined resistance is lower than the one of the parts being welded, IIRC. Even then, they get deformed and worn out, and need eventual replacement. Usually these tips are copper alloy.

BTW, that very site started my curiosity on these devices :)

Reply to
plisandro

There is a reason why you don't see this done. And Awright explained what it is. But I will also explain and maybe you will understand one of the explanations. With a diode in series with the primary, the current will only flow one way thru the transformer. So all the magnetizing force will be in one direction. And the core will magnetize in one direction until the transformer is fully magnetized in that direction. After that there will be no flux change, and without any flux change there will be no secondary current. Even worse there will be no back emf to counter the voltage applied on the primary. So the only thing that will limit the current will be the dc resistance of the primary. So the current will be quite high and your fuse will blow. Your fuse as in the diode you have in series.

Dan

Reply to
dcaster

Caution please. Sounds like you are not into Electronics.

A cap in series of the transformer would likely - depending on size would reduce the impedance of the input and cause a short when power is applied.

A cap - who has a reactance equal to the reactance of the transformer would drop it, but a Transformer with a tap switch providing the mains power to the welder might help - but another large core transformer.

The best way is to put a high wattage resistor in the output and adjust it.

Transformers that electronics can control are typically RF or higher frequency than 50/60 Hz. In older machines they controlled the amount of metal in the core - the coupling between primary and secondary was movable and this provided the 'loss' of lines or power and the output would drop.

Just not an easy thing to do without a lot of bench testing and experiments.

Martin

Mart>> Not sure that I understand exactly what you are referring to here, but

Reply to
Martin H. Eastburn

Yeah, i see it now. Bad idea, thanks.

Reply to
plisandro

My bad - i was thinking about a parallel cap (as in power factor correction :), which of course will do nothing to reduce current through the transformer. I need to stop posting late at night...

An inductor as ballast could do, but it'd need to be near the primary in value in order to produce any noticeable effects - which means a big iron core or an even bigger air one. Still, an option.

A switcheable tap would be ideal, but i don't think i'll be able to find such transformers in those power ranges...

Not feasible in this design - the resistance would need to be comparable to the weld resistance, which is small fractions of an ohm and means a lot of power to dissipate. Switching it in and out of the system would be a problem too in order not to increase secondary resistance by much. The only practical way to control power is in the primary.

Anyway, it was just a thought. I also considered using a solid state relay, which has much smaller switching times, but SSRs able to handle

3kVA loads are expensive and require a lot of heatsinking.
Reply to
plisandro

Need to watch the tank frequency - less the wrath of the power grid overlords.

Sure it is - you just don't use the resistors I used. Carbon piles that are the size of a tower computer. The resistance is variable by turning a screw that squeezes the stack. We dumped > 1000 amps through ours - thought we were in trouble as we exceed the maximum suggested current of 800 amps steady. Our sense wires were on the pile so the wrist size of thousands of wires (copper) wouldn't drop voltage. The stack was smoking!

Not feasible in most shops due to carbon catching on fire :-)

I think the best way is to generate (like a switcher) it in HF and rectify it to be the DC you want. AC would have to be generated from the powerful DC or perhaps the internal AC.

Hope you saw the 5bears design - that seems reasonable and knowing him for some years and I bet it was in full working order before a web page was developed.

Reply to
Martin H. Eastburn

A method of reducing the size of a transformer used to control the input voltage of your welder transformer is to use a transformer with a high-current, low-voltage secondary and put the secondary in series with the welding transformer primary in bucking mode. In this configuration, the input control transformer power is only the bucking voltage times the input current of the welding transformer - way less than the full input power of the welding transformer (depending upon the magnitude of the bucking voltage, of course). You could enhance this setup by using a small Variac to control the primary voltage of the input control transformer - at least smaller than would be necessary to directly control the welding transformer input voltage.

An alternative to an expensive very high current SSR is a surplus dual SCR module. You place the two SCRs in inverse parallel and provide a low current switch or relay with series resistance between the two SCR gates to create a bilateral static switch. I don't remember details at this moment but could look it up in my 1960s era Westinghouse SCR manual if you are interested. Don't build anything based only on my vague description from memory above.

High current, dual SCR modules are a dime a dozen these days. One that I just happened to buy a few of on a family visit to the Los Angeles area last week is:

"DUAL SCR POWER BLOCK SW Semiconductor (ltaly) #STT 091/08. Current rating is 91 amps. Peak reverse voltage is 800 VDC. Dimensions 3 5/8" wide X 13/16" deep X 1 3/16" high. It contains two diodes which have a common anode/cathode conneciton and separate connections for the opposite anode and cathode. This allows the unit to be used in bridge type applications. All units have heavy duty screw type terminals and an electrically isolated thermally conductive baseplate. The mounting bases of all the units have two mounting holes. The connections to the gates of the device is via 1/8" quick connect tabs. RFE Stock #PD2202A $5.00"

Search for the stock # in the online store at

formatting link

With a little more sophistication you could turn this into an input power phase control, rather than merely a static switch. Note also that you have to be very careful about current surge in applying SSRs to control of inductive loads like motors. I don't know much about this myself, but was advised once long ago by an SSR manufacturer's tech rep. that an SSR used to control a motor had to be way over-rated to tolerate starting surge.

awright

Reply to
awright

A method of reducing the size of a transformer used to control the input voltage of your welder transformer is to use a transformer with a high-current, low-voltage secondary and put the secondary in series with the welding transformer primary in bucking mode. In this configuration, the input control transformer power is only the bucking voltage times the input current of the welding transformer - way less than the full input power of the welding transformer (depending upon the magnitude of the bucking voltage, of course). You could enhance this setup by using a small Variac to control the primary voltage of the input control transformer - at least smaller than would be necessary to directly control the welding transformer input voltage.

An alternative to an expensive very high current SSR is a surplus dual SCR module. You place the two SCRs in inverse parallel and provide a low current switch or relay with series resistance between the two SCR gates to create a bilateral static switch. I don't remember details at this moment but could look it up in my 1960s era Westinghouse SCR manual if you are interested. Don't build anything based only on my vague description from memory above.

High current, dual SCR modules are a dime a dozen these days. One that I just happened to buy a few of on a family visit to the Los Angeles area last week is:

"DUAL SCR POWER BLOCK SW Semiconductor (ltaly) #STT 091/08. Current rating is 91 amps. Peak reverse voltage is 800 VDC. Dimensions 3 5/8" wide X 13/16" deep X 1 3/16" high. It contains two diodes which have a common anode/cathode conneciton and separate connections for the opposite anode and cathode. This allows the unit to be used in bridge type applications. All units have heavy duty screw type terminals and an electrically isolated thermally conductive baseplate. The mounting bases of all the units have two mounting holes. The connections to the gates of the device is via 1/8" quick connect tabs. RFE Stock #PD2202A $5.00"

Search for the stock # in the online store at

formatting link

With a little more sophistication you could turn this into an input power phase control, rather than merely a static switch. Note also that you have to be very careful about current surge in applying SSRs to control of inductive loads like motors. I don't know much about this myself, but was advised once long ago by an SSR manufacturer's tech rep. that an SSR used to control a motor had to be way over-rated to tolerate starting surge.

awright

Reply to
awright

Heheh, no, not really. An easier way would be to use a patch of heavy gauge cable and soaking it in a water container to avoid needing to cool it after a few welds. Still, if it ain't portable it's not very... mm... handy :)

Reply to
plisandro

No sweat. Use two or more microwave oven transformers in a row so the secondary can be straight and run thru all the transformers cores, make a bend and run back thru all the transformers cores again. Then to vary the power just decide on how many transformer primarys to use.

Dan

Reply to
dcaster

What a cool idea! So simple and easy & so powerful. You're adding-up the kva of the transformers.

If I understand correctly, the magnetic field creating the voltage in the secondary is the sum of the fields in the separate transformers. You would use 1/2 the number of secondary windings to get the same secondary voltage, but twice the secondary current.

You could get a 5 kva spot welder by using 4 1250 "watt" microwave transformers, 2 pair of 2 in series, in parallel, with 240v input (240v at 20a = 5 kva+-).

Bob

Reply to
Bob Engelhardt

of the transformers.

It is neat, but physically arranging the transformers so the cores line up might be an issue, specially when short, low resistance winding is needed. Still something i'll definitely keep in mind though!

secondary is the sum of the fields in the separate transformers.

It is - you're harnessing the flux of both transformers in the secondary. The primaries appear (electrically) as a single one and you basically get an output with the sum of the power capabilities of both.

Reply to
plisandro

You know, for some reason i never considered this... a triac (dual- scr) rated at 600V 60A can be found for like $5 and will do the job nicely... in fact, more reliably than a relay, which i kinda worried wouldn't last long under the stress of quick switching.

Indeed - since the welder will be controled with a uC, adding a zero- cross detector in the mains and firing it in mid cycle (phase control) is rather trivial and allows for very precise and fine timing.

And yes, triacs are *very* sensitive of inductive loads, because back- emf when the current is swicthed on can (and will :) fry the device rather quickly. I don't quite remember the details, but this can be solved adding a capacitor & zenner in parallel to the primary winding so the voltage spike is clipped. Of course, everything connected to mains (z-c detector & triac) must be optoisolated.

I'll definitely go this route, thanks for the heads up!

Reply to
plisandro

On Jul 30, 3:24 pm, Bob Engelhardt

You got it.

Dan

Reply to
dcaster

I don't think that they need to be aligned. It's all a matter of the total flux enclosed by the secondary winding. The cores could be in different rooms and the flux still enclosed by the winding, to take an extreme and totally impractical example.

Bob

Reply to
Bob Engelhardt

Yes, but you'd have 20mts of secondary winding between them :) What i mean is that the configuration most likely to result in shorter wiring for the secondary would be something like:

+-------------------+ | +----+ +----+ | | | t1 | | t2 | | | +----+ +----+ | +-------------------+--- | | Secondary loop(s)

Like i said, still a good idea. Just a trickier implementation, but two 1,5kva transformers are cheaper than one 3kva...

Reply to
plisandro

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.