Diode question

[jonrev]

Wonder if one of the techie types might be able to answer a question about the rectifier diodes in a MIG welder.
My welder has a centre tapped secondary main transformer and each end of the winding has a rectifier diode. The outputs are then joined together and supply the torch etc. The centre tap goes to the earth clamp via the choke. Since each diode is half wave rectifying it's phase of the supply from the transformer could someone tell me:-

1) Is each diode handling the full output voltage of the welder (I assume it is?)
2) Is each diode handling the full output current of the welder (I can't decide!)

I admit this is more out of interest than a need to know but it occurred to me that if the output rectifier should fail I'd like to upgrade with a couple of beefier components but am not sure what specs I should look out for. ie would a 140 amp welder need 140+ amp diodes and would they need to be at the highest output voltage the machine runs? I know it would be safest to assume the highest figures but cost of parts rises very fast with voltage and current so lower figures would mean cheaper parts!

Got my SIP 140 apart again to play with the wire feed- PWM speed controler parts have arrived so that will keep me busy for a while. Just had hernia op and too much time on my hands???

Cheers to all

 

[shenion]

Each diode is handling the full peak-to-peak voltage. That is if the transformer is 17VRMS, the diode needs to withstand 48V.

Each diode handles the full current but only for 50% of the time. So the average is half.

For a 140A welder I'd want at least 100A diodes. They need to withstand large surge currents. Preferably 200V or more. you will need to add clamping if the voltage is a less (I'dd ad it anyway.)

See if you can find 120A diodes. I think the next step up is 160.

 

[scarecrow]

I doubt if it's a half wave rectifier it might look like there is only 2 as they might be back to back
is as far as I know the full voltage and amps go through the rectifier

 

[rtbcomp]

Two diodes with a centre tapped transformer give full wave rectification.

 

[PTvor]

1) Is each diode handling the full output voltage of the welder (I assume it is?)
2) Is each diode handling the full output current of the welder (I can't decide!)

From what you describe, and bearing in mind Shenion's example of a 17V transformer, (17-0-17) I'd say an absolute minimum of a Peak Inverse rating for the diode of 17x1.41=23.97V (It's a sine wave and has an RMS value and a peak value, the 17V is going to be the RMS (sort of average) value). However, no one would cut things that fine and there are mains spikes, so say double it at least for comfort - 48V is the practical minimum by my estimate, the same as Shenion said. You can't buy a 48V PIV diode and 50 is the next best choice so it's 50V, if you can get higher rated ones for not much more go for them.

As for current, diode ratings are for continuous current. Short pulses can be 10 x the average rating. On a continuous basis, at 140A total, each diode would take half the current, but once again you wouldn't cut things that fine. If the output was shorted you'd want the transformer to warm up and the trip to go rather than the diodes fail. 100A seems like the absolute minimum to me, higher is better if it won't break the bank. Heatsinking must come into it and diminishing returms.

Practically, note the type numbers on the diodes they used and look them up. Write the types and their rating inside the case with a felt tip. If they fail, get the same or better with regards to current and voltage, bearing in mind the form factor. If they fail, you may not be able to read the type numbers.

 

[shenion]

The diodes need twice the peak voltage. The output would swing to +23.9v and then -23.9v.

I have seen 100V diodes used, i prefer 400V as there are large spikes especially on low-end welders with little filtering or clamping.

 

[PTvor]

There's a centre tapped transformer with the centre tap to earth. Consider one end of one winding open circuit going from +24V to -24V peak at 50-60Hz. Now consider a diode connected as in a full wave rectifier arrangement. with a load back to earth, cathode to earth. On one cycle it does not conduct and has -24V across it, on the other it conducts and has 0.65V across it.

That's with a purely resistive load.

But, with a load which maintained the O/P at 24V, and there's a choke in this and at least stray capacitance which could be significant under no load conditions....

You're right, or at least as near as makes no odds. Each diode has to take at least twice the peak to peak.

Mains spikes on top and 50 V for the diodes is looking far too low. 100V or higher if you can afford it,

 

[jonrev]

Thanks guys. Amazing isn't it that what would seem to be a fairly simple 'current/voltage in a circuit' type question actually involves so many parameters when you look at all the components and variables in a welding machine. Hadn't even considered the inductance or capacitance although the idea of adding a capacitor set to smooth the weld had been in my mind after reading Scotts web pages.
Was actually tempted by a pair of Schottky diode modules (200 amp, 45 volt) on the 'Bay' and did bid a few quid but wasn't sure if Schottky's were really suitable so lost out to a more enthusiastic bidder. Dare say someone will now tell me they would have been the 'dogs' and worth a mint?
The other thing I've been reading up on is diodes in series and it seems they can indeed be used in this way to increase voltage capability. Shame paralleling them is a no-no as a cluster of little ones would probably be way cheaper than big beasties.

 

[PTvor]

If the diodes work, why mess about with them? It just seems like needless expense and maybe new heat sinks with some risk of getting it wrong, for no benefit.

I can't see that fitting Schottky diodes would make much of a difference.

Just had hernia op and too much time on my hands???

Hands itching on the soldering iron is likely one of the side effects of recovery. Fight it!



.

 

[jonrev]

No my hands are always itching to take things apart and play with them. As I said I wasn't desperate to change the diodes but just thought it would be worth looking out for spares while I'm scouring ebay. Sometimes real bargains do come along. My SIP is actually welding pretty well but I still study the auction listings 'just in case'.
Interesting feedback from you guys and thanks for that. I enjoy scanning these forums and have learnt a lot in the process. Isn't this interweb jobby amazing? Almost nothing you can't get information about with a few minutes clicking as long as you can sort the wheat from the sh-t. Links to UTUBE instuctional vids are fascinating ranging from sensible and serious to downright wacky.
Ah well back to the soldering iron.
Thanks to all.

 

[Wedg1e]

I'm with you Jonrev; just because something isn't broken doesn't mean you shouldn't 'fix' it