Smoothing caps advice please

[johnrev]

I want to experiment further with my Topmig 150 by fitting capacitance between transformer and choke as the better welders have. I have 12 of these caps which are 4700uF and 63v but not sure about the other specs and if they would be suitable.
I would like to connect all 12 in parallel for a total of somewhere around 55000uF.If I understand previous posts it is suggested using a number of smaller caps is 'kinder' to each one so would these be likely to survive?

Thanks
John

 

[voipio]

Connected between rectifier and choke, of course. The combined capacitance would be in the right ball park, but those look old, so it would be worth checking the capacitance.

 

[Rayf]

If they have been on the shelf for a long time reforming would be a good first call.
Reforming Electrolytic Capacitors

Manufacturers
claim that most old electrolytics can be saved if the correct procedure is followed, regardless of how long they have been unused. Such capacitors must be "reformed". This process consists of applying rated voltage through a resistance (about 30,000 ohms, five watt) for five minutes plus one minute for each month of storage (see figure 6). As the capacitor reforms, the voltage across the resistor will drop (measured at the Xs in Figure 6). If that voltage will not drop below 10% of applied voltage after one hour, the capacitor is probably beyond help.

More info here:https://www.6v6.co.uk/vcomp/tech_tips/reform_caps.htm

 

[johnrev]

voipio thank you. I was concerned about the specs on the caps where it says 1/100Hz 12.1 A/50'C. Does this mean that at 100hz it will handle 12.1amps up to 50degreesC temp? Just out of sight it says 3.2amps at 85 degrees C.
Have no idea how that translates to an input from a rectifier to 12 caps in parallel being asked to deal with (up to) 150amps???
My guess is that I'd be asking a lot of them but maybe as they are probably from a time before the Chinese took over the electronics industry, they may be robust enough?

 

[johnrev]

Thanks Rayf that's really interesting. Will read the full info. No idea how old or time unused these caps are. I got them off fleabay probably ten years or more ago. Just tested one on a (Chinese?) capacitance tester and got 4963uF and ESR 1.3ohms although ESR is likely to be way out apparently due to large cap testing errors.

 

[Rayf]

They are/where a good quality component
12.1A is the ripple current they are rated for and a good spec. these caps would easily cost north of £25 each.
Just monitor the temperature after use and if it stays under ~45 deg C you should be OK, if they get significantly warmer they will eventually fail or become ineffective.
With the intermittent use seen in welding they should cope well. Do not bunch them up together allow for a decent passive airflow around them.
As for parallel connecting them, consider using some copper bus bar drilled to suit and some good beefy 10mm2 plus cable to the welder internals.

 

[johnrev]

Just read that article- fascinating stuff. Thanks for the advice and I'll do as you say re connecting them. Think I'll mount a second fan in the Topmig as it's getting more crowded in there with the other mods. The original is a fairly decent one but was mounted well inside the case and served to stir the air around rather than blow it out of the grill. I've put a ducting round it to concentrate airflow to the outside which it does well now. A second one sucking air in the front should make it even better.
Thanks again
John

 

[Rayf]

Your welcome.
just noticed we are in the same part of the country! I'm in Poringland.

 

[johnrev]

So we are. I'm in Wingland near Kings Lynn. Always loved Norfolk. Used to come up here to Lowestoft area with our caravan regularly from Essex and alway said I'd like to live here. Bought this place to renovate, met my current wife and now (almost) retired so life is good.
Take care
John

 

[a111r]

If they have been on the shelf for a long time reforming would be a good first call.
Reforming Electrolytic Capacitors

Manufacturers
claim that most old electrolytics can be saved if the correct procedure is followed, regardless of how long they have been unused. Such capacitors must be "reformed". This process consists of applying rated voltage through a resistance (about 30,000 ohms, five watt) for five minutes plus one minute for each month of storage (see figure 6). As the capacitor reforms, the voltage across the resistor will drop (measured at the Xs in Figure 6). If that voltage will not drop below 10% of applied voltage after one hour, the capacitor is probably beyond help.

More info here:https://www.6v6.co.uk/vcomp/tech_tips/reform_caps.htm

Thanks. I was pondering this very question earlier, as I have replaced a single 33,000 MFD smoothing cap on my very old MIG with a used but better rated component.

Would you say there's a 'lazy' version that's viable: e.g. turn on at low volts, feed the wire, stop and repeat, before arcing?

 

[Seadog]

That was the usual way, start with the voltage low then work up. However, the current thinking is that you start at the rated voltage but current limit, increasing the current in steps until cooked.

 

[johnrev]

Well caps are fitted (roughly) and I switched on the welder and pressed the torch trigger with my eyes closed?? No bangs and everything seems to be ok........ except!
......When trigger is released the wire feed just carries on churning away untill the caps voltage drops below a few volts when it clicks off. Takes maybe a minute to reach this point as I haven't fitted any bleed resistor yet.
Obvious really since the caps take over from the transformer and hold in the feeder relay. Obviously once the charge is depleted everything stops and once I fit a bleed resistor the time it takes for this to happen will be reduced to a much shorter value but it will still happen and even if only for a second or so must surely be a problem when welding.
Been trying to sort this out in my head but brain has seized so wonder if you can help me decide what to do.

The system on my machine is that the PWM motor drive circuit is constantly live and it's output to the motor is fed through a relay which is closed by the output from the original SIP speed controler which is now no longer doing the motor feeding. The welding transformer is still switched by the old PCB relay and is then rectified and the DC is connected across the cap bank through the choke then out to the business end of the welder.

My thoughts are that the cap bank will always be a problem if it's switching the feed relay as there will be some delay which I don't want. It should be possible to take the relay signal from before the rectifier by using a seperate small rectifier direct form the transformer output and operate the relay from here so isolated from the caps. A alternative would be taking a feed from the welding transformer input to a 240v relay to operate the feed motor but this would mean swapping the relay I have fitted to a 240v type. Am I covering the simplest ways to go or is there another way I'm not seeing?

Thanks
John

 

[Seadog]

Put a small PSU in that's dedicated to the wire feed motor. It's quite a common upgrade.

 

[a111r]

Mine didn't go bang either .... yet (a single 100v, 33,000 MFD cap). I've only fed the wire through, not welded.

Would a (bigger) resistor over the motor help?

 

[johnrev]

My feed motor runs from a PWM speed controler fed from an independant 20v PSU. These are both permanently 'on' when the machine is switched on and the output switched to the motor by a relay. The relay is powered from the old SIP pcb when trigger is pulled so powering on the motor. The caps are keeping the relay latched on unitill they discharge so keeping the motor running.
I appreciate that the relay alone is putting only a small load on the caps so taking a while to discharge them but even with a bleed resistor there will be a delay before the motor stops which will give wire stick out etc. The relay has a NC contact to short the motor for a brake but obviously only works when the relay drops out.
Got about 55,000mfd of caps wired in and no bleed resistor yet as I wanted to be sure nothing horrible happened when I powered on before going too far. Not yet tried welding as feed motor has it's problem.

John

 

[johnrev]

Well that's sorted. May not be the simplest way but I've used what I have in my bits boxes and it makes sense to me. Took a supply from the welding transformer primary feed and ran it to a small 24v 'brick' transformer then used this to supply the motor feed relay. Motor now stops instantly again and the cap bank holds about 35v at power setting '4' dropping maybe a volt in ten minutes.

Next job is the bleed resistor and it's relay. Would be easier if the welder had a nice big contactor but the potty little power switch relay doesn't look up to the job of draining the caps down to me so will use yet another one to do this job. I have some automotive relays rated at 80amps which look to be sturdy enough. Now looking for high wattage resistor/s to plumb in for this job. Back to the bits boxes.

After all this I'm just hoping it works to improve the welding. If not I'll be pulling it all off again which would be disappointing to say the least.

 

[hotponyshoes]

On a lot of welders the fan(s) keep running for a while after mains power is removed so I presume they bleed off the caps to the case fan?

 

[johnrev]

That's a nice idea- two birds with one stone. Only prob is that you need the caps emptied after each weld run or you've got a live torch which will arc and (in my case) make you jump when you touch the weld bench or work. Think it's mainly the inverters which run the fan after the machine is switched off (my little Banggood stick inverter does).

 

[johnrev]

OK a little puzzling to me now so wonder if any of you can explain the following-

All hooked up with the capacitors. There are 12 of them at 4,700uF each making around 55,000uF total. I have a relay dumping the charge through around 60ohms when torch trigger is released and this seems to work fine allowing the voltage to drop from 40v to 5v in about 5or6 seconds. I figure this is a fairly gentle discharge rate both for the caps and the relay that's connected to them. The caps have been wired into the machine using the wire which used to be the earth lead before I changed it for some thicker stuff. Probably about 6mm diameter. I used this in the hope it would introduce a small resistance to the shock of the instantaneous short across the rectifier when pulling the trigger.

Welding is OK with this setup and the sound is now more like an angry wasp than the sharp crackle it used to be. After doing some welds one after the other the resistors were just about warm, the caps were cold and there was no sign of anything getting hot or being stressed that I could see or feel. After a longish trial weld at various settings of maybe 18" in all I felt the rectifier and it was warm to hot. Then put my hand on the transformer and it was b.....y hot. Too hot to hold onto and so was the choke. Seems the original innards don't like the caps much. Having said that I've no idea if these were getting that hot before as I'd never felt them.

Now the thing I'm a little surprised about. It feels that the power is lower than it was before the caps. I'd set the welder at power 4 and wirefeed 5 to get a good weld before. Now it feels like I need power 5 and wirefeed 5 to get the same sort of result. All settings seem to need a higher welding current than before. I admit I haven't done much welding with the Topmig but I'm sure I'm not imagining the difference.
I also find that the weld is tending to 'pulse' at times. Again I'm no expert but I'm sure it didn't do this before the change. Perhaps the choke would need to be different with caps fitted. If so then that's beyond me entirely

Right now I'm thinking I will remove the bits and go back to pre-cap setup. The welds are not really any better and though the sound is different it's nothing like my old Eland or my Britweld. These are just so smoooooth in comparison. The fact that I need to use a higher power setting obviously means more stress on the cheapo SIP innards and the inrush when the caps are working must stress the rectifier more. All in all I think I'll be better off leaving this sows ear as it was. The wirefeed is now excellent and that's the bit that made it so useless before. The experiment with caps was interesting but the results don't warrant the possible drawbacks.

So if anyone has any thoughts on this I'd be delighted to hear them.

Thanks
John