There are some articles written by more technically competent folks on converting an alternator to a welder. 

It seems that the built in diodes are often rated for only 32V, and will fry if used for the welder, which will peak over this voltage both directly and as a result of inductive kickback when the arc is broken.  Thus the need for replacing the diodes either in the existing fan cooled plate design or externally.

The usual application for an alternator type DC welder is for emergency (typically off road)  field welding as the duty cycle can be quite low and the weld quality expectations are equally low.  Clearly, you can't expect to sustain the welding power you need for regular work from an auto alternator but for emergency use it makes sense as the weight and size of adding an extra alternator to the off road vehicle is better than the other options.

You can beat this performance with your existing CS and welder set for low power, I think.  A friend ran his small gasless MIG welder on my Listeroid 6/1 without a problems; he welded  the thin wall steel tubing for my solar racks. 

Hi Stef,
The CS 6/1 is a bit puny for a welding source.  3/32 rods will be right on the edge; they run about 26 volts at 95 amps,  2470 watts (just over the sustained output of the 6/1) but you must also allow some losses in the inverter so you'd really need near  3000W continuous; that's more than the 6/1 can typically do.

No harm in trying since you're engine and not generator limited. You might try bumping up the rpms slightly if your flywheels can handle it safely.

That is an important safety issue, Veggie, and one I missed.  Thanks for catching it.  Important for tools which are older and may not be double insulated, so a safety ground is important for shock safety. The safety ground does nothing for modern double insulated tools, which have no ground connection at all.

An ohm meter to confirm internal neutral to ground bonding(connection) in the welder set is warranted. 

A decent capacity welder will require much more power than any hand tool,  likely well over 50 amps, so you could easily have two 25 amp 115V oulets on breakers, if you like.  A GFCI  type outlet would be a more valuable safety feature, in my mind, as on a job site, things can be rushed so some extra shock safety is a nice feature.  GFCIs save people, breakers save equipment.

Hey Veggie, Incandescent bulbs have heat persistence of the tungsten elements-  more with higher wattage as its thicker wire.  This is why larger wattage incandescent bulbs have less flicker.  My mains and fast reacting AVR are OK on 250 watt incandescent heat lamps but 60 watt incandescent bulbs are still bothersome to me. 

Some  LED bulbs have a regulated switching power supply in the base which will eliminate flicker IF there is adequate capacitance to store some excess energy to ride through the lows; thus some electronic persistence as there is virtually no persistence of the LEDs.  The problem is that the space is so tight that adding capacitance is hard to do.  So what we mostly get are minimalist LED bulb electronics designs that work adequately on grid power or inverters but have visible flicker on generator power.

I found some LED bulbs with SMPS in the base (I took them apart.) with full SMPS that operate well on my 120VDC (126 to 146VDC). I suspect they would not show any flicker on generator power.   I didn't use them as I couldn't tolerate the quality of light or the radiated emissions.  Alas, I just looked and cannot find them in my shop.  I think it was an older Phillips design.  Avoid any LED bulb which is marketed as "dimmer compatible" for generator power as that is going to be a flicker monster. 

With a bit of effort in finding a model of LED bulb with a true SMPS in the base,  it should pretty much eliminate the flicker on generator problem for those who tolerate the unnatural light spectrum of glowing phosphor (white LEDs are really miniature fluorescent lights; they have a dab of phosphor over a UV LED.

Bob, I concur, starting from scratch is the best way. Much efficiency, simplicity, and reduced home power EMI can be gained by having a single high quality SMPS for providing regulated DC to all home/office lighting fixtures.  You can even use some of the stock cheapo LED bulbs; some will often operate on say 160VDC or other DC voltage.  This gets you the advantage of eliminating regulating circuitry in bulbs, a common failure point, and allows for the best energy efficiency.
A couple companies in the US promote such systems for commercial/office lighting systems, but I don't think its really caught on. 

I expect most LED bulb power electronics are operating below 75% efficiency, and have 0.6 or below power factor.  The single regulated DC supply for all lighting addresses approach that, and of course would also eliminate generator flicker.

 

I would check the voltage of the generator output before planning to add outlets.
I seems likely that it would be a lower voltage at higher amperage for the welder, thus the big wires you report. 

Pulsed DC is not a panacea of compatibility, but it is an interesting solution.  The method of doing  PWM of the raw PV input to get RMS 120VDC is clever and minimalistic, since no filtration is required.  I appreciate learning of this method, Veggie, it's something I've never read about before. 

Induction motors will not run on pulsed DC.    Some switchers have bootstrap aka startup circuits that are AC dependent, and they will not start on pure DC.  About 80% work fine.  Pulsed DC should work for many of the ones that don't, depending on the bootstrap circuit.  Near 95% of switchers should be fine with pulsed DC.

LED lights/bulbs with switching supplies in the base that I've tested work fine on 120VDC.  I don/'t use them because I find the light spectrum and EMI bothersome.  I have a lifetime supply of soft white incandescent bulbs instead.

Hi MikeNash.  I'll try to answer your questions first.
"The SSR you sent the link to is, I presume, normally closed and the thermostat (eBay UK item number:394041622609) closes at the set temperature thus disconnecting the supply."

No, all the higher voltage DC SSR's I've seen are normally open. You need a thermostat which opens on high temperature, closes on cold.  Just as a standard water heater thermostat does.  Of course with some simple electronics we can change a thermostat output to NO or NC (normally open or closed if you must, but keeping things simple is nice.

"My three heating element ports are 2 at the top, one at the bottom, is there such a thing as a SSR that, instead of switching the power off, diverts it? That would allow the upper element to raise the temperature of the top third and then use the lower one as a 'dump' for the rest of the day?"

This is easy if you have a standard electric water heater thermostat which already has just such an dual temperature element diversion from upper to lower elements (at least here in the US). You'll have to add a second SSR, one for each element, and use the water heater thermostat to just switch the  12V to the SSR control.  Replacement thermostats of any rating will be fine, since you will only be swiching about 6 ma of 12V via the contacts. 

Again, many ways to do this electronically, like a couple thermisters, a zener reference and a single quad op amp plus "glue" (resistors and capacitors) but I'm trying to avoid custom electronics.  If you can't find what you need, I can help with plan B.

Veggie,
It is wise to oversize the ampacity of a DC SSR and there is no penalty for doing so.  The higher rated once will have lower on resistance, and thus for small currents will not need much heat sinking, if any.  I agree that some skepticism in the rated ampacity should be applied, as they are made of Chinesium and specification were written by the marketing department.  I have used them well under the rating and they do seem to work and hold up fine.