Variable DC power supply for alternator welder

[phoenix]

I have been trying to make a car alternator into a welder for my military truck. I have everything done except the variable power supply to power up the rotor coil in the alternator. I found someone that used a circuit that he said worked, but when I looked at the actual specs of the devices he used, I found out that it is overkill.

OK, first, I have a rotor coil on my alternator that is 1.4 ohms. I am using battery voltage, so I will say 14 volts max. so I should have a max amp draw of about 10 amps right?

My circuit is using a 10K pot that feeds the base of a 3055 transistor. The original circuit was built using a TIP 3055 which is a TO220 case. I dont know the amps out specs on this, but I had a TO3 cased 3055 laying around and the specs on that are 15 amps out. He was using the emitter of the TIP 3055 to drive the bases of two 2N5885 transistors in TO3 cases. From the emiter of each 2N5885 there is a 12 inch long coiled wire to act as a resistor so that the two transistors emitters can be connected together. each of these transistors according to the specs will output 25 amps each. To me, this is a waste since the rotor coil can only use 10 amps. Why could I not just use one TO3 cased 3055 transistor to drive the field coil directly? I am trying to make this circuit as simple as possible, because I will be showing a bunch of people how to make one. It sounds too good to be true that I could just use one 10K pot and one transistor to make a high amperage variable power supply.

I thought I should add that this supply should be variable from 0Volts to 12Volts DC and It should be able to put out at least 10 amps.

can anyone help me figure this out?

 

[(*steve*)]

Perhaps if you can post a scan of the circuit, or a link to the information you found...

 

[Resqueline]

At first glance his circuit may look overrated but there are some reasons to do it that way.
The 2N3055 may be popular and seems to be powerful, but looking closer on its datasheet it becomes clear that it's a really bad performer.
At 10A collector current you may need as much as 3.3A base current to get the voltage drop as low as 3V. For 4V drop you may get away with 2A. That's bad.
You won't get much more than 7V/5000/10=0.14mA from the pot. In total you'd need a current gain of 10A/0.00014A=71430. Thats a lot even for a darlington connection.
It can be done real simple but then you need better parts. I'd suggest using a 1k pot for starters. That brings the needed current gain into the realm of darlingtons.
Then it's "just" a matter of locating a darlington that retains a gain of 7000 at 10A and a low Vce sat (like 2V). I'm hesitant to spend my time doing that.
Maximum power loss: 7V*5A=35W which is no problem for any transistor but will still need a substantial heatsink.

Just qurious: do you remove or replace the rectifier bridge in the alternator?

 

[phoenix]

The alternator is used almost as is, but you have to remove the internal voltage regulator and use this circuit to power the rotor. alternators are 100% duty cycle so whatever the amps of the alternator is, will be full welding power. Some 3 phase rectifiers will burn out, but it is because they are "fused" to not go over 32 volts. At full rotor volts, the alternator can put out over 140 volts. I can dial that in to 120 volts so I can run tools that have brushes, like grinders and drills, as long as they dont have varible triggers.

 

[Resqueline]

Ok, I see. Yes, I'd expect an unregulated alternator to easily surpass the voltage rating of their rectifiers. Very useful application btw..
So once dialed in the voltage will keep fairly stable with different loads? I guess the rpm's will need to be kept constant or else the output would change?

 

[phoenix]

Ok, I see. Yes, I'd expect an unregulated alternator to easily surpass the voltage rating of their rectifiers. Very useful application btw..
So once dialed in the voltage will keep fairly stable with different loads? I guess the rpm's will need to be kept constant or else the output would change?

yes, if the RPMs are kept constant, the voltage will be constant. You can really adjust the output two ways, the RPM of the engine, and the voltage being supplied to the rotor. The alternator takes about 1 HP per 10 amps of alternator output. you can get some heavy duty bridge rectifiers or you can just start out with an older alternator. The old Fords had external regulators and are a good choice for this setup. It takes less amperage to weld with an alternator since it is high frequency DC voltage being used. I have been told that this high frequency agitates the welding puddle and takes more porosity out of the weld. In effect, it allows a novice welder, weld as good as a professional. I have been told that a 85 amp alternator works very well on 1/4 inch steel. Obviously if you want some real power for welding, you need to get an alternator from an ambulance or find one on Ebay. I bought a new Ford alternator on Ebay for my military trucks normal charging alternator that is 200 amps and it cost me 99.00. That is less than half of what a rebuilt one cost at Car Quest. The one I am going to use for this welder I got for free, and is a chevy style CS144 that has been modified for an external regulator.

I am thinking that I want to put a small switch mounted onto the stinger of the welder so when I press the switch, I will instantly get full voltage to the rotor coil, therefore having more voltage/amperage to start the arc then once started, I can let the button go and let it back down to the setting on the dial. I also am thinking that I could use a power window motor from a car and put on it a set of metal drive wheels, then, buy a store bought mig gun off of Ebay, and make a box to hold a spool of wire. then I can clamp my welding stinger onto a terminal on the box so that I have an instant mig welder.

I am still having issues figuring out what to do for this variable circuit. I dont think the circuit will work to the full extent I want it to the way it is. I called DigiKey and they were not any help either. I am not knowledgable enough to figure this part of the circuit out myself unfortunately. I have looked on the internet and have not found any circuit that would provide the 0-12 volts at 10+ amps. I thought of even doing it my way, which is any way that I know how, and this is what I came up with: Using about 20 diodes, I can solder them all in series and then use a rotary switch to short the diodes one by one raising the voltage since each diode would drop the voltage .7 volts. The only problem with this Idea is that I would need a rotory switch that can handle 10+ amps. I cant find one. so I am stuck not knowing what to do.
From what you have told me, and that I can understand, I would be getting about 7 volts max out of the original circuit that I discribed.

I can do a full write-up on this welder when I finally get it working. I may just build it with the original circuit, but I just hate to do things twice when I know that it is not what I want.

 

[Resqueline]

Thanks, that's very useful information. I know it's an old concept but it's certainly one that stands being brought to attention.
I found a transistor selector over at st.com and that was the best one I ever saw. There I found the TIP142 to be the best choice for your application.
It will drop only 3V @ 10A and requires only 40mA base current to do that.
It's has a big case so it should be sturdy and easy to deal with.
A potentiometer of 470 (or 1000) ohms and (at least) 1W should be sufficient to drive it. Then you only need the usual diode across the rotor field.
Three parts plus a heatsink is not bad for a welder. Oh, you'll need an insulating mica sheet or something similar for the transistor also.

 

[phoenix]

Thank you very much for all your help. I can just use any old standard diode for the protection diode right? I would think that the surge voltage on this coil should be very small. I just bought all the parts except the diode, I have a ton of those around. At Digikey there is another of the tip142s--the tip142T this one has 15 amps out, but not as much gain. I went with the tip142gos.

I have a friend that is into hydoelectric dams, he owns several of them. He told me that all you need is a low RPM motor to have the water turn to make electricity.If you bolt on a 110 volt AC motor to your vehicles engine, as long as you have the motor going faster than its rated RPMs, it will produce 110 volts AC. All you have to do is have a push button to momentarily excite the field with 12 volts to start the motor producing electricity. I thought this was a really cool trick, but the weldernator will do a good enough job for 110 volt tools even though it is not AC.

This is the welder so far, I am going to install a double outlet in the front cover for the 110 volts and will have an on off switch for the main power. the first meter is for the DC voltage going into the rotor coil. the second meter is for measuring the voltage output of the alternator so I can dial it in to 110 or see where I am at for welding voltage. The third meter is the amp meter for the output of the welder. I have a shunt that will handle 500 amps. The meters are cheap Ebay items, but they should work

here is the truck I will be puting the welder in. I will sometime get all the parts I need for the winch I have for the truck and I will probably put an on board air compressor in it as well using a York air conditioning compressor.

I really appreciate the help. Thanks again.

 

[Resqueline]

Cool case! Just hope it'll work as intended & w/o problems.

I was unable to find a datasheet on the S variety but it sounds like you chose right anyway. The T type is beefier but not as good in other aspects.

If you ground the pot via a resistor 1/10th of the value (max) of the pot you'll get rid of the 0-1 dead-band in the start (where nothing happens) btw.

The diode will only get a kick when you release the trigger on the handle, or if the pot glitches while adjusting, so a 1N540X or even a 1N400X should do the job.
You could even get away w/o a diode, since there's diode in the transistor that will protect it (in this configuration), but it's best to short the kick where it originates.

I've heard that the field of an asynch motor will collapse quickly w/o some active excitation circuit to make sure the phase relationships stays right.

Cool truck also! That one gets around.. Hood reminds me of the '70s Jeep Cherokee? I have a '44 GMC 2-1/2ton 6x6 & a '43 Dodge 3/4ton 4x4 btw..

 

[phoenix]

Is this what you were thinking about? I want to make sure that we are on the same page as far as design. I may not know what I am doing, I have not done any component electronics for 15 years.

The military truck is a 1967 Jeep M715 and uses the civillian cab (modified for a flat windshield and convertible top) and front end. Jeep used the same parts from about 1963 until they stopped the full size Jeeps in 1991. The M715 isnt that popular or common since most people dont even know about them, but Jeep did make 30,000 of them from 1967-1969.

 

[Resqueline]

Yep, that's just the way I had in mind. Just beware that the actual transistor pinout is B-C-E.
I't not until I see it on paper paper though that I notice that use of the momentary switch could be hard on the pot if it's turned way down low. The 100 ohm resistor however does a fairly good job of keeping the potential short in check. I guess it'll be used for only a second or two and that also helps.
But do you think it would be problematic to use a changeover switch and three wires going to the handle? Or possibly the switch could drive the field directly.
I hadn't seen that jeep before, but then I haven't seen'em all..

 

[phoenix]

I had it a three wire switch in my original schematic, I forgot about that. I would think that when you push on the switch, the coil will get a on-off-on bounce that may not be too good for it. I am also concerned about the lenth of wire for the switch. I would be using something like 18 gauge wire but it would be around 100 feet, since my welding leads are at least 50 feet long. also, I bet I would have to worry about all the induced voltage having the wires running along taped to my main power cables. Then I would have to go with shielded cable------- I think I will just use it without the switch for now and see how it is. I just have never seen an arc welder with this feature and thought it would be cool just for the cool factor. I will keep you updated.