Selecting Capacitors

[I'mClueless]

Hello,

I have an automotive alternator that's rotating at 6000 RPM and the 3 phases are fed through a Bridge Rectifier. That creates DC current, but with a 18kH (6000 x 3) ripple. THe amperage will be 270 but I do not know exactly what the voltage will be, however I know that it will be less than 24V. The application is a welder and my questions are below:

Would I calculate the capacitor uF saturation time by dividing 18,000 by 1,000,000?

How do I select a capacitor that will not fail from heat?

Would it be better to use a couple of Hi Amp 12V batteries wired in a series?

There will not be a conventional voltage regulator, however I will control the voltage by controlling the current to the fields. When I strike an arc than then the fields will be excited and the motor RPM will increase.

Yep,
I'mclueless

 

[mneary]

6000 rpm does not produce 18KHz from three phases. It's more like 300 Hz. (revolutions per minute, cycles per second)

Your ripple in three phase rectification might not be large enough to bother filtering if all you're using it for is welding.

 

[crutschow]

I agree with mneary. The small amount of ripple from a 3-phase generator should not affect the welding process. There's no reason to filter it.

 

[tcmtech]

In regards to the ripple frequency most bigger alternators are 12 pole designs so, (12 poles x 6000 RPM x 6) / 60 seconds = 7200 Hz

To get a fair weld from an alternator you will need a reactor coil inline with one lead. Its basically just a big choke that helps smooth out the welding current and stabilize the arc.
If you are just using the field current limiting method to control your output amperage you may likely see over 100 volts open circuit. Some of the bigger commercial alternators have surprisingly high open circuit voltages at higher RPM's while unloaded if they dont have some form of voltage regulation system which is good for welding purposes.

As far as a capacitor goes if its a stick welder design you really dont need it. The reactor choke is far more important.

 

[BrownOut]

A six pole alternator ( 3 pole pairs ) would give F = 6000X6/120 = 300 Hz

A 12 pole alternator (6 pole pairs ) is F = 6000X12/120 = 600 Hz

A 24 pole alternator (12 pole pairs ) is F = 6000X24/120 = 1200 Hz

 

[I'mClueless]

tcmtech, BrownOut,

You guys have provided some very enlightening information. The Leece Neville alternator does in fact have 6 north and south poles. So I'll take another stab at calculating the frequency of the ripple.

Alternator revolving at 6000 RPM / 60 = 100 Revolutions per Second
100 * (6 N & 6 S poles) = 1200
1200 * 3 phases = 3600Hz

The welder will be used for both Stick and TIG welding. If my calculation is correct then with stick welding I agree that the ripple should not be a problem. However with the TIG welding I cannot formulate an opinion because I have no experience in that field. I should easily pick it up because I have years of experience with acetylene welding. Both types of welding require the ability to work a puddle.

tsmtech,

You stated: "you will need a reactor coil inline".

My CNC machines use line reactors on the incoming service just prior to the Variable Frequency Drives and I was under the impression that the purpose was for cleaning up incoming line noise.

I don't relate to the usage in the welder power line.

Yep,
I'mclueless

 

[tcmtech]

Alternator revolving at 6000 RPM / 60 = 100 Revolutions per Second
100 * (6 N & 6 S poles) = 1200
1200 * 3 phases = 3600Hz

Your right. I did my math in my head without reasoning everything out. 3600 Hz ripple is right for a 12 pole 3 phase alternator at 6000 RPM.

As far as the regulation goes you will have a bit of power drift with a simple current limited exciter system. I built a 100 amp one back in high school that worked surprisingly well once I found out about the need for the reactance coil.
I just ran the output through the secondary windings on an old 30 amp battery charger. It was a dual half wave with a center tapped secondary so I just reconfigured it to have both halves parallel for 60 amps.
Being the duty cycle was low and its DC current it worked well enough. The old battery charger windings would overheat about the same time the alternator did.

I dont recall the exact manufacture but many years ago (1960's or 70's maybe) there was a factory built portable welder unit built out of a small two cycle engine and a slightly modified Delco alternator with a big inductor coil on the output.

They where an off red color but where not made by Lincoln, I dont think. They welded surprisingly nice but screamed like a raging bee when you struck and arc.
One of my friends had one I tuned up for him many years ago and as far as I know it still works!

Most TIG work is done with AC and is very fussy about its power source and control so don't expect much from it if you are going to use an alternator for a power source. Ideally TIG needs an additional high voltage and high frequency superimposed on the main arc to help stabilize it and thats not so easy to add on to a homemade alternator welder.

 

[I'mClueless]

Your right. I did my math in my head without reasoning everything out. 3600 Hz ripple is right for a 12 pole 3 phase alternator at 6000 RPM.

I can easily excuse my ignorance when it comes to electronics, however I embarrass myself when I make an error in a simple calculation such as the one that I made in the original post.

As far as the regulation goes you will have a bit of power drift with a simple current limited exciter system.

One of the more desirable features of TIG welding is the ability to modualate the power while welding. That is accomplished by using a foot pedal much like used with some musical instruments.

My power pedal utilizes the internal mechanism out of a Casio keyboard volume pedal. I replaced the 50k Ohm POT with a 10k Ohm POT and then installed the mechanism in an aluminum housing that I made to replaced the Casio plastic housing. I copied this concept from the Weld Talk forum and then Kicked It Up a notch by building my own housing.

I built a 100 amp one back in high school that worked surprisingly well once I found out about the need for the reactance coil.

I failed to mention that I'm running the alternators AC current through a 300A/1000V 3 phase bridge rectifier. Does that eliminate the need for the suggested line reactor?

I just ran the output through the secondary windings on an old 30 amp battery charger. It was a dual half wave with a center tapped secondary so I just reconfigured it to have both halves parallel for 60 amps.
Being the duty cycle was low and its DC current it worked well enough. The old battery charger windings would overheat about the same time the alternator did.

I chose the 270A Leece Neville so as to avoid running against the maximum output capability. Several professional welders have stated that while their machine is rated at 250A they seldom use more than 160A. There are several companies that offer 160A and 200A machines so I can only assume that the experienced welders prefer not to buy machines that would require being ran at it's maximum output.

With an auxiliary fan for cooling the alternator combined with a 50% demand on it's capability I hope to have a 100% duty cycle.

Most TIG work is done with AC and is very fussy about its power source and control so don't expect much from it if you are going to use an alternator for a power source.

Welding aluminum requires the use of AC current along with several variable parameters and yes they say that it can be very fussy. On the other hand welding steel rquires the use of negative DC current with power being the only variable parameter.

Ideally TIG needs an additional high voltage and high frequency superimposed on the main arc to help stabilize it and thats not so easy to add on to a homemade alternator welder.

I have already built a High Frequency arc starter and stabilizer which consist of France 15kV transformer, a couple of spark gaps, 590pF.30kV capacitor, 1 Ohm/10W resistor and the inductance coil.

Now that I have switched to the alternator I will have to build another HF box that ultilizes an automotive coil for the high voltage source.

I don't mind building another HF box because the existing unit scares the heck out of me.

Yep,
I'mClueless

 

[tcmtech]

I failed to mention that I'm running the alternators AC current through a 300A/1000V 3 phase bridge rectifier. Does that eliminate the need for the suggested line reactor?

To get a good weld puddle the reactance system is nessisary. How much inductance you need will depend on what and how you weld.
If my specs sheets are correct the Leece Neville 270 amp alternators have diodes rated for considerably more than 300 amps output anyway and I believe the stock diodes may already be 1000 volt rated. I have ran the ones on my wind generator against my 1000 volt megger to see what the PIV breakdown was and it couldn't get them to individually reverse conduct on the 1000 volt range so I figured that was good enough for a system that could possibly produce 400 volts open circuit.

One of the more desirable features of TIG welding is the ability to modualate the power while welding. That is accomplished by using a foot pedal much like used with some musical instruments.

My power pedal utilizes the internal mechanism out of a Casio keyboard volume pedal. I replaced the 50k Ohm POT with a 10k Ohm POT and then installed the mechanism in an aluminum housing that I made to replaced the Casio plastic housing. I copied this concept from the Weld Talk forum and then Kicked It Up a notch by building my own housing.

SO what method of controlling the field power are you using being the pot is a 10K? If you are just biasing the feedback on the regulator the high voltage spikes and higher voltages that welding needs will likely kill it rather quickly.

Welding aluminum requires the use of AC current along with several variable parameters and yes they say that it can be very fussy. On the other hand welding steel rquires the use of negative DC current with power being the only variable parameter.

As far as how someone welds and what they weld with I have learned to basically just not care. Half the "expert certified welders" I have ever met and known weld one way and say thats the only correct way to weld and the other half of the "expert certified welders" weld a completely different way and say thats the only right way. My opinion is that just use what ever works for you.
What ever works works and how you do it is your business not mine.

I used to work for Praxair as the welding service tech until corporate shut down the service department for this region because it looked good by the numbers on paper.
I liked that job and was good at it too but at least I made copies of all of my service files when they shut down so I have a great data base (about 5.7 gigs) for almost all of the common welders and plasma cutter systems ever built.

Yep,
I'mClueless

You dont sound to clueless to me if you got this far.

 

[I'mClueless]

tcmtech,

WOW! I really hit the jackpot when you signed on.

To get a good weld puddle the reactance system is nessisary. How much inductance you need will depend on what and how you weld.

Well I have some small 3 phase line reactors that I salvaged out of an old Fanuc control and then I some large TCI 3 phase line reactors that were amongst some components that I purchased at an auction. The small reactors don't look like they would be capable of handling 160 Amps and the large reactors look like they could loose 160 Amps somewhere between the input and output.

If my specs sheets are correct the Leece Neville 270 amp alternators have diodes rated for considerably more than 300 amps output anyway and I believe the stock diodes may already be 1000 volt rated. I have ran the ones on my wind generator against my 1000 volt megger to see what the PIV breakdown was and it couldn't get them to individually reverse conduct on the 1000 volt range so I figured that was good enough for a system that could possibly produce 400 volts open circuit.

Prior to encountering you my only source for information has been reading articles on the Internet which were detailing similar projects by people that probably know just about as much I as I do. That's somewhat of an intellectual loop.

All but one of DIY builders replaced the standard automotive diodes with non-avalanche diodes from Transpo. One of the seemingly more advanced DIY builders gutted his alternator and used six 300A/1000V diodes to build a bridge rectifier so I thought that my 300A/1000V 3 phase bridge rectifiers would be right for the task. Come to thing of it he was using a 160A alternator so perhaps I should use two BRs wired in parallel. The BRs that I have are mounted in a 6 x 6 heat sink with large cooling fins for heat dissipation.

SO what method of controlling the field power are you using being the pot is a 10K? If you are just biasing the feedback on the regulator the high voltage spikes and higher voltages that welding needs will likely kill it rather quickly.

I'm using a 25HP Kohler engine to power the alternator. I planned on using 12V from the engines electrical system to excite the alternator fields. That way I have a realatively constant voltage to work with. The only regulator will be the power pedal.

As far as how someone welds and what they weld with I have learned to basically just not care. Half the "expert certified welders" I have ever met and known weld one way and say thats the only correct way to weld and the other half of the "expert certified welders" weld a completely different way and say thats the only right way. My opinion is that just use what ever works for you.
What ever works works and how you do it is your business not mine.

Yes you are correct, the end result is all that counts, but only if it will pass a penatration and stress stress test.

I used to work for Praxair as the welding service tech until corporate shut down the device department for this region because it looked good by the numbers on paper.
I liked that job and was good at it too but at least I made copies of all of my service files when they shut down so I have a great data base (about 2.5 gigs) for almost all of the common welders and plasma cutter systems ever built.

I spent 45+ years in all aspects of the machinist trade. My status was retired but that changed when Fannie May & Freddy Mac crashed the US economy which resulted in my retirement account loosing 43% of it's value. I had to re-open the shop and work is hard to find now and when found it's so cheap that it's hard to rationalize taking it.

I'm an old man that doesn't have many years left to live so I won't suffer long if our First Colored* President manages to RAM his socialism down our throat.

* Colored: RED

Yep,
I'mClueless, but I sure as hell did not vote for Barack Obummer

 

[tcmtech]

I'm using a 25HP Kohler engine to power the alternator. I planned on using 12V from the engines electrical system to excite the alternator fields. That way I have a realatively constant voltage to work with. The only regulator will be the power pedal.

Not quite following that one. The 10K pot obviously can't handle the 12 volts at around the typical 4 -5 amps that the field will take so I assume you have some sort of voltage or current regulator circuit it controls.

 

[I'mClueless]

The 10K pot obviously can't handle the 12 volts at around the typical 4 -5 amps that the field will take

That was a simplistic explanation. I have circuit schematic that illustrate the use of transistors with a potentiometer to avoid frying the POT. At this time I cannot intelligently describe the circuit so I reduced it down to using a POT.

On another note: I found a web sight where the DIY welder builder is using both a reactor coil and capacitors. DIY welder - Build your own portable MIG/TIG/ARC welder I e-mailed the guy to see if he would share his engineering. Mean while I have been doing online research for how to size a capacitor and so far I lapse into a mental overload about half way through the first paragraph.

Yep,
I'mClueless

 

[tcmtech]

For MIG (wire feed) you need a constant voltage and rarely need to go over 40 - 45 volts peak. This type of welding may benefit from the capacitors to some degree.

SMAW (Stick) and TIG prefer a highly variable voltage and a fairly stable amperage and typically run around 25 - 40 volts at the arc unless the rod sticks.

Here is the service manual for the Miller Legend 301G engine driven multipurpose welder generator. It uses an independent 120 Hz three phase alternator system to produce the welding power.

This will give you an exact referance to how one of the top alternator based engine drive welders is designed and what is and is not used on the output stages.

 

[I'mClueless]

WOW! HOLY CRAP BATMAN I GIVE UP!

Yep,
I'mClueless, totally that is!

 

[tcmtech]

Here is the simplest engine drive type multi function (SMAW, TIG and MIG) welder system service schematic I have that is based on a single phase alternator and basic rheostat type power control system.
The Miller Bobcats are well rated and make good solid welds in all ranges plus I own one and have always been happy with it.
If you can't handle the circuitry and theory of operation for one of these then maybe you should reconsider building your own welder and just spend the money and buy one.

The Miller Bobcat 225's and 250's usually sell for $400 -$1000 for fair to good used ones Plus you get 8000 or 10000+ watts standby power as well. You might even pay less if you can get one that has an engine problem or electrical glitch!

The units with the Onan engines have a notorious problem with the solid state ignition module going intermittent. They start up and run good for a few minutes then just shut off until the module cools down. Then they will start right up and run again for a few minutes until it overheats again.
I got a machine in near new condition a few years ago with only about 450 hours on it because of that for only $280. Then fixed it and resold it two weeks latter for $1500 on a one hour and $95 repair.
I also have an older one sitting by my shop with a complete engine overhaul that I picked up for $100 because of the same problem but I think I am going to keep that one.