Bridge rectifier with capacitor and choke filter

tfleming

H-M Supporter - Gold Member
H-M Supporter Gold Member
Joined
Nov 21, 2015
Messages
226
This probably (maybe) is a wee bit off-topic for machining, however, it does cover basic electronics, and might help others when needing to do similar things. I want to get into TIG welding for giggles (already stick, gas, and MIG weld), but I also don't want to drop a load of cash (just yet). So, I am building a rectifier with filter to plug into my trusty old Lincoln tombstone 225 AC welder and turn it into a scratch-start TIG machine. Here is the schematic for it:

1532343710394.png

It is a basic full bridge rectifier, with 2 capacitors and a choke to smooth out the DC waveform. Will use at least 1, possibly 2 muffin pan cooling fans to keep the old girl comfy. Also, I will not be running this at "full throttle". Will probably never go over 150 amps on the input.

Input and comments are welcome, as I haven't started this yet.
 
Last edited:
Not knowing much about the power requirements for tig welding, I can say that the rectifier circuit that you've come up with should perform to the requirements you've listed. The caps on their own would give you a ripple on the D.C. output of only a few volts p-p. You didn't give a value for your inductor but I would think that anything over a few henries would do the trick.
 
Here little data

Dave New-relayHF-170-wiring-diag with capacitor.jpeg
bd978940c8ae7fa26db764a6fe1aac9c.jpg


Sent from my SAMSUNG-SM-J320A using Tapatalk
 
Not knowing much about the power requirements for tig welding, I can say that the rectifier circuit that you've come up with should perform to the requirements you've listed. The caps on their own would give you a ripple on the D.C. output of only a few volts p-p. You didn't give a value for your inductor but I would think that anything over a few henries would do the trick.

Tony, the choke came off a 150hp dc drive at 460volt. I am not sure what the henries rating is, but since they were surplus and sized for that application, I think it should work just fine. As far as TIG output requirements, they are pretty straight forward. Most welding I will do will be in the 80-150 amp range, so I oversized things a wee bit just to keep things from "smoking" (the listed input and output values are the specs of what this setup should be able to do). I might even put a switch in to eliminate the filter portion, as pulsed DC current has applications in TIG as well. Again, everything I bought for this is well under $100, and it will give me some flexibility with the AC welder without dropping multiple C-notes...........
 
This brings back memories of the good old days when all power supplies were designed like this. Entire books have been written about linear power supply design.

As I recall, choke (inductor) input filters were used for high current applications. A choke input filter has better voltage regulation than can be attained with with a capacitor input although the voltage will be lower. Both my Miller welders use a choke input filter.

A full wave bridge rectifier has a cycle period of 8 msec. When the voltage from the bridge drops, good filter will supply current until the voltage rises again. A capacitor does this by storing charge and discharging into the circuit when the voltage drops. The equations governing the charge/discharge are q = CV where q is the charge in ampere-seconds, C is the capacitance in farads and V is the voltage and i = dq/dt where i is the current and dq/dt is the time rate of change in the charge ( a little calculus here LOL).

The take-away here is the to sustain high currents, a large stored charge is required. The exact calculations are fairly complex but a back of the envelope calculation shows that to sustain a 50 amp current over an 8 msec. time would require capacitances on the on the order of a half a farad. Most of the farad class capacitors that I have seen are rated for lower voltage than would be found in a welder. Capacitors can be wired in series to increase the operating voltage at the expense of capacitance. Two 1 farad 12 volt capacitors in series would have an operating voltage of 24 volts and a combined capacitance of .5 farad. For a 75 volt maximum, you would need six capacitors in series for a combined capacitance of .17 farad.

Chokes, on the other hand, store energy as magnetic flux. Magnetic flux is proportional to current through the windings and it is fairly easy to design a choke which can handle the high currents that welders experience. When the voltage drops, the magnetic field begins to collapse developing a back emf which boosts the dropping voltage.

I have used old transformers by replacing the original windings with a single winding of wire capable of carrying the required current. An old battery charger transformer could be used effectively as a choke by just using the secondary winding.
 
Your theory is sound RJ. I taught electronics engineering courses for 32 years until I retired a year ago May. One of my 'specialty' courses was linear applications including power supplies, transistors, etc. One way to remember what the filters do is to remember that capacitors delay change in voltage while inductors delay change in current. As far as the amount of change in voltage on the caps, which is usually referred to as the a.c. 'ripple voltage', it boils down to a simple approximation of the DC current draw divided by (frequency X capacitance) where the frequency is the output pulse frequency of the bridge, in most cases 120Hz (double the line frequency). For the OP's circuit, assuming 150A nominal draw, this becomes 150A divided by (120Hz X 1F). Note the two .5F caps in parallel combine to give 1F total capacitance. So the peak to peak ripple out of the capacitor section becomes about 1.25Vpp. This relatively small change in voltage then produces a relatively small change in current which is then further filtered by the inductor. With this amount of capacitance the inductors job becomes easy.

Of course then all of this assumes a steady current draw of 150A. The welding process itself will cause large fluctuations in current, so oversizing the inductor is prudent in an application like this in order to account for the system variables. That's my story and I'm stickin' to it. ;)
 
RJ and Tony, thanks. One thing to keep in mind here is that I am looking to do some pretty basic TIG work to see if I want to invest the $$$ for a decent, TIG welder. This is definitely a "backyard" setup, but it am hoping it will produce a relatively stable DC current. Plus, it is a fun little project. If there are slight pulses, that is actually not a bad thing, as some of the high end TIQ welders have variable DC frequency capabilities (they are also $3k machines). Also, again, I plan on the 150A being the upper end of how hard I will push this. It will spend a lot of it's active time around 80a-100a. Consider this a "proof of concept" setup. I don't own an oscilloscope, but I would love to see how "clean" the DC output from this when complete. For sure, the capacitors are the expensive part of this, and if I need to add a couple more, I can certainly do that.

I do have a question though, for these caps, is there any suggestion on whether to put them in series or parallel. They are rated for the 75 OCV that the welder produces. Voltage drops off to 25 VDC when the arc is initiated. Lastly, I designed this prototype from looking at how some others have approached this, so the choke was added in addition to the CAPs. Does it make any different to run the choke on the + leg or the - leg? I would presume not, but have to ask.
 
Hi Tom, the choke can be in either leg. As far as caps, they add in parallel but divide in series. Two 100uF caps in parallel are 200uF, but in series only 50uF
Basically the same circuit is used to convert an ac buzz box to a dc welder, without the caps, since dc stick welding doesn't require quite as constant a voltage source, you can get by with just a rectifier and choke, and even the choke is optional for dc stick if you aren't too picky about your welds
Mark
 
Last edited:
The higher the effective capacitance the better the filtering, so I'd leave the caps wired in parallel. It makes no difference which leg the inductor is on as long as its in series with the load.
 
Back
Top