View Full Version : SCR Control and Breaker Tripping

02-12-2006, 06:01 PM
Hello again... I built the scr 'light dimmer' circuit that Mike W posted a while back ( or on a different thread ), and need some help with it. It appears to work, by the way my welder reacts to it. I can run it either full ON or full MIN, but when I try to vary the power, my circuit breakers trip.

This is just IDLE, not while welding, I have yet to put a load on it.

I took some amp readings, full power setting, idle about 8 amp; ( fan running at full speed). Minimum setting, idle about 25 amp, fan barely turning.

For the Time being, I dont want to put in bigger breakers, wire, etc... just incase there is something I missed.

My welder is a Hobart TR250. If I can't adapt the INPUT of the welder to be controlled, how about a control for the OUTPUT using the same technology ( scr ).

Thanks In advance

02-18-2006, 04:27 AM
slamdvw, can you post the link to the SCR light dimmer circuit that you used? I tried a search on "Mike W," but didn't find it. I'm curious how it is used with your welder and what Mike W's experience with it was.

I'd be very cautious about using it on your welder, based upon your description of the 25 amp idle current. That would indicate that something is definitely going wrong and you could damage your welder.

Speaking in broad generalities, an SCR "light dimmer" is not very well suited to controlling the power into a transformer-input device or an inverter type welder unless both are made to work together. One problem with a transformer load is that the phase shift between voltage and current caused by the inductance of the transformer messes up the phase-shift control of a dimmer intended for incandescent lamps (resistive, with no reactive components). This can be overcome, but it takes special circuitry in the dimmer.

Another problem is that your dimmer on the input may be fighting with the internal current/voltage control built into the welder. (I am not familiar with the Hobart TR250 or its control technology.)

If the load is an inverter-type welder, another problem can arise. (All this is speculative. I'm not an electrical engineer.) An inverter welder input will rectify the input power and store the energy on capacitors. A simple lamp dimmer operates by delaying the instant of turn on of the SCR or triac after the zero-crossing of the voltage waveform. This is fine for control of the power to a resistive (non-reactive) load like an incandescent lamp. However, it could just lead to extra stress on the rectifiers and minimal control of power for an inverter welder because the energy storage capacitors in the inverter welder will still try to charge up to the peak voltage on each half-cycle. (This neglects the influence of any power factor modifications to the input circuitry of the welder).

All this is not to say what you are trying to do is wrong, but I would be very careful to be sure that your dimmer is compatible with the input and control circuitry of your welder.


02-18-2006, 05:06 PM
I can't remember how I came across it, but I have found this schematic on converting a Stick machine to a TIG... http://img.photobucket.com/albums/v110/tek798/Welding/Welder_SCR_control_gif.gif

Dont know if it will be an active link or not, ( still kinda new to this web board posting bit :D )

The Resistors are the sizes listed, the DIAC's are NTE6407's... The caps are 100,000 pF's. I asked for the size listed and the electronics shop gave me those.

The welder is a 100% transformer unit, no lightweight inverter stuff here! The control circuit is a saturable core reactor on the output side of the xfmr.

From what I have read about the circuit, people are having good luck with it. I dont doubt that it is, I feel it is something that I'm missing. ( IIRC it's also discussed on the 'hobartwelders' board ).


02-18-2006, 07:14 PM
Interesting! Looks like a pretty standard bidirectional phase controled circuit. Since the whole circuit is floating and in series with the welder primary, it's hard to see how it could be causing more current demand than the bare welder primary UNLESS IT IS RECTIFYING THE INPUT TO YOUR WELDER by having asymmetric firing. Since it is essentially two independent power switches (except for the control pot loop), the two could be less than perfectly matched, leading to asymmetrical triggering and rectification.

The best way to observe that triggering is symmetrical is with an oscilloscope, but it is tricky and most welders do not have a 'scope. Do you have the ability to observe the line voltages on an oscilloscope? You have to be very careful measuring line voltages with a 'scope because without special isolation measures, one side of the scope input is usually hard ground. If you can find a matched pair of current transformers, you might be able to safely observe the current waveforms, but most current transformers are not very good at high frequencies present in an SCR chopped waveform. You might also try a matched pair of Radio Shack step-down transformers (with 240VAC primaries) to isolate a 'scope from the hot line, but if there is a strong DC component (which I suspect to be the case) on the welder primary, the step-down transformers could get fried.

OK. Here's a thought. Disconnect your welder and substitute a power resistor, like a 240V water heater element (with the heating part submerged in water), connect a multimeter across the load, and see if there is any DC voltage across the load as you sweep the dimmer control pot. If there is, your dimmer is not triggering symmetrically. Don't turn your multimeter down to a range less than your AC line voltage because you can burn out your meter even though you are not seeing a meter response. Since the resistor responds the same to AC or DC voltage, you probably won't pop your breakers (as quickly) with this test setup (but the utility transformer might not like it). Your welder input looks like a short circuit to DC voltage.

If you find that there is DC present across the load, it could be causing high DC current to flow in your welder primary, hopefully popping breakers, but possibly frying your welder transformer.

Getting a little more exotic, if you do have access to a 'scope, you can use a 220VAC to low voltage transformer to observe the voltage waveform across the resistive load, as the transformer will isolate the hot lines from the grounded 'scope input. Don't do this until you have shown that you don't have a lot of DC across the load because if there is DC present, it will burn out the step-down transformer primary.

Have fun, and be very careful.

Oh. Wait a minute. If your welder has a saturable reactor in the output, why are you not using a foot pedal to control there instead of controlling the power input? Do you have a xchematic of the welder? Have you contacted Hobart for advice on adding a foot control to the saturable reactor control circuit? It seems to me that that is a much more practical place to do your controlling, rather than in the welder input.


02-19-2006, 05:41 AM
Further thoughts on the dimmer circuit.

Asymmetrical triggering can be caused by defective component or even by normal tolerances in the device parameters causing triggering at different levels.

One way of minimizing the possibility of asymmetrical triggering is to use only one trigger circuit that sends trigger pulses to both SCR gates on each half-cycle via dual secondary windings on a pulse transformer. The extra pulse to the reverse-biased SCR has no effect.

Another method is to put a single SCR and trigger circuit inside a full rectifier bridge and put the bridge in series with the welder primary. That way, the SCR circuit sees the same waveform on each half-cycle. Extra care may be required to assure commutation of the SCR on each half cycle. This is the most secure approach, as failure of a component (other than a single diode in the rectifier bridge) cannot result in DC on the welder primary.

All this being said, I still would try very hard to effect control of the output current at the saturable reactor in a manner recommended by the manufacturer, not at the welder power input terminals.


02-19-2006, 07:52 AM
Funny you should mention an O'scope... as part of another hobby I do have one :D. Never thought of using it. I'll try seeing what my wave forms are. There is a schematic on the inside top of the welder case.

I was thinking of using the Sat.reactor for control, but as it is Minimum setting is still too hot. ( unless it's a technique problem... 20 ga mild steel, argon shielding ). I'll try the scope... and see what i come up with...



02-19-2006, 02:25 PM
Is it feasible to post the schematic? I'm interested because I have a similar problem with my P&H WWII era AC welder with saturable reactor control.

Generally, a voltage-dropping resistor in the output circuit is not desirable due to the lack of desirable smoothing action of a smoothing choke. But, depending upon the configuration of your output circuit, it may be a feasible means of reducing your minimum voltage/current.

By pure chance I was at C&H Sales in Pasadena last week and looked over their high-power resistor assortment. No, I'm not talking about something that is soldered onto a circuit board. I'm talking about something weighing about 5 to 10 pounds that would be mounted under a transit vehicle. As I recall, they had values from about 0.1 ohms to about 1 ohm, which may give you about the voltage/current drop you need for your thin sheet.

These resistors are typically a flat band of resistive metal loosely wound edgewise around a ceramic support for perhaps 20 turns. Typical lengths are about a foot and outside diameters from about 3" to 5". Also some oval units.

I think it would be feasible to remove the ceramic core and insert an insulated iron core to add inductance. It would look a lot like some of the reactors you see in welders, except bigger and with resistor material instead of copper. I think the ceramic cores of some of them were segmented and could be removed without damage. In any case, they could be fractured and dumped out. Rated current is in the 100 to 200 amp range, so you should have no problem at your desired low currents.

I have no idea how feasible this is, but it intrigues me. I'm also not sure how the saturable reactor control loop would react to current limiting outside of it's control loop. I THINK it would be OK because at minimum setting the entire saturable reactor inductance is in series with the welding current and the control loop may be essentially inactive because I think the system works by INCREASING welding current as you INCREASE saturable reactor control current. But I have had a lot of difficulty finding any accurate technical description of how these things work. Most of the descriptions in welding books just describe the hardware without any analysis of what is actually happening in the control loop.

Keep in touch.


I almost forgot to say the high-power resistors are all $25.00. I think their website is www.candhsales.com. They have an on-line catalog, but some of the units are not listed due to limited stock.

Tell us a little about your electronics interest and experience. I am a noise and vibration engineer (mechanical) with a lifelong hobby interest in welding and electronics (since the '50s).

02-21-2006, 04:37 AM
I was reading my Westinghouse Electric SCR Manual in my tiled reading room this morning and ran across a discussion of a phase control consisting of a single SCR and trigger circuit within a full rectifier bridge, the configuration I had recommended above as a means of assuring symmetrical triggering.

The author pointed out that at low output, a phenomenon of cycle skipping can occur that would cause DC on the output. It occurs because triggering on one half-cycle affects the voltage stored on the capacitor across the diode that results in locking out triggering on the next half-cycle. The result is that you get conduction on alternate half-cycles, thus producing DC. Turning the pot up to a higher output position resolves that particular problem.

This does not refer to your dual SCR circuit, but does show that there are booby traps out there.


02-21-2006, 05:35 AM
Aaahh, ok. Well tomorrow ( tuesday ) i'll fire up the 'scope and hook a big resistor across the output and measure for DC. I figure i can avoid the ground problem by running the scope off DC or an inverter... just make sure IT don't touch anything grounded.

I will try and post you a copy of the schematic of the welder... soon as i figure out how to scan it. It's a sticker attached to the top panel of the machine. But in the mean time, it's basically the same setup as a Dialarc.

I'm a 'jack of all trades' type of person, particularly if it hooks to line voltage and makes a whirring noise. I find electronics interesting, though my experience is a bit rusty. Covered a lot of theory and stuff in college ( im basically an apprentice electrician with a little over a year of exp. ) but it's been a while sinc then. Welding came about as a hobby... started because Im building a hydronic heating system in my house and needed to weld copper tubing. First time actually touching a TIG torch was about 6 months ago, kind of a last minute deal. Was at the welding supply store, figured "what the heck"... gave it a shot, and been trying to improve ever since. ( college courses i have attended: autobody repair and refinish, electrical maintenace tech, and refrigeration and a/c ).

02-22-2006, 04:37 AM

link to my welders schematic,... scanned from top cover

02-26-2006, 01:55 PM
Hi again, Robert:

Here are some additional thoughts on your effort to get fine control of the welding current of your Hobart TR250. Be aware and forwarned that I am not an electrician or electronic engineer. I am an electronics/welding hobbiest and have dabbled in both for too many decades.

First, the schematic on the upper right corner of the label pasted inside the unit cover has some interesting info. Assuming this is an actual schematic, not a simplified representation of the circuit (except for leaving out all the switching details), It shows that the control circuit does not include any welding current feedback to the control loop. That is, it appears that the control circuit is not sensing the welding current and making changes to maintain it at the desired current but is, rather, an open loop current setting system. The control current to the saturable reactor sets the effective inductance of the reactor without sensing and correcting for the actual welding current. (The "Z" vertical bars through the "CONTROL REACTOR" are graphic representations of the hard, rather than soft magnetic saturation characteristics of the metal core of the saturable reactor.)

This simplifies setting up your own supplementary current control mechanism to further reduce welding current because there will not be any destabilizing or troublesome conflict between your new current control system external to the saturable reactor and the internal saturable reactor current control system. Note that the "RANGE SWITCH" merely selects more or fewer windings on the reactor. I propose that you add an external resistor/reactor to reduce the minimum current you can select.

The saturable reactor controls the welding current by varying the inductance in series with the AC portion of the welding circuit. When the magnetic core of an inductor saturates, magnetically (all the magnetic domains in the core have been aligned and a further increase in current through the coil does not result in any more domains becoming aligned), it's inductance suddenly appears to drop to a very low value, thus providing less opposition to the welding current. By applying a DC voltage to the control winding (Terminals E and 5 (?)), the operator pre-aligns some of the magnetic domains, that is, moves the reactor closer to saturation. Thus, saturation will occur earlier in the waveform, allowing more current to pass. (This is a very amateur effort to understand and describe the operation of a saturable reactor control, so don't quote me as an authority. I eagerly welcome any more learned explanation from other posters.)

There are two important implications I take from this limited understanding of your welder circuit: (1) I think you can add external resistance/reactance to the AC welding current loop to reduce the DC welding current to levels lower than provided in the stock configuration without problems, and (2) I think you can easily provide finger or foot control by means of a small potentiometer (variable resistor) providing base drive to a high current darlington transistor that will allow you to vary the current over the range now provided by your front panel current control.

This is an exciting realization for me because, as I said before, I have been puzzling over how to equip my WWII era Air Force surplus P&H/Harnishfager AC 450 amp stick/TIG welder that I also hope to provide with a rectifier. Your schematic is much more straightforward and comprehensible than the one I think I remember for mine and makes me hope that the same technology I am suggesting for you can be applied to my welder.

Now that I've led you on and hopefully whetted your appetite, I have to go finish renovating a rental apartment unit that has to be ready by the end of the month. What I'd like you to do if you are interested in this approach is to determine the resistance and power rating of the "FINE CURRENT CONTROL" pot and the range of currents through the reactor control loop from the pot wiper. Judging from the 15 amp fuse feeding the "CONTROL RECTIFIER," the pot resistance is probably less than 10 ohms and the pot is probably a fairly large ceramic power pot of 100 watts or so rating. It would be really useful - even vital - to know the actual range of control current through the saturable reactor from the "FINE CURRENT CONTROL" pot wiper because that is the current the power darlington will have to pass.

I believe that we will be able to provide output current control equivalent to what would have been provided by the manufacturer with a foot/finger control without the use of SCR control of the power to the welder. This has the advantage of not messing with the power voltage/waveform to the fan or the "FINE CURRENT CONTROL" pot, as does the SCR "dimmer." The main welding current control element is there in the saturable reactor. All we have to do is insert a means of remotely controlling the control current to the reactor.

More later. Please let me know if this is all of as much interest to you as it is to me. I don't want to flood the thread with extraneous ruminations that is not of interest to you.


03-19-2006, 04:25 AM
Hi Robert:

Haven't heard from you for a while and was wondering if you (1) lost interest in my ruminations on getting low current from your welder, (2) haven't had an opportunity to make the measurements I requested, (3) really want SCR control of your inpt voltage and output current, or (4) killed or injured yourself making the requested measurements.

If you are just not interested in pursuing my approach, that's fine. I'd just like to know so I can stop pondering the problem.



03-22-2006, 06:49 AM
awright: sorry!... im interested, just haven't had the time. It never fails, If i have nothing to do, i have all the time in the world.. but if something comes up, seems EVERYTHING goes wrong. but if it helps, i still have the top of the welder :D so it reminds me everytime i go to the shed, that it's still on the to-do list.

I should know here in the next day or so
Thanks again


04-30-2006, 05:37 AM
awright: Sorry for the long delay... had a project that came up and I was able to use the MM135 on it. I haven't been able to get actual amperage readings from inside the welder, however... I did a simple test... If you remove the 'control fuse' ( no voltage to reactor ) you get minimum output ( around 35 amp). The Fine Control IS a large ceramic wire wound resistor, bout 5 or 6 inches diameter.

On the SCR control circuit... I was able to probe around on it while operating a part inductive / part resistive ( large space heater ) and found that it IS making DC. It'll sure stop the blower on the space heater! It seems that only one of the two SCR's are firing like they're supposed to. They're chopping the positive waves, but not the negative.

I reversed the control circuit connections, what fired scr1 is now firing scr2 and vice versa, and not it's chopping the negative instead of the positive.

I have a friend of mine helping me work on it.

For the time being... I can MIG the joints... then when i get the time, hopefully before I need the TIG I can get this thing working. ( isn't the internet wonderful... amazing the information one can find! )

I just wanted to say thanks for the help. sorry for the delays in responding.


04-30-2006, 02:41 PM
Good to hear from you again, Robert. I had almost forgotten about your project.

As you must have observed, your test reversing the SCR trigger leads proves that the SCRs are both OK, but your trigger circuit is defective. Have you given the trigger circuit a close examination? Check for burned components, cold solder joints, component values, etc. I still don't like the idea of two mostly independent trigger circuits instead of one circuit feeding both SCRs.

Be very cautious about using your space heater as a load, as the DC component can burn out the fan motor. It's interesting that you say, "It'll sure stop the blower on the space heater!" Applying DC to most types of AC motors, including shaded-pole motors used in many small fans, actually acts as a brake on the motor. This is used to greatly reduce coasting on some machines, like table saws and radial arm saws, in which coasting can be dangerous because people get impatient and don't wait for the saw to stop before reaching around the blade for the workpiece. But the application of DC is done in a very controlled manner to avoid burning out the motor winding. For this reason (possible motor burnout), I still recommend a water heater element cooled in a bucket of water (dont submurge the contacts!), because there is no fan to worry about. The elements are around $15, I think.

Will you still be interested in achieving fine control of your welding current using the internal controls, even if you get the SCR dimmer circuit working? Remember that controlling the input power to the machine with the dimmer affects the fan and saturable reactor control circuits, also, which the internal controls would not. Of course, at very low welding currents, both those circuits become less critical.

If still interested, try to make the measurements of pot resistance and current that I suggested earlier.

Keep us posted.


04-30-2006, 03:35 PM
Ok... took some readings on the resistor / control circuit inside the welder.

Res: 32 ohm
Voltage varies 0 - 24 volts

I had a brainfart... forgot that the circuit is DC and not AC... and I measured AC current in the control circuit ( 0 - 16 amp ).

As far as the DC burning out the fan motor... that's why I used that heater. If i kill it, a quick trip to the local mall ( read DUMP ) fixes it :). We're going to do more tests on the SCR circuit today, i'll let ya know what I find.


05-01-2006, 03:34 AM

A friend and I, we disected the circuit 7 ways to Sunday... trying various things. Changed components, rewired things, even ran the circuit on 110v to make it a little safer ( that, and a light inductive load was messing with the wave forms (fan motor on heater)).

We ran a 500w halogen light off of it, wired for 110... come to find out a resisitor was what was causing the problem.

On the o'scope... the SCR's would chop the 'negative' part of the sine, then the positive... it would kind of look like halfwave DC.
After removing a resistor, it chopped both positive and negative evenly ( was kind of cool to watch on the scope). After bypassing various components, we discovered that R5 was causing the uneven triggering.

Various loads, ( motors, heaters, fans ) all seemed to have different wave forms. The heater gave the 'truest' of them all ( being 99% resistive ). We hooked it to the welder, was able to vary the power from high to low. Low causing the most current drawn, xfmr grunting pretty hard. Then the moment of truth... struck an arc with two carbons. The transformers seemed to like being loaded and dimmed rather than unloaded.

Some more experimenting with 2 scr's, 2 diodes, and a light dimmer switch. Im thinking... instead of chopping the power coming in ( where it messes with the fan ( cooling )) i'll make a variable rectifier. any thoughts?

Enough rambling... but so far, it works. now needs fine tuning... but IT WORKS!! Thanks!


05-01-2006, 02:53 PM
Robert, I worry when you say;

"We hooked it to the welder, was able to vary the power from high to low. Low causing the most current drawn, xfmr grunting pretty hard."

That doesn't sound like a healthy situation for your welder transformer. Can you think of a good reason that the rig would draw highest current on lowest power setting if things were working together well? Did you look at the waveform from the chopper while powering the welder? The welder is a highly inductive load, especially at low welding currents.

One protective item that comes to mind is the DC output monitor incorporated into many high power audio amplifiers to disconnect the speakers if DC is detected on the amplifier output. You might consider adding such a circuit to your welder if you continue to use the SCR chopper on the welder input.

I think your 'scope observation of the various types of loads having different waveforms is in line with my observation way back in this thread that a chopper circuit must be specifically designed for inductive loads if it is going to work well with them.

Transformer "grunting pretty hard" could be a sign of transformer stress to be wary of, or could merely be that the high frequencies present in the chopped incoming waveform excite the transformer core into vibration (due to magnetostriction - change in dimension of the steel under the influence of a magnetic field), and you hear the sound radiated by that vibration.

I'd advise at the very least that you monitor transformer winding temperature carefully when you use the input chopper. Most line frequency transformers not specifically designed for them don't like large harmonic content in the incoming power, which is just what results from chopping the incoming power. The harmonics excite in the core high frequency flux reversals that do no useful work (other than to maintain the chopped waveform), but generate heat due to hysteresis (energy loss on each flux reversal) in the steel. Probably no problem, but good to keep an eye on it.

A digital VOM with thermocouple readout would do the trick, and they are not very expensive these days. I'd try to get the thermocouple tip as deeply inside the windings as feasible without trying to move any wires in the winding, which could cause insulation damage. You could also consider providing a thermostatic switch of the type imbedded into windings of motors with thermal protection, and hook it up to an alarm or to the main contactor control loop. You should be able to get such a thermostat from a local motor rewiring shop. You could also use an HVAC manually reset snap-disk thermostat like the Grainger 2E362, although it would be more difficult to get close to the hot parts of a transformer and windings than the motor thermostat, which is a more compact, flat package with leads.

You didn't say why R5 was causing problems. Was it open? Shorted? Wrong value?

So, how does it work with the thin sheet stock you were trying to TIG?

Sounds like you are very pleased with the situation. Glad to hear it. Keep us posted.


05-01-2006, 08:45 PM
err...Where IS R5? I only see R1 through R4 in the schematic at http://img.photobucket.com/albums/v1...ontrol_gif.gif.

Still mulling over your comment about the welder drawing high current and grunting at low power from the SCR "dimmer." Doing some more recreational reading of my 1963 Westinghouse SCR Designer's Handbook, the simple two SCR dimmer they show for resistive loads (incandescent lights) has the following comment:

"The second (advantage) is that by deliberate slight mismatch of the gate firing sensitivity of the two SCRs, transfer from twice per cycle firing to once per cycle firing can be made to occur as the control setting is reduced to very low values."

The author describes this as an advantage because, for a resistive load that can use DC or AC, the transfer to once per cycle firing does no harm and extends the control to lower powers than is easily achieved with twice per cycle firing. But you cannot (or should not) tolerate DC that results from once per cycle firing because it could damage your machine.

While they mention "...deliberate, slight mismatch...," in reality normal tolerances in manufacturing usually results in a significant spread in gate sensitivities and Diac firing voltages, so you would expect rectification below some low power setting. Fortunately, fixing the mysterious R5 seems to have resolved your main problem with breakers popping, but I wouldn't be surprised to find that rectification is still happening at low power settings.

Does the "grunting" of the transformer and excess power consumption happen suddenly at some control point as you turn down the power, or does it increase gradually over the range of the dimmer control? An abrupt onset of high current demand and noise might indicate the transition to once per cycle firing and the resulting application of DC to the welder.

It sounds like your repair of "R5" may have improved the asymmetrical firing situation, but not eliminated it altogether, and that the improvement causes the remaining rectification to occur at a power setting that is low enough for the transformer and breakers to tolerate.

I'd recommend taking a closer look at the power waveforms into the welder with your 'scope.

Do you have a sense of just how low you would like the current to go for successful TIGing of the thin sheet you want to work with?

Hope I am not putting a damper on your pleasure at getting the rig working.


05-02-2006, 12:05 AM
Nah, I'm finding the information helpful... causes my one track mind to jump off and go on a tangent. Helps me think outside the box.

I have two versions of the schematic... the one I sent you, the resistor is R3... it's on one leg of the control circuit.

I had another thought of the larger current drawn... My machine does not have PFC installed. I have a pile of motor run caps, I thought about putting a couple in to see if it helps.

Yet another thought was to replace two of the four diodes in the bridge, with SCR's... Instead of varying the power going in, I'm doing it on the output where there should be less chance of damage. I'd rather smoke an scr than a transformer.

(( see told ya it gets me thinking)) Im finding that scr's don't like inductive loads... on the light bulb test they did their job VERY well... both positve and negative parts of the waveform were chopped, evenly. ( i should pull the lid off my MM135 and see how they did it)).

I'll try actually welding with it, and see how it goes. have someone keep an eye on my current meter while it's loaded. It did seem to work better while the machine was producing power.

Thanks for the info.. I'll keep ya posted


05-02-2006, 02:10 AM
OK Robert, help me out here. How can a problem with R3, the resistor in series with the potentiometer in the control loop common to both SCR trigger circuits, cause triggering problems in only one SCR? Unless there is something not shown in the schematic at http://img.photobucket.com/albums/v1...ontrol_gif.gif, or unless the defect in R3 was asymmetrical (behaving differently in the two directions), the loop with R3, F1, F2, and R4 would have exactly the same effect on the two SCRs. Conceptually, the four components could be replaced with one resistor variable from 1K to 251K. The physical proximity of R3 to one of the SCRs has no bearing on the behavior of the loop.

By the way, what was the defect in R3? If you didn't, in fact, find a defect in R3, but rather found the circuit working after you changed R3, you have a problem somewhere else that happened to be corrected (temporarily) by handling and flexing the board. I'm always concerned when a problem simply disappears without my having found and corrected a specific problem because it implies that the problem could reappear at any time.

The immediate problem is not necessarily that SCRs don't like inductive loads (although they CAN lead to triggering and dissipation problems), it's that trigger circuits don't like inductive loads. Actually, no power semiconductor devices like inductive loads because the phase shift of current relative to voltage leads to higher power dissipation in the device. The way these problems are handled in SCR circuits is by designing the trigger circuits so the SCR is triggered when the designer wants it to be triggered, not when the load forces it or allows it to be triggered. I'm definitely not an SCR circuit design expert, but I think the very simple trigger circuit shown in your schematic is designed for resistive loads like incandescent lighting, not for inductive loads. Unfortunately, I'm not smart enough to know how it SHOULD be designed to handle inductive loads.

Another point: shifting the SCRs to the output side of the welder transformer will not necessarily lead to better performance if you have not corrected the triggering problem. Asymmetric triggering of SCRs in the output circuit will lead to the same problem of having DC currents circulating through a transformer winding with the associated possibility of saturation of the core. Doesn't matter if the DC is in the primary or the secondary - saturation is saturation. The fact that the SCRs are in the output rectifier bridge does not alter the effect on the transformer windings of asymmetric triggering.

This is not to say it is not feasible to control the secondary current this way - lots of commercial welders use this technique. It is just to say that shifting the SCRs with their inadequate trigger circuit to the secondary is not a fix to the problem of DC currents in the transformer. Coming up with a reliable trigger circuit is the solution (or enhancing the range of the saturable reactor control). Additionally, my gut feeling is that it is more costly to get higher current SCRs for the secondary than it is to get higher voltage SCRs (which you already have) for the primary. (Just a guess.)

PFC is always a good idea, but I do not think it is the appropriate fix for the excessive current with the SCR dimmer at low settings. But try it out and see. I don't know the required values, but I think it takes a lot more capacitance than you will get with a few motor RUN caps. And, of course, you can't use motor START caps in a continuous application like this. Taking a look at the PFC caps in a welder similar to yours might give you the approximate capacitance required. Or ask the manufacturer of your welder.

Good luck.


04-22-2013, 11:17 AM
Hi everybody, im posting the schematic for" DESIGN OF A FULL WAVE CONTROLLED
CONVERTER USING DC DRIVE" this scheamtic im intend to use to drive four power SCR that i have to convert my AC arc welder in to DC, mayby it can help to others to,
im calling everyone for discusion about this circuit. ( included the cosin wave generator circuit from sine wave)

04-22-2013, 12:39 PM
The issue with the original design does not appear to have anything really to do with his firing circuit - it has to do with the fact that phase control basically hates inductive loads. The reverse EMF coming back from the chopped waveform hitting the large inductance plays hell with the firing of the SCRs. I note that you do not have any kind of output filters or snubbers in the design, so this one is also likely destined to fail in the same way. Do a search for "SCR snubber" and you will get lots of good information.

- Tim

04-22-2013, 03:00 PM
How about this unpolarized snobber circuit?

04-23-2013, 06:12 AM
That's a basic circuit . . . without knowing the details of your load, values are going to take a bit of work. I'm still trying to figure out *WHY* you want to do something like this . . . . the other downside of phase control into an inductive (transformer) load, is a $h!tload of heating in the transformer . . . . not good for them at all . . . . Last I tried it was on a toroidal track light power transformer . . . figured no iron core, maybe I could get away with it . . . 10 minutes and it thermalled and shut down . . . can run for days on normal power. The other thing needed to run inductive loads tolerably is a large series inductor with the SCRs to limit the rise time of the waveform the load sees . . . (and you see these in quality dimmers for resistive loads, since otherwise, the filaments "sing" with the harsh waveform and sudden current changes . . . can get quite noisy with large wattage lamps!).

- Tim

04-23-2013, 10:11 AM
Thanx Tim for your advices i really apriciate that, and yes you got a point, i im intend to use an inductor capacitors to smooth down more the output DC,
the reason im doing this is to get a good smothed down and fully controlled DC at the output of my AC welder ( The DC welding is much more quality then AC welding) more variety of welding materials ( for example aluminium) etc. etc.

04-23-2013, 02:33 PM
What's your welder? If it's small (120volt, 20 amps or so) you would be better served finding a surplus 20A variac (variable autotransformer) - pure AC out, and nothing solid state to fight back . . .

- Tim

04-24-2013, 08:01 PM
Hi Mendimano. Are you sure that this is the correct schematic? Where is the sine to cosine convertor that you are mentioning?

From what I understand of the cricuit, the input is the raw AC wave that is fed into a non-inverting follower. This just buffers the signal. Then it goes into an inverting comparator, which goes low when the voltage at its input exceeds the set point. This is not quite correct for an SCR controller, which is supposed to activate gate drive when voltage exceeds the setpoint. The comparator is followed by a differentiator whose purpose is unclear. At this time, the signal is a step and a differentiator does not convert sine to cosine. Again, could you point out this converter. The 555 is not functioning as a timer, since there is no timing capacitor. Instead it is operating as a window comparator, which does not seem reasonable for a gate driver. What is its intended function?

Finally, for welder applications, especially with an output choke, the load is somewhat inductive. In that case, a continuous gate drive is usually called for, which consists of a higher frequency pulse train. Also, this circuit probably also requires a sunbber. There appear to be no specs on the gate drive transformers. These will be very different depending on whether they are fed by a pulse train or not.

Where did you get this circuit from? Is it from a welder schematic or some kind of app note, or did you prepare it yourself?

What is the purpose of the SPICE like graphics below the circuit diagram? The purpose of the second one seems to show the phase shift of a capacitive load. There is no shift on the oscilloscope, and a capacitive load is not appropriate for welding anyway. Where did this thing come from? The third graphic shows the gate drives. This does not look correct. As I pointed out above, there is no pulse train. Also, lower SCR drives should be blanked during the negative cycle and vice versa. Again, where did this graphic come from? Was it a circuit simulation of the schematic presented?