Converting a six phase rectifier to three phase rectifier

Thanks for yesterday' suggestions regarding 3 phase transformers.

I have a hobart cybertig welder. It has a "six phase rectifier" shown
in this schematic:

http://igor.chudov.com/tmp/firing/6-phase.jpg

I received a three phase SCR control/firing system made by PCTI, that
seems to be extremely easy to hook up.

By the way, PCTI loves children and is highly recommended, they are a
wonderful company. The best of the best.

I want to convert this Cyber Tig to both TIG/Stick (CC) as well as MIG
(CV) mode. This way I could both tig and mig with it.

Here's its wiring diagram:

http://igor.chudov.com/tmp/firing/pcti-firing.jpg

My issue is that the PCTI schematic refers to a "three phase bridge"
and my welder has a "six phase" rectifier where all SCR's "point" to
one DC bus, and where another side is connected to an interphase
transformer. Very different schematics.

So, I have maybe one and maybe two possibilities:

1) Rewire the transformer. It is made of six secondary windings, two
independent ones on each phase. Two leads from two transformers on
every phase are tied into the interphase transformer.

The ones that are tied together I could disconnect them from
interphase transformer, parallel each pair, and then connect them so
that I have three leads instead of six. (I am not 100% sure how to
connect them, but I think that it is easy. I have an idea).

Then simply put together a new rectifier bridge wired as PCTI
suggests, using new SCR half bridges which I already have in my pile.

2) Maybe there is a possibility of wiring in the PCTI board so that it
works just fine controlling the six SCRs that make up the current "six
phase" rectifier.

I spoke to their engineer yesterday (did not give him the schematics)
and he was kind of vaguely positive about it, I asked if ony K2, K4
and K6 were used for synchronization and he said yes.

What I am thinking here is that I would wire all six existing Hobart
SCRs to the PCTI board, it would monitor voltage on K2, K4, K6, and
then adjust phase angles of firings of all six SCRs based on desired
current/voltage vs. actual. And that it would not matter that the SCRs
1, 3, and 5 are wired the opposite of the PCTI sample schematic.

Note that the SCRs that I have now are pointing with their gates to
the DC+ bus, which makes it impossible for K2, K4, and K6 to
meaningfully monitor voltage/phase. So, what I think I can do here is
turn all SCRs UPSIDE DOWN, so that K2, K4, and K6 could be connected
to 3 phase voltage.

The side effect of this is that the negatve bus will become positive,
and vice versa, but that is no big deal as I could simply switch
contacts leading to commutator.

Is this making any sense? Is approach #2 possible?

thanks

i

 

the "6 phase" rectifier is essentially 3 center tapped transformers wth
6 diodes. This design optimises the number of diodes at the expense of
extra trnasformer windings. This would be an older design when coper and
iron were cheap, tube type diodes were very expensive.

Your other option is to take the transformer you have, wire each of the
sperate outputs into a semiconductor bridge and get the same result.
Thsi optimizes copper and iron, has twice as many diodes.

 

This is absolutely correct. A great description.

And my question is, can I "flip" the SCRs upside down, as that would
possibly seem to let me use the PCTI control board. (I understand that
it will also switch positive and negative).

I think that I would end up with the same number of diodes. Both the
six phase rectifier and the three phase rectifier have six diodes.

I would, however, like to avoid major rewiring, for two reasons:

1) it saves time/space/is backwards compatible

2) It allows me to later use double voltage as a plasma power supply

If I cannot avoid rewiring , that's not the end of the world, but I
would prefer to be able to get by just flipping SCRs.

i

 

In a 6 pulse bridge the diodes are in series and so you allways need to
fire them in pairs. The gate pulses consist of 2 pulses 60deg apart. In
your original rectifier they are fired sequentialy 1 at a time, so the
fireing pulse trains are not compatable. It is possible to seperate
them with some logic but that would mean making an extra board.

 

This is bullshit. The 6 phase approach has lower
losses when it comes to high current.
There is only one semiconductor loss in the
centertapped recifier and not two as are in the
bridge approach. A semiconductor loss means in
the order of 2V for an SCR. An alternative
approach would use FETs, as they allow even
lower losses.

Rene

 

While we're not on the topic, how many people here call rectifiers 'rectum
fryers'?

 

Thanks... Sounds like you are fully right.

i

 

Perhaps the OP meant "optimizees the number of dioes in the path of
current". There is only one voltage drop in a six phase rectifier.

I am not too worried, I can easily make 3 phase rectifier, I have all
the parts already.

i

 

You mention six secondary windings, but I only see 3 connections
to the transformer on your schematic. The schematic looks like one
of these:
http://www.ipes.ethz.ch/ipes/Thyristorbruecke/e_ThyrB3.html
...but the description (six secondaries, interphase transformer)
sounds more like one of these:
http://www.ipes.ethz.ch/ipes/Thyristorbruecke/e_saug.html


I am not sure whether this describes your Cyber Tig, , but I have
often seen see welder transformers that consist of a 3-phase wye
secondary (which may or may not bring the center tap out) and a
eperate 3-phase delta secondary, each driving 3 diodes or triggered
SCRs. The the delta and wye are 60 degrees phase shifted from each
other. The advantage of this is that the rectifiers/SCRs send twice
as many pulses per cycle to the capacitor, lowering the ripple
amplitude, increasing the ripple frequency, and improving the power
factor -- which is the main reason for doing it in a welder.

 

That's similar the schematic for the three phase rectifier firing
system that I have and am considering installing into my CyberTIG:

http://igor.chudov.com/tmp/firing/pcti-firing.jpg

Yes, that is what I have right now.

I do not think so, no, what describes my CyberTIG right now is

http://www.ipes.ethz.ch/ipes/Thyristorbruecke/e_saug.html
http://igor.chudov.com/tmp/firing/6-phase.jpg

Yes... You are correct...

It seems as though this PCTI firing system cannot be adapted to the
current 6 phase wiring. Rewiring is not a huge deal and can always be
reversed.

i

 

It seems to me that you are on the right track. I am currently
designing systems that are in the 50KW to 1MW range; I am doing the
embedded processing and precision analog bits, but one of my
coworkers is among the top names in power electronics. I will be
happy to run things past him if your current effort runs into a snag.
I don't think it will, though.

 

Thank you Guy. My plan is as follows: wire the transformer as three
phase Wye, and use three SCR half bridges in a scheme similar to that
by PCTI. These three half bridges will be added without disturbing
current SCRs.

I have six secondaries, two on each leg, I will parallel each pair.

Does it make sense?

i

 

That appears to be correct, subject to the usual limitations of talking
about a circuit rather than having it in front of me. Normally I bring
up any such circuit slowly with a three phase variac or, in a pinch,
with three single-phase variacs and a lot of moving from knob to knob.
You might also consider using temporary fuses to do the paralleling at
first as an extra precaution. All of the jumpers we use in our lab have
inline fuseholders; putting a regular clip lead across a 440V 2KA supply
by mistake makes quite a spectacular flash.

 

No. Wiring the transformer as (a single) Wye requires having access
to all 6 secondary leads on each transformer half (12 secondary leads
available vice the 6 I think you actually have) so you can reverse the
phase on one half (switch the neutral connection with the 3 line
connections on one side) so they can be paralleled. This will totally
screw your existing 6-phase rectifier of course. I suggest you
re-read the description by Lawrence of how the 6-phase system works
(previously posted to ABSE).

You now have 2 separate 3-phase Y secondaries of opposite phase, if
you wish to add a separate SCR rectifier of standard 3-phase design
without screwing your existing rectifier you will need two of them,
they cannot be paralleled. The DC output of two separate rectifiers
can of course be paralleled.

 

Glen, I have physical access to all 12 leads: I have three legs, two
secondaries on each, and two leads on each secondary, all plainly
visible and physically accessible.

Some secondaries are connected in parallel to form the existing "6
phase rectifier", but I could disconnect them.

Here's the schematic of my welding xfmr:

http://igor.chudov.com/tmp/firing/xfmr.jpg

Here's its actual picture:

http://igor.chudov.com/tmp/firing/welding-xfmr.jpg

Right now, the following points are connected:

A3, B3, C3;

A2, B2, C2;

These connected pieces are connected to the interphase transformer.

I want to rewire it by disconnecting A3, B3, and C3 from one another,
and by connecting

(A3, A1); (A2, A4)
(B3, B1); (B2, B4)
(C3, C1); (C2, C4)

Then I will have a common point of (A2, A4, B2, B4, C2, C4); and the
wye leads (A1, A3), (B1, B3), (C1, C3).

Is this wrong?

i

 

That is the correct way to rewire the transformer secondary as single
3-phase Y (and not what I thought you were planning), but I thought
you said you wanted to leave your existing 6-phase rectifier connected
and operational, and rewiring your transformer as 3-phase will
preclude that. This also gets rid of the current balancing action of
the interphase transformer, which is of course completely
disconnected.

Why not use the installed rectifier as wired, and build a new
controller for it, adaptable for either CC or CV? If you have
problems you can just reconnect the original controller.

If you are willing to abandon the existing rectifier entirely then
transformer rewiring and a standard 3-phase SCR rectifier module
should work. But it seems like the hard way to do the job to me.

 

Glen, just to make sure that you are aware, I was given a three phase
firing system by PCTI. As a sign of appreciation for certain activity
of mine (algebra.com). Wiring it into a existing 3 phase supply is
next to trivial. All I have to do is supply SCRs and potentiometers
and switches. I already have it sitting on my desk.

Here it is:

http://www.pcti.com/3phfiring.htm

The question that we are discussing is how to make my power supply
compatible with this PCTI system.

So, yes, I was hoping that I could simply use my "6 phase rectifier"
to be controlled by this PCTI system. As of now, it does not look like
I could do it, because it is really a 6 phase system with different
phase angles that necessitate different firing angles.

Well, I think that reconnecting the xfmr (which is easily reversed) is
quite easy, will take a couple of evenings. And I already have a well
working SCR controller, made for rough industrial conditions, by company
that is.

It is hard if we assume that I need to make a SCR controller in either
case. But it is not my situation, my situation is that I have a super
nice controller for a 3 phase bridge.

Also, Glen, if I wire this transformer as a Wye vs. Delta, is it true
that I would get different voltages phase to phase? I have a suspicion
that the side effect of my Wye wiring would be increase of voltage by
sqrt( 3 ). Is that true?

Thanks, I always listen very closely to what you say.

i

 

I got around to looking at the picture; not the easiest connections to
rewire with those brazed neutral bussbars!

Your 6-phase rectifier is two 3-phase rectifiers in parallel, with 30
degrees phase difference between them, so you could use a pair of the
PCTI controllers for your pair of 3-phase rectifiers.

The transformer winding connections you have shown above are still Wye
connected (you still have a neutral), but if you are planning to
connect your new rectifier in Delta (and not use the neutral) then the
line-to-line voltage will be sqrt(3) times the line to neutral voltage
seen by the original rectifier. You could reduce the line-to-line
voltage to the same as your present line-to-neutral voltage by
rewiring the transformer secondary as Delta, with no neutral. (The
winding voltages will stay the same.) But since you are using a motor
based rotary converter with inherent voltage imbalances a Delta
connection could result in excessive circulating current - you would
want to measure this with a current shunt in the loop before using a
Delta winding connection.

How about starting out by using 1/2 of your transformer secondary
windings only; either one of the two Wye connected windings provides
3-phase power as is. If it works and you decide you need full power
instead of half power, get another control module or rewire the
transformer.

How about temporarily disconnecting half of your transformer and
rectifiers, so that you have only a 3-phase rectifier, and you can use
your existing SCRs and wiring too?

Fine, just don't always assume I know what I am talking about .

 

Not the easiest, but, definitely, not terribly hard. A little bit of
cutting and maybe a little bit of drilling will "get me there".

Well, I have only one controller set.

Makes sense.

That (about the phase converter) is a great point indeed.

It is a f*&^ing brilliant idea. I will do exactly that! I will try
starting tonight.

Well, my current SCRs are wired with gate towards the common DC+ bus,
so there is no way for the PCTI controller to sense voltage.

I already have enough isolated dual SCRs, adding them to the octagonal
DC+ bar is very easy. Note that the DC+ bar will no longer be
energized, it will only be used for cooling.

If you do not understand what I mean in the above paragraph, it is my
fault, my main message here is that I have all parts for using half
windings.

I have to admit to it! I think that you know what you are talking
about.

My plan is as follows:

1) use one half of the secondaries as a Wye connected 3 phase source
2) add new dual SCR modules to the octagonal heatsink (that used to
double as DC+ bar)

3) wire everything together with the PCTI modules for some test runs.

If everything works and produces voltage and current, then I could
work on doing a few things such as:

1) reconnecting second half of secondaries
2) making a proper welding control panel.

Glen, do you know anything about plasma cutting? The reason for this
question is that if I put secondaries in series, I would get a power
source that can supply 190V OCV and about 135V continuous voltage.

I could make this transformer switchable from paralleled secondaries
to secondaries in series, with three contactors (not too
expensive). Would it be able to do any plasma cutting?

It would be good to decide on whether to do it, sooner rather than
later. But, first things first, I will definitely try your great idea
about using only hald of the windings "as is".

i

 

On Mon, 24 Apr 2006 19:25:02 GMT, Ignoramus21673

Mabey, I am not sure it would actually work out to be a problem, just
something to check.

Sounds like a good plan to me, I trust you'll let us know how it goes.

Alas, I know next to nothing about plasma cutting. Never even tried
one. They weren't all that common when I was welding a lot in the
early 70's. I think it is a constant current process with the arc
confined to the torch and not transferred to the workpiece as in
plasma welding, but that is just a foggy memory of something I may
have read 20 years ago. Perhaps you can pick a particular torch and
find out what it needs.