What's new
What's new

Phase Perfect dpc-a10

Hardplates

Stainless
Joined
May 8, 2019
Location
What once was a free country
I was given a 10 hp DPC that is broken and am attempting to repair it. So far I have replaced a bad IGBT and a hex buffer that had one of the outputs shorted to ground. I am getting a flashing red LED and the contactor is not closing. It appears to be trying to make the third leg at the correct phase angle but well under voltage. The previous owner replaced the AC caps once when it stopped working. Years later it died again and he tried replacing the caps again but it didn't work that time so he bought a new model that they support. Going by the number one rule of troubleshooting (never trust the guy that came before you) I am beginning to question the basics like all the wires being connected it the correct location. I flag labeled everything but who knows if it was correct when I got it since it wasn't working. There seems to be very little about repairing them and I would love to start a thread to help people troubleshoot these since the manufacturer chooses to no longer support them. I have some very basic electronics knowledge and an o-scope and the means to do component and board level repairs. I can't imagine I am the first person to attempt to understand and repair one of these and imagine someone far smarter than I has and could be of some help.

Paul
 
Screenshot (19).jpgScreenshot (20).jpg
This may be THE problem, but it is certainly A problem.
I believe this used to be a hall effect sensor inside the current sensing unit for L3. As you can see in the pic the only markings are an odd A and 16L. I cant find any hall effect sensors with similar markings. Anyone have and ideas?
 
I have found that trying to repair a complex device that you have no service data for is usually a money and time sink that does not result in a trustworthy operating unit.

I am not saying it cannot be done, just that you generally need enough savvy on the particular general type of device to identify what the circuits should be doing, and recognize them as you trace out your own schematic (since there is normally nothing forthcoming from the manufacturer).

Just as background, you ought to know that I have designed a lot of power electronics, including VFDs, SMPS, and the like. I would probably not bother to try to fix the PP, unless I had to for some reason, and even then, I'd have little confidence in the result.

This is the result of hard experience, I was once much more optimistic than I am now, but back then, circuits were simpler, and parts were far easier to see and identify.

If you want to do it, go ahead. I will mention that the first rule of power electronics is that if the IGBT is bad, then so is the gate driver, whether or not it SEEMS bad. That one bit of data saves much frustration. BTW, failures may get right through the gate driver into the micro that drives it, which is a real issue, due to programming. Does not always happen, though.

The second rule, is that there is a reason for all failures..... you need to be pretty sure what that reason is before finishing the job of repair.

The last bit of info I will leave you with is that you ALWAYS check out the operation without the power devices (IGBT, etc) connected to the drives. That way you can see if the drive is working right. The problem you face is that you will not have any service data to show you what "right" really is.

Other than these things, it is all just finding and replaceing every bad part before connecting power.... "easy peasy", right?
 
Thank you for your response. I don't doubt that things could get very complicated very quickly but this is something I want to attempt to repair. Some people sit down for a cold beer at the end of the day.....I rack my brain with projects I don't fully understand haha. I believe the previous owner wired the unit incorrectly and that is the reason for failure. I have taken your advice about the drivers and ordered all new ones, as well as every NAND gate, NOR gate, AND gate, OR gate, buffer, opto and so on that is on the board. This stuff is far to cheap now a days to not throw some parts at it. Provided the DSP is alive and well the rest of the board is simple.....right? haha. The DSP has a clock signal and is outputting to some of the gates and buffers so I am "assuming" its alive and well. Thank you again.

The hall effect sensor appears to be an allegro A3516LUA. I also believe it is for over current protection, not current sensing like I previously thought.

Paul
 
Be nice to know if they are worth bothering with. Mine soldiers on issue free in its 15th year. Of course it never had to really work for a living. Only thing that craps out in this system is the seldom switched BR270 breaker
 
Be nice to know if they are worth bothering with. Mine soldiers on issue free in its 15th year. Of course it never had to really work for a living. Only thing that craps out in this system is the seldom switched BR270 breaker

Just a guess but I would imagine they are like anything else, fairly easy to troubleshoot if you understand how its supposed to function. The difficult part is learning how it is supposed to function with very limited documentation and support. Just for the fun of it I recently decided to go back to school in my spare time for a BS in electronics engineering technology so I see this as a functional opportunity to learn more about electronics and hopefully end up with a usable item. Everyone has their hobbies, mine just so happens to sometimes be fixing things that people say aren't worth fixing.

Paul
 
Go for it.

I do not mean to discourage you, as long as you work safe, as there are rather high voltages inside.

The big problem is that there are programmed devices in the unit, and I very much doubt that any of those can be obtained now. Your best bet for obtaining a working device might be looking for another unit that went bad for a reason not involving the PC board in question. Then you could make a good unit from parts of two, and still have fun working on the bad PC board.
 
Go for it.

I do not mean to discourage you, as long as you work safe, as there are rather high voltages inside.

The big problem is that there are programmed devices in the unit, and I very much doubt that any of those can be obtained now. Your best bet for obtaining a working device might be looking for another unit that went bad for a reason not involving the PC board in question. Then you could make a good unit from parts of two, and still have fun working on the bad PC board.

You have not discouraged me at all :D You are right about high voltages as the board ground is a floating AC mains line. I believe the only programmed device on the board is a Texas Instruments 320LF2406APZA DSP controller. I would absolutely love to find another broken unit that is cheap. Problem is they seem to be difficult to find in working order let alone broken. I would imagine there is a way to retrieve the programing from a working unit. If it was built by man then it can be unbuilt by man.
 
I am interested in hearing what you find out, and I am sure many others here are as well. Please keep us informed. If you could make a start at creating a schematic that would be most helpful, both to you and others as well.
 
You have not discouraged me at all :D You are right about high voltages as the board ground is a floating AC mains line. I believe the only programmed device on the board is a Texas Instruments 320LF2406APZA DSP controller. I would absolutely love to find another broken unit that is cheap. Problem is they seem to be difficult to find in working order let alone broken. I would imagine there is a way to retrieve the programing from a working unit. If it was built by man then it can be unbuilt by man.

Well.. hmmm.. you may have already noticed, the PM community has been fretting over the "impossibility" of repair of the older "Blue Case" Phase-Perfect for quite some time.

Anecdotal evidence that (at least) one crucial component has gone unavailable. A common, and very valid, reason for any maker to terminate service is actually for legal liability exposure reasons of older gear, as much as technical reasons, or even MORE so. BTDTGTTS, decision-maker, different industry..

What I call the "pass elements", because I have seen so many different technologies for output stages in more than 60 years of pushing electrons - differ by HP.

AFAIK, the control logic board is the same. And that is where the barrier lies.

Even so, I haven't thought to fight that if/as/when my older, bought-used, blue-case 10 HP was to wake up one morning and been found to have shat the bedclothes and drawn its last breath, so to speak.

The "White Case" 10 HP, bought NEW, full warranty, has been here for over a year. And an RPC. And a 3-P capable 10/12 kW (altitude thing..) MEP-803a NATO diesel gen set.

Statistically, there just CANNOT be that much life left in the old P-P. But it does still work.
Or it DID.

ISTR I had $1200+ in it, Inbound freight - from the Dakotas, was it? - Included.

That said, it no longer owes me a dime. As no one else has - so far - delved as deeply into possibility of repair as you have, I'm of a mind to gift it to you, "for the cause."

I shall be compelled to leave testing of that to you. Too much Old iron. Too little time.

Where in Pennsylvania are you located?

Born Pittsburgh, know most of the 2-lane "alleged" blacktop AKA velcro'ed together potholes - across the State without need of a map. Town & Country even has tall-sidewall high-profile high-load-range tires two sizes over OEM max. "Agony County tires". From when Mum was still living In Bethel Park. I did say I knew PA alleged-roads?

Use PM's "email" feature on the member profile page to contact me. I quit even TRYING to keep the PM-PM box clear long ago.

Bill Hacker
 
Statistically, there just CANNOT be that much life left in the old P-P. But it does still work.
Or it DID.
That said, it no longer owes me a dime. As no one else has - so far - delved as deeply into possibility of repair as you have, I'm of a mind to gift it to you, "for the cause."

I shall be compelled to leave testing of that to you. Too much Old iron. Too little time.

Bill Hacker
Those of here who are interested in the Phase Perfect dpc-a10 thank you for your generous offer to the OP of this thread.
 
I have sent an email to thermite in response to his generous offer. I would like to thank everyone for their input and please keep it coming. Last night I downloaded Autodesk Eagle with the intentions of using it to create a schematic. I'm sure it will take me some time to learn the software let alone trace and draw out the board, but I am working on it in my spare time.

Paul
 
I use Eagle version 6.6.0, don't think it was all that hard to learn, have used it to design several PC boards that I had made. Not that I am an expert. It will think you want to design a PC board, so you have to select packages for all the schematic symbols you want to use. Kind of a hassle if all you want to do is draw. There may be a way to turn this off, I never really looked.

Just a very simple hand drawing showing AC line input circuit, parts and their interconnection to the IGBTs and their drivers, then components from the IGBTs on out to the 3 phase connection would be a very good start IMHO.

My favorite method for drawing less complicated schematics that don't get turned into a PC board is drawing on a vector layer in an old photo editor, made by Jasc Software, Paint Shop Pro version 7.02 You can use use images of component packages you find on the web on a raster layer of your drawing to get pin-outs and graphics easily for ICs and such.
 
I use Eagle version 6.6.0, don't think it was all that hard to learn, have used it to design several PC boards that I had made. Not that I am an expert. It will think you want to design a PC board, so you have to select packages for all the schematic symbols you want to use. Kind of a hassle if all you want to do is draw. There may be a way to turn this off, I never really looked.

Just a very simple hand drawing showing AC line input circuit, parts and their interconnection to the IGBTs and their drivers, then components from the IGBTs on out to the 3 phase connection would be a very good start IMHO.

My favorite method for drawing less complicated schematics that don't get turned into a PC board is drawing on a vector layer in an old photo editor, made by Jasc Software, Paint Shop Pro version 7.02 You can use use images of component packages you find on the web on a raster layer of your drawing to get pin-outs and graphics easily for ICs and such.

I will draw something simple up then.

What I can tell you off the top of my head is that the DC bus is in relation to one of the AC legs and not earth. If you check it to earth then it is a DC sin wave (+ or - 330ish volts depending on which rail). Also all the drivers signals are opto isolated from the DSP part of the board as well as power being isolated through transformers. I believe the only reason it has two IGBTs is because it is regenerative. Under normal load I believe one is acting as a half bridge rectifier and the other is generating the third leg. Please don't take anything I say as gospel as I am just going off the top of my head.
 
The "old blue" here is serenely comforted to know that at least TWO resident GENIUSES are on the job :D

Don't look at ME! Don't even know what IQ I have. Neither does DoD.

Faked IQ tests so I had free time to mess with girls. And motor cars. And MORE girls!

I at least knowed what was important!!

Besides.. I do waaay too much stoopid s**t to be truly "smart". PM is but one witness.

That's....at least ..."The American Way" ain't it?

:D
 
Same here, I often appear more intelligent that I am. I'm only ok at fixing things because I'm MUCH better at breaking things.

Funny you should mention that.. there's wisdom in it, and a higher percentage than average in a machine shop. Trick is.. to make more of the mistakes between our ears BEFORE we make them between our hands.

My "theory" is it being about the extra "dimensions" humans work with, infancy onward. Parallel-but-not-quite alternative futures.

X, Y,Z, and T-for-time the four obvious dimensions, but...

.. joined by the need of compensating for our senses having a delay, our reaction time to analyze inputs and choose a course of action another delay, our nerves, muscles, and body-part inertia adding more delay, yet ... to actually implement a decision.

.Despite which... a child - or a mouse - learns to run through a gnarly forest, a person can stone, blowgun, or point-shoot a bird on the wing, even with a handgun, fly a jet in a "dogfight", or pick wise financial choices when millions of others are "in the opposition" of the grand casino of a stock market.

How can that be? Any of it?

Small child, shrew, deer, bear, cheetah - running broken ground without falling down, birds doing the damndest of aerobatics, even mating on-the-wing, cetaceans storing a multi-dimensional simulcrum of their entire local zone in memory, updated even as they sleep but half the brain at a go - clear out to "Top Gun" jet jocks, olympic gymnasts ..spacecraft already OUT of the solar system..or a fortune in the stock market?

The future. Predicting it. Selecting of it. Changing it.

Time machine? We ARE one.

"Change agents", one and all.

Massively parallel, trillion-plus bit "CPU" and a distributed I/O system with local autonomy isn't just working ONE future X,Y,Z+time-line, weighing and discarding options to select what works best.

Dozens. Hundreds. Perhaps thousands.. or even more yet, rather. Sorta blurred as bits and pieces are emphasised, others discarded, faded away to smoke, never fully built-up. "Ghostlike", even.

Works well. Has lasted a long time. See-also Cetaceans. Cephalopoda. Arachnids, even.

Don't get TOO damned cocky, brethren.

A single ignorant Mosquito with a "brain" the size of a dust mote, if even ... will humble your arrogant ass in a New York Minute! Worse than just "humble", if carrying a virus in the bomb-bay.

Count on THAT much!

Miz God has a right Devilish sense of humour, if somehow you had missed that part.
 








 
Back
Top