Phase perfect DPC 20... Not so perfect at the moment

I need some help if there is any help possible. I purchased a dpc20 on eBay from a machine reseller that was untested as is. I receive it in well packed very clean condition( not so clean that it looked suspicious).. I pulled the covers and gave everything thing a once over, looking for noticeable damage or disrepair. All looked good. Wired it and checked my voltage before I flipped the breaker. Getting 120v to each leg and 240v across. Good so far. Crossed my fingers and flipped the breaker. Nothing for several seconds then fans kick on. Check input voltage still 120v,120v, checked t1 120v, t2 120v, t3 nothing. Little red led blinking. Red only. Check online and it says internal short. Contacted phase technology and the sent a trouble shooting guide. Guide says problem with IGBT’s or the drivers of the IGBT’s on the control board. I contacted phase technology again and asked what to check and where the igbt’s are located(I found them) they said the dpc series is no longer supported and could offer no technical support..

Has anyone had this issue? Any help and guidance would be greatly appreciated. I’m not an electrical wiz but I can read, research, and figure things out. Just need a direction to move

Sounds like one of the IGBT modules has lost its mojo. A good electronics repair shop should be able to diagnose and fix it.

saintscustomwork said:

I need some help if there is any help possible. I purchased a dpc20 on eBay from a machine reseller that was untested as is. I receive it in well packed very clean condition( not so clean that it looked suspicious).. I pulled the covers and gave everything thing a once over, looking for noticeable damage or disrepair. All looked good. Wired it and checked my voltage before I flipped the breaker. Getting 120v to each leg and 240v across. Good so far. Crossed my fingers and flipped the breaker. Nothing for several seconds then fans kick on. Check input voltage still 120v,120v, checked t1 120v, t2 120v, t3 nothing. Little red led blinking. Red only. Check online and it says internal short. Contacted phase technology and the sent a trouble shooting guide. Guide says problem with IGBT’s or the drivers of the IGBT’s on the control board. I contacted phase technology again and asked what to check and where the igbt’s are located(I found them) they said the dpc series is no longer supported and could offer no technical support..

Has anyone had this issue? Any help and guidance would be greatly appreciated. I’m not an electrical wiz but I can read, research, and figure things out. Just need a direction to move

Click to expand...

The issue is that the basic task they do is silly-simple. JST has posted a block diagram on PM.

Downside is that by itself, that simple task has near-zero stability as far as responding - as a P-P does do, and rather well - to changes in load. Or just no-load to load to no-load.

No electromagnetic inertia as there is no transformer nor the windings an RPC has as rotary transformer. No heavy rotor storing mechanical energy.

Dynamic monitoring, feedback, and control are the REAL core of a P-P. As if it were a living beast.

That's what most of what is on those PCB's is about. The boards are proprietary. Some of the components are as well.

It takes a well-equipped shop, lots of decent equipment, and relevant, not just generic, experience to fault-find and repair in an affordable (read SHORT) amount of time.

What the tech has to find doesn't stop at a blown pass-element. (S)he has to find the CAUSE of its failure and correct that as well. And test. And test. And test again.

The greater problem yet?

There aren't enough of them failing often nor regularly enough that are not simply replaced for there to be much of a business made out of repairing them once the time has arrived that Phase Technologies has decided to cease.

That P-T had repaired their P-P for as many years as they had done sort of sucked all the Oxygen out of opportunity for a third-party repair shop to even get familiar with them, years ago. Not much meat left on the bones, now.

I didn't wait for my DPC-10 (blue) to fail. The PT330 (white) is already here.

Many, many, MANY things I would rather have spent the money on, but there it is.

The seller probably got it for nothing, or a scrap price. So that's how some people feed off of other honest type people....

Are the IGBT's in a module or are they individually installed on a heat sink?

Does Phase Technologies monitor PM? There are enough incidents like the OPs reported on PM that I am leery of buying a PP, even though I would love to ditch my noisy RPC. It's not so much a concern with low reliability of the unit, as it is with the "tough noogies" attitude when they no longer support their product. They need some real competition. I wonder why there isn't any. Has anyone looked at their patents?

I don't think a PP is any less reliable that other similar devices (VFDs, drives, etc). But, it does seem odd that they won't repair them. And, they are low volume devices, so there's not going to be a lot of independent repair shops with experience fixing one.

rons said:

The seller probably got it for nothing, or a scrap price. So that's how some people feed off of other honest type people....

Are the IGBT's in a module or are they individually installed on a heat sink?

Click to expand...

Individually on gigantic heat sinks. They are going to be a bear to get to. Behind the board with all its numerous wires. I’ve sourced some new igbt’s and read an article on testing the old ones. I think I will give them a try unless someone chimes in with a definitive solution

Before getting excited about IGBTs, you should know that the driver system that controls the IGBTs may also be damaged. Those circuits are "usually" damaged when IGBTs fail, and the PCB that carries the control circuit is likely the part that they no longer have available.

If it were me, I would verify that the control signals were working before I spent money on IGBTs.

JST said:

Before getting excited about IGBTs, you should know that the driver system that controls the IGBTs may also be damaged. Those circuits are "usually" damaged when IGBTs fail, and the PCB that carries the control circuit is likely the part that they no longer have available.

If it were me, I would verify that the control signals were working before I spent
money on IGBTs.

Click to expand...

how do I go about doing that?

saintscustomwork said:

how do I go about doing that?

Click to expand...

You may not be able to. If the PP is in an alarm state, the driver signal is probably not functioning.

Just my experience, but I've very rarely seen the IGBT take out the drive circuit. I've seen the drive circuit fail and take out the IGBT, which is basically the same result. Usually the IGBT fails as a dead short.

ewlsey said:

You may not be able to. If the PP is in an alarm state, the driver signal is probably not functioning.

Just my experience, but I've very rarely seen the IGBT take out the drive circuit. I've seen the drive circuit fail and take out the IGBT, which is basically the same result. Usually the IGBT fails as a dead short.

Click to expand...

so would you chance 50 bucks for a replacement igbt and see what happens?

ewlsey said:

I don't think a PP is any less reliable that other similar devices (VFDs, drives, etc). But, it does seem odd that they won't repair them. And, they are low volume devices, so there's not going to be a lot of independent repair shops with experience fixing one.

Click to expand...

Warranty in the booklet is only for one year.

They supported them for-fee for a lot longer than that before finally ceasing.

saintscustomwork said:

so would you chance 50 bucks for a replacement igbt and see what happens?

Click to expand...

Yes, if your testing shows it is bad. Most times the IGBT just suffers a premature death and replacing it fixes the issue. In more rare cases, the control has issues as well that can take out the new module or leave you no close to fixing the issue.

saintscustomwork said:

so would you chance 50 bucks for a replacement igbt and see what happens?

Click to expand...

I would chance the $50, if it fails, sell it on ebay, as-is, or for parts

You should be able to replace the igbts with 5$ replacements (600v,20 ish amp rated) and it will still work, but without any real load capacity. After you pull the originals, check that they are not shorted, and with a 9 volt battery, turn on. If they do then they are probably not bad.

I believe the blue units are at least ten years old and have not been supported for several years. I also bought one and got no help from the seller but did get a better price on a replacement from PP. The old ones had a single board vs several boards on the new that evidently are cheaper to replace and still supported. Paying more than a few hundred for a blue unit is asking to become poor. Dave

When an IGBT fails, it usually shorts. That means the silicon melts.

When that happens, the gate insulation fails (melted) and "some amount of" the high voltage is applied to the gate drive circuitry. Most gate drive circuits cannot "take" having a couple hundred volts applied, as they operate on maybe 24V. The usual case is killing the gate drive.

Sometimes it does not happen. Sometimes the gate drive is just "damaged", and fails to pull down hard, or becomes slow, heating up the IGBT.

I do not know how the PP drive works, some types are less likely to be damaged. But standard practice is to just plain replace the drive circuit if the IGBT(s) fail. PP themselves probably did that, although they may have just tested the drive output and not replaced if good.

You don't have a replacement, so you just need to know if there is a good shot at the thing working if IGBTs are replaced.

Without knowing the circuit, I would load the drive with a smallish capacitor, around 0.0033 uF or 0.0047 uF, and see if it has what I would consider "reasonable" rise and fall times as seen on a 'scope (one with isolated inputs).. If it passed that (with no IGBTs), then I would probably risk it, knowing that it could be a waste of effort and money.

I doubt it will trip out for not having IGBTs. It may be possible to fool the circuit into thinking it has the right output volts, if it errors out as under voltage.

JST said:

When an IGBT fails, it usually shorts. That means the silicon melts.

When that happens, the gate insulation fails (melted) and "some amount of" the high voltage is applied to the gate drive circuitry. Most gate drive circuits cannot "take" having a couple hundred volts applied, as they operate on maybe 24V. The usual case is killing the gate drive.

Click to expand...

It costs maybe a dollar per kilowatt of inverter capacity to put two diodes at the output of the gate driver, and a TVS diode or some other surge protection device capable of handling the short circuit current of ~300 volts flowing through the gate drive resistor. the last time I witnessed this was with a 10 ohm, 1/4th watt gate drive resistor and the fets blew up which took out the gate drive resistors. the $0.50 sod123 tvs diode successfully protected the gate driver. Granted the short circuit current would also include the resistance of the melted silicon, but i was pleasantly surprised to see a sod 123 diode cause a 1/4th watt resistor to become a fuse, and suffered no damage.


Anyhow, there is another often overlooked problem regarding inverters. usually when one igbt fails the other still turns on like it was supposed to, because it is just not practical to build protection circuitry that can act within the dead time normally given to IGBTS. not that it can't be done.. its just more than a couple extra dollars.

When the both of the igbts short out the current will rise.. and rise.. and rise, according to the resistance of the igbts, the dc bus capacitors, and the inductance of the dc bus and the igbts, which is usually as low as possible. Lots of igbts now can handle 5 to 20uS of short circuit without catastrophic failure. (very significant thermal stress though) and it IS practical to build circuitry that can detect an igbt failure and turn the next one off within that 5-20uS window, its just that also costs a few dollars.

anyhow.. when the IGBT successfully vaporizes; that current continues and the voltage will rise to whatever it takes to push all of the energy into the plasma, (which is a significant shrapnel hazard!) before or just after the arc finally extinguishes this voltage can be in the tens of thousands of volts. this voltage will not destroy a fully floating gate driver.. but usually they aren't fully floating.

not many inverter manufacturers use separate, optically coupled fully floating gate drivers that can handle 1000 or more volts with a rise time of hardly anything, but it is practical.

saintscustomwork said:

Individually on gigantic heat sinks. They are going to be a bear to get to. Behind the board with all its numerous wires. I’ve sourced some new igbt’s and read an article on testing the old ones. I think I will give them a try unless someone chimes in with a definitive solution

Click to expand...

What are the IGBT part numbers?

Do you have a 2 channel oscilloscope?

rons said:

What are the IGBT part numbers?

Do you have a 2 channel oscilloscope?

Click to expand...

eupec ff300r12ke3
No scope.