Can anyone repair a Huanyang VFD

[Bob Willson]

Malibu suggested that I start a new thread for this, so...Can anyone help me?

I recently bought a Huanyang HY03D023B inverter, which is a 220v input 3.0 KW output VFD.

I could not get it to work, and told the supplier in China that it was a dud. After many days of measuring and them assuring me that it was a minor problem that they could fix easily, they finally said that they couldn't fix it and that they would send me a new inverter (I have to pay another $35 freight. :(

The thing that is wrong with it seems to be no voltage at all on the output side. It measures very high resistance between all output terminals:
The resistance measurements are:
U-V = 4.7 Megaohms (4M7)
U-W = 12.44 Megaohms (12M44)
V-W = 4.82 Megaohms (4M82)

and then, when power is applied:

Measurements across the U-V terminals was 10.6v, Then after pressing start it was 11.1 then after pressing stop it was 10v

Measurements across the U-W terminals was 4.3v, Then after pressing start it was 4.0 then after pressing stop it was 2.6v

Measurements across the V-W terminals was 8.5v, Then after pressing start it was 8.3 then after pressing stop it was 10.6v

As soon as I press the start button I get a fault dL which according to the manual is an output short circuit.
Their suggested course of action for this fault is:
1) Check whether the connection wire of the motor has a short circuit. (The motor is not connected)
2) Check whether the insulation of the output wires is good. (No output wires attached)
3) Send for repair.

So, does anyone have a circuit diagram for this inverter, or better still can anyone advise me what component has failed?

The manual also says that the VFD is fuse protected. I cannot find a fuse anywhere in the box.

Your replies would be much appreciated.



Bob Willson

PS Humphrey9999 suggeted that it may just be a matter of replacing the IGBT transistor. (At least, I think it is a transistor)

[sjm]

The manual also says that the VFD is fuse protected. I cannot find a fuse anywhere in the box.


I'm assuming you're looking for a glass tube with a wire running through it?

Most modern PCB mount fuses will look like a large SMD resistor. Could be a 1206, 1210, 1810 etc. If it's a through-hole part, it could resemble a flat disc, like a 5 cent coin, although not necessarily round - it could be square.

As suggested in the other thread, please post a high resolution photo of the board.

[Bob Willson]

251705251706251707251708251709

[Humphrey9999]

Here are the pinouts, from a manual. (Does this look like the same one Bob?): -

251710

Bob, what's on the back of the bottom board?

[Malibu]

Malibu suggested that I start a new thread for this, so...

(Actualy, I thought it might be good for Neil to start a sub-forum under electronics for anyone that might like their stuff fixed.. :) But this is good anyway )

If it helps, my inverter is the same brand of Hangyung-whatever, but a 2.2Kw instead. I just took a run downstairs and measured the output resistance (Motor unplugged, wires still connected to the drive) and were as follows -
U-V = 3.474 Meg
V-W = 3.536 Meg
U-W = 3.550 Meg

Near enough to balanced for me. At a glance on your measurements, one of them is way out so maybe as Humphrey suggested, an IGBT is shot (open/shorted)?
As a side note, interesting enough it was 55 Meg to earth across all phases.

Hopefully this helps your cause :2tsup:

[Bob Willson]

Very similar but not exactly the same.251712

[Bob Willson]

Looking at the photo vfd bottom half -2

At the top right corner there is a transistor that seems to have a short across the top two legs but not across the bottom two legs about 0.5 k and 2.0 meg respectively. Or is this normal?

[Humphrey9999]

Very similar but not exactly the same.251712

Thanks. (Mine was fairly close, but could well be a different or earlier model.)

Thinking about it, and going by the 'output short circuit' indication, there's possibly a short before the physical output connections, which wouldn't show up measuring those connections because the U, V and W leads appear to be capacitively coupled. ie. A capacitor in series with the output.
Measurements before those capacitors could show drastically different results.
Otherwise, I'd track each output line back, as far as the transformer if necessary, looking for shorts including solder bridges between tracks/pins from assembly.
Current sensing is probably done at the primary of the transformer, so the fault could lay with the switching element, transformer or anything afterwards.

[Humphrey9999]

Looking at the photo vfd bottom half -2

At the top right corner there is a transistor that seems to have a short across the top two legs but not across the bottom two legs about 0.5 k and 2.0 meg respectively. Or is this normal?

What transistor? Is it on the other side of the board? Q22? What are those transistors on the other side of the board, Q17 to Q22? (Type) (The heatsink probably needs to come off.)

[Bob Willson]

Sorry, I meant top left. :-:-:-:- Next to the red LED.

Again.. :-:-:-:-

[Humphrey9999]

Sorry, I meant top left. :-:-:-:- Next to the red LED.

Again.. :-:-:-:-

Ah, now I can see it - in a TO-220 case, (metal tab).
What numbers are printed on it?
(I'd still be looking at Q17 to Q22, too.)

[Bob Willson]

Those are beneath the bottom PC board, fastened to the cooler. as such they are quite difficult to get at. Also, some of the values change a little as I read them.

However, given the above, then:

Q17: E - C = 5M0, C - G = 7M0 E - G = 10K0
Q18: E - C = 1M2, C - G = 1M2 E - G = 10K0
Q19: E - C = 6K0, C - G = 1M7 E - G = 10K0
Q20: E - C = 000, C - G = 10K0 E - G = 10K0
Q21: E - C = 5M0, C - G = 5M0 E - G = 10M0
Q22: E - C = 5M0, C - G = 10K0 E - G = 4M5

These seem to me to be all over the place. The only really bad one is the Q20 E - C= 0

[Humphrey9999]

Those are beneath the bottom PC board, fastened to the cooler. as such they are quite difficult to get at. Also, some of the values change a little as I read them.

However, given the above, then:

Q17: E - C = 5M0, C - G = 7M0 E - G = 10K0
Q18: E - C = 1M2, C - G = 1M2 E - G = 10K0
Q19: E - C = 6K0, C - G = 1M7 E - G = 10K0
Q20: E - C = 000, C - G = 10K0 E - G = 10K0
Q21: E - C = 5M0, C - G = 5M0 E - G = 10M0
Q22: E - C = 5M0, C - G = 10K0 E - G = 4M5

These seem to me to be all over the place. The only really bad one is the Q20 E - C= 0

Yep, that 0Ω reading between the emitter and collector of Q20 looks sus. A schematic would be handy right now, but I can't get my hands on one. I'd remove the heatsink for access, remove Q20 and see if it still measures the same. These are probably 3 pairs of push-pull drivers, driving the output.
For the other measurements, do they change if the DMM leads are reversed?

Some values will appear to change during measurement if there's a capacitor connected. The resistance will increase as the cap charges.

Enough for me for tonight - a snack and off to bed.

[Malibu]

These might be helpful. They came from the Mach 3 forum

[Humphrey9999]

These might be helpful. They came from the Mach 3 forum

Just what the doctor ordered, John.

Well, my guess of capacitive coupling was wrong, but not to worry. Didn't realise it had to go down to 0.5Hz, too low to capacitively couple with such a small cap. Those caps are just across the supply for smoothing.

The output is 3 push-pull stages, though, and measurements taken on Q18, Q20 and Q22 should be roughly the same, so it looks like Q20 is cactus, unless one or both of the zener diodes connected to the gate have failed short-circuit. (These zeners are 18V, 1W, to prevent the gate voltage rising above about 18.6V.) Even if this is the case, the IGBT could still be dead. Shame the schematic doesn't say exactly what the IGBTs are.
The other measurements might have varied due to the polarity of the DMM leads.

The heatsink must come off next. (Bob, you'll need some heat transfer compound when you replace it, to ensure good cooling.)


Bob, does your DMM have a diode test function?

[Malibu]

Shame the schematic doesn't say exactly what the IGBTs are.

Yeah, but the guy that posted on Mach 3 forums provided a link to an e-bay seller from where he got his replacements from. I traced it out, but the link was dead, so no help there.

Here's the forum link where I got all this from - Huanyang VFD controller plugin (http://www.machsupport.com/forum/index.php?topic=14182.230)
It deals with communications and it's 32+ pages long!

Found it... How's this? :)

[Humphrey9999]

Yeah, but the guy that posted on Mach 3 forums provided a link to an e-bay seller from where he got his replacements from. I traced it out, but the link was dead, so no help there.


Doesn't really matter, since the number will be visible once the part is removed.

After a quick search, I think they're FGA25N120ANTD 1200V 50A NPT Trench IGBTs. (A mouthful.) Can be verified as above.
In Australia, I think they're stocked by RS Electronics.

[Humphrey9999]

Yeah, but the guy that posted on Mach 3 forums provided a link to an e-bay seller from where he got his replacements from. I traced it out, but the link was dead, so no help there.

Here's the forum link where I got all this from - Huanyang VFD controller plugin (http://www.machsupport.com/forum/index.php?topic=14182.230)
It deals with communications and it's 32+ pages long!

Found it... How's this? :)

That's where I've been getting my info too. Good to confirm that the IGBTs are FGA25N120ANTD. I just checked. RS have them for $6.80 ea before GST, here: -
Buy IGBT Transistors Transistor,IGBT,with Diode,N-channel,50A,1200V,TO3P Fairchild Semiconductor FGA25N120ANTDTUX online from RS for next day delivery. (http://australia.rs-online.com/web/p/igbt-transistors/6715395/?searchTerm=FGA25N120ANTD&relevancy-data=636F3D3226696E3D4931384E4B6E6F776E41734D504E266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E5B5C772D5C2E2F252C5D2B2426706F3D313326736E3D592673743D4D414E5F504152545F4E554D424552267573743D46474132354E313230414E54442677633D424F544826)

[Malibu]

I just checked. RS have them for $6.80 ea before GST

That's a better price than buying a new inverter! Seems awfully cheap for RS too, unless they've gotten more competitive in the last few years?

It occured to me while I was looking around internet for the schematic and other bits and pieces, just how many of these drives must be out there! Some good reports, some bad and most of them to do with parameter settings. (half the hits were from the CNC forum here on WWF), but they look to be everywhere!
They seem to be an essential part of any CNC system, which makes sense for the price verses function features. I know when I've bought drives from SEW or Siemens over the years as an electrician, sometimes I'm staggered by the cost of the drives (which have only ever been 50hz outputs, except one, which was a high speed spindle drive from Germany... and the list price was $39k for that one...)

[Humphrey9999]

That's a better price than buying a new inverter! Seems awfully cheap for RS too, unless they've gotten more competitive in the last few years?

Might be even cheaper yet - it's possible that only one or both of the zeners is dead and that the IGBT is OK.

[jhovel]

I found a source on eBay for 10 of there IGBTs for $25. Free postage.
I would like to have at least three on hand for any (inevitable) mishaps with one of my three inverters....
Does anyone want me to split this lot up and post some to you? $2.50 each plus $0.60 for local postage.
Just an idea.

[forp]

I found a source on eBay for 10 of there IGBTs for $25. Free postage.
I would like to have at least three on hand for any (inevitable) mishaps with one of my three inverters....
Does anyone want me to split this lot up and post some to you? $2.50 each plus $0.60 for local postage.
Just an idea.

Please don't buy from eBay unless you know the seller personally. Electronics component are fake a plenty on EBay. Check prices at digikey or x-on.com.au they have reasonable pricing.

[sjm]

Electronics component are fake a plenty on EBay.

It's not just ebay. The company I work for recently wrote off 200k worth of assembled product because the manufacturer had sourced grey market components that didn't meet the datasheet specs (high speed BGA ADCs). Lawyers got involved, and it turned out cheaper to get them rebuilt, rather than reworked.

[Malibu]

Might be even cheaper yet - it's possible that only one or both of the zeners is dead and that the IGBT is OK.

I'm the first to admit that this sort of stuff is a little above my knowledge base, although I might be able to muddle my way through it.
But looking at the circuit, if the zeners are open or shorted, would that give the 0 ohms reading across the E-C of Q 20? It would make a little sense for the voltage readings taken if that were the case, but I would have thought it more likely that the internal diode would have copped a belt, hence the 0 ohms.

[Humphrey9999]

Gotta agree with sjm and forp - for the cost of a few bucks, better to stick with the Fairchild originals.
For fast delivery, RS can't be beaten, but Mouser or Digikey are probably cheaper.

[Humphrey9999]

I'm the first to admit that this sort of stuff is a little above my knowledge base, although I might be able to muddle my way through it.
But looking at the circuit, if the zeners are open or shorted, would that give the 0 ohms reading across the E-C of Q 20? It would make a little sense for the voltage readings taken if that were the case, but I would have thought it more likely that the internal diode would have copped a belt, hence the 0 ohms.

My eyes are at fault, I guess. Been having trouble with them lately, that's why I don't do any hands-on electronics any more. I could have sworn the short was E-G, not E-C. Sorry :-
Won't be the (zener) diodes. Either the junction or the inherent protection diode, as you say.

[Malibu]

I found a source on eBay for 10 of there IGBTs for $25. Free postage.
I would like to have at least three on hand for any (inevitable) mishaps with one of my three inverters....
Does anyone want me to split this lot up and post some to you? $2.50 each plus $0.60 for local postage.
Just an idea.

3 inverters with 6 per unit is 18 IGBT's.
25 bucks on 10 drivers for peace of mind is pretty cheap, even if half are below quality. I'd buy them as a last resort, but would prefer RS or Mouser

[Malibu]

My eyes are at fault, I guess. Been having trouble with them lately

Yeah, I know what you mean!

Hell, I'm having trouble keeping up with this thread! I hardly get a reply posted and there's another 3 postings done already! :)

[jhovel]

Please don't buy from eBay unless you know the seller personally. Electronics component are fake a plenty on EBay. Check prices at digikey or x-on.com.au they have reasonable pricing.

Thanks for the warning! On further search, they are available from all sorts of reputable electronics dealers at around the same price and closer to the RS price.
In any case, they are easily and cheaply available. No need to have stocks on hand after all.
I expected them to be difficult to get for some reason.

Also, I didn't twig that there were 6 in each inverter - I could only see three and assumed that there were 2 in each package. Wrong guess....

[Bob Willson]

My thanks to everyone who has kept this thread going. I wasn't able to reply because of the flooding. I only just got power back at 1 am last night.

I have some things to do, but soon as I can I will take off the IGBT and measure it again and replace it if necessary.

[Humphrey9999]

My thanks to everyone who has kept this thread going. I wasn't able to reply because of the flooding. I only just got power back at 1 am last night.

I have some things to do, but soon as I can I will take off the IGBT and measure it again and replace it if necessary.

Hi Bob. Just noticed where you live. Did you cop it very bad? (I used to live in Goodna, glad I don't any more.)

I was battening down the hatches ready for when it hit the Illawarra, but we were lucky and it fizzled out early.

[Bob Willson]

Hi Humphrey

No, I was OK. I didn't really expect to go under again, because I live half way up a slope which raises me about 2 metres from the poor buggers below me.
As it turned out, the water heights in Brisbane were not as bad as had been predicted. Unlike Bundaberg which is really copping it hard.

Really all that happened to me is that I lost power and internet for a few days and I think that I only lost those because SEQEB was playing it safe.

[Humphrey9999]

Hi Humphrey

No, I was OK. I didn't really expect to go under again, because I live half way up a slope which raises me about 2 metres from the poor buggers below me.
As it turned out, the water heights in Brisbane were not as bad as had been predicted. Unlike Bundaberg which is really copping it hard.

Really all that happened to me is that I lost power and internet for a few days and I think that I only lost those because SEQEB was playing it safe.

Good to hear. My brother lives in Brisbane, too, but wasn't seriously affected.

[Bob Willson]

OK, I have removed and measured the resistance of the IGBT.

G - C ∞
G - E ∞
C - E 3M8

[Malibu]

G'day Bob

If you have a diode test on your multi-meter you're probably better off using that.
Do the same tests on diode setting, and each time, swap the two leads around - so, first off red to C, black to E then black to C, red to E, etc, etc...
As a final comparison, check the readings against another IGBT, or maybe two. They should be approximately the same readings.

[Bob Willson]

Tester set for diodes.
With the positive lead to the emitter and the black lead to the collector I get a reading of 0.425. No other leads give any signal at all.
The other transistors give the following readings in circuit.

Q17: E - C = ∞, C - G = 0.431 E - G = 1.7 to 1.9
Q18: E - C = 1.4, C - G = 0.431 E - G = 1.7 to 1.9
Q19: E - C = 1.084, C - G = 0.000 E - G = 1.7 to 1.9
Q20: E - C = 0.425, C - G = ∞ E - G = ∞
Q21: E - C = ∞, C - G = 0.431 E - G = 1.7 to 1.9
Q22: E - C = 1.4, C - G = 0.431 E - G = 1.7 to 1.9

The E-G readings were very similar in all cases as also were the C - G
Q19 has two anomalies.
I hope that I didn't stuff up any measurements, it all gets a little confusing after a while. :rolleyes:

[Humphrey9999]

OK, I have removed and measured the resistance of the IGBT.

G - C ∞
G - E ∞
C - E 3M8

Hmm. No C-E short there, now.
Measure between the C and E pads on the PCB, where the IGBT came from.

[Humphrey9999]

Tester set for diodes.
With the positive lead to the emitter and the black lead to the collector I get a reading of 0.425. No other leads give any signal at all.
The other transistors give the following readings in circuit.

Q17: E - C = ∞, C - G = 0.431 E - G = 1.7 to 1.9
Q18: E - C = 1.4, C - G = 0.431 E - G = 1.7 to 1.9
Q19: E - C = 1.084, C - G = 0.000 E - G = 1.7 to 1.9
Q20: E - C = 0.425, C - G = ∞ E - G = ∞
Q21: E - C = ∞, C - G = 0.431 E - G = 1.7 to 1.9
Q22: E - C = 1.4, C - G = 0.431 E - G = 1.7 to 1.9

The E-G readings were very similar in all cases as also were the C - G
Q19 has two anomalies.
I hope that I didn't stuff up any measurements, it all gets a little confusing after a while. :rolleyes:

With an IGBT, due to the insulated gate, a diode test will only be able to measure the internal diode, E-C, but any measurements to the gate will be infinite. 0.43V is about right for the diode, I'd say.

[Humphrey9999]

Bob, I remember you saying that the supplier would replace that unit, but that you'd have to pay $35 postage. Returning it is really the best option. Unless you don't have to return this one to them, in which case you could end up with two.

[Malibu]

Tester set for diodes.
With the positive lead to the emitter and the black lead to the collector I get a reading of 0.425. No other leads give any signal at all.


That's about what I'd expect for a forward biased diode, unless you get the same reading with the leads swapped over.
The 1.7 to 1.8 on the G-E would be the path of the 10k resistors and zeners.
It's a bit odd that now the 0 reading is on the G-C of Q19, unless it's the opto driver at fault - PC2?

[Bob Willson]

Hmm. No C-E short there, now.
Measure between the C and E pads on the PCB, where the IGBT came from.

C - E 3M3 resistance both ways (+ and -)
C - E diode tester: 0.487 with + to E and - to C

[Bob Willson]

Bob, I remember you saying that the supplier would replace that unit, but that you'd have to pay $35 postage. Returning it is really the best option. Unless you don't have to return this one to them, in which case you could end up with two.

No, what I said was that they are supplying me with a new one, but that it cost me an extra $35 postage. This saves me returning the old one to China. BUT I would like to be able to fix this one so that I then have a spare. .. Quite apart from that, I hate throwing away something that may be able to be fixed relatively easily.

[Humphrey9999]

No, what I said was that they are supplying me with a new one, but that it cost me an extra $35 postage. This saves me returning the old one to China. BUT I would like to be able to fix this one so that I then have a spare. .. Quite apart from that, I hate throwing away something that may be able to be fixed relatively easily.

Right. Fair enough.

In that case, if all else fails, the other option is to wait for the new one to arrive, then do comparative measurements between equivalent points on both.

Did you measure between the C-E pads on the PCB, with the IGBT removed? That C-E short that you measured must be there, since the IGBT itself isn't shorted, unless there was previously a solder bridge across the pins on the PCB causing the problem.

[Humphrey9999]

C - E 3M3 resistance both ways (+ and -)
C - E diode tester: 0.487 with + to E and - to C

These aren't with the IGBT removed, are they? This is getting confusing.

[Malibu]

G'day all,

I've just stripped my drive down so I could give it a good blow out to get rid of the layers of MDF dust in it. While it was apart, I took a few measurements and noted down the results.
The red line in the picture is the red lead of the DMM, the blue is the black lead. GCE markings are from the datasheet specs. The resistance readings were typical of all IGBT's, no matter which way the leads were put on the pads.
All readings were an in circuit test on a Fluke 189 DMM

[Bob Willson]

It would appear that my VFD is a little different from the 2.2kw drive.
Underneath the main body of the VFD, buried in the cooling fins, are three large capacitors, labelled C1, C1 and C2
I had to disconnect a wire to one of the capacitor connectors when I removed the lower board, and that may be what is causing the difference between the readings.

[Malibu]

Yours might be a little different (For some reason I have 4 red power wires on the board, you only seem to have 3), but I didn't remove the driver board. I only took out the keypad (and extended the wiring to mount it outside the cabinet while I was at it), then the control board. What the picture shows is the front of the board as mounted in the inverter - mine will still have the chunky caps buried deep down, but I didn't need to get to them.

[RayG]

Hi All,

I've got 2 Huanyang 4KW vfd's to repair, I'm planning on swapping out the IGBT's and zeners, has anyone done this repair? and is there anything I should be watching out for.

The symptoms are drive logic powers ok, but as soon as the output stages crank up I get a dead short and the RCD's all drop out.. :no:

So, I'm pretty sure it will be a failed IGBT.

Regards
Ray

[Humphrey9999]

has anyone done this repair? and is there anything I should be watching out for.

Yep - soldering irons are hot - don't burn your fingers. :D

Don't forget the heat-transfer paste, either.

[Malibu]

Hi All,

I've got 2 Huanyang 4KW vfd's to repair, I'm planning on swapping out the IGBT's and zeners, has anyone done this repair? and is there anything I should be watching out for.

The symtoms are drive logic powers ok, but as soon as the output stages crank up I get a dead short and the RCD's all drop out.. :no:

So, I'm pretty sure it will be a failed IGBT.

Regards
Ray

G'day Ray

Reading through the previous posts on this subject, you'll probably be pretty safe in new IGBT's all 'round. Maybe zeners, but they're only to limit voltage to the gate, so unless they're shorted, they will probably be OK. Easy enough to see if they have a short by using the diode test on the multi-meter. Leads one way across the zener, take note of the reading... leads swapped over should be a completely different reading. If you have a low reading both ways, you might have a problem (There's more to it than that, but it's a good start!)
For the price of IGBT's from what's already been posted, it's pobably worth the change-over and at the very least, a good exercise for information for anyone else that may have these inverters. A little knowledge shared goes a long way!

Oh... and things you should watch out for? hmmm... turn the power off first before sticking your fingers inside! :)

Speaking of which.. Bob? How did your repairs go, if at all?
It would be good to hear that yours is alive and breathing again! :2tsup:

[Bob Willson]

Hi John
I got the new VFD from my Chinese supplier, plugged it in and all went well. So very happy there and also happy that China has again proven good on its warranties.

I desoldered one of the IGBTs but that is as far as I have gone so far as I wanted to get on to the rest of my very long and protracted fencing project.

The new motor/VFD combo works great, and I have no problems cutting 1mm of aluminium at 14000 rpm and up to about 1500mm a minute using a 6mm dia 2 flute cutter.

I will return to the VFD repairs a little later when I hear how Ray went with his repairs:D:wink::rolleyes:. The problem sounds as though it is exactly the same as mine. I will just need to source the IGBTs at a good price.

[Corwin]

Hi,
Any progress on the VFD repair? I just purchased mine (2.2kW) from a chinese seller. It passed the initial check, but after I first tried reverse run it started showing "dL" error. The symptoms are very similar to the mentioned above.
So far the seller says "we check all before shipping, it's good" and refuses to exchange. I keep asking but I'll possibly find myself with a broken VFD :~ Need to prepare for repair.

[Bob Willson]

Hi Corwin
I haven't actually followed through with any repairs yet.

However, we generally say who the suppliers of these VFDs etc are so that they know that we know who they are, and this helps to protect others from dodgy Chinese Ebayers.

Conversely, we give good rep to those who give good service and warranty, such as love-happyshopping (http://stores.ebay.com.au/love-happyshopping/_i.html?_nkw=spindle&submit=Search&LH_TitleDesc=1&_sid=749694567). Many people on this board have bought from him because of recommendations from this board and been very happy with the whole experience.

[Corwin]

Hi Bob, thanks for your reply.
Love-happyshopping don't ship to my country, so I had to search for other sellers. I bought from Chai (linearmotionbearings2008), and that kind of service is a surprise for me as I had purchased a few times from him before with no problem. I still hope he solves the problem. Otherwise I'll report on VFD repair :U

[Bob Willson]

Tell Chai that you are writing to this forum. Again, we have mostly had nothing but excellent service from Chai and he gets lots of business from the forum, so any bad reports can have an immediate negative affect on his business.

[Corwin]

We've worked things out with Chai. I received a partial refund and I'll try to repair the VFD. Finally Chai is a great seller, particularly considering that the time to open a case was over (some shipping problem, the parcel was returned to China twice according to the tracking).

What I have: the resistance between UVW is different, just similar to your VFD. The logic seems to work, the VFD runs OK without a spindle.
Need to open a cover and check that 2.2kW model also has FGA25N120ANTD. If so, the difference between models is mostly in power supply capacity.

[Corwin]

Need to open a cover and check that 2.2kW model also has FGA25N120ANTD.
Confirmed. Without unsoldering I can say that one phase has much lower resistance, at least one IGBT is shorted. I've ordered a couple FGA25N120ANTD off ebay $1.90 each. On the one hand, this is too low price and doubtful quality. On the other hand, don't they use the same cheapest components for Chinese VFDs? :)

[Bob Willson]

Good luck Corwin
Please let us know how it all turns out.

[Corwin]

So, I've just replaced the broken IGBT. The inverter works now - that's good. But the spindle rotates jerkily, particularly at low speed. The spindle is OK, at least it works well with other Toshiba VFD.
What else can possibly be wrong? I don't think I broke anything while repairing. Is it worth to replace the second IGBT in the phase? I'd try to replace them all, but I bought only two...

[Corwin]

Or it just can be a vector control mode in Toshiba VFD that gives a smooth rotation :) I need to check other modes

[Corwin]

Well, no such jerky rotation in any mode with Toshiba VFD.
I guess it's not normal that spindle body (lying on a table) is twitching when accelerating from rest? I don't remember that with my previous Huangyang either.

I should probably watch the VFD output with my O-scope...

[jhovel]

Hi Corwin, Bob Wilson, RayG and others,
has anyone finally successfully repaired a Huanyang VFD?
I'm still interested to find out all I can about these drives, mainly for future reference.

[Bob Willson]

Hi Joe I was going to try it, but the cost of the components put me off and the supplier sent me a new one under warranty anyway. I do still have all the parts though, so it may change at some time in the future.

[stal]

hI bob ..youre vfd is 4kw output if I understnd well.
please tell me the type of igbt you have
thankyou

[NCArcher]

That was 6 years ago. I'd guess that the IGBTs are in the bin by now.