Good evening,
I am not an electrical engineer, I am a tool and die maker who runs a shop full of old Yaskawa electronics. I have had success fixing spindle and servo drives in the past, now I’m kinda stuck. Need some help

I have a Yaskawa 626 mk2 spindle drive which blew up 2 weeks ago. It’s basically a VFD . It came up with an over current alarm , smoked most of the V leg of the output forward drive circuit. I replaced the parts . It also blew one side of a power darlington ( Toshiba mg100g1al3) and shorted the other side of the same phase on the H bridge. These modules are hard to get, as they have been discontinued. New old stock is pricey. I can’t just replace the drive with a new VFd, it has circuitry on it that orient the spindle for a tool change( Cnc mill). Upgrade from Yaskawa is 15k plus.
So, what can I replace these unobtainable power transistors with? Something modern?

thanks..

 

Is this what you have?
There are many in the market. I cannot find a data sheet. It might be " MG100G1AL3-5 is a(n) 2500V, 80A IGBT Power Transistor Module ". Without better data I hate to suggest a replacement.

 

It will be difficult to find a replacement for the old custom-made module.

 

That is exactly what I have. Most of what you will find on a Google search are in China . Quality is sketchy and I already have a box of “ new “ units that are obviously pull outs. See data sheet. Thanks.

 

I should add, after fixing the firing board and replacing the two fragged darlingtons, the drive runs. It motor knocks like bad gears and it pulls A TON of current accelerating .Once at speed , current draw is reasonable and the motor runs quiet. I’m kinda suspecting the B-E in the first of the pair in some of the modules are shorted. Low gain. I do t want to blow this up again, it was a lot of bench time to figure out. Thanks for all advice.

 

Just doing a quick Google search turned up several vendors selling them. New Old Stock and warranted pulls? Apparently it was a custom made part for TEAC/Toshiba so there are no equivalent parts without getting into reverse engineering a replacement module and testing it.
MG100G1AL3 equivalent - Search (bing.com)

Here is a list of vendors sourcing it.
MG100G1AL3 Price & Stock | DigiPart

 

If you have the specifications of the Darlingtons, especially current, voltage, and gain, it should be possible to duplicate them with two replacement transistors, although there will probably be some revisions to the mounting scheme.
Unfortunately I have not been involved with that particular aspect of power electronics using that style of transistors. I hope that you have the circuit diagrams of the drive because that will help a lot in coming up with the replacement arrangement.
I seriously suggest investigating what caused the failure, so that the destructive event does not happen again when things are repaired. I have seen that happen more than once.

 

600V, 100A (200A peak), 600W!!!

 

Looking at Datasheet Archive MFE4010 datasheet download which is a cross reference sheet there is this which appear to be equivalents?:

 

We are talking about a 40 year old drive and even when these were the height of technology, they were known for issues. In my case , just worn out 40 year old silicon. Fix one channel, 3-18 months later, another blows. The boards are old and the traces break, some from heat, some from vibration. Right now, I kinda need it running. The ultimate solution is a new control , then I can use any spindle drive I want, I’m leaning towards a WJ-200 Hitachi, and replacing the servo drives with AMC . It just takes a while to wire the I/o and getting the toolchanger .working consistently. A while team wrote the logic for it over months, I’m one guy doing it all. It’s a six week project. Need it running. Thanks.

 

Out of curiosity, Jayare, where are you located?

Paul
KI5VNH

 

600V, 100A (200A peak), 60KW!!!
I am greatly impressed with the specifications of that module. 0.2°C/W is that achieveable?

 

Many of those very high power devices have a much larger mounting area, so that there are 3 or 4 square inches to couple to a heat sink. And the base of the package is a big slice of what appears to be copper. So the answer is "quite probably yes, under perfect conditions." "Your results may differ."

 

600V, 100A (200A peak), 60KW!!!

I am using 1200V 300A parts that drive car/truck motors.

It clearly has a C-E diode inside. The current gain is very high for an old HV, HC transistor. Probably is a Darlington.
12uS of storage delay is bad. lol 40 years ago part.
One place called it a IGBT but clearly it is not.
I have many high voltage transistors but none that can take the 100A. sorry.

I have some modules that look like this. Not my project but I am using them at 1200V. I will get the part number today. They are IGBT and NOT the same thing. Can I get the schematic to see how the Base is driven. I might be able to modify things to work with IGBTs.

 

Out of curiosity, Jayare, where are you located?


I am in central Florida, Melbourne area.

 

Back to post #16.
What I have is oGCNS080A120S1 1200V 20A SiC MOSFET Not going to work.
Also IXXN110N65B4H1 650V 240A IGBT It might work but it needs a different drive circuit.
The MOSFETs and IGBTs need 0 to 15V of drive at 0 current while the old part needs 1V at 2A of drive.

 

Unfortunately, schematics are not available. Yaskawa isn’t nice that way, or any other... I have a maintenance manual pdf I can post. It has some clues. Below are a couple pictures of the drive power section with and with out the transistors installed and one with them complete. Wires are marked.

If I understand this correctly, one of the issues being faced is a modern igbt uses 5v switching, and this uses around 8v. This causes the transistor to lock on, “ saturated “. Maybe a small board could be made to switch an opto coupler to drive the 5v base voltage? Just thinking with my limited knowledge.

 

Unfortunately, schematics are not available

Unfortunately, that is all too common due to a certain country building what they like to call "Replicas" of technology they did not put the time and effort into developing themselves...

 

Would've come given you a hand if you were in Texas. Best of luck with this...sucks that you're out of commission until you can find a replacement. It's too bad it isn't an IGBT module; plenty of those around. You may be able to do an IGBT module, but advising you on how...I would be afraid to do that given the power involved and being so far away... Bowing out... -Paul KI5VNH