Repairing a power hybrid module

[mikeselectricstuff]

When hybrids go bad, it doesn't always mean it's Game Over

[Psi]

That's a pretty awesome repair.

[vk6zgo]

Great job Mike!
It reminds me of when I was in the TV Broadcast game.
Everything had to be fixed now!!
As we were on the other side of the world to many of our suppliers,we did many an unorthodox fix,though never anything quite as creative as this.
Some people of little imagination call it "butchery",but I think it shows a deep well of knowledge, imagination & adaptability.

[Kremmen]

Absolute brain surgery 

Actually that module is a familiar one - last year i made a proto driver for an ancient Fujitsu-Fanuc 5 phase stepper using the sister hybrid PMM5303 which is apparently identical but for the plastic housing shape. This driver was to replace a set of original stone age (huge!) linear drivers for said motors. (here on top of an Omron logic).


In the end i decided to scrap the motors in favor of true AC servo solution.
Got a set of 2 kW Yaskawa Sigma II servopaks as a donation - slightly older but very able devices and more than capable of moving my Matsuura 1000 vertical mill (this is a 6 ton CNC mill able to handle engine blocks and bigger). The only problem was that the servopaks did not fire up due to the aux power supplies shutting down immediately after start. Like the OP, i did not feel like giving up because these are devices worth thousands of bucks/euros so i started to investigate.

Sigma II out of its enclosure. At the bottom is the aluminium housing for the power IGBT module, rectifier, filter caps and fans.
Middle PCB is main bridge circuitry, IGBT drivers and external connections
Top PCB is controller - microcontroller + FPGA and aux circuits.


Finally the culprit turned to be a rotten hybrid controller for the flyback power supply. This is the middle board underthe controller PCB.



Fortunately one servopak did start with drastically reduced load so i was able to verify the feedback voltage from one of the flyback transformer coils. After that it was relatively straightforward to design and implement a replacement controller. This one is based on Fairchild FAN7554 and works like a charm 

The prototype powering a servopak. Display shows alarm code 9 - missing encoder which is as expected.


The replacement board. I ordered them from ITead. Out of 10 board 2 had a short between power and ground which was slightly disappointing, but 8 worked OK and only needed 3. Otherwise no problems.


Verifying functionality with a random flyback transformer with similar feedback coil.


Checking flyback waveforms with the test setup


The controller in its new home.
Total cost: <100 €; saved: up to some 4500 € depending how lucky one would be locating working ones in eBay and elsewhere (the equipment suppliers have ridiculous prices  ), or alternatively 3000 € + mail (Yaskawa Europe service center fixed price for repairing 3 units). Bliss 

[digsys]

Great job. Had to fix an ABS module for a Merc once (they wanted $4K to replace it). NO type of solder or flux
would take, so I drilled a couple 1.2mm holes, tapped to 1.5mm, found some 1.5mm screws in a head assembly
of a DVD drive, made tiny eyelets and linked them that way. Whole operation was under a large 3D boom
scope, which made it quite easy. Then I was inundated with these bastids :-)

[firewalker]

Grate repair.

Alexander.

[KTP]

Awesome video.  You are starting to give Dave a run for his money...if you start sniffing stuff I think you could put him out of business 

I use 28 gauge 7 strand teflon coated wire for repairs.  The stuff is awesome (but expensive).  I noticed when you touched the solder iron to your wire the insulation melted right off.

[MikeK]

Very nice.

Is that yellow tape special?  It reminds me of the yellow tape on lithium-ion batteries.