Hybrid IC IPM YPPD-J017C J018C repair

[paulstef]

I had this very old Plasma TV with a Z-sustain failure. It uses the YPPD-J017C YPPD-J018C IPM. Two transistors were blown, but their anti-parallel diodes were ok.

Assuming they were either IGBTs or n-channel MOSFETs I replaced them with 2 common MOSFETs in a D2PAK.

No need to buy expensive IPMs for old TVs.

In these simpler IPMs (YPPD-J014C, YPPD-J018C, YPPD-J017C, STK795-821) there only seem to be two half-bridges, gate driver and buffer ICs, the rest are discrete components. Since this is a single layer board , probably with an aluminum substrate it's easy to deduce the schematic.

In my case the half-bridge is connected to VS and GND. VS is 185V.

I chose two MOSFETs with a VDS higher than VS (actually way higher) and higher continuous drain current rating than the VS fuse rating.

I'd also choose one with low gate charge.

There is not much space but a D2PAK fits ok. https://www.youtube.com/watch?v=UfMOFi7aQN8

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

I've wondered about this myself. Great tutorial.

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

MOSFETS wont last long in this application--been there, Done that! The Originals Are IGBT's--thats why there are parallel diodes across 'em, If they were MOSFETS they wouldnt need 'em....

You need to use IGBT's of at least 330V ratings and at least 200A pulse....

Look at the Panasonic DG302 or DG502 types for replacements, or 30F124 summit of that type....

--Lots of common ones to choose from, although it needs to be remembered that the peak current in the sustain circuits can be in the order of over 180A for a 42" and 300A for a 50" Plasma panel..... .

Mount direct on the heatsink with the gate resistor actually on the Gate pin of the device, 2.2 ohm carbon.

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

MOSFETS wont last long in this application--been there, Done that!

You need to use IGBT's of at least 330V ratings and at least 200A pulse....
--Lots of common ones to choose from.

Mount direct on the heatsink with the gate resistor actually on the Gate pin of the device, 2.2 ohm carbon.

What MOSFET did you use and in which module? It's true I wondered why they used diodes too.

I agree, once it worked I could measure the half-bridge output peak current in order to determine if the FET will support it.

330V sounds good to me eventhough I wasn't able to measure voltage transients at the IPM module, it was a clean 185V.

There are probably differences between each IPM model but at least one floating on the internet seems to indicate they used only FETs, the YPPD-J014C.

Anyway, I didn't have SMT IGBTs of that size, only throughole and I wasn't in the mood to mount them on the heatsink. :-)

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Still unsure if IGBT or MOSFET. I have seen designs with diodes in parallel with FETs , apparently because the bodydiodes characteristics weren't good enough. The only sure way to know if the FET will support it is measuring its drain current....

200A would be too much in the long run for this MOSFET

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

The set was a 42PC1DA with the YPPD-J018C module.

--In my case (as seems common on these) it was the X-SUS module, the ERC circuit IGBT's had become intermittent.
--Normally when they blow big time short they destroy the driver transistor pair and the driver chip too, the intermittent ones which sorta tick dont kill the drivers....

I used one of the FGPF series--cant recall exactly which, but the ratings were 330V 200A pulse.
--The set lasted me 6 months--Then I sold it on. I haven't heard back from it so assume its still going....

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Just checked my notes--It was a FGPF4536, 220A/p 360V IGBT by Fairchild semi....

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Just checked my notes--It was a FGPF4536, 220A/p 360V IGBT by Fairchild semi....

Ok. Thanks for that info.

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Ok. Thanks for that info.

I see the one you chose are the cheapest available. Also available in a DPAK package.

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

I measured about 70A peak in my case on the output of that halfbridge.

I'm using other more powerful MOSFETs now because I had removed the diode and didn't want to buy a new one.
But I guess you are right these were actually IGBTs. Thanks again.
https://www.youtube.com/watch?v=CARdm0o0FyM

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

You guys are smarter than me for sure and you would make a whole lot of people happy including myself by expanding this tutorial to include the J015x, J016x and J018x.

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

J015/16C is from the 50 inch sets....

Same basic applies--but each of those IPM's is more simple than the J018C we experimented with as they only have half of the bridge, ie--only two IGBT device--although in practice I think they used two sets of parallel pairs Plus the diodes...

IF the originals have actually blown short--as indicated by black silicone etc, then the driver devices will need changing too.....

Should therefore--apart from choice of part and the drivers--be a simpler process to repair....
--Leave the diodes and remove just the IGBT's, wire up some suitable ones in their place, attached by as short leads as possible and mount direct to heatsink.

Only problem I had was the position of the IGBT fouled some item on the board, so had to extend the pins of the IPM a few mil to clear....

With all these IPM LG sets--definitely worth adding a cooling-fan to those IPM's modified or original--They get bloody hot!

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

My confusion is, say, a little more basic. Here is the ER ipm from the ysus of a 50" LG. It is the J016E.

As you note it is simpler in design. This was ticking so I nothing blew perhaps the igbt went open. But I see where these diodes and transistors are labelled but the components are not visible because they are in some non-standard form factor bebeath the board ?? More confusing is that Q2 and Q4 are a single pad with jumper wires so I can't imagine the actual connections. Am I making sense ?

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Oops forgot the ipm pic

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

My confusion is, say, a little more basic. Here is the ER ipm from the ysus of a 50" LG. It is the J016E.

As you note it is simpler in design. This was ticking so I nothing blew perhaps the igbt went open. But I see where these diodes and transistors are labelled but the components are not visible because they are in some non-standard form factor bebeath the board ?? More confusing is that Q2 and Q4 are a single pad with jumper wires so I can't imagine the actual connections. Am I making sense ?

The bright squares are the actual components, that would be similar to what you would see inside a transistor or diode. The jumper wires are the terminals bonded to the die.

If one of these was blown you would remove the wires, then the die is already isolated. Then just place the IGBT / diode terminals were the jumper wires were bonded to (the golden-like traces)

[paulstef]

Re: Hybrid IC IPM YPPD-J017C J018C repair

This module looks good though. Why do you think it's blown?

You can check the diodes directly with a multimeter. to see if the IGBTs work you would need to drive them and put a power supply on the gate driver and a little bus voltage and load.

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

On the dead IPM's Ive had--That were intermittent, --(Greyish pink dim flickering picture, with patchy 'solarised' colour)

The actual die devices Looked physically Perfect in every way.

I tested them by using various meters and component-testers stabbed through the silicone onto the gold-plate tracks and found they were O/C.

I chipped away the die and came to a hard crusty black layer under the die-chip, that was very hard to remove, Under that was a bright soft solder to the base copper track....

It appears the die soldering fails and carbonises--probably due to the removal of the Lead in the solder. RoHS Rubbish strikes again--Damnable Tree-hugging Greenies--Idiots the lot of 'em......

As Paulsteph says--To Properly test an IPM, You'll need to rig up a 15V (12V works too) supply to the driver-chips, But you'll need a 5V P-P square-wave at around 10-20KHz to drive one of the IGBT's--There's a capacitive/resistive network on the Gate of one of the IGBT's in this and in the J018C.....

I used a car headlamp wired in parallel as a load so pass around 10A (12V) through the IGBT's to test....

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

These fail open too as far as I know. Not sure which component exactly but have had this happen in couple of sets and replacing the ipm fixed it.

In the above pic are the emitter and the gate jumpers not touching the same spot ? And where is the collector pointing to ? Lol sorry for being obtuse.

[Alastair E]

Re: Hybrid IC IPM YPPD-J017C J018C repair

These fail open too as far as I know. Not sure which component exactly but have had this happen in couple of sets and replacing the ipm fixed it.

In the above pic are the emitter and the gate jumpers not touching the same spot ? And where is the collector pointing to ? Lol sorry for being obtuse.

Gotta think in 3D!

--The Collector is actually the underside of the die device--attached to the solder/track the device is fixed to.

The Gate appears to connect--just like the Emitter, to the top of the die--But Very close examination by powerful Loupe will show that the surface of the die is divided up into specific regions that form the device terminals....

[mmartell]

Re: Hybrid IC IPM YPPD-J017C J018C repair

Aha! Thanks for the explanation!