When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
GS450H with a blown inverter. Got a used one and still won't start.
Hybrid TechnologyUnique topics related to the GS450H model hybrid drivetrain and other features/options found only on the GS450H. Please use the main 3GS forum for discussion about shared components with other third generation GS models.
I opened the bottom part of the inverter and it looks pristine. Is there anything else to look for?
this looks ok but to be sure .remove the bard u can see on top in this pic and check the chips under for pieces blowen out of them (if they look ok then its a good assumption that the igbts are ok ) . But to be sure Then look under the next board(it cant be removed unles u take the whole board and igbts out ) to remove them look at my how to on the change thermal paste thread
Well I took out that top board and the igbts. That top board has a few burn marks but that may have been because I used a heat gun to help pry the base off. I assume everything else looks fine?
There's usually a ton of... ehh... magic smoke when an IGBT fails---I mean, blows up. Mine was especially spectacular. Yours looks pristine.
I'd check the harness - you know where to find the diagrams. Replacing the hybrid ECU is bound to be a major pain, and I don't think I've ever heard of one going bad.
If you know somebody with such a car, ask if you can try plugging the inverter in theirs - you just need to plug in the inverter 40pin connector - bonus points for fooling the battery disconnect plug with a paperclip. If there are no DTCs present - the inverter should be good. You don't need the hybrid system to be online for the "lost comms with inverter" codes to be set.
Thanks Xenith and Lwerewolf. I am quite busy so don't really get much time to mess around with the car. I might just take it to a hybrid technician so they can find the problem.
Time to look at this sort of problem is the key .
i find that most so called hybrid tecs dont really have that good an understanding of the finer points of or systems . And usally just come up with i"ts the inverter mate "
even the lexus tecs . I have come across 3x gs450h that have blowen an inverter and wont goagain when replaced . 1 of which i had no time to test . And the other two i only had a breff look afyer fitying a new inverter to and then they wetr sold as is . All seam come up with this lost coms . I have a feeling that its the hybrid ecu but have not confirmed this at all . When i get time ill pull mine and fit it to my mates one . But as u say its the gettung the time
All you need to start with is finding the wiring diagrams (they list pin numbers and connector layout), disconnecting the inverter and the hybrid ECU harness and probing the related serial lines - anything that's HTM, MTH, CLK, REQ, HSDN (not really necessary for this).
DO NOT break the connectors, as they are pretty much unobtainium at this point - as in, you cannot order them... at least the 40pin inverter one. Same with the fusible link. Anyways - just take it easy on them.
Relevant diagram cutout:
Connector pinouts:
Check all the listed connections between the HVECU and the MG ECU (basically the inverter) and report back
Keep us posted if you decide to pursue this.
Last edited by Lwerewolf; 04-30-21 at 08:51 AM.
Reason: pasting windows "screen snip" directly into the post leads to weird results
Thank you so much for your help Lwerewolf. I used a multimeter and followed the repair manual (except the steps that require an oscilloscope) and came to the conclusion that the Hybrid Vehicle Control ECU was bad. The wiring harness checked out fine. So I bought a used ECU from the wreckers, installed it, did the ECU COMMUNICATION ID REGISTRATION (jump pins 4 and 13 on the OBD2 port for 30min with ignition on) and it started right up. I used Techstream to also write the car's VIN to the new ECU. All the codes are now gone.
Now the plan is to replace the thermal paste in the inverter (easy enough following Xenith's guide), do a coolant flush and hopefully the car will run reliably for at least another 10 years. Thanks again for the help guys, really appreciated.
I highly advise that you DO NOT floor it if the car is sub-100kph and mg2 is in 2nd gear - this isn't indicated anywhere, upshift happens at 95kph (non-US/JDM-spec diff), downshift should be felt at 55-ish kph (unless the car is in S-mode and engine braking). Alternatively - monitor the inverter MG2 IGBT temps under those conditions - if they don't spike above, say, 80 celsius... should be fine.
On another note, this is the first hybrid ECU that I've read about that has been judged as "failed".
If you do decide to change the thermal paste - get a good torque wrench and follow the rx400h instructions. I would NOT use thermal grizzly whatever - you're looking for long-term stability, not for a 5% drop in temperatures. I used Electrolube HTSP. I'd look for something that is as close in terms of properties to the original thermal grease (noted in the rx400h IPM TSB) as possible.
The OEM grease (at least for the rx400h) is Shin-Etsu G-747.
From what I've read, the failure mechanism is deterioration of the solder layer, not of the thermal grease. Double-side cooled IGBTs as first used in the ls600h (and subsequently, rx450h, 4th gen gs450h, 4th gen prius, etc) appear to mitigate this problem completely. Same thing (as in - solved) with Semikron SKIN and other sintered but otherwise traditional IGBT packages.
Re: long term reliability - I'd look into rebuilding the brake pump, more specifically the DC motor commutator, brushes and bearings.
I highly advise that you DO NOT floor it if the car is sub-100kph and mg2 is in 2nd gear - this isn't indicated anywhere, upshift happens at 95kph (non-US/JDM-spec diff), downshift should be felt at 55-ish kph (unless the car is in S-mode and engine braking). Alternatively - monitor the inverter MG2 IGBT temps under those conditions - if they don't spike above, say, 80 celsius... should be fine.
On another note, this is the first hybrid ECU that I've read about that has been judged as "failed".
If you do decide to change the thermal paste - get a good torque wrench and follow the rx400h instructions. I would NOT use thermal grizzly whatever - you're looking for long-term stability, not for a 5% drop in temperatures. I used Electrolube HTSP. I'd look for something that is as close in terms of properties to the original thermal grease (noted in the rx400h IPM TSB) as possible.
The OEM grease (at least for the rx400h) is Shin-Etsu G-747.
From what I've read, the failure mechanism is deterioration of the solder layer, not of the thermal grease. Double-side cooled IGBTs as first used in the ls600h (and subsequently, rx450h, 4th gen gs450h, 4th gen prius, etc) appear to mitigate this problem completely. Same thing (as in - solved) with Semikron SKIN and other sintered but otherwise traditional IGBT packages.
Re: long term reliability - I'd look into rebuilding the brake pump, more specifically the DC motor commutator, brushes and bearings.
Thanks. I've ordered some Arctic mx-4 which should last at least 8 years. Also, is it okay to floor it from 0-100 kph?
I highly advise that you DO NOT floor it if the car is sub-100kph and mg2 is in 2nd gear - this isn't indicated anywhere, upshift happens at 95kph (non-US/JDM-spec diff), downshift should be felt at 55-ish kph (unless the car is in S-mode and engine braking). Alternatively - monitor the inverter MG2 IGBT temps under those conditions - if they don't spike above, say, 80 celsius... should be fine.
On another note, this is the first hybrid ECU that I've read about that has been judged as "failed".
If you do decide to change the thermal paste - get a good torque wrench and follow the rx400h instructions. I would NOT use thermal grizzly whatever - you're looking for long-term stability, not for a 5% drop in temperatures. I used Electrolube HTSP. I'd look for something that is as close in terms of properties to the original thermal grease (noted in the rx400h IPM TSB) as possible.
The OEM grease (at least for the rx400h) is Shin-Etsu G-747.
From what I've read, the failure mechanism is deterioration of the solder layer, not of the thermal grease. Double-side cooled IGBTs as first used in the ls600h (and subsequently, rx450h, 4th gen gs450h, 4th gen prius, etc) appear to mitigate this problem completely. Same thing (as in - solved) with Semikron SKIN and other sintered but otherwise traditional IGBT packages.
Re: long term reliability - I'd look into rebuilding the brake pump, more specifically the DC motor commutator, brushes and bearings.
well done as for the first blowen hybrid ecu i have not seen someone diagnosed eather but seen afew 450hs not go after the inverter was replaced and had a hunch it was the hybrid ecu but never had the chance to test it . Like Lwerewolf said the thermal paste longevity is the most important not the slight better cooling . I monitor my temps all the time when driving with hybrid assist and a tablet and run my a/c all the time with my rear vents facing yhe parcel shelf and the middle front to facing stright back and along the roof to help the battery cooling.
well done as for the first blowen hybrid ecu i have not seen someone diagnosed eather but seen afew 450hs not go after the inverter was replaced and had a hunch it was the hybrid ecu but never had the chance to test it . Like Lwerewolf said the thermal paste longevity is the most important not the slight better cooling . I monitor my temps all the time when driving with hybrid assist and a tablet and run my a/c all the time with my rear vents facing yhe parcel shelf and the middle front to facing stright back and along the roof to help the battery cooling.
oh and were is the hybrid ecu located by the way
HV Control ECU is in the engine bay at the front next to the fuse box. Not too hard to replace.
Monitor your temps under various conditions. Usually 0-100 is not a problem unless you're stuck in 2nd gear for some reason (maintenance mode comes to mind) - the car also gets out of the.. ehh... "high temperature" speed zone of MG2 really fast in 1st gear, and the upshift also keeps mg2 in a happy spot. I'm more worried about any sort of passing when you're in 2nd and you're close to the shift-back point - only way to shift down that I'm aware of is dropping speed to sub-55. Funnily enough, my sustained inverter IGBT temperatures during a 0-140+ kph run are lower than during cruise. Given that you guys can use right hand drive parts, I'd imagine that a relatively fresh Crown GWS204 inverter would work nicely and might not be as degraded as the... ehhh... available gs450h ones.
mx4 is far from dedicated long-term-stable thermal greases like the g-747. Keep that in mind for potential future change intervals. Also, read up the rx400h TSB and also read up on how grease is put on IGBT baseplates - it's not a CPU, application does tend to matter. I used a film thickness gauge and a paint roller. Not super critical, but not fun.