I blew up the inverter????

[bestmapman]

Who is the dealer. I am sure a lot of us would like to stay away from that one.

[hobbit]

Wow, this thread went totally out of control before I found it
and caught up... While I sort of like the idea of a reverse
protection diode, even a .4 volt drop might be enough to fool
a charger into doing the wrong thing or terminating early or just
not pushing enough charge into a dead prius battery from another
source. Example: I used my car to charge a friend's car [not
jump, i.e. crank his starter, we knew better]. Even with mine
in READY at 14V [no, not 14.8] I could only get a couple of amps
to flow from mine to his. I had to leave things that way for a
couple of hours to finally push enough charge into his battery to
get his car started, and that was direct without any intervening
diode. To really *charge* another 12V you really need a source
north of 15 volts to get any appreciable current across.
.
Now, take a look at
http://techno-fandom.org/~hobbit/cars/ginv/615dcdc.jpg
and follow along. Near the upper right, which corresponds to
the left rear corner of the inverter when it's mounted in the
car, you see black and orange wires coming up through a rubber
gland in the case. Just below and to the left of that are two
things with curved metal straps bolted down across them. The
rightmost one of those is the output filter for the 12V, and
the left one is the DC/DC transformer that gets driven by some
of the power transistors along the left edge of the circuit
board. In between the two is the diode module. It's a typical
center-tapped transformer/rectifier circuit. Below the inductor/
diode assemblies, on the circuit board, are some small orange
capacitors that are the final noise filtering for the 12V output.
The DC/DC output comes out the large grey connector to the
right of that final section, that sticks farthest to the right
outside the inverter case right there. You can find this with
a little groping around behind the inverter [taking the cowl
off obviously makes it much easier but it's visible/reachable
just under the big orange hose feeding MG2]; it goes to a thick
wire which heads back toward the big harness at the firewall.
.
If you were to backpower that circuit, i.e. reverse jump with
all that connected, you would send high currents through the
final filter, diodes [forward biased, oops!], and transformer.
They are already high current parts and might be able to sustain
that peak load until something else blew, you saw a fat spark and
yanked it, or whatever. But it might not. The most likely result
is the diodes would short and then the whole thing would present
a permanent, fairly direct short to ground once 12 volts were
externally applied even in the *correct* polarity.
.
Now, note that the picture also shows an area of small, low-
power components that visibly blew up. I'm not sure if that
was from a reverse jump -- galaxee/DH might know since it came
from them in the first place. Since most of the control
electronics and low-power stuff in the inverter are powered
*off* when the car's off, it doesn't really seem likely since
there's a conduction barrier across the DC/DC main transformer
that would prevent backpowering any of the rest of the board.
But who knows, there may be another path and it wouldn't take
much wrong-way current to fry the small parts.
.
One test to do is pull that grey connector behind the inverter
and isolate the DC/DC from the rest of the 12V system. Give
the car a known good 12V source, and one might be able to
temporarily jumper the 100A main fusible link [which it indeed
is, and that's the correct spelling] with a thin piece of wire
so that will also act as a fuse if there's still a problem,
and see if anything will power on. Then, use a test light or
other protected source to apply +12 to the inverter's DC/DC
output -- maybe even an ohmmeter would work, to at least
determine if the DC/DC output is shorted or not. It *is*
possible to break the system down into individual sections
and isolate the problem. Now, most of the circuit diagrams
from the EWD don't actually show where the jump terminal is
connected and frankly I'm not exactly sure where it sits and
I'd go out and look were it not for the dark and mosquitoes
right now. I *think* it terminates the wire that comes up
from the 12V in the back, and the DC/DC is downstream of the
main link from there.
.
Generally, the 30A encapsulated, soldered-in fuse inside the
inverter [visible near the center of the board] wouldn't blow
in this case because that protects the 200V coming in from the
hybrid battery, not the 12V.
.
All of these sub-modules inside the inverter can be replaced.
Not from Toyota, because they want to sell you a whole new
inverter and their Prius tech training doesn't really go into
what's inside it anyways, but from an independent like Steve
or Lusciousgarage or Art's who actually knows what they're
doing when romping around in this area. I can't help locate
one in your immediate area, but if you can manage to find a
shop that is known to be good with the higher-end electronics
and find a tech who works there who's really jonesing to dig
into hybrid technology, please give them this opportunity.
.
I hope it works out okay, and I hope the final and very
specific diagnosis/fix will be posted. Possibly in a new
thread since nobody's likely to wade all the way through this
one to find the answer when they go looking two years hence.
.
_H*

[Patrick Wong]

Originally Posted by hobbit

Now, take a look at
http://techno-fandom.org/~hobbit/cars/ginv/615dcdc.jpg
and follow along.

Hi Hobbit,

Thanks for joining the discussion, since you've previously dissected the inverter.

• What disassembly was required to take the photo?
• Isn't another layer of the inverter installed above this layer?
• If yes, how hard would it be to remove and reinstall the top layer?
• Are the diodes custom or are they off-the-shelf?

You've introduced another interesting issue with your point that other components besides the diodes appear to have failed. It looks like the most cost-effective solution is to purchase a salvage inverter.

[galaxee]

hobbit, come to think of it i'm pretty sure that one was blown in a reverse jump. we didn't even think about the stuff that looked that screwed. there are very few ways to get your hands on a non-warranty bad inverter, and i can only think of one. toyota takes back all the warranty parts.

[patsparks]

You know if I was a person who couldn't fix the inverter, which is how I describe myself to strangers, I'd go the route I went with the MFD and hunt out a electronics tech to have a go at fixing it. I'd gather as much information as I could find and invite an electronics tech over to have a look then extract the guts of the inverter to try to repair it. Either that or seek out a change-over unit if they exist. To me it's hard to believe that a fuse in the inverter which has blown can't at least be bypassed with another fuse either in the casing or external. Isn't the idea of a fuse to protect the other components? No point in having a fuse if that fuse failing writes off the whole inverter.

It's my understanding (perhaps mistaken) that most of the component cost is in making them compact so if a single component has failed inside the case is there a reason it cant be bypassed and a similar but perhaps larger component mounted externally?

Sorry if I'm speaking sh** I just thought maybe I was thinking outside the box and would say what I was thinking. Feel free to say I'm wrong.

[Seamus]

Originally Posted by galaxee

hobbit, come to think of it i'm pretty sure that one was blown in a reverse jump. we didn't even think about the stuff that looked that screwed. there are very few ways to get your hands on a non-warranty bad inverter, and i can only think of one. toyota takes back all the warranty parts.

I think if I was Ginny I would want all replaced parts returned to me.

[hobbit]

For the rest of the inverter teardown pages, knock the picture
filename off the above URL. It shows the whole sequence of
opening it up. Remember that we're talking about the UNDERSIDE
of the unit here.
.
The diodes say on them: TS906 C2 53 with no particular company
logo on them. Nothing really substantial Googles up for that
although there's a hint that it might be an equivalent custom part
number for a 35 amp automotive diode. That vague reference comes
from one of the offshore data-sheet pirates, so don't take that
as anything like certain. They're ganged up 3 and 3 on each
transformer leg for a 100 amp capable output and quite well
heat-sunk, so that might make sense.
.
The clearly smoked parts in the picture are near the control
input connector for the DC/DC converter, so it's possible the
reverse jump did manage to back-bias parts right near there.
I haven't looked to see if there's an always-on power connection
between the 12V output and that area of the circuitry.
.
For those that missed the relevant exchange a couple of years ago,
I should once again publicly thank galaxee and her DH for making
this unit available to me and the Prius community. I'm still not
done studying this thing, in fact! It's been a great demo for
both indy techs and the public alike, as well as a good teaching
tool for how Toyota implemented their motor control.
.
_H*

[Patrick Wong]

Originally Posted by hobbit

...Generally, the 30A encapsulated, soldered-in fuse inside the inverter [visible near the center of the board] wouldn't blow in this case because that protects the 200V coming in from the hybrid battery, not the 12V...

Hi Hobbit,

I had forgotten to ask you about the above excerpt from your prior post. Is the 30A fuse just for the high voltage side of the DC/DC converter? If so, since the DC/DC output is fused at 100A (at 14V, hence 1,400W output), I am surprised that the input fuse would not be much smaller, say 10A at 200V (for 2,000W input).

[Scruge]

Originally Posted by Patrick Wong

The dealer techs do not have the facilities or training to replace individual inverter components in the field. Hence the fact that the cost of two 100A diodes might be $10 or $20 is irrelevant unless you can find someone to successfully take apart, repair, and reassemble the inverter.

And we know Toyota certainly doesn't want to spend a couple extra bucks to prevent a $4800+ repair..

I assume that most of the ECUs have series diodes to prevent the flow of incorrect polarity current. A couple of the ECUs may have crowbar diodes to force popping of the associated fuses in that event. This may be intentional, so that Toyota has a way to prove owner misuse.

Unfortunately, other causes can be the result of dealer, road side assistance, repair dept. etc. If they are going to set a trap then it better be able to ID the varmint.

In my opinion its a crap design.

Its like not putting a fuel filter on the fuel system.. and expecting the customer to buy only gourmet bottled fuel.

[andyprius]

Originally Posted by Patrick Wong

Hi Doug,

I would say that the 40A current rating is insufficient, given the DC/DC fuse link is 100A and the fuse in the fuseblock over the positive battery terminal is 120A.

All great posts, to get these posts to Toyota Headquarters may get them moving. PATRICK: If the jumper cables are reversed, then current flow direction is reversed, right? so the current flows from ground thru the Electronics, thru the fusible link to the positive post and back thru the donor car. With NORMAL current direction flow shouldn't the fusible link protect the inverter and any other electronics downstream. Perhaps the fusible link is way overrated. I believe that tepid battery is rated at about 32 Amps, so why is a 100 or 120 Amp protecting an inverter that would blow at a much smaller current draw???? The total sum of this and the thread started by Ginny is that the dealership was presumptive, arrogant, rude and just plain ignorant in the treatment of this customer. The whole situation is totally without reason and Toyota should be made aware. I think you and ICARUS are correct, it's highly possible that the inverter is still good and only the fuses and battery need to be replaced. So in Ginny's case it's worthwhile to have a new battery installed and fuses checked. Since her dealer has not actually done the work...................THEY ARE GUESSING.