P3120 and P3125 with subcodes

The resolver is a fascinating part. I've got the engineering specs (got a data package from the vendor, Tamagawa Trading Co., LTD, 1020 Kega, IIDA, Nagano-Pref, Japan, 395-8520.)They call their chip "Smartcoder AU6802n1" and I remember seeing it in the HV ECU.

It works on the principle of a pair of coils 90 degrees apart an a two-lobe, ferrous disk connected to the shaft. A stimulus signal, 8-20.4 kHz is applied to the two coils and they measure the variable reluctance to know exactly the position of the shaft at all times. It was use a variable reluctance that led me to some experiments in an alternative encoder for the NHW11 accelerator pedal.

Bob Wilson

 

Interesting! It looks as if the chip converts the result into an absolute position encoding, where for RPMs less than 60,000 (plenty for us) it can encode in 12-bit resolution, or about 5 minutes of arc. My practical question would be about how the resolver itself is mounted to the endplate in MG2: whether it bolts in a fixed position, or is slotted for some range of adjustment and if so, how critical the adjustment is to the functioning of the car? How close is close enough?

-Chap

 

I think good Prius Friend Hobbit had posted some photos and details about the time he was working with "Up The Voltage" the NY hybrid training group. I suspect everything is slotted on the shaft and registered by the case.

Bob Wilson

 

this resolver, does this have a certain way that it has to be set when reassembling the trans in order for the mg2 to function or no.

 

That's more or less what I was trying to figure out in this thread. It isn't stated in the shop manuals, but some things I have read seem to suggest that it has some range of adjustment; 3prongpaul mentioned recently that the bolts holding it in place should be left alone when removing the end plate. What I don't have is any information on how wide the adjustment range is. In other words, if those bolts ever get loosened, can the resolver be rotated over a 5-degree adjustment range? 10? 1? Just barely turned due to slop in the bolt holes?

My just-informed-enough-to-be-dangerous guess would be that the motor will still run as long as the resolver is pretty close to correctly aligned. Now, off by 5, 10, 15, 20 degrees I could imagine trouble, possibly even unhealthy currents through the inverter, etc. It's interesting to wonder if there's some number of degrees such that being misadjusted by more than that would make the motor go backwards. (I think the rotor poles are 45 degrees apart.)

Thanks to Bob finding the datasheet for the resolver chip, we know that the smallest change the resolver can even report is about a twelfth of a degree, so any misadjustment smaller than that is probably no big deal. Somewhere between error-too-small-to-detect and error-too-big-for-the-motor-to-work is probably a range where the motor will work, just not as efficiently as it could.

It's just a question of exactly how sensitive the adjustment is. So far, I've never seen a post here from anybody with more information than that.

-Chap

 

i hope its something that is either easy to do, or that it is easily avoidable (meaning that there is a specific set of bolts to loosen to play with that and as long as you don't do that your fine) if i ever have to replace my mg2