Hey sledbuilder,
I’ve been digging pretty deep into the old Bixby documentation, BixCheck, and the actual MaxFire firmware for the OpenMaxFire project, and I think there are a couple things here that may help.
First, you remembered the reversing feeder motor correctly.
The factory Bixby service manual says the original feeder gearmotor is designed to mechanically reverse if it jams. It’s a 120VAC gearmotor rated at 120 in-lbs at the output shaft.
From everything I can tell, the motherboard itself is not reversing the motor electrically. The reversing behavior is part of the original gearmotor/mechanism.
Because of that, I would be a little suspicious of the eBay replacement motor. It may spin the wheel and look fine, but if the speed, torque, or internal reversing mechanism is different from the original, it may not behave the way the stove expects.
As far as the current-monitor idea goes, I don’t think that is what is causing light 8 on this generation of MaxFire.
We have traced the feeder logic in firmware 2.02, 2.06, 2.70 and 2.71. The board commands the feeder motor and watches the feeder-wheel sensor at J9. The firmware actually times the wheel movement based on the sensor transitions.
That value is available through the computer port as CR07.
BixCheck’s automatic “Feed motor and sensor” test runs the feeder and checks that CR07 falls in an acceptable range. All three BixCheck versions we have analyzed use a range of 0x10 through 0x68.
So if your problem stove is actually passing that test, that is interesting. It means that at least during the BixCheck test, the stove is seeing believable feeder-wheel movement.
One thing I would definitely check next is the feeder wheel itself.
The factory service manual says there are FOUR magnets pressed into the feeder wheel. The prox sensor sees those magnets as the wheel turns.
Since you already swapped the prox sensor into your working stove and proved the sensor itself is good, I would inspect all four magnets on the problem stove very closely.
Look for one that is:
- missing
- loose
- pushed farther into the wheel than the others
- damaged
- or noticeably weaker than the others
It is possible for BixCheck to pass a quick feeder test but still have one position become unreliable during actual operation.
That would also fit pretty well with what you’re describing: startup begins normally, pellets feed, the stove lights, and then the controller decides the feeder wheel isn’t moving correctly and goes into shutdown with light 8.
I would probably check things in this order:
- All four feeder-wheel magnets.
- J9 connector and wiring, especially for an intermittent connection.
- Compare the original motor to the eBay replacement for speed, torque and the mechanical reverse feature.
- Make sure the feeder wheel turns freely when actually loaded with pellets and there isn’t a fines/alignment/cover-plate issue.
- Then start looking harder at the motherboard feeder output.
A board problem is still possible. For example, the feeder output could work during a BixCheck test and then become intermittent during a real startup. But if that is happening, I think it would be more of a motor-drive/output issue than the board rejecting the motor because of current draw.
There actually are later Bixby systems that report motor current, so the current-monitor theory isn’t totally out of left field. But on the MaxFire firmware we have preserved and disassembled, light 8 is tied very clearly to feeder-wheel movement feedback through J9.
Another interesting clue is in the old 2.06 firmware notes. One of the changes says:
“Feed wheel error (#8) blinks starting with the second move attempt.”
That matches what we’re seeing in the firmware pretty well. Light 8 is basically the controller saying, “I commanded the feeder to move and I did not see the wheel movement I expected.”
I also would not update the firmware just to try to fix this yet. If the stove is losing feeder-wheel feedback, 2.71 still needs that feedback too.
If you can post the firmware version in the problem stove, the BixCheck version you’re using, a picture of the original motor label from the working stove, and the eBay replacement motor, that would help a lot.
Also, since you already have BixCheck communication working, we can probably go one step further.
OpenMaxFire can watch the feeder command and the J9 feedback during a normal startup. We could log CR02, CR03 and CR07 and see what happens right before light 8 comes on. That should tell us whether the board is still commanding the motor, whether the prox signal is actually changing, and whether the feeder movement timer stops updating.
That would probably narrow this down a lot faster than swapping more parts. If you'd like to try to troubleshoot it, shoot me a PM or an email. (
contact@openmaxfire.com)