[Apologies if I screw up the quotes...]
wilsonintexas wrote:sorry bad at quotes.... It throws the error at power up. I assume it does a self check....
OK. When I drive a BLDC motor, the first thing I do (in the Power On Self Test) is to verify the motor is attached, the sense leads connected and it "makes sense" in its current state. Then, (depending on the nature of the mechanism to which it is attached), I rotate the motor a bit in each direction -- just to cycle through the phases and verify it seems to respond, as expected. In some applications, a motor cable might be left unplugged, or, plugged in "backwards" (so, if I drive the motor in a clockwise direction, I expect to see the feedback signals confirm this direction of rotation).
I suspect the Segway is designed with the "no user serviceable parts inside" mentality so they are expecting only "authorized dealers" to be servicing it. OTOH, they may be paranoid enough to plan for the hobbyist tinkerer (easier to perform an extra test in software than to risk a lawsuit if a tinkerer climbs on and gets thrown because the motor wiring was backwards -- because he screwed with it!).
One thing, for sure, is you can look at the three sensor signals and erify that they represent one of the six possible states: {101, 100, 110, 010, 001, 011}. In particular, 000 and 111 are "not possible". So, if the controller sees one of these two conditions, it knows that something is wrong. (1 being ~5V and 0 being ~0V)
In the 000 case, it could be that R81 has "opened" and isn't sourcing power to the three "pull up" resistors (R49, R50, R51). I can't tell if R81 is supposed to be a "0 ohm jumper" (
i.e., a piece of wire) or a relatively low resistance to limit the power going off board from the onboard power supply. Assuming it is a resistor and wasn't designed to tolerate a prolonged short on its load (pin 2 of the connector), it could have simply opened from excessive power dissipation. This could occur if you had pin 2 grounded for some reason.
Even if R81 is intact, if pin 2 is presently grounded, then you'll also see 000.
If pin 2 (or the wire connecting it to the motor/sensor assembly) is
open, you'll see 111. You'll also see this if pin 1 is open
or if L4 truly is a wire inductor and is open.
You'll also see 111 if one or more of the hall effect sensors has been fried (
i.e., because you shorted a power line -- even the 5V -- directly to any of pins 3, 4 or 5. Or, if the cable is unplugged!!
[Note, I'm not being pedantic in my logic, just trying to imagine the sorts of things that MIGHT have happened while you were tinkering]
[I'm trying to think of how I could get you to hack together a "test probe" to examine the outputs -- on the left side of the R43's -- to see if they are stuck high, low or pulsing with motor motion. What is the specific error? Or, is it so generic that it just says "Motor 1 defective"?]
TED: It would be tough to get test leads there.... The error is a series of flashing lights (3 then 9 or 3 then 6) I fond one site that groups the range as bad motor sensors. One person on the other forum mentioned which sensor it was. BUt I suspect that et stops at the first error. sometinmes when I have wires pinched it throws an eror about a foot pad sensor.
If you know what a 'scope is, then you probably know what a VOM/DMM is, as well? If you have any reasonable DMM on hand (even one of the free harbor freight variety), you can use this on the DC Volts (20) scale and probe the voltages present on those signals.
With the cable plugged in (the P16 in the schematic), you should be able to access the signals at the motor or the circuit board. Ideally, you want to look at the signals while the cable is plugged in as that's what the controller is seeing. If you unplug the cable, the controller should throw an error because the 111 state will appear on those inputs. What you want to know is the signal levels when plugged in but not "in motion" (
i.e., they should be static/unchanging so you don't need the "speed" that a 'scope provides to your observations).
AT one point I had resoldered the leads bad put the motors back in their original configuration and everything was fine.
Meaning you had removed your "extension cords" and all was as it should be?
TED: Correct (well almost the original I had a short section of wire I let in place because the leads ere getting short.
But after the last failure, I out it back and I am still getting the error code.
So, it is now failing while in the
original (no extensions) configuration?
TED. correct
This suggests you've got a broken wire or flakey connector (I can't see the parts so can't tell you how likely possibility each would be). Note that if you've plugged and unplugged the cable(s), you may have mangled/bent a pin/contact in a connector and may not see it -- unless you have evidence that it is mangled (because you see bogus values on the signal leads!) Or, may have fatigued a wire at the point it enters a connector shell and can't see the break because its inside the insulation.
It would have been nice to move the motors.
Why couldn't you ALSO move the segway's controller (and get rid of the segway, altogether?)
Ted: I was trying to so this with as little investment as I could. I had the Segway and I do not have another controller to put in there.
Yes, but my point was that the Segway has a controller inside it (somewhere). The motors ON THE SEGWAY were wired to that controller. You tried to lengthen the wires so that you could "remote" the motors (and sensors). Could you, instead, have removed the controller from the Segway and sited it on your chair, as well? Or, is it not easily accessible IN/ON the Segway?
[And, remind me again as to the reason for your choice of the Segway... is it because it has a higher top speed than most powerchairs (for that
Horse that you claim is a dog
) Or, is it just "what you happened to have on hand"?]
No, the issue I was fearing was that noise glitches were getting on the signal wires (they act as antennae) and those glitches were being interpreted as valid signals by the controller. The top drawing ('scope trace) shows what the waveforms from the three Hx signals should look like while the motor is in motion (moving continuously in ONE direction):
http://www.patentsencyclopedia.com/img/20080272765_03.pngTed: I agree with this...
What if I ran the 3 signal leads in coax, and grounded the brading.... that may shield them.
My initial suspicion was rooted in not having any information about the circuit and signals in question. But, the component values in the schematic you've presented suggest that "noise" won't be a problem.
I would love to get them, but do not have a scope with me right now. Maybe in a few months when I get back to Dallas, and have a change to pick one up from my dad.
But I am not sure what shape they are in. They have been in a back room for years, not sure if they have the leads, or what features they have
A DMM would be more than adequate. You just want it to be relatively high impedance --
not "a light bulb and a piece of wire" (
i.e., if the light illuminates, you have voltage present -- this is how you troubleshoot old pinball games without the fuss of a fancy meter!
)
Ideally, if you could get a probe on each signal (one at a time) and note its level with the cable connected AND with the cable removed.
Removed, you should see 5V on each of the 3 signals (pins 3 4 and 5).
Connected, you should see 5V on no more than two signals (but AT LEAST one!) and 0v on no more than two signals (but, also, at least one!) I'm guessing this isn't the case and the controller is just throwing up its virtual hands in despair...
[You're in Calif presently? I'm sure if I look upthread I could answer this for myself based on your earlier comments re: daughter (?)]