EagleTree erratic RPM readings - RC-Monster Forums

kazuaki · **EagleTree erratic RPM readings -** 09.13.2009, 06:05 PM

I just received my eLogger V3 with the brushless RPM sensor, I have no previous experience with EagleTree units. I threw it to my truck to just see how it works. I used both sensor wires from the brushless RPM sensor and stuck them in the motor bullet connectors. I setup the unit in the software for 2 poles (stock VXL motor). It seems to work for the most part, but I am getting some odd RPM spikes, that are obviously incorrect. Anybody have a clue why this is happening?

Thanks

whitrzac · 09.15.2009, 01:01 PM

Quote:

Originally Posted by kazuaki

I just received my eLogger V3 with the brushless RPM sensor, I have no previous experience with EagleTree units. I threw it to my truck to just see how it works. I used both sensor wires from the brushless RPM sensor and stuck them in the motor bullet connectors. I setup the unit in the software for 2 poles (stock VXL motor). It seems to work for the most part, but I am getting some odd RPM spikes, that are obviously incorrect. Anybody have a clue why this is happening?

Thanks

wheel spin??

BrianG · 09.15.2009, 02:33 PM

Seems possible. Looks like it only happens as you accelerate from a slow speed, so that makes sense. What vehicle is this in?

It is kinda weird how it is almost a perfect square wave though...

kazuaki · 09.15.2009, 02:40 PM

I just showed that part of the graph. It did it while I wasn't accelerating too. This is the 1/16 e-revo with the stock VXL esc/motor. This was just a test before I get to running the serious power. I'm 99.9% sure this is not wheelspin.

BrianG · 09.15.2009, 02:47 PM

I guess I'm too lazy to go look, but how is rpm being measured? Via magnet/coil or does it attach to the motor wires?

kazuaki · 09.15.2009, 03:37 PM

Sorry for not being more detailed. The EagleTree brushless RPM sensor hooks up to two of the motor wires. Then you configure the logger to know how many poles the motor has.

BrianG · 09.15.2009, 04:03 PM

Hmm, so no magnetic interference to worry about. Unless the sensor wires are running parallel and too close and is picking up induced signals from the other wire at the same time...

fastbaja5b · 09.15.2009, 09:25 PM

I get the same on my Eagletree, some crazy rpm spikes up to 60k rpm, which is far fetched when it's a 3000kv motor on 2s lipo.

No idea why or how, but it happens during a run or on the bench and seems to happen more when you're in reverse if that makes sense.

J57ltr · 09.15.2009, 10:06 PM

I have had a lot of problems with the brushless rpm sensor. Before I set the upper limit it would spike over 100K. I have only had it give accurate readings once. Sometimes it reads 0 rpm and it's a straight line up to 30K or so Then it follows motor rpm. The wires for the sensor are only about 1.5" long and go to a board with a few componets on it (haven't looked under the shrink and paper to see, go figure), out of the board is a standard servo cable. I have done everything except wrap brass mesh around all the servo, motor leads and battery leads, and ground it to the - terminal of the battery.

I even soldered a pair of wires to two of the motor leads and put a plug on it and a mate on the leads of the rpm sensor. I still get the same kind of readings. I have seen others with pretty good graphs so maybe I am overlooking something, or something is wrong with my sensor. I think they may have a forum at eagle tree. It seems to come up a lot, I have heard it from at least 2 other people around the LHS. I also bought just about every sensor for it (made my own temp sensors, after I bought the first one), So I still have the hall effect one. I plan on mounting it in the diff. so I'll let you know how it comes out.

Jeff

BrianG · 09.15.2009, 10:09 PM

I would try to isolate the wires as much as possible away from the motor wires, especially not running parallel to them. They could be picking up EMF from all three phases at once. Or maybe it is getting "confused" between each phase firing and the PWM pulses that make up each pulse...

kazuaki · 09.15.2009, 10:33 PM

I sent an email to EagleTree support a few days ago. We'll see how that goes

I'll try to isolate the wires a bit too.

J57ltr · 09.15.2009, 10:39 PM

Done it I have tried everything.

I coiled all the wires on the servo end and kept them at right angles to any wire.

I braided the motor wires.

I coiled all the servo wiring for the esc and steering servo.

I have added a new switch and ran 24 ga wire and did the old drill twist for the wires.

I also added the 2 wires for the quick connect for the brushless sensor soldered to the terminals directly at the esc, which is also twisted pair.

I changed the direction the wires how they come out of the controller (Bottom now) and did all the other changes at the same time. Mine has done this even when the MM was factory fresh, and has not changed at all the operation of the sensor. I really haven't fooled with it much lately, I'm already logging 3 temps, and I just got a fresh diff so I want to try that one. I also have not been able to show mph with the motor sensor, but I probably have a setting wrong. It asks all kind of questions trans ratio, spur/pinion, # of poles and a few others I can't think of right now. But no matter what settings I have or what I have set to log mph doesn't show.

Jeff

pinkpanda3310 · 09.15.2009, 10:54 PM

Quote:

Originally Posted by BrianG

I would try to isolate the wires as much as possible away from the motor wires, especially not running parallel to them. They could be picking up EMF from all three phases at once. Or maybe it is getting "confused" between each phase firing and the PWM pulses that make up each pulse...

This seems to make the most sense. At lower power, like in reverse, the power is only sent to the motor briefly and then the motor spins a little from momentum and hence acting like a generator. There is obviously something resonating with the sensor.

BrianG · 09.15.2009, 10:55 PM

This might be a bit complex and take a little math to figure out, but here goes. Add a diode to one lead, and another diode to the other lead, but reversed. Then, add a capacitor between the two. The diodes keep the cap from messing with the ESC EMF feedback signal, and the cap helps filter the PWM pulses chopping up each phases. If done right, the ET will see cleaner signal indicating when each phases is firing, but not the pulses. You would have to know the PWM frequency to size the cap so it filter the small time between PWM pulses, but not the whole commutation signal. If that makes any sense...

J57ltr · 09.15.2009, 11:13 PM

I think I heard mention of 13.5K as the switching frequency for the MMM I'll have to get out the O-scope (have to wait till I get back on friday). The board actually has 7 caps, 2 diodes, and at least 1 IC. It's half the size of the SOT23 diodes.

The sticker on the back doesn't come off easily but when I get back I'll see about making a schem for it.

Jeff