Invalid Hall Pattern -- but only in one of 4 configurations?!

[chuyskywalker]

So, I've got these 4 components:

• MXUS XF19FAT Motor (geared, 1200 watt)
• RadMini4 Stock Motor (geared, 750 watt)
• Frankenrunner
• Phaserunner v6

Now, both motors require an adapter to L10 for the phase/franken, so let me details a bit about those. For the most part, match the colors and we're good.

On the MXUS adapter, there are two exceptions.

1. The motor's white wire is a temp+hall combo signal. The adapter takes motor white -> L10 gray (which, according to the manual, the Runner with demux into pulse and temp readings). When connected to the Frankenrunner, it does just that perfectly -- I get reasonable temp readings and the speedo on the dashboard works when spinning the tire by hand.
2. The white wire on the L10 is connected to ground as I found some odd behaviors when connected to the Phaserunner (weird speed reporting; floating voltage?)

For the RadMini motor adapter, there is no temp sensor, so it's just 3 phase, 5 hall, 1 hall for speed.

I attached a bunch of photos of the motors, cables, and other such items at the end.

Here are the configurations that work:

• Frakenrunner -> RadMini
• Phaserunner -> RadMini (I've been riding the bike in this mode, works flawlessly)
• Frakenrunner -> MXUS

However, the Phaserunner -> MXUS is failing in some odd ways.

1. Most obvious, is that it fails autotune with an "Invalid Hall Pattern" error during autotune.
2. It shows a 10.38Kv compared to the Frankenrunner detecting a 9.35Kv (it runs noticeably faster during detection).
3. I noticed that speed is always reported as zero, as if the demuxing from the temp signal is not actually working.

My goal is to have the Phase+MXUS working together. If the hall detection on the MXUS failed with both controllers, I could blame the motor/adapter. If the Phaserunner failed hall detection with both motors, I could blame/rma that. However, neither of those is the case, so I'm a bit at a loss as to what's going on and why in this one specific combo things don't work.

Any thoughts, ideas here? I'm not sure what troubleshooting steps to take next.

 

[amberwolf]

So, I've got these 4 components:

• MXUS XF19FAT Motor (geared, 1200 watt)
• RadMini4 Stock Motor (geared, 750 watt)

In a quick search around the web, I couldn't find their kV / inductance / etc data that could be used to directly enter into the PR setup program (since the controllers each detect different values, it's likely that this has something to do with the issues you're having). If you can locate this info anywhere, you can put it in directly. Alternately, you can directly enter in the info that the working controller detects into the one that doesnt' work.


It could also be a connection issue...even though the connectors all work in some combinations, you could still have an issue with spread pins on the barrel sides of one or more pins in such a way that they work with one connector (because that connector's pins insert in a way / at an angle that creates a good connection despite the loose-fitting barrel).

We had this kind of problem with aircraft electronics connectors back in the 1980s when I worked at Honeywell CFSG. Stuff would work fine with some units, and not with others, and sometimes only under low temperature condtions (like when freezing down towards -40C in the burn-in chamber), and I've seen it with plenty of connectors since then, especially those designed with round barrels with no spring clip in them or around them to cause pressure/tension on the inserted pin to force consistent connection resistance. None of the "ebike" connectors I've seen are designed with that.

On the MXUS adapter, there are two exceptions.

1. The motor's white wire is a temp+hall combo signal. The adapter takes motor white -> L10 gray (which, according to the manual, the Runner with demux into pulse and temp readings). When connected to the Frankenrunner, it does just that perfectly -- I get reasonable temp readings and the speedo on the dashboard works when spinning the tire by hand.
2. The white wire on the L10 is connected to ground as I found some odd behaviors when connected to the Phaserunner (weird speed reporting; floating voltage?)

Does the PR correctly detect speed and/or temperature when connected to a different motor's sensor? If not, there is likely a problem with a wire anywhere up to the PR PCB inside it.

If the PR does work with a different sensor, there is probably a connection problem between the one that doesn't work and the point it connects to the PR (for the portion of cable not used with the one that does work).

 

[chuyskywalker]

It could also be a connection issue...

When on the PR/MXUS combo, I can spin the wheel very slowly and see on the PR dashboard all the halls light up; even in (what I think is) a sensible order. It feels more like, for some reason, during auto-tuning it's going too fast and can't properly detect for some reason.

Does the PR correctly detect speed and/or temperature when connected to a different motor's sensor?

Yeah, when connected to the RadMini motor it's fine and dandy on the speed sensor.

 

[amberwolf]

A connection issue doesn't have to be a go-nogo thing. Sufficient resistance in a connection can cause changes in the signal that can cause problems; these are more likely to manifest at higher currents or faster timing (higher frequencies).