Question Nine Continent FH212 & CA V3 -> what are sane temperature values ?

[HolgiB]

Hey there,

I have been searching ES for a while but not really found an answer. So there you go

I have a Grinfineon 4825 controller paired with a Cycle Analyst V3. Now I bought a fast Nine Continent RH 212 from a German vendor.
This is identical to the motor which is sold by Grin: FH212 Fast Wind

I will switch out my existing cheap DD hub motor to the RH212 this weekend and I have a temperature sensor now. So the question that arises to me:
Which temperature borders do I chose for thermal throttling with my Cycle Analyst ?

I asked over at the German Pedelecforum but didn't receive any answer. One guy over at the forum did use 90°C as starting point and 120°C as upper thermal limit.

Any ideas ? Otherwise I will start conservative with 90°C lower and 110°C upper limit.

TIA,
Holgi

 

[E-HP]

I have had mine set to 90C/115C which has worked well, but I monitor the temps while riding and above 90C and manually roll back/feather the throttle to keep the temp under 100C and stop to let the motor cool once it hits 100C. The rise in temp can be very fast once you hit the high 90s, and I've had my temp go from 90C to 110C in a few seconds so it's best to keep an eye on it. I've hit thermal rollback a few times unexpectedly when I was distracted, and it did what it's supposed to do, but I've never hit the 115C limit, since I caught it in time. My Statorade needs replenishing, which makes the temp rise more quickly, so I may drop the max to 110C.

 

[HolgiB]

Nice, thanks for your reply.
So 90°C lower limit and 110°C cut off sound like a good starting point for me.
I guess I should reconsider applying Statorade although the motor will mostly see flat surroundings with only a few mild hills.

 

[E-HP]

Nice, thanks for your reply.
So 90°C lower limit and 110°C cut off sound like a good starting point for me.
I guess I should reconsider applying Statorade although the motor will mostly see flat surroundings with only a few mild hills.

Statorade and temp monitoring are cheap insurance against walking a dead ebike home.

 

[amberwolf]

I have a Grinfineon 4825 controller paired with a Cycle Analyst V3. Now I bought a fast Nine Continent RH 212 from a German vendor.
This is identical to the motor which is sold by Grin: FH212 Fast Wind

I will switch out my existing cheap DD hub motor to the RH212 this weekend and I have a temperature sensor now. So the question that arises to me:
Which temperature borders do I chose for thermal throttling with my Cycle Analyst ?

Since others already covered the settings:

What temperatures do you actually see when riding?

If they're not very high, you dont' have to worry about this.

I'm using a mismatched dual motor setup**** with my heavy heavy-cargo SB Cruiser trike, and during my commute with many stops and starts and using regen braking much more than physical brakes, it's usually below 70-80C even in our hot summers. In winter, like now, it's less than 50C.

**** ex-A2B "500w" Ultramotor with Phaserunner v6 on the right side, and MXUS 450x (either 3 or 4, can't remember which) with ancient grinfineon. both running off same 52v pack.

 

[HolgiB]

Statorade and temp monitoring are cheap insurance against walking a dead ebike home.

True, but this knowledge does not help if you are running a cheap 140€ direct drive hub motor without temp sensor which I had in this bike for several years.

What temperatures do you actually see when riding?

I just started mounting the FH212 yesterday and still need to rewire the Phase wires & hall sensors in the first place. So asking about temperature limits is more of a precaution.

If they're not very high, you dont' have to worry about this.

Actually I had a MTB with a 12T GMAC, 30 Amps Infineon controller plus CA V3 for two years which I sold a friend of mine and this is what keeps me cautious. I have seen high temperatures up to 120°C IIRC running the CA into thermal throttling when riding in the hills. So I would not be really surprised if I run into similar issues with the FH212 on the steeper climbs.

I played around a bit with the Grin motor simulator and in theory the 212 should never overheat unless I hit inclines with more than 5%. But the same simulator told me that I never run into overheating with the 12T GMAC which was not really true. So I am really curious for the first rides in the hills but this will take a while since most roads in the woods are blocked by fallen trees right now.

I'll report back my findings and thanks you all for your answers !

 

[E-HP]

True, but this knowledge does not help if you are running a cheap 140€ direct drive hub motor without temp sensor which I had in this bike for several years.


I just started mounting the FH212 yesterday and still need to rewire the Phase wires & hall sensors in the first place. So asking about temperature limits is more of a precaution.

Actually I had a MTB with a 12T GMAC, 30 Amps Infineon controller plus CA V3 for two years which I sold a friend of mine and this is what keeps me cautious. I have seen high temperatures up to 120°C IIRC running the CA into thermal throttling when riding in the hills. So I would not be really surprised if I run into similar issues with the FH212 on the steeper climbs.

I played around a bit with the Grin motor simulator and in theory the 212 should never overheat unless I hit inclines with more than 5%. But the same simulator told me that I never run into overheating with the 12T GMAC which was not really true. So I am really curious for the first rides in the hills but this will take a while since most roads in the woods are blocked by fallen trees right now.

I'll report back my findings and thanks you all for your answers !

I almost bought the 212 but ended up with the Leaf motor. Both motors modeled very closely in the simulator for performance and heat/efficiency, but I went with the Leaf. I've done a lot of temp and voltage sag testing with the motor, and I've found that you can identify a sweet spot with respect to climbing speed that allows longer climbs without overheating. Its really about how many watts you're running since 10% or more of that turns to heat. For my bike, climbing at 19 to 20 mph seems to be the sweet spot, being careful to hold the throttle steady, and works for a 1 mile 20% hill staying just under 100C at the top, no hubsinks, only Statorade.
I was doing voltage sag testing on this video, and wasn't paying any attention to the temps. I started the video about 1/2 way up the hill so the motor is already warm (when I do temp testing, I try starting with the motor around 35C or less). Fortunately the limits kicked in and dropped the power enough for me to eventually notice. From the video, you can see how slowly or quickly the temps rise. At one point, after a curve, you can see the watts jump up to 5kW (since I'm testing voltage sag), at the steepest spot on this route at around 17%. The temps move up pretty quickly at that point, and the current roll back starts happening shortly after that.

 

[HolgiB]

Ok, I got my wiring cleaned up today and went for a short ride. The FH212 runs silently but doesn't provide much better acceleration compared to my old direct drive but this was to be expected since the old motor is also pretty fast on 36V.

One thing bothered me though / left me puzzled and this was the temp sensor. Since I have a ratther old Grinfineon 4825 controller plus an old Cycle Analyst V3 There is an adapter cable involved which uses Anderson for phase wires, 5 pin JST for hall sensor and 2 pin JST for the temp sensor. I configured the CA V3 to the standard 10k NTC temp sensor.

My E-bike was hanging from the ceiling in my basement with a temperature around 18-20°C.

The CA V3 reported a freezing -36°C

I entered the CA setup menu and the temp sensor showed 4.98V and -36.5°C

I unplugged the temp sensor and measured with my multimeter. I got 15.95 kOhm. Not shure if this is a valid value for the NTC based on 20°C environment temps. I also measured the temp sensors resistance directly at the adapter cable to enshure that my extension cable to the CA at the handle bar is not damaged somewhat but got nearly identical 15.91 kOhm. After some pluggin, unplugging the temp sensor and spinning up the motor without resistance the temp sensor shows 1.54V and 63.3°C. Not very reasonable since I can not imagine that the motor heats up without load. During a few circles in front of our house the CA sometimes showed more reasonable values and often the -36°C.

What is happening here ? Do I have a new RH212 with defect temp sensor ? I would be interested in the beta constant which the 10k NTC has. Does anyone know it ?

OK, so far (as often) a bit hit and miss.....

 

[amberwolf]

When I've seen that low negative temperature, it has always been a connection fault to the thermistor somewhere along the line, usually at the connector.

For the too-high temperature, it's usually been when the beta for the sensor isn't the default one the CA uses. You can determine the beta of a thermistor. While I don't recall which method I used, a search on that phrase brings this as the first hit
https://amwei.com/how-to-calculate-ntc-thermistor-beta-value/
and the relevant part is copied below. There are other useful pages in that list if this one doesn't work out.

But the CA doesn't have a field for the beta itself, so you have to use the settings it does have for scale factor and offset to fiddle with it until it works.

Beta Value Calculation Formula​

NTC thermistors beta calculation formula

4 key Components of NTC thermistor Beta Calculations​

1. RT1 = Resistance at Temperature 1
2. RT2 = Resistance at Temperature 2
3. T1 = Temperature 1 (K)
4. T2 = Temperature 2 (K)

 

[HolgiB]

@amberwolf: Thanks for your fast and detailed reply !
So I guess I need to go back and check the complete extension cord from sensor to the CA V3. I have the bad feeling that perhaps the two pin RST connection on this "cable spider" below the the CA is somehow broken / disconnected. Will have to recheck this as well and resolder a dedicated connector to the CA V3 if needed. May be I find something out if I wiggle the cables manually when watching the CA V3 setup menu.

 

[amberwolf]

Remember the CA display updates are not instantaneous, so you have to go slowly or you may miss a change. I'm not sure if it applies any averaging to the setup menu values but if it does you'll miss short value changes.

Usually for this sort of thing, you're better off watching a multimeter display if you can connect it to the last connector prior to the CA itself, once you eliminate the cabling up to that point. You may even be able to use the continuity/diode test function and listen for glitches; this is essentially instantaneous for faults that exceed the limit for dis-continuity for the specific meter used.

 

[HolgiB]

When I've seen that low negative temperature, it has always been a connection fault to the thermistor somewhere along the line, usually at the connector.

Dude, you saved my day. You have no clue how this comment helped !

OK, I'll explain a bit in a minute. First off, I found a segment in the CA documentation in regard to NTCs, motors and the beta value.
"The Cycle Analyst has a 2-pin input intended for motor temperature sensing. Many motors supplied by Grin have a 10K NTC thermistor (Beta constant of 3900) prewired into the stator which can hook up to the CA3 through a 2-wire extension cable. "

So we have the Beta constant and know that 10 kOhm means roughly 25°C.

OK, I build my own "test thermistor" using a 2-pin JST connector and a fixed 10 kOhm resistor


Unfortunately the CA still showed still -36°C. So I opened the CA itself:

I couldn't measure any connection between the yellow wire for the thermistor on the circuit board and the JST connector. So I took a closer look and things came together quickly. If you look closely you can see that all wires which are connected to the cable spider (E-Brake / Potentiometer / etc) coming from the second cable in the middle are soldered correctly. Only the yellow cable comes from another cable (seems to be the power cable feeding my light). So I soldered the spare yellow cable protected with shrink tube to soldering pad of the thermistor and *BAM*:

24.5°C as expected with the fixed resistor.

Now the motor thermistor is working as well.....

Issue solved..thanks to you all....