TSDZ2B OSF - Random Thermal Throttling

[first.loser]

I have a TSDZ2B with OSF.

When it is behaving normally, the temperature would be something like 20*C up to 38*C.

However, I have instances when the temperature would shoot up from 20*C up to 60, 80, 120+ *C even while I am fully stopped at an intersection.

I have tuned the motor down to draw a maximum of 8amps from 11 amps thinking it is drawing too much and heat builds up.

I have also turned down my assist in power and torque to provide 10% more in level 1 up to 20% more in level 3. (I only use level 1 assist most of the time).

Initially this seemed to help since my temperature no longer shoots up. It was like this for almost a month. However, yesterday it started to temperature throttle like crazy at 120+*C.. even at level 1 assist. I also noticed progressively weaker assist halfway into my commute even thought it says I am getting an additional 70 watts of electric assist.

I am trying to figure out what is going on here. According to my 860C display I normally draw 2.6amps and never went over 7.9amps..

Could a battery cause this?

Here are my specs:

TSDZ2B with OSF and 860C display 48v 750W
Brass Gear
Premium cooling Kit from ebike stuff (plate, fins, everything)
Temperature Sensor (LM35 bought from digikey)

Battery is a RADMISSION 1 48v 10.5aH 504wH which I got from my old bike. I've only been using the supplied charger for it.

I limited the TSDZ2B to provide no more than 300 watts of electric assist. I really under powered everything to try and eliminate the heat problem.

 

[Zambam]

With temps of 20*C up to 60, 80, 120+ *C , does the motor case feel correspondingly hotter as the temp rises? 120C would have fried your motor and be too hot to touch with bare hands.

 

[first.loser]

To be honest I haven't touch it when it shows that temp.. I wear gloves when I ride.

On the way home today I will make sure to touch it.

Alright I am charging my bike at work and I noticed that my battery now will not charge pass 90%. The rad charger says it is done charging even when I just plugged it in.

I arrived at work with 87% left and I plugged it in and it shows it's finished charging. (Charger has 2 LEDs and when it is charging it usually shows 2 red lights. When it is finished it shows 1 red 1 green).

It has been 3 hours since and my battery is stuck at 90%.

 

[first.loser]

It didn't do it for a prolonged period on the way home today. Temperature shot up for a second 3 times in my 45 min commute.

I found that it seems to do it when I spin over 90 rpm for a prolonged period even at level 1 assist.. flat road or incline. Power cuts off then temps shoot back down almost instantly and power comes back on.

I'm really trying to find more info about my RADMISSION battery but nothing comes up other than the specs I listed above. I do have a spare one.. maybe I can try that tomorrow.

I attached a photo of my battery voltage from the 860C display over 45 minutes from 90% charge. Maybe someone can help me interpret it.

I am already eyeing a custom battery at the same time

 

[Gruesome]

Loose wire and/or a short in the thermal sensor connection, would be my guess. I'd try to see whether I could change the temperature reading while stationary by wiggling the temperature sensor wires.

 

[first.loser]

Oh man... I was hoping not to open the motor up and everything again . Crankbolts and Bottom Bracket thread are red loctited in place and feels solid. Maybe I'll do that Friday night I guess. It might give me some chance to look at the grease. It's been almost a year / 900 kms since build.

I swapped to my other RADMISSION Battery today. So far nothing unusual happened. Temps were between 20*C- 26*C. I even did level 3 assist for a good 10 mins.

Is there any chance my Radmission battery's BMS is proprietary and won't like to talk with my TSDZ2B properly? I've been running this setup for about a year.

 

[Zambam]

BMS only talk to the battery, not TSDZ2B. Is the new battery charging to full?

Where is the LM 35 thermally connected to? When you said the temperature “shot up”. How fast was it in C / sec?

 

[first.loser]

The spare battery is charging to full 100%.

The old battery yesterday only charging to 90%.

The temperature shoots up in a second. Sometimes it stays up like that anywhere from a second to a few minutes. Temperature shoots back down to normal temps in a sec or gradually lowers down sometimes.

LM35 is connected to the same spot as tutorial video below.

 

[Zambam]

LM35 is in contact with motor lamination per video. With that iron mass, motor temp can't rise and fall in a sec. Should be more gradual. Something's wrong with the temp circuit. Needs investigation. Take motor cover off and have a look at LM35. Good solder joints? No intermittent shorts? Test it with a hair dryer? Replace LM35?

 

[first.loser]

Sorry I couldn't provide an update yesterday night. I was unable to replicate thermal throttling on the ride home. I also wanted to open up the motor and check it out but family obligations kids, wife etc means I can only work late at night and into early morning.

Here is a short 15 sec video of the thermal throttling on my way to work this morning. This time it does it when I hammer hard on level 5 assist and it generates anywhere from 250-360 watts of assist. I touch my motor right after and it is coldish warm. It is 16*C outside this morning.

I noticed battery amp shot up to 11amps (red line) even though my max setting is 8amps.

I double checked my settings for the motor and battery in OSF:
*On 100% charge*
BATTERY:
Battery 8amp
Low cut off 39
Voltage calc% 100
Resistance 200
Voltage est 54.4
Resistance estimate 0
Power loss estimate 0

SOC:
Reset at voltage 54.2
450wH (504wH rated minus 10% as suggested by OSF manual)

MOTOR:
Motor Voltage 48V
Motor Power max 375
Motor Acceleration 10
Motor Deceleration 100
Field Weakening Enable

I set the settings like that because although I don't have info on the rad battery, it was used with a RADMISSION 500w motor. So I felt it can at least do 8amps continuous.

Hopefully I have time to open it up tonight and poke it around a bit. Testing will be as follows:

1.) With motor on, move wires to see if temps shoot up. If temps shoot up look for short.

2.) If temps do not shoot up, blast it with a hair dryer and monitor temperature.

 

[Zambam]

Watched your video. Motor temp is displaying a lot of noise! Before opening up the motor, try reseating the throttle connector a bunch of times then go for a test ride.

 

[first.loser]

Sorry forgot to mention my motor is noisy due to brass Gear!

I had a nylon gear before that burned out at 2k kms before I had OSF.

 

[first.loser]

Alright I am starting to feel like it's the battery again. Before I left work, I decided to tune everything down and try to hammer it hard:

BATTERY:
Battery 6amp
Low cut off 39
Voltage calc% 100
Resistance 200
Voltage est 54.4
Resistance estimate 0
Power loss estimate 0

SOC:
Reset at voltage 54.2
450wH (504wH rated minus 10% as suggested by OSF manual)

MOTOR:
Motor Voltage 48V
Motor Power max 250
Motor Acceleration 10
Motor Deceleration 100
Field Weakening Enable

It was raining so I even had lights on. First part of my commute back is a bridge and I set it to level 3 and hammered hard providing me with 140-160 watts of electric assist for just under 10 minutes. Battery amp draw was 6.1 and turning red on the graph but she did not throttle down on me. I hammered hard at level 1 assist giving me 60-70 watts electric assist for 30 minutes. At the last part of my commute I have to climb 13% grade for 200m and it decreases to 8% grade. I hammered hard at level 5 providing 240 watts of assist for 10 minutes. It did not throttle down.. but my amp draw was 9.1amps when it was maxed at 6amps in the settings.

Anyway it was the smoothest ride home in a while and no throttling down!

 

[Gruesome]

Sorry forgot to mention my motor is noisy due to brass Gear!

I had a nylon gear before that burned out at 2k kms before I had OSF.

I think 'noisy' meant 'electrically noisy', as in 'reading jumping up and down a lot'.

 

[Zambam]

I think 'noisy' meant 'electrically noisy', as in 'reading jumping up and down a lot'.

I was wondering why he brought up the brass gear and motor noise, which I did not even notice when watching the video. I was looking at the temperature display jumping from 30C to 94C and back down to 30C instantly which I called noise. Just learned he rides in the rain which could corrode electrical contacts, causing a lot of weird problems.

I would suggest he check and clean every electrical connection and water proof them with dielectric grease.

 

[first.loser]

Lol. First of all thank you everyone for your patience and clarifying .

I reread Zambam's post "Motor temp displaying a lot of noise!". I was reading all of this quickly and my brain omitted the displaying part.

Here is a picture of my setup and maybe give thoughts on if this is causing noise.

To clean up my wiring I cut open an old water bottle on both ends and stuffed everything there.

The red crimped connectors are my lights. I could probably use more electrical tape to not expose the metal.

Then there is the connector for the battery.

At the top of the downtube is my connector from the display to the motor.

At the back closest to the rear wheel are my speed sensor and lights Y split cable. *When you flick it hard enough, the temp goes +/- 1*C* Is that noise? When I pedal sometimes my shoe kicks it accidentally when I am hammering hard.

Lastly, there is the 8 pin connector. Is this what you want me to reseat? Because I don't have a throttle. However, I believe the 8pin connector is for TSDZ2Bs that came with throttle--which mine did along with the old VLCD 5 Display which connects to the actual throttle from the display.

I do need to let you guys know that all the connections are solid and click nicely in place. Except for the 8pin. This one feels like it doesn't go all the way in no matter how hard I push. I eventually just pushed it in as far as it will go (while aligning all the pins of course) and held it in place with more tape.

I was about to start cleaning and putting grease on the connectors tonight but I realized I had no dialectic grease in my tool box. Just lithium grease and a dozen small tubes of various loctites lol, and Mobil 1 Grease. Gut feeling tells me I can't use any of those in place of dialectic grease. In short I had to abort mission and resume either tomorrow night or Sunday night.

Oh and what's the best way to do this? First turn the battery off, spray with 99% Alcohol scrub with cotton swabs, dry with compressed air, and just apply grease on all connections liberally? Some of those pins are really tiny...

Yes it rains a lot here. Pacific Northwest. It was hot up until 2 weeks ago but this week it's cloudy and rainy.

 

[Zambam]

At the back closest to the rear wheel are my speed sensor and lights Y split cable. *When you flick it hard enough, the temp goes +/- 1*C* Is that noise? When I pedal sometimes my shoe kicks it accidentally when I am hammering hard.

Motor temp should not go +/- 1 C when you kick the speed sensor cable. That is also noise!

Did you ride in the winter with wet salted roads? Salt is conductive and can cause lot's of electrical problems. Wires do not like sharp bends (they can break and become intermittent, I can't tell from your pics if this is a problem). Using electrical tapes on wires generally not good practice in wet environments since it traps moisture.

As for cleaning electrical connections, Brake Clean spray + tooth brush works well.

Sounds like your 8 pin connector has issues and can't be seated all the way? Take a look with a magnifier to make sure none of the pins are damaged (bent, pushed in, corroded). The mating rubber connectors have a lot of friction to get a tight seal and needs to be pushed hard to seat all the way. Some lube on the rubber mating surface may help to get it to seat all the way.

Did you charge the battery (off the bike) that only reached 90 % (on the bike) and measure the voltage with a DVM and not rely on 860C display? What's the results?

Sorry about my comment on reseating the throttle connector. I didn't know what I was talking about!

 

[Gruesome]

Lol. First of all thank you everyone for your patience and clarifying .

I reread Zambam's post "Motor temp displaying a lot of noise!". I was reading all of this quickly and my brain omitted the displaying part.

Here is a picture of my setup and maybe give thoughts on if this is causing noise.

To clean up my wiring I cut open an old water bottle on both ends and stuffed everything there.

I think you need to strain relieve the wiring entrances and exits to the enclosure, otherwise the enclosure and the sheathing might do more harm than good. See comment below.

The red crimped connectors are my lights. I could probably use more electrical tape to not expose the metal.

In my experience electrical tape a) does not stay in place and b) leaves an unholy mess from the adhesive.

Then there is the connector for the battery.

At the top of the downtube is my connector from the display to the motor.

At the back closest to the rear wheel are my speed sensor and lights Y split cable. *When you flick it hard enough, the temp goes +/- 1*C* Is that noise? When I pedal sometimes my shoe kicks it accidentally when I am hammering hard.

Both (change in reading when hit; routed so it can be kicked) are bad.

Lastly, there is the 8 pin connector. Is this what you want me to reseat? Because I don't have a throttle. However, I believe the 8pin connector is for TSDZ2Bs that came with throttle--which mine did along with the old VLCD 5 Display which connects to the actual throttle from the display.

I do need to let you guys know that all the connections are solid and click nicely in place. Except for the 8pin. This one feels like it doesn't go all the way in no matter how hard I push. I eventually just pushed it in as far as it will go (while aligning all the pins of course) and held it in place with more tape.

I was about to start cleaning and putting grease on the connectors tonight but I realized I had no dialectic grease in my tool box. Just lithium grease and a dozen small tubes of various loctites lol, and Mobil 1 Grease. Gut feeling tells me I can't use any of those in place of dialectic grease. In short I had to abort mission and resume either tomorrow night or Sunday night.

Oh and what's the best way to do this? First turn the battery off, spray with 99% Alcohol scrub with cotton swabs, dry with compressed air, and just apply grease on all connections liberally? Some of those pins are really tiny...

Sounds good to me.

Yes it rains a lot here. Pacific Northwest. It was hot up until 2 weeks ago but this week it's cloudy and rainy.

A general comment on wiring: the 2nd-to-last picture (name ending in 591) shows several spots where stresses are concentrated, with wires exiting from sheathing or wrapping and making sharp bends. It's easier said than done, but it would be better to immobilize cables against frame members, and have no forces or movement transmitted from the cables to the connectors and solder and screw joints. The corrugated sheathing looks sturdy, but concentrates stresses on the wires where they exit. Ideally you'd terminate (strain relieve) all sheathing at an enclosure (box, can), but that becomes bulky and cumbersome. Maybe replace the corrugated tubing with something less bulky, like heat shrink tubing? I'd look into stiffening the transitions from sheathing to bare cable (which should be avoided) and from connectors to bare cable (can't be avoided) and also the connections themselves with different diameters of heat shrink tubing, and avoid electrical tape. Automotive cloth tape as used under the hood for wire looms in cars might work, but is not really made for exposure to the elements.

 

[first.loser]

I have relocated my speedometer and lights Y splitter so it is completely out of the way. It had some sharp bends as well from before. I have rerouted it and consequently relaxed it a bit to eliminate the sharp bend. My shoe should not catch it now for sure.

I cleaned the connecting points with alcohol and dried them with an Air Duster. I applied some dielectric grease. All connectors including controller to screen, screen to 8pin motor connection, battery connection, y splitter speedometer and lights.

Currently temperature does not have noise even when I flick the y splitter like before.

Noise shows up when I turn on lights as per video below.

I think the weak point is my crimped on red connectors. Too much metal is showing and the female feels loose when the male goes inside. Touching it slightly causes my headlight to turn off even though lights are on.

Will be replacing it with a connector but gotta Amazon it and will take a few days.

I ride my bike in the rain but not so much. I do not ride in the snow. Not that brave lol.

I will change connectors and report back.

So far the bike seems stable at 250w and 6amps. I am able to ride at level 5 assist for a good 5 minutes. Won't be able to take her for a good ride til Monday when I go to work. But it is raining in the afternoon... Will see how that goes.

I am eyeing a 52v 14aH battery from a local supplier. When my noise problem is resolved perhaps I'll pull the trigger.

 

[first.loser]

I rode her again to work today. 15 km in the morning over cast 16*C. 15 km night time raining 18*C

The bike feels good and actually I feel less tired even though settings are maxed out at 250w and 6amps.

I was able to use level 5 assist for 5 mins straight providing 200-240 watts electric assist. My temps also did not go over 28*C. Most importantly, there is still no throttling.

Thursday and Friday this week will be interesting because we are getting summer weather and temps again. It will be sunny with highs of 24*C and 25*C. So I will see how the motor likes it when it is a few degrees hotter.

I am more sensitive to noise and it occurs during the following conditions: when lights are on (not moving) and when the motor is providing assist and moving.

Although I like the look of the wiring on the bike right now because it's clean, I will try to spread the wires over the frame as per the suggestion above to see if noise is reduced or eliminated.

By the way can someone help me interpret the graph? What does it mean when the battery amps shows red on the graph? This is when motor was providing over 200w of power. This is on the bridge and on the 8-13% grade hill.

 

[Zambam]

I am more sensitive to noise and it occurs during the following conditions: when lights are on (not moving) and when the motor is providing assist and moving

Although I like the look of the wiring on the bike right now because it's clean, I will try to spread the wires over the frame as per the suggestion above to see if noise is reduced or eliminated.

Are you talking about mechanical noise (sound from brass gear) or electrical noise (jumpy motor temperature readings that are not real)? The cause of both are likely not related.

 

[first.loser]

Are you talking about mechanical noise (sound from brass gear) or electrical noise (jumpy motor temperature readings that are not real)? The cause of both are likely not related.

The electrical noise with not real jumpy temp readings. It occurs when lights are on and when motor puts out some assist.

 

[Gruesome]

How are you detecting the electrical noise, if not from the jumpy temperature readings?
By the way, if the temperature display jumps just between two adjacent numbers (as opposed to readings further apart), it's not necessarily an indication of larger than normal noise, the real temperature could be just sitting at the boundary of those two readings, and normal noise moves it back and forth over the edge.

The red segments in the battery current graph are probably just meant to signify 'high current'.

 

[first.loser]

How are you detecting the electrical noise, if not from the jumpy temperature readings?
By the way, if the temperature display jumps just between two adjacent numbers (as opposed to readings further apart), it's not necessarily an indication of larger than normal noise, the real temperature could be just sitting at the boundary of those two readings, and normal noise moves it back and forth over the edge.

The red segments in the battery current graph are probably just meant to signify 'high current'.

Thank you.

Currently it is just two adjacent numbers for example: 20/21.

When lights are off it sits at 20.
When lights are on it jumps to 20/21.
When there is electric assist it jumps to 20/21.

Noted about the battery graph. The battery isn't hot during those times when I touch it. It was just concerning when I see red as I don't want an explosion during my ride.

Motor seems smooth and happy with current settings limited to 250w and 6 amps.

 

[first.loser]

I thought I would post an update.

My old battery (RADMISSION Battery Pack) had unknown specs regarding these and I could only top out at 5a setting in OSF and limited to 250w. Setting it higher would give funny readings and loss of power over an extended period.

I just bought a new battery to experiment with and it outputs 20a continuous and 30a peak with Panasonic Cells. I upped the amps in OSF to maximum 12a and 350w peak.

At the same time I also redid my wiring so it goes over and around the frame when I changed battery. I eliminated sharp bends etc. No fake temp readings. Temps stayed below 35*C. I rode like this for a week. I just didn't like how there's a bunch of wires showing on my bike...

Last night I did my wiring as before thinking my old battery was the problem and I liked to the clean look. Near then end of my commute I got the fake temp noise again and no power.

I will be redoing my wiring again to go around the frame..