Dave's 80:1 RC mid drive kit build log

8kW burst, 4-5kW continuous (w/heat sink and airflow, PCB will be potted somehow). I'd like to see phase currents in the 200A burst range. These systems seem to be much like the e-skates, high peak currents for 2-3 seconds during acceleration then a more tolerable current in the mid-upper rpm range.
 
And with Bluetooth modules you can change settings via cellphone. Even monitor stats on iwatch/Android Wear. The current issue with electric skateboard size ESC I think is everyone wants more output but are limited by space. Most boards barely fit 12s5p or less even. So that limits output to 100a ish based on cells used. Not to mention unless your board has bindings it's real hard to stay on the board while applying full throttle. But lucky for everyone people keep pushing the limits. VESC6 is supposed to have 80a continuous. But with better cooling might be able to increase. Not tracking anything in 4-5kw cont range yet. I'd say the A200s but that's one expensive monster and still in initial production.
 
A little more on the A200s VESC, seems like some will run it @18s - they better be careful to avoid meltdown.
Seems like a tough build, with a solid heat sink. Designed to be stacked if you fancy that. More info in the links below pic.

g7BdU34.png




https://vesc-project.com/node/339

http://www.electric-skateboard.builders/t/making-the-a200s-16s-200a-high-powered-vesc-hw6/50654
 
Those VESC look neat but they are quite sizeable. Not sure they would be the ideal candidate for our setup. That said I can envision some fun uses for them.

I remember telling Dave I wanted more than the talon 120... I didn't know what I was talking about. It bursts so hard when uncorked that it's somewhat scary on a bicycle. I think adding the additional mass via the heatsinks is going to add plenty of thermal capacity to the ESC. If you want more cooling then add a fan. The talon faces inward when using the heatsinks and helps protect the fan from debris (mounting holes and power already provided). Even easier if you have a triangle mounted kit. As long as you are geared properly for the terrain you are riding you shouldn't have any thermal issues.
 
knurf said:
Giant Glory Tangent
I’m starting a glory build (on a 2013) soon, on a gen 2, did you have any problems getting everything aligned and pretty?

Nothing serious. I did tap the fixed BB cup over just a hair on the provided bottom bracket but it's no biggy. Depending on your gearing you may want to rearrange the rear cassette to get the best chainline. This is primarily for endurance and wear than it is in actually getting a running kit. I ran a 9 speed cassette for the last half of the riding season last year. When I put on a new wider range sunrace cassette and chain this year I pulled a couple of the sprockets to get a more ideal chainline. I found myself riding in my low gears a lot (I'm a clyde and love hills).

Watch out for wheelies your first ride. Antisag setup of the glory seems to like to pop that front wheel in the air 8) Post up when you get it together.

What are people using for fenders on these DH bikes? I've seen the little plastic front fenders that go on the fork but they don't have me convinced...
 
The little plastic fork fenders are a great, don't keep the bike clean but keep a lot of spray off your face.

These BB's from Luna allow us to easily shift the spindle in the frame (or another way to say it is we can easily slide the bearings on the spindle to accommodate different widths and chainlines).


20180509_170337.jpg20180509_170432.jpg


With a spindle stickout of ~23.5mm, plus our Cyclone flanged freewheel and 1/4" thick spider mounted to the inside face, the motor chain should line up with the outer chainring position. Moving the chain side bearings (the fixed cup) can fine tune chainline since the spider is located via crank/ISIS interface, then simply move the left side bearings (removable cup) to the correct width for the frame. 83mm BB is shown in the pic, set the distance between flanges on the cup to the actual width of the BB, the Tangent mounting plates will add 2.5mm per side and allow us to tighten against the frame rather than bottom out on the bearings.
 
tangentdave said:
8kW burst, 4-5kW continuous (w/heat sink and airflow, PCB will be potted somehow). I'd like to see phase currents in the 200A burst range. These systems seem to be much like the e-skates, high peak currents for 2-3 seconds during acceleration then a more tolerable current in the mid-upper rpm range.

I finished soldering some hw6 vescs last week. I've been ok so far with 14s although just like with the current rc escs there is no headroom to fully charge 14s. We should be ok to push voltage a bit further than the eskaterssince we dont regen. I'm on 150a phase and might try a bit more once I get some spare ics. Might be ok since this new design has current measurement on all phases and cycle by cycle current limiting on each fet. Most importantly, the sensorless start feels pretty good (much more powerful than phase runner) and stator saturation compensation helps with the the low speed high amp jitters that sensorless foc often suffers from.
 
Hey Dave,

I sorted out the throttle esc beep, but still haven’t been able to put too many km’s on.

My trouble is with the temp sensor. I don’t get 5 minutes down the road without it triggering the CA to stop. It reaches a very high temp in a very short time so I’m thinking something is faulty. Even if I put it on a stand and throttle with no load for a few minutes the CA flashes over temp (1188C).

I’m sure I have the CA setting correct as you have outlined in your setup guide. It even reads ambient accurately. Just under power (again even no load) temp climbs fast and it triggers over temp. The water jacket is full and the case is barely warm so I’m sure nothing is hot.

Is there a way to trouble shoot the sensor in the 3220? Any suggestions? Does the “negative”,”positive” make a difference in connecting the two wires to the CA?

Thanks!
 
Leeleeducati said:
Hey Dave,

I sorted out the throttle esc beep, but still haven’t been able to put too many km’s on.

My trouble is with the temp sensor. I don’t get 5 minutes down the road without it triggering the CA to stop. It reaches a very high temp in a very short time so I’m thinking something is faulty. Even if I put it on a stand and throttle with no load for a few minutes the CA flashes over temp (1188C).

I’m sure I have the CA setting correct as you have outlined in your setup guide. It even reads ambient accurately. Just under power (again even no load) temp climbs fast and it triggers over temp. The water jacket is full and the case is barely warm so I’m sure nothing is hot.

Is there a way to trouble shoot the sensor in the 3220? Any suggestions? Does the “negative”,”positive” make a difference in connecting the two wires to the CA?

Thanks!

Is it possible you have a 10k ntc thermistor but with an incorrect beta value? That would read the correct 10kohm at ambient and then get way off at higher temps. Since it's just a resistor polarity should not matter
 
I just checked again, it's actually it's a bit off even at ambient. It's reading 33C when ambient here is 15C right now. I also tried changing to 10k in the setup and ambient reads ~103C so there's some other scaling I'd need to input if that is the case. Not sure what to try next?
 
Hmmm. That thermistor was installed by Astro Flight. They use a KTY84/130 by NXP semiconductors. The polarity of the connection shouldn't matter (somebody chime in if I'm wrong, when I swap the polarity there is no difference). If the CA reads 1140C, that means the connection to the thermistor is open. The CA will flash a thermometer in the upper left but should still display the measured value if the connection is closed.

Leelee, you're saying the temp indication is reading hot when the motor is actually not hot to touch? If that's the case, you can temporarily disable the temp protection while we troubleshoot; keep the max amps setting 120 or less. Astro uses like 30ga single strand wire on their temp sensors...if you enter the CA menu and cycle to the SETUP TEMP SENSR main screen, it'll show you the measured value in real-time. Watch this screen while you bend all the temp sensor lines from the motor to the CA, any change may indicate a physical issue at that point.

-dave
 
I'll try bending some of the wires. In the meantime, I just measured resistance at 630 at 16C on the two wires coming off the motor. Can someone else let me know if they are close to this value?

Correct, the motor is barely warm when I get the temp to trip.
 
I'm measuring the same resistance on a bare thermistor, room temp air is about ~590ohm, when I squeeze between my fingers to warm it up I get about 630ohm.
 
I checked all the wires, it all looks good. Even get a consistent 630 at the CA connection.

I just tried putting it on a stand and running it approx. 50%-75% throttle for about 3-4 minutes and things look like they are getting hot. The aluminum that holds the freewheel is very warm. Would that be normal? Temp. raised to approx. 90C (again no load on stand).
 
When the unit is not loaded, the amount of throttle commanded isn't super important as it will just spin full speed and use whatever power that requires. 3-4 minutes at full speed without airflow will result in a unit that's warm/hot to touch, especially the aluminum output piece (it's the recipient of all the work being put in by the motor and there is a rotary seal between this piece and the stationary cover behind the freewheel). In order to keep the gearbox this small, there are no rolling elements between the eccentric gears and the output- the eccentric gears are sliding on the pins (the amount of individual sliding is very small but the motor is spinning very fast, so the tiny sliding occurs frequently). Before I ship these, they're run on the bench at slow speed for a couple 10 minute sessions. At the end of ~10% speed for 10 minutes, the units are pretty warm, so 3-4 minutes at full speed before they're pretty warm seems normal. There's about 300W going into the gearbox without any heat leaving...so...yup.

Now, the temp sensor. I'm not sure why it would be indicating hot when the motor is not hot to touch. Can you watch the temp shoot up during use? Does the temp fall rapidly after you close the throttle? The thermistor is embedded in the motor windings and will react very quickly to powers over about 3kW, and should fall quickly as heat spreads to the rest of the gearbox. Are you seeing similar behavior?


-dave
 
So the mechanics sound ok then with heat at the freewheel. It never really gets “hot”, just warm to very warm but nothing you couldn’t leave your hand on.

As for the temp, yes it does cool down very quick. When I watched the temp climb under power it climbs steady but then at some point spikes... so imagine a steady climb 30, 40, 50, 60, 70, then shoots to 100+ then back to 70, 80 etc. there’s almost a spike at some point sending it past the shutoff, but if I instantly reset the CA it’s back to 80 ish. Is there some setting to signal average? Any idea why ambient starts higher than actual? 20+ degrees more at start (even after an overnight cool down). 38C on the CA vs 16C measured in the garage.

Was Googling... Do you think I have this noise problem?

https://www.youtube.com/watch?v=xihDc4YHAZ8
 
I’d try and take the unit/thermistor out of the equation, just to divide and conquer.
Do you have any old potentiometer or electric junk you can scavage for one? Then wire it to the CA and see what it does. Measure the resistance, compare and see what the CA makes of it.

As per the CA manual, a small capacitor (0,47uf) can be soldered to reduce the jitter, but I don’t think this is the issue here. Simple enough to try if you have any around I guess.
 
I tried plugging into the KTY84/130 in the 3220 again this morning (CA settings as Linear; .45V, 262V etc.). Ambient is 16C, 3220 reads 43C (see bottom pic). Wiggled all the wires and no real change.

I then tried a 10k Grin thermistor I had (top pics) at the same connection point as the 3220 to eliminate any wiring questions. Changed CA settings to 10k and I'm reading ambient correct ~16C.

Might just be a problem with the 3220 temp sensor.
 

Attachments

  • 5A2711EA-6CDC-47CC-B0E6-306BDAF34A63.JPG
    5A2711EA-6CDC-47CC-B0E6-306BDAF34A63.JPG
    56.2 KB · Views: 2,091
Last attempt on my end... and yes duct tape can fix everything :)

The good thing about collecting ebike stuff over the years is that you never know when you'll need something you already have. I had some KTY84/151's I dug up to test the Linear mode in the CA. Scaled to .44/267 and ambient is correct and temp seems to read just fine. The fill port was the only place I could think of to stuff the new temp sensor until we can figure out what to do with the one in the 3220?

Sad to say it looks like it's defaulting to the internals of the motor.
 

Attachments

  • 10EA8E6A-2D3C-4FB6-B470-85185E72C8DC.JPG
    10EA8E6A-2D3C-4FB6-B470-85185E72C8DC.JPG
    66.6 KB · Views: 2,074
Whelp. I used this particular motor testing the v1.5 gearboxes, possible I damaged the cables or something. We can install a new sensor inside the motor by pulling off the end cap, we can get straight to the windings and pull the cables out the same exit used now. Might be a good idea to remove the rotor so the magnet doesn't suck the epoxy into the gap. I can do it here in just a few minutes, I'll post the procedure for opening the gearbox too. You could epoxy that 10k thermistor against the motor coils and call it fixed (assuming it'll handle high temps- is it coated in plastic?)
 
Back
Top