Tangent Build...V10...Astro 3220

[joostj]

Notice that HV series from Castle are rated 12s, so that using a 14s pack is risky, even with precautions...plus the 18650 have more intrinsic sag/ripple than lipos....At 160A with an 8p pack, you are at 20A per cell....probably I would go for a 12s-13s but 10p pack or even more. Remember to keep the battery/esc cables as short as possible and to use a bigger capacitor bank (sold even by castle) if your cables are longer than the stock coming from the ESC and/or your battery are not so big in terms of max discharge rates compared to the max expected peak amps your system needs.

That's great advise.
I realize i am exceeding the max voltage, and it is a risk. But it seems to handle it relatively well.

My CAv3 setting limit the amp to 100. It feel like plenty of power my for current usage. By using 30q, I can technically get 120amp from the battery pack (15amp x 8p).

If you look at the Talon 120 vs. the Edge HV120, you can see that the capacitors are much larger in the Talon...that can be a big benefit...as well as the better heat dissipation.

 

[tangentdave]

The HV80 logging software shows about 0.5V of ripple with the extra Castle capacitors and about 1V ripple without. This includes the extra several feet of battery cable under riding conditions. Castle recommends keeping ripple voltage below 10% of system voltage. This is why I don't add the extra caps anymore. Joost is the only rider to damage an ESC (that wasn't due to a pressure washer).

Ham's broken gears were my fault, I took too much material out of the gear which concentrated stresses, caused a fatigue crack through the narrow parts and broke off a chunk of the gear face. The gears ain't narrow no more, but I'm still running the lightened gears in my bike without issue. So is Joost. Saving 10g isn't worth a long term reliability issue.

-dave

 

[panurge]

Hey Dave.....The tangent is a piece of art, and has already shown to take the 3210/15 torque/power levels very well.
And I'm sure it handle the power of a 3220, eventually with some minor refines for the increased torque/power and the demand of some heavy and/or skilled rider (solid gears etc.), I'm more focused on how the increasing power affects the reliability/durability of the bicycle drivetrain.
Anyway I would like to get my hands in one soon! and you'll be surprised by the application I have in mind, a true, BIG BIG wheel....

The HV80 logging software shows about 0.5V of ripple with the extra Castle capacitors and about 1V ripple without. This includes the extra several feet of battery cable under riding conditions. Castle recommends keeping ripple voltage below 10% of system voltage. This is why I don't add the extra caps anymore.

Yes, using extracap should be a need for some systems, or simply a form of prevention for others, like a general fuse for the batteries.....would be good obviously to have datalogs of these disaster events....and commuting data with many battery models, capacity and chemistry...
Personally I would have both onboard (extracap and fuse) in the event of disasters....

Sure you've played more than me on a comparison among the 80, 120 and 160 versions of the HV ESCs but I can't see the reason to use a 120A instead of a 160A model, with a 3220 4t, costs and a minimal form and weight factor seem very marginal to me. I would feel better running a 120A on a 3210 and a 160A with a 3215 instead ....

Just to ask: have you guys castle datalogs along the whole discharge cycle, hard driving a 3220, with one of those 14s8p 25r packs?
That would be really, really, interesting.....

As we well know, 25r and 18650 in general have a wider operational voltage range than lipos: a 14s Li-Ion could be discharged at the level of a 12s LiPo starting 8.4V higher (that's one of the contra in any case for Li-Ion batteries on performance machines).
For these reasons, although I have Lipos on my bike, I've used with success for a time a 13s 18650VTC pack .....there where 130 of them (10p) top charged @52.5V (or 4.05V per cell) Not any problem indeed....and virtually it's totally safe (advice from my E.I. RCguru) if your first throttle hit is wide and long enough to consume the juice needed to drop the voltage @50-51V no load, that's the Castle rate......(I.E.a race start or a time attack climb). I've also tested HV lipos that are 4.35 per cell or 52V@12s without any problem.

That said....running 14s, is not sustainable I think.

A >200A continuously capable Lipo pack and an HV 160 are the requisites IMHO for a system that would run the continuous and the peak power levels the 3220 sustains, without constrains, with a safety margin, and with an acceptable performance drop along the DOD (from 55 till 46 kph in my case with lipos, compared to 55 - 39kph with Li-Ions, or, in a 100A CA limited system: from 5000W to 4100W with lipos and from 5000W to 3600W or even less with 18650.

 

[motomoto]

Panurge wrote:

Hey MotoMoto, nice to hear again about your V10, Is it still on the road? how many miles if I can ask? Feedback about how the freehub handles the job?

Everything seemed stressed with all the torque. I didn't have a lot of failures, but they would come with time. It would be a high upkeep situation.

It just sits in my shop. It was my first experiment with electric and since then all I have been thinking about is a bike to produce. I am collecting parts for a production
prototype that will be belt drive with a Revolt 160 short. I will start a thread when I am far enough along.

 

[tangentdave]

The heat does accumulate in the drive. More testing is in order of course; the 3210 limited to 60A battery current (~100A phase currents) means you can pretty much ride hard through a 20Ah battery pack and not think about overheating. The 3220 at 120A battery current (~200A phase) is the larger equivalent. I'm not concerned about the durability of the drive unit being affected by higher input torque (120A only generates like 6ft-lbs of motor torque), I am concerned with overheating the bearings and motor windings. All the motors going forward will have the temp sensor from Astro built in.

My goal for the system is a race-ready setup; it has to put out power for a 2hr race without slowing down the rider. 120A and a 3220 is crazy fast on the trail and unless you're full throttle hill climbing, you can focus on the race, not your motor temperatures. Everybody in the forum here knows the components and can choose their own setup- please do, let's find the optimal arrangement. What other speed controls compare to an HV160 or the TalonHV120? I'd love to get my hands on bare ESCs and bond them inside a custom heat sink and screw it down to the mounting plate to waterproof the whole shebang.



-dave

 

[Nathan]

When the esc blew the cav3 was only at 70amp. I think the issue was likely due to heat.

This is no surprise and many people from the start said this was the main weakness of this kit and other RC motors on a ebike. ESC's are not meant for stop start loads that you get on an ebike. The same for the motor ! Yes it will work for a short time but long term reliability is not one of rc gears highlights.... so expect to keep forking out for these replacements.

This stuff is designed to be used in a plane at full throttle with low loads compared to the high loads and shock loads of stop/starting at full power that a bike sees.

Get a proper heavier ebike controller if you want it to last on an ebike but then you may have similar problems with the bearings on the motor being underrated for the loads of an ebike.

 

[litespeed]

The peaks on R/C helicopters far exceed those of a plane. I've peak 200Amps on a 100 amp controller hundreds of times.

You are correct though, peaks last milli seconds verses minutes. I would think on the bikes we need something that easily doubles the peaks the bike would need, 200 amp rated controller on a bike that peaks at 100 to start.

I wonder if the cap pack would help controller longevity.......http://www.castlecreations.com/products/cc-cap-pack.html Or maybe two?

Tom

 

[panurge]

Not totally related but might explain a lot:

Extraploated from the attacched PDF, got it somewhere on the web time ago....

From the Bernie’s Rant du Jour november 2005 - Castle Scribe

Ripple Current Dept:

No, this has nothing to do with cheap kid oriented wine (if you know this reference you are too old for hi-tech stuff anyway.)
You all know that a speed controller works by switching full current on and off really fast, the percentage of on to off in each
pulse being averaged by the motor, so I don’t need to go into this. But what happens to the battery when the controller asks for large current
in small pulses like this? During the on phase the battery’s voltage drops. To make up for the loss of current the controller gets the rest
of it from those big capacitors on it’s end. When the controller slams the current gate shut at the end of the pulse the battery refills the
capacitor then waits the long and dreary microseconds for the process to repeat.
But suppose the battery pack is sorta wimpy and cannot put out the amount of current the controller is asking for. The controller gets the
current the battery cannot supply from the capacitors, but now the battery is still trying to refill the capacitors when the controller wants
the next current pulse. The capacitor gets completely drained and the battery voltage drops even further.
Current drops in the controller. Now the battery cannot even start to fill the capacitors before the cycle repeats and on the next pulse the
battery voltage drops even further. Electrons are running around trying to fill empty capacitors at same time they are trying to meet the
controller’s frantic effort to keep the ravenous motor system’s electron appetite filled.
Voltage fluctuation goes up on the battery side untilthe capacitors give up. They then do what everything does when it gives up, they explode. They might not actually explode but just start to think about it really hard.
This is a bad thing. Parts start melting off the controller, wires burn, batteries get hot and bothered and general mayhem ensues.
The point here is simple. One of the most reliable ways to put your electric system at risk is to use batteries that are not up to the task. Too big a battery never hurt anything but too small a battery has destroyed many systems.
With this in mind, ponder the significance of the first part of the rant.
Happy flying, and keep the smoke
in.
Bernie Wolfard
Support Specialist
Castle Creation

 

[macribs]

The peaks on R/C helicopters far exceed those of a plane. I've peak 200Amps on a 100 amp controller hundreds of times.

You are correct though, peaks last milli seconds verses minutes. I would think on the bikes we need something that easily doubles the peaks the bike would need, 200 amp rated controller on a bike that peaks at 100 to start.

I wonder if the cap pack would help controller longevity.......http://www.castlecreations.com/products/cc-cap-pack.html Or maybe two?

Tom

Not sure how many you would need for e-bikes, but my guess would be far more then 1 or 2 to help out.
Years ago we did custom car stereo builds, and to keep feeding those hungry low ohm amps we used 1 farrad caps the size of 1/5 gallon bottle or so.
Now I don't know the physics behind it as it was not my job to do that, I was just building and wiring, but the more amps used in bi-wiring multi subs systems the more caps was used. Cos the there was 15-20 feet of 2 or 4 gauge wires but still, those caps where huge and this cc-pack has only three tiny little ones.

I think e-bike use with start/stop many times in a row, or sometimes even uphill with heavy load will put a lot of strain as max current will be drawn probably much longer on an e-bike then what people do with planes and probably helicopters too. So as e-bike usage will draw more current the number of caps needed will increase is my guess.

Maybe someone with physics or EE knowledge can do the math and try to estimate how many of those cc-caps will be needed? How could one calculate the size or numbers of caps needed?

I could be totally off here, and if I am let me know. It's not my intention to post garbage, just think that 1 or 2 cc-caps will be too little to work for e-bike use.

 

[recumpence]

A bit of input from my experience running 3220s with a 160 ESC...... (sorry Dave, I'm not trying to hijack your thread)

I am currently running (pun intend :wink: ) a 3 turn wye 3220 and an Edge 160. The battery is a 12s lipo 40ah 35c pack. I am running HUGE 2 Guage wire and a 500 amp capable shunt for the CA. My system pulls 285 amps repeatedly.

My observations;

#1 The battery is BY FAR the biggest area of importance when it comes to ESC survival (followed closely by wiring)
#2 Heat is a bigger issue with the controller than the motor, at least when discussing the 3220. My controller hits 165f occasionally, but the motor runs 120f and under all of the time.
#3 The difference in price and size between the Castle controllers is minor. Therefore I say it is best to run a 160 on every system just for durability sake.
#4 With my crazy battery and wiring, I only see .15 volts of ripple at the max even pulling nearly 300 amps. I haven't run any additional caps in a very long time. Keep the power system (battery and wiring) capable of providing the power needed and no additional capita are needed.

Game on......

Matt

 

[litespeed]

I agree with everything Matt said but most e bike batteries are up to nominal currents just not peak currents. That was why I recommended the cap packs.

These Astro systems are what intrigue me the most here and what I plan to do next....finally.

Tom

 

[macribs]

A bit of input from my experience running 3220s with a 160 ESC...... (sorry Dave, I'm not trying to hijack your thread)

I am currently running (pun intend :wink: ) a 3 turn wye 3220 and an Edge 160. The battery is a 12s lipo 40ah 35c pack. I am running HUGE 2 Guage wire and a 500 amp capable shunt for the CA. My system pulls 285 amps repeatedly.

My observations;

#1 The battery is BY FAR the biggest area of importance when it comes to ESC survival (followed closely by wiring)
#2 Heart is a bigger issue with the citrine than the motor, at least when discussing the 3220. My Corley his 165f occasionally, but the motor runs 120f and under all of the time.
#3 The difference in price and size between the Castle controllers is minor. Therefore I say it is best to run a 160 on every system just for durability sake.
#4 With my crazy battery and wiring, I only see .15 volts of ripple at the max even pulling nearly 300 amps. I haven't run any additional caps in a very long time. Keep the power system (battery and wiring) capable of providing the power needed and no additional capita are needed.

Game on......

Matt

If my math if correct a 12s16p will weigh in at around then 10-11 kilos all welded up. 240 A battery current. Would that be enough battery current or could pack be even smaller? Guess what I am asking is what is the battery current limit one must surpass if using 18650 30Q's for battery pack for a 3220 Hv 160 setup? Would such a pack be stable enough for drawing high current repeatedly in such setup without crippling ripple and banks of caps?

That C rate and battery Vs phase current still puzzles me I must admit.

 

[tangentdave]

If the bike is in close to the right gear, the motor is accelerating under power. That means current is falling.

Joost is the only rider out of 36 kits so far to damage one of these RC ESCs, and he wasn't riding it at the time, so who knows what happened. The ESC does get hot if the gearing is very tall and the motor RPMs are very low and the throttle input is very large. I mean like full throttle up a steep-ass hill near top gear just crawling along. Everyday trail riding does not stress the ESC. Maybe repeatedly drag racing and starting halfway up the cassette could get the ESC hot, but you'd have to do it 10 times in a row and not let it rest in between.

There is a noticeable difference in performance between a LunaCycle 11.5Ah PF pack and a 20Ah 25r pack. The PF pack is the smallest I would run 60A settings with, I see 5V of sag under the full throttle acceleration, but this is only a 14s4p configuration. The pack is noticeably warm after riding through it hard. The 25r pack isn't phased by 60A settings and performs great with 100A settings. I'm saying settings because that's the max battery current setpoint in the CA- this is never seen when riding because the motor is already moving.

-dave