Spinningmagnets' RC Friction-Drive Build-log

spinningmagnets said:

I get sync-loss if I try to accelerate suddenly, (screeching noise, loss of power) but if I accelerate at a gradual rate, no sync loss. I will try some hills and also different voltages and kVs, and a different throttle..

Click to expand...

Ouch !! .. you hit a nerve there !
keep us updated on this situation and any solutions you can find.
I am still bugged by this issue and cannot decide which was to jump next as there doesnt seem to be a proven solution..
EG..
CC ESC ?
different motors ?
Halls in the motor + sensored controller ?...( ahh but, which one ?)
Keplers "magic" throttle ?
AO suggestions ?

Hillhater said:

AO suggestions ?

Click to expand...

Copy EVTodd's design. :wink: He never reported any issues other than motor "repair" and/or potential failure prevention upgrades, and he & his bike are no light-weight pairing. His motor choice is less than $50, and I can verify it is small in size. I own two. He did have to re-glue the magnets, and I don't recall whether he replaced any bearings?

At 36v he got excellent performance, no ESC burn-outs, high top end high-speed starting from stand-still stops...

What more could you want?

You could buy a more expensive motor & probably avoid the upgrades to the motor, but that's about all I can think of for improvement with his design & components.

Yes, the more I experiment, the more I find I must compliment Todds initial intuition about his component selection. There may be other minor differences, but the major ones between the two right now are that he is using 36V, and he is using a Castle-Creations HV-85A ESC. He had gotten the ESC as 'used' from Ebay at a good price, and he'd also gotten the three 36V cordless drill-packs at a VERY good price. (three of them would guarantee that he never suffered voltage sag from a too-small battery, and the extra range always helps).

When I became interested in making something similar, I priced 3 new drill-packs + charger, and a new CC-HV ESC (yikes! http://www.google.com/products/catalog?q=Bosch+36V+battery&hl=en&cid=3784706462075852214&os=sellers#).

http://www.castlecreations.com/products/products_fly.html Zibby used the $140 CC "ICE-100A" (5S-8S) with no issues, and right now for higher voltage, CC has a $190 HV-80A available (5S-12S).

Snath used a conventional E-bike controller with halls attached to the motor, Gwhy also. Based on the voltages and kVs used, and the resulting speeds attained, I think my Miking 120A ESC isn't working right (it was cheap, it can now be my back-up ESC).

If you plan to use a 2-1/2" diameter roller, such as the motor-shell (like Adrian and Kepler) Kep has reported that an 85A gets very warm, but the 100A ESC handles the loads fine. So I am confident after reading Todds results using an 85A on the smaller 1.0"/1.25" rollers, that the CC HV-80A would work great for this. Of course, for anyone who wants to drive the axle with an RC/ESC system, Matt has established that the Castle-Creations HV-160A is the best choice, and no other comes close.

Obviously i could replicate Todds set up, and hope to get the same result, ..but that would mean starting from scratch.. motor, controller , batteries, and drive design. ...!!
That is a totally different drive (more complex) and not a favoured solution for me.
Also, as SM's drive is similar in concept to EVTodds ( but he now has seen sync issues) i am keen to see if we can isolate the cause... motor, ESC , throttle etc
My issue is that all the components i have used, ( 63mm 200kv turnigy motor, 100a KForce ESC , throttle, 5s Lipo etc) have worked successfully for others (Kepler etc) in various combinations
Also the Castle ESC failed with sync issues for Oatnet ..so no guarantee there.
The conventional sensored Ebike controller is not readily available for a 5S , 80A system without some major mods (unproven also), so a sensored system is also unproven answer.
No,.. i guess i realise the answer is a different motor and ESC, but the question is ... which will be the best bet. ( costly gamble !)

Hillhater said:

Also the Castle ESC failed with sync issues for Oatnet ..so no guarantee there.

Click to expand...

Was that with the Kepler drive & EVLogix's "throttle-izer" <--That can't be counted as a negative for Castle or the motor.

We need an update from Kepler on his new throttle control...

I hope I'm not side-tracking your thread SM. I think your design will eventually work with the proper adjustments too.

deVries said:

Was that with the Kepler drive & EVLogix's "throttle-izer" <--That can't be counted as a negative for Castle or the motor.

Click to expand...

No. He first had the problem when just running with a servo tester as throttle.
I have also found no difference with either a hall based throttle system or when using a servo tester to control the ESC.
Recumpence has also mentioned he has in the past had sync issues when using a Castle ESC on certain motors (AXI's ?) :?

Spinner, sorry to hear that you have sync problem also but I am not surprised, as you know I have tried just about every friction drive configuration and always have the same sync problem. I have tried two motors, turnigy 50 and 63 size and two ESC's , Mystery and Turnigy but have only tried one voltage - 5 cell LI-PO. I have an 8 cell LIFEPO4 pack that I use with my Cyclone powered bike so I might try that sometime, but that would mean a different ESC so maybe not. I have a sensored controller that came with a second hand hub motor so maybe I should try fitting some Hall Sensors. I just can't decide what to do so I will keep watching with interest . Good luck
Robbo

I have been scratching my head trying to work out what is different enough between out designs such that you seem to be having the sync issues, when I haven't. We are using very similar ESC, motors, batteries, so I started to focus on the other differences. The main ones I see are:

1) The geometry of the progressive engagement
My motor physical translates by an inch or so between engaging the tyre, and fully loaded. This will tend to smooth out the motor acceleration, and load. This could help maintain sync. But I assume yours does something similar, so maybe it is just the profile of this "dampening".

How much translation of the motor is there when loaded up?


2) Potential sources of increased resistance
Your motor to roller coupling is something my drive doesn't have. Is it possible that when this is loaded by the tyre that something moves, flexs, binds up. Any misalignment may cause increased resistance to rotation.

Is the motor as easy to turn by hand when coupled to the roller, compared to un-coupled?

When tyre pressure loads the drive axle, does it close up the slot clearance to the motor drive "tongue"

What are the no-load power requirements for having the motor
- spin in free air
- coupled to the drive roller
- driving the wheel (no-load) with minimal engagement
- driving the wheel (no-load) with enforced full enagement (ie. sliding the drive by hand to an amount typical of a loaded situation on the road)

What limits the amount of engagement you get into the tyre? Maybe you are getting too much causing extra rolling resistance.



3) ESC Settings
Not sure of what I have my settings at, but I have the the slowest throttle response set. This was primarily to slow down the initial engagement with the tyre on my design, but would definitely be a factor in sync loss.


These are all just random thoughts, but hopefully they might help get to the bottom of the problem.

Just for the record I have only tried combinations of the following motors, ESC, and batteries.

Motors:
HXT 63-74 200kv Brushless Outrunner
TGY AerodriveXp 90 SK Series 50-65 270Kv / 2100W

ESC:
Turnigy Brushless ESC 85A w/ 5A SBEC
TURNIGY K-Force 100A Brushless ESC

Batteries:
Turnigy 5000mAh 5S 20C Lipo Pack
Turnigy 5000mAh 6S 20C Lipo Pack

Best of luck,

- Adrian

I don't have too many sync issues either but I can sure make the motor loose sync if I don't manage the power input properly at low speed.

Few things are working against the guys using the smaller 50 diam motors. Firstly, these smaller motors are more susceptible to sync loss then the larger motors. Second, the cheap ESCs loose sync much more easily then a good quality Castle. Third, is that simple servo tester throttles are quite touchy to use especially down low and as such, just a small over correction on throttle application can invoke an out of sync event. Add these 3 together, and you are working hard to manage sync.

However, the aim of a friction drive is to be an economical light weight solution so we need to make the most of what we have.

I have zero sync issues at the stage I am at. However to achieve this, there are compromises.

1) I would love to use a 50mm motor and save 500 grams over the 63-74, however, the smaller motor just doesn't have enough copper in it to be a reliable solution. I regularly get the 63-74 too hot to touch under hard testing conditions and still never had a failure, not even a loose magnet. I suspect the higher magnetic flux of the larger motor allows the rotor to get to the next pole more easily and hence less chance of sync loss. (Just a theory though I am not a motor expert by any stretch )

2) The K-force 100A will do the job. The setup needs to be optimized to minimize sync problems. (These work for me but are only suggestions)
Set the brake to off
Set cut off to soft
Set cut off thresh hold to High
Set start mode to soft or very soft
Set timing to max (26.25)
Set governor off
Set PWM to 24kHz
Set motor Load to very heavy

3) Throttle control.
Never apply throttle without the bike moving. (I set a electronic minimum start speed of 5mph or 8kph with the interface)
Keep the throttle input gentle at low speeds. Once the bike is moving at a reasonable speed, throttle can be applied much more aggressively without fear of sync loss.
Bottom line is that this type of setup should never be pushed hard at low speed.
The mythical Kepler interface forces this to happen by mapping a maximum current draw for a given speed range. This in conjunction with a minimum start speed, large motor, and optimized esc ensures sync is never lost no matter how hard I try.

Just like to add that its not that hard to ride around sync issues especially when component selection and setup has been optimized. The Interface just stops people doing the wrong thing and is a way of making an RC friction drive user friendly enough for the general public to use. It can be a little frustrating also for the end user because it forces them to pedal and get the bike over that start thresh hold plus power levels are set quite low at these slow speeds. However, this is what it takes for this sort of drive to be reliable.

Yes, I know, so were the hell is the interface. All I can say is that its close. The software engineer is working on profiles and a user programming interface at the moment so that it can be optimized for different setups. Although optimized for the eboost, we are keen to make it available to other users in the near future. I will startup a separate thread on the interface soon. Not sure which section to start it in though as it not ready for sale yet.

I know you guys are looking for cost savings. However, I have never had a sync issue with a Castle HV controller.

Again, I know cost is a factor here. However, there sure do seem to be alot of sync issues mentioned in threads lately, when the solution is simple (though more costly).

Matt

recumpence said:

... I have never had a sync issue with a Castle HV controller.
Again, I know cost is a factor here. However, there sure do seem to be alot of sync issues mentioned in threads lately, when the solution is simple (though more costly).
Matt

Click to expand...

Matt, I respect your extensive experience on this , and i am convinced the CC ESC is one of the best available, but i fear that even it may not eliminate the sync issue in all situations. I is surely a combination of several factors...ESC, motor, throttle, voltage ?, etc, etc..
As you said yourself....

recumpence said:

I have not run the HV100. But, I do run 160s and they are flawless for syncing. Bear in mind, though, that the motor itself can be an issue. My early AXI ourunner was easy to lose sync with on my HV110 ESC, but the Astros and my Plettenberg never lose sync no matter how hard I abuse them.Matt

Click to expand...

PS:- do you think the Castle HV ESC's are better at sync than the Castle ICE range ( assuming the higher voltage capacity is not required ??

"I hope I'm not side-tracking your thread SM..."

Click to expand...

Not at all, deVries. No matter where the conversation takes place, the sync-loss issue seems like it 'should be' solvable, and here's as good a place as any. I'm glad for any input.

My initial thoughts were that the Castle-Creations ESCs are more expensive for a reason. They clearly have a more sophisticated design which I'm certain not only allows the added features (data logging, wider choice of performance settings, etc), but it is also certain to be better at filtering and analyzing the motors back-emf in order to determine the best timing [compared to the average ESC]. This is no small capability with our motors spinning thousands of times a minute (6,000 RPM = 100 R per second).

I only hoped to find components that were 'adequate' for the low-end of the market. However,...since Matt has shared that even with some of the CC ESCs, the AXI motor sometimes experiences sync-loss, perhaps the motor is a major contributor in the chain? I remember that Todd uses a Towerpro motor with his CC HV-ESC.

I am weak at electronics (still learning something new every day on this forum). I know various motor designs have differing 'inductance' (haven't understood what that is yet, perhaps it contributes?). CC ESCs are clearly better at avoiding sync-loss, and I have ordered a Hobbywing HV-ESC, and I have one of Matts throttles as the last two experiments before I accept that CC is my only option (if they ARE the solution, then so be it).

The Miking 120A ESC has a high amp-rating at a cheap price, the catch is: it has no adjustability for the settings, the LVC didn't work, "might be" contributing to sync-loss.

Exact same issues with 6S LiPo and 24V SLA.

Adrian, thanks for posting. I'm a big fan of your build. I adjust the drive so that when pedaling, the roller is barely touching the tire. When powered on, it slides back a little less than an inch, and is stopped by a bolt in a slot in the top-plate. Initially I started with a fairly shallow angle. Because it slides in a straight line onto the curve of the tire tread, the more it slides back the deeper it presses into the tread. I gradually increased the angle until there was no more slippage under brisk throttle. Slippage was determined by a change in roller sound (scratching rubber?) and motor RPM 'blipping' up and down.

I'm fairly certain the motor shaft is not experiencing any side-loads or binding in any way, and also that it is remaining centered on the drive-shaft. I believe that all the downforce loads are carried by the the two side-plate bearings of the upside-down U-channel. The roller-clutch bearings are very snug on the shaft, and the shaft is very snug in the sideplate bearings. Everything continues to roll freely and show no 'slop' when off the bike.

When I accelerate gradually, everything performs quite nicely. I'm even now imagining a snubbed mechanical throttle linkage. I suddenly pull the trigger down, but the servo-tester shaft only moves at a pre-determined rate. It would just be a band-aid instead of a true solution, but it would work (for now?...).

Please understand, I am not trying to criticize at all. I just figured I would point out that which is painfully obvious to me. Then again, I would love to find a sub $100 controller that will work for bikes. That would be awesome!

Yes, I had sync issues with my weird 14 pole AXI. However, they were different sync issue than you guys are having. The issues I had were specific to one RPM at a certain throttle setting at a certain load. I found out later that this is a known issue with the AXI and (at the time) the only issue, and supposedly, that was related to the old style HV firmware and ONLY with that one motor. It had something to do with the resonant frequency of the can at that RPM and load.

Anyway, I do not want to hijack this thread.

You guys know me, I am a big fan of CC controllers. However, I tend to be open to spending more on my bikes than others. I understand that completely.

Matt

Anyway, I do not want to hijack this thread

Click to expand...

No hijack Matt, I want to hear anything you have to say!

Just remembered: Also ordered a 50mm motor at 400-kV to get higher speeds at 6S (why not?...)

Is it possible that when this is loaded by the tyre that something moves, flexes, binds up?

Click to expand...

Anything is possible, but the 1/4" thick aluminum channel seems VERY rigid. Your design and Keplers are the most compact and lightest configurations possible, so on this style I leaned towards unneccesary added strength.

I think I'll jump in for just a second ... My drive is actually spinners , with mods to the way it contacts the tire.. No problem at all with motor to roller excel ...I did have Voltage sag , but no problems at all since going to 12 gage wires all the way from bat turminals to the 6s 85 amp ESC w/2caps.. two 12 volt SLA batteries .. Hot off the charger 25.6.. Over volted !! Yes.. I think thats the trick .. Haven't noticed any faze problems ...I do peddle to 5-10 mph... The low end is really good , top speed says 25 :wink: . I really do believe over volting, is a good thing with in reason.. Friction drive is the way to go ...Thanks to you pioneers on this forum ... Bill

Kepler said:

Yes, I know, so were the hell is the interface. All I can say is that its close. The software engineer is working on profiles and a user programming interface at the moment so that it can be optimized for different setups. Although optimized for the eboost, we are keen to make it available to other users in the near future. I will startup a separate thread on the interface soon. Not sure which section to start it in though as it not ready for sale yet.

Click to expand...

Ok, thanks for the update & ESC suggestions. ETA

What's the new "Halloween-FD Avatar"

Friction drives by nature are not the prettiest things around. Looking at ways of making it more appealing to the general market.

Make a contoured sharky-looking housing for it, with radio-active-looking color accents on black, and rename it the "Boost'em Boojum". :lol:

Hmm...that's catchy. I think I'll use that for the next drivetrain I come up with.

Some good news, I have at least one combination that I couldn't get to lose sync (not 'yet',...I will keep trying, haven't tackled a big or long hill). The constants have been: 26" fat tire, level road, 180-lb rider...

I recieved a new Hobbywing HV-70A ESC and program card last week, so of course it snowed. $85 + tax & shipping
http://www.hobbypartz.com/07e-c-platinum-70a-hv-opto-v1.html
No BEC in this one (to provide 5V to the throttle), so I bought a cheap tiny Radio-Shack battery box that held four AAA NiMHs to equal 5.4V (which is how Todd did his) http://www.radioshack.com/product/index.jsp?productId=2062254 Also, it comes with a small extra power wire with a no-spark resistor (its use is not required).

Hobby-Partz has a warehouse in California, so it took less than a week and shipping was not expensive. I bought the 295-kV motor from them when it was on sale for $45, and I have been pleased with both the motor and ESC. The Turnigy HV-70A is reported to be a direct copy (also the 120A version), but the savings would have been only $8, shipping would cost more, and there's always the 2-3 week wait from China to consider (plus the Turnigy 'copy' might not be quite as good?). http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=9130

Waiting for LiPo, so I got some cheap SLA (yes, I've seen Hyena's battery pics), and when the SLA was at 24.6V (equal to a full 6S LiPo) I got sync-loss when I tried sudden acceleration (instead of gradual) when I used both the Exceed 295-kV or the Turnigy 250-kV. The Hobbywing ESC was a fail at 6S.

The Turnigy Servo-Tester (ST) was acting hinky on the test stand, so I only used the E-Sky ST. I thought perhaps a lower kV with a bigger roller would be more likely to suffer sync-loss than the other way around, so I put on the 250-kV motor and the 1.25" roller to get right at it (it could rain at any moment). The SLA pack was configured for 37V to replicate 10S, and the ESCs power-ramp was set to very-soft (per Keplers post), and no matter what I did it wouldn't lose sync.

The very-soft power setting was odd at first. Slam the throttle all the way on, and the speed comes on gradually all the way up to the top. Not sure if the improvement has been the ESC, the higher volts, or the power-ramp setting. I have taken Adrians suggestion from his thread and have ordered one of these $8 servo speed regulators to test: http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=8863

Managed to get that test done today because the sun finally shined long enough to dry-up a strip down the middle of the road, and yes...the roller slips badly when its wet. The 250-kV with 1.25" roller at 37V provided 20.x-MPH.

No sync-loss...Using a Castle-Creations Phoenix-ICE 75A at 24V (Ebay $90)
http://www.castlecreations.com/products/phoenix_ice.html

Its nice that this ESC has the data-logging, and it also has a built-in BEC to power the E-Sky ST as a throttle (no separate 5V power supply needed), and also it can run 8S (28V)...but the big question for me was will it suffer sync-loss at 24V? and I couldn't get it to stumble under any conditions (sudden full throttle being the worst culprit). I used the factory default settings, worked right out of the box.

I don't know if its the hardware or the software, (also it has an donut-inductor on the throttle wire), but their secret sauce works. Its almost a moot point, because the available motor-kVs for the 63mm diamter motors work better at 36V, also I could hear and feel that the drive ran better on hills at 36V. In fact, in an earlier post I posted that the Hobbywing HV-70A ESC ran good at 36V, but had sync-loss at 24V.

If you live on relatively flat terrain and don't mind only 20-MPH using the highest kV 63mm motor (Exceed 295-kV), this model of CC ESC will do the trick for a simple 6S system. However, any interest that has been expressed around here has been for help on the hills, plus there's always someone who wants to go a little bit faster (higher volts than 6S will help both of these).

At 36V (10S) there are many motor-kV choices in the commonly desired speed ranges. At 22V-24V (6S) theres really only a couple that are useful, unless you want lots of choices that only go at the slow end of speed ranges. edit: The smaller 50mm motors are likely to be well suited to the task of driving a small 1.0" roller on flattish terrain, and kVs are available from 270-to-400+ for up to 30-MPH on 6S. However, the shafts are only 8mm, but they are an option...

edit: also had a thought today, another option might be to use a 63mm motor with a fatter 1.5" roller (which will fit my drive) but the fatter roller might require a higher 100A Phoenix ICE ESC (only $25 more than the 75A). The fatter roller would give the available kVs of 63mm motors a good top-speed selection-spread while using only 6S.

I haven't tried the inexpensive servo-signal speed-adjuster yet, it sounded like Adrian had enjoyed some success with that, and it may prove to be beneficial here.

Kepler and Adrian use the 63mm motor-shell as the roller, which is about 2-1/2" in diameter (the smaller 50mm motors run a little warm as a shell-drive, and most avoid them). Their 85A ESC ran a little hot, but several makes of 100A ESC have worked well for those drives. 75A may sound light for my drive, but I am only using a 1.0" diameter roller (optional 1.25"), and Todd has reported his similar drive was using 60A peak, with about 10A during cruise, so I am confident the 75A (100A peak?) is more than adequate (never got warm). The data-logger link will be here soon so I can verify my guesses.

A while back I read up about 36V (Todds original pack) and 28V cordless drill packs (Makita, Milwaukee) and several of the manufacturers have a long term committment to continue producing the 28V packs. Oddly, the 28V packs are configured to output more amps than the 36V packs, and the 28V system was designed as a response from customers who like the 36V power, but found the packs too bulky. They are expensive, $150 each and $100 for the charger, so three packs would be around just under $600 for only 9aH of range (3 X 3aH). Each battery has a 5-LED "fuel gauge", and the CC-Phoenix-ICE range of ESCs will run at 28V.

Now that I've settled on 36-ish volts, I'm ordering two 5S LiPos to continue testing. I'm really glad the weather got better this week!

(edit: just found out Zippy is now producing 4S, 5S and 6S packs in 8000MaH, so two of these bricks would provide 8aH of 28V for $120, or...8aH of 37V for $148) http://endless-sphere.com/forums/viewtopic.php?f=14&t=25700

Just figured out the Castle Creations USB data-link. It is free (if you buy a new ESC from a legitimate retailer) by using a product-code found inside the package ($24 if you don't have the new product code). The code wouldn't work at first, until I figured out that one of the code digits was a capital letter "OH" and the other was a number "Zero".

I downloaded the program from the Castle website. The Windows program-install wizard window made the installation easy, (although I had to right-click and select "install as administrator"). The Castle data program is easy to access and read. These runs were made on the factory default settings, but its nice to know I can change them now if I want.

The website lists the 8S (33V max) Phoenix ICE-75A as 75A continuous, and 130A temporary peak. I made eight back-to-back hard acceleration runs on a slight uphill using 24V of SLA with a 1.0" diameter roller and the 295-kV motor, 180-lb rider on fat 26" tires.

All the runs were 60A peak, 30A to maintain top speed on a slight uphill, and downhill amps were almost unreadable because they were so low.

Watts were peaking at 2000, temporary voltage sag during accel with 12aH of SLA was about 5V, no sag during cruise. Started with 27V and average volts logged were 24V.

It was a warm day, and the peak temperature logged was the the last run, it zigzagged up and down on each run with the average temp sloping up, and final internal ESC temp was 100F (no fan installed yet).

Voltage-ripple...Fortunately, I ordered the ICE line (8S/28V) instead of the ICE-Lite (6S/ 22V). They both come with power-input capacitors, but they are sized to help a model aircraft instead of a heavy E-bike. I was getting enough voltage ripple (4V-ish ?) that I need to install a couple low-ESR caps ($5), and the extra ESC voltage capability probably helped me avoid popping something inside the ESC.

Since I am powering a 1:26 friction-drive instead of an axle-drive, I had hoped to avoid adding caps, but...it is what it is. I purposefully did a couple runs at 1/4 throttle, to make it as bad as it could be (Thanks to Matt for his generous assitance and advice!)

I don't regret having bought the generic no-feature 120A ESC (it still works), because it allowed me to sort out the mechanical end of the drive without worrying if I would fry the ESC (only $40). It will work as a back-up, and it doesn't have sync-loss as long as I don't accelerate hard.

Since I live 1.5 miles from work on a flat commute, I will use 6S and the 1.25 roller on the prototype drive. I will continue to test the production drive at 10S (37V) using the Hobbywing-ESC and 1.0" roller because the area around the college has lots of small hills.

edit: I want to clarify that I am recommending 36V for hills because the college students overwhelmingly don't want to use LiPo. Kepler/Adrian have proven these drives can climb well on 6S (22V-24V), but there is a lot of heat, and they use a 100A ESC at a minimum.

Without the high C-rate of LiPo available, I would need a bigger/heavier battery of LiFePO4/LiMn to avoid voltage sag. 36V doesn't bog down as much on hills, and as a result, it doesn't draw as many amps to maintain a long climb. If you like LiPo, this drive climbs well on 6S, but I would add a fan and a temp-readout...

spinningmagnets said:

Just figured out the Castle Creations USB data-link. It is free.

Click to expand...

That's great news.

Voltage-ripple- Fortunately, I ordered the ICE line (8S/28V) instead of the ICE-Lite (6S/ 22V). They both come with power-input capacitors, but they are sized to help a model aircraft instead of a heavy E-bike. I was getting enough voltage ripple (4V-ish ?) that I need to install a couple low-ESR caps ($5), and the extra ESC voltage capability probably helped me avoid popping something inside the ESC.

Click to expand...

what happened to the ripple voltage after the extra caps were added?

-Adrian

what happened to the ripple voltage after the extra caps were added?

Click to expand...

I havent added them yet. I was so excited to finally get real data, I posted right away. Since I will be accelerating gently, and will not be using 1/4-throttle on an uphill, I am not worried to use the ICE-75 for my flat-road commuting as it is. I have two caps left over from when I ordered some and added two low-ESR fatties to a Turnigy 85A. When rearengine wanted to buy a drive from me, he opted to buy the 85A I had, so I don't have any data on that, other than it ran fine for me, and rearengine says he's happy with its performance.



The Hobbywing 70A that I have used at 36V (of SLA, soon to be LiPo) was $85. If someone wanted an HV (18V-50V) Castle Creations ESC (from Ebay) the 60A can be found for $115 (the next-bigger 80A can be found for $170). If the 60A has a temporary peak amp capacity of over 100A, it should be adequate for a 1.25" roller or smaller if you have light hills (with added caps, of course, and perhaps a fan).

For hills, I feel that 36V performs much better, and when considering the higher-voltage ESCs, the 70A Hobbywing 50V-ESC is $30 cheaper than the 60A Castle Creations 50V-ESC. I will beat on the Hobbywing ESC like it's a rented mule, and if it survives I won't have to spec the Castle-Creations $115 ESC as the minimum. Although, I will always recommend the Castle Creations unit as the better option if funds allow.

I will take it easy on the voltage ripple until I get the extra caps on, and then I will retest under hard acceleration (also 1/4 throttle on an uphill) to see how much it improves. AJs postings have suggested that doubling the uF (while making sure the extra caps are low-ESR) should cut the ripple in half, and Matt has said its very important to keep the ripple under 5V (the lower the better). Here are my preferred low-ESR cap choices for voltages "up to" 6S (22V), 10S (37V) and 12S (44V) http://endless-sphere.com/forums/viewtopic.php?f=30&t=22194&start=15#p325273

Yup... No problems at all with the 85amp ESC with the 2 caps.. I'm pretty much running the drive like I do my hub motor ... usualy all out or, not at all .. Should be even less worry with the one" roller ...

by the way I got the roller in the mail today.. thanks... I'll see how this works with the launch ramp, and report the out come ... Bill

Hey Spinner ... Have you moved in , and working in your garage yet?
I just wanted to post and say the drive is working as planned , but I took the launch ramp off ... I found that when it wasen't eccelerating it would set on the ramp and spin, because i would still have the thottle wide open.. Ya follow? The one way roller has to be in contact all the time now :| .. 85 amp esc with caps is still going strong ..How is your drive working ? By the way did you get your 1" roller back ok?

Kepler! if your out there ... How did your launch ramp work out ? Adien's seems to work fine with out it... Happy Easter to all... Bill