14kw continuous 94% efficient motor.......

I know John got his running with internal halls. But, his probably has a smaller ballance wad on the rotor. :)

I am excited to hear your finished, running setup!

Matt
 
I have yet to load test mine beyond what a leather flap could induce, but it was running smooth and cool through the entire throttle range.
 
coleasterling said:
We've gotten more testing done and the actual Kv is around 102...Nice! The controller isn't in yet, but according to Kelly, should be in this week. After that, we can really start putting thing thing through its paces. I'll be sure to update as soon as the controller is in. The hall sensor mount is also made and the halls are ready to go in...Just waiting on the controller.

Now for something funky. I used the 4-wire method to measure the resistance between each set of phase wires in delta and it came in well under 20mohms...Seems really low to me. Does that seem feasible to yall? The setup was a power supply with the motor and a small lightbulb in series, with one multimeter measuring current and another measuring the voltage drop across 2 sets of phase wires. I let the system get to somewhat steady state, then took the voltage drop divided by the current to get the resistance. I did it with and without the lightbulb and got the same resistance, also. The equipment is top notch (Fluke 289, Fluke 43B), so that's probably not the issue if there is one.

Thanks muchly!
-Cole


Resistance varies at the inverse square of Kv, so if John's was say 50Kv and 100mOhm, it would be logical your 100Kv version would be 25mOhm. Some of the RC hobby motors are 8.5mOhm phase resistance, so your 20mOhm number seems entirely reasonable.

If it's getting hot, it's likely a timing issue (as you know). If you could test with something sensorless just to spin it up and check no-load current and compare it with the no-load of running sensored, you would have a pretty good idea where the heat is coming from.
 
Wow, that'd make one hellova Friction Drive motor! :mrgreen:

Anyone know where I can buy 3inch wide 700C tires and a bike that will fit them? :p

Sweet stuff Matt, maybe someday I will be able to afford one of your builds, I'd take a Tadpole with that motor and get lots of tickets on the freeway! :twisted: :twisted: :twisted:
 
I spoke with Mark over at Thunderstruck for a couple of hours today and it occurred to me that I may have the controller for you guys...

Its the Sevcon Gen 4 it is an ac / bldc controller that is supposed to be the cats meow in efficiency, etc. its range goes up to 116v (HVC) should pull 31 cells.

Heres a link to pics at electric Motorsport, Thunderstruck doesent have it up on their page yet. Thunderstruck went WAY out there to give me a good price, also, they said they would program it for free (The programming module is like $900 :-( Electric Motorsport wanted $100 "engineering fee" to program it.

http://www.electricmotorsport.com/store/ems_ev_parts_controllers_sevcon_gen4_72-80_550.php

If you like-ee tell them you want "Toms price"
 
recumpence said:
I plan on puting these into production once fully developed.
Can length--- 4.5 inches
Can diameter--- 3.75 inches
Shaft diameter--- 1/2 inch (keyway shafts will be used)
Expected efficiency--- 94% (more on this later)
Continuous power--- 14kw
Burst power--- 25kw
Weight--- Under 7 pounds (expected)

Basically, this is a radically redone and expanded Astro motor that we make 50% in-house and Astro winds and installs the laminations for me.

The 94% efficiency is a number Bob came up with based on the 3220 motors. Most of them peak at well over 94% efficiency. Also, this is essentially one and 1/2 3220 motors. As such, it should be good for 14kw without active cooling, or 20kw with fans (even more in short bursts). This motor is 1.5 times the power of a 3220 with 3 times the cooling surface area giving a net 50% increase in cooling area per lamination size versus the 3220.

Anyway, this is not a one time thing. I have plans to make many of these motors.
Since we are at the New Year, I thought I would bump this thread for any updates about this motor's testing & production. I know Bob retired, but the new ownership ought to be going in some direction by now... and, hopefully, they are committed to getting this motor built too? :?:

Can we learn anything new with an update, please? :wink: :twisted: 8)

Happy New Year! :mrgreen:
 
I am waiting on reports from the two prototype users I have sent motors to. Also, Astro got a huge government contract. So, they are backlogged. Plus, we just got a big Okuma CNC that we are trying to get figured out. So, our shop is backogged too.

It will happen, though. :)

Matt
 
Hi Matt,

We're also waiting for the Hall Sensor Kits, which should be good for the entire Asto 32xx model line. Combined with Jeremy's controller, which he mentioned combining with Burtie's timing module we could have a really nice solution.

It sounds like it would not be that hard to set up the controller (mostly higher powered fets) to handle the power of this new motor which would make the motor a much more affordable option.
 
I don't know how big your Okuma is, but from the pictures, they look like full-featured machines. I'm guessing with full flood coolant, tool carousel/changer?

Niiiiiice.

Katou
 
katou said:
I don't know how big your Okuma is, but from the pictures, they look like full-featured machines. I'm guessing with full flood coolant, tool carousel/changer?

Niiiiiice.

Katou

I do not know the model number off the top of my head. But, yes it has a tool changer, spindle coolant feed, and a X axis only table. The Y axis is on the head casting along with the Z. We got a crazy good deal on that mill.

Alas, I am a mere mortal. Dave is the shop owner and my CNC machinist. I am just a client of his who happens to rent space from him and give him lots of work. Hmm, I also somewhat got him started in his shop.

Maybe I should take a week or so off and learn the machines with him. :)

Matt
 
Um, yes? I would in a heartbeat. At the very least, you would learn enough to aid yourself in designing stuff to be made on that mill. I had a great deal on a manual big bridgeport clone, but a backdoor deal cut me out of the loop. Dang.

Another deal will present itself I'm sure, hopefully after I've had the opportunity to rebuild the little one-car garage from the 1940's to something a bit more suitable.

Katou
 
recumpence said:
Dave is the shop owner and my CNC machinist. I am just a client of his who happens to rent space from him and give him lots of work.
Funnily enough, that is the exact same position i'm in in 2011, renting workshop space for my new fabrication business off my CNC router mate who will be cutting out my frame parts...

recumpence said:
Maybe I should take a week or so off and learn the machines with him.
Hell yeah!!! :mrgreen:

Can't wait to see how the 2 prototypes of this motor get on.


Paul :D
 
Our progress has been relatively slow, but will be picking up considerably starting tomorrow. This will be the torture test everyone has been looking for...probably much more. I can't thank Matt enough for his help. If you're considering doing business with him for this or his gearbox, DO IT. You won't regret it.
-Cole
 
I think our Kelly is doing some funky stuff. Out of the 36 combinations, we have 24(24!!!) that spin the motor. Each of these 24 spin the motor approximately the same...We get to 4500-4600 rpm on each of those on 95V. I am thinking part of it could be the electrical rpm as well as timing. Our Kelly is supposed to have the super high speed option which gives us 100k e-rpm, but who knows. I am going to test that by machining a magnet holder that only has 2 magnets in it to trick the controller into thinking the motor is spinning a quarter of the speed it really is. If this increases our rpm, then we know what part of the problem is!

I am also doing another hall sensor mount since the last one didn't seem to be adequate. The hall sensor wires were a little more fragile than expected. This one will be a custom CNC'd circuit board that will better support the halls and allow us to solder larger wires to them. It will also allow easier timing.

Progress is slower than expected due to some very unfortunate events that I can't detail, but IS continuing.

-Cole
 
coleasterling said:
I think our Kelly is doing some funky stuff. Out of the 36 combinations, we have 24(24!!!) that spin the motor. Each of these 24 spin the motor approximately the same...We get to 4500-4600 rpm on each of those on 95V. I am thinking part of it could be the electrical rpm as well as timing. Our Kelly is supposed to have the super high speed option which gives us 100k e-rpm, but who knows. I am going to test that by machining a magnet holder that only has 2 magnets in it to trick the controller into thinking the motor is spinning a quarter of the speed it really is. If this increases our rpm, then we know what part of the problem is!

I am also doing another hall sensor mount since the last one didn't seem to be adequate. The hall sensor wires were a little more fragile than expected. This one will be a custom CNC'd circuit board that will better support the halls and allow us to solder larger wires to them. It will also allow easier timing.

Progress is slower than expected due to some very unfortunate events that I can't detail, but IS continuing.

-Cole
What is the no load amps at that rpm ? Should be less then 10amps im betting. We found that a very small timing adjustment made a huge difference on the collossus! There is a guy by the name brutie who made a soft ware base hall timing adjustment tool. So you can set a base rpm timing offset and a rpm based timing offset!
We found that if the no load Amp reading was hi this helped a lot!
 
No load current is most definitely well above 10A. We haven't bothered setting the o-scope up, but the motor gets hot really quickly. I've found that small adjustments on this thing make a big difference, as well. I have thought about burtie's software, just not sure how to implement it for our project. It seems like an AWESOME tool, though.
 
Either your timing is off or the combination is bad. I will have to look back and find it, but my no load was an amp or two on 36v. Although my KV is half of yours, you shouldn't have such a high idle current.
 
Well, assuming that the controller isn't doing something stupid, then it has to be hall timing. I don't have much confidence in Kelly, however.
 
Our hall spacing is 120 degrees and the Kelly Auto-Detects it. I have tried setting it just for 120 degrees and it performs the same.
 
Turns out it was a single defective hall that still switched alright, but whose wave(when viewed on an o-scope) was really short...as in the wavelength was half of what it should be. We got it replaced and the motor is running great! With a pretty good timing advance, we were running at 13700...in WYE!! We'll most likely be running in Delta with little to no advance because of the lower resistance, but this is GOOD. I still can't reveal project details, but I'll say there were several people who told us there was no way it would go that fast, no way it produces the power it does, and on and on...Today shut them up. Great job Matt, and great job, Astro!!!
 
Back
Top