Another RG250 with Colossus Motor

[SplinterOz]

sparkvark, thanks.

There is a noise from the bearings at about 3000rpm, even with the motor free. You can hear this more as the motor winds down due to the fact I accelerated through this range quickly so as it wouldn't stay that way under load.

I agree that these bearings may not be the best and I am going to replace all 5 in the motor over the next week or so.
Here is the internal setup of the motor: note the two sets of two bearings on the shaft and the one bearing for the skirt.

I am thinking of roller bearings for the shaft (maybe even one to replace each pair) if I can get them the right size.
Opinions welcome.

 

[Hillhater]

A nice big double row ball bearing set into the motor mount as close to the sprocket as possible ..would be the best solution.

 

[sparkvark]

A nice big double row ball bearing set into the motor mount as close to the sprocket as possible ..would be the best solution.

I agree this will allow for a much lower amount of slack. I wonder if that wound put allot of drag on the motor though? What is the expected motor rpm under load? Dose the motor vibrate a bit or dose it have a rpm were there is a stronger vibration ( other then at wide open throttle)? I dont know much of anything about this motor but I do have a good bit of experience running large rc planes on brushless motors so I would be glad to be of any help.

 

[SplinterOz]

Ok the motor revs to over 6000rpm no load... I expect to get 5000rpm under load. The bearing noise only occurs around 3000rpm. The motor gets quieter (and less vibration) over that value.

There is very little vibration even during the noisy range.

Bearing in the motor mount would not be too hard... might talk to the fabricator about putting in a recess for that.

 

[Arlo1]

I got a pack of ten of the inner bearings for 20 bucks. I will get you the number tomorow

 

[SplinterOz]

I got a pack of ten of the inner bearings for 20 bucks. I will get you the number tomorow

Thanks Arlo1 that will help. I hope to get all the stuff done this week in the evenings and have it assembled again next weekend.
Were the bearings hard to replace?

 

[Arlo1]

I got a pack of ten of the inner bearings for 20 bucks. I will get you the number tomorow

Thanks Arlo1 that will help. I hope to get all the stuff done this week in the evenings and have it assembled again next weekend.
Were the bearings hard to replace?

No they just hammer out and hammer in lol. They are a very common number 6004 i think and they are used in motorcycles for wheel bearings. I found them in a pack of 10 for 15 pluss tax from a motorcycle parts catalog at work!
You can find the outer bearing on ebay as well!

 

[markobetti]

It has been a very productive day.

I a 13 tooth front and a 54 rear... so about 4.15 ratio. Rear wheel is 196.5cm circumference approx 62.5cm diameter. So that would give a speed of 142Km/h at 5000 rpm. Probably don't have the right rear sprocket.... however the sprocket is from the KLR range of bikes and I can get them up to 72 teeth.

The chain in place and the motor is spinning up the wheel. Note based on the above I was pulling approx 5400 so the rear wheel speed was approx 153 Km/h (95 MPH). This was pulling 8.6 amps for the full speed no load test to the back wheel so about an extra 3 amps (228 watts at 76 volts) of loss due to chain, bearings and any brake drag. Seems to be reasonable to me.
[youtube]6jaJ9IP9keg[/youtube]

I JUST CAME BACK ... WOw Splinter , really wow.... I cant wait to see the load and working motor..... ::

 

[SplinterOz]

I JUST CAME BACK ... WOw Splinter , really wow.... I cant wait to see the load and working motor..... ::

Thanks... it is working really.. really well. Hopefully I will be able to put some load on it next weekend (damn this day job) then we will see what it it made of.

 

[Burtie]

Hey Splinter,
This is all looking pretty good!
I am keeping my fingers crossed that your controller holds up ok during the load testing.

I wonder if it might be an idea to run it at a reduced voltage for the initial, load test so it is not so stressed. Then move up to the intended voltage after you have aquired a few test results from the bike.

 

[SplinterOz]

Hey Splinter,
This is all looking pretty good!
I am keeping my fingers crossed that your controller holds up ok during the load testing.

I wonder if it might be an idea to run it at a reduced voltage for the initial, load test so it is not so stressed. Then move up to the intended voltage after you have aquired a few test results from the bike.

Burtie, agreed... I have a battery pack that will put out 1/3 of the planned voltage (so 8S) set up I will have to box it up somehow to protect it during testing. That should give me some safety margin.

 

[dangerzone]

considered the big sevcon brushless controllers? expensive and reasonably large but good for 30kW+ continuously

eg http://www.electricmotorsport.com/store/ems_ev_parts_controllers_sevcon_gen4_36-48_650.php

apologies if this has been considered/discussed before

I was also wondering whether the Colossus could use all the power of a G8055, but that seems a bit too much... Aren't there any cheaper controllers which use the same FETs and have similar specifications..?

 

[Arlo1]

considered the big sevcon brushless controllers? expensive and reasonably large but good for 30kW+ continuously

eg http://www.electricmotorsport.com/store/ems_ev_parts_controllers_sevcon_gen4_36-48_650.php

apologies if this has been considered/discussed before

I was also wondering whether the Colossus could use all the power of a G8055, but that seems a bit too much... Aren't there any cheaper controllers which use the same FETs and have similar specifications..?

Sevcon makes smaller controlers!
But I am very confident Collossus can take all the G8055 can put out.
The sevcon controlers mesure phase current so they limit the low end torque quite a bit and they use 150 volt fets so the rdson is high in comparison to other 100 volt fets!

 

[SplinterOz]

If I were to get a Sevcon controller then one of these would do
[pre]Model Nominal Operating Current Boost
G8 72-80 39.1-116
G8018 180 215
G8035 350 420
G8055 550 660[/pre]

Now the middle one would do me fine... BUT and it is a big BUT...
The middle one is priced at 925USD based on the link and they Highly recommend buying the programming device.. 900USD....

So that is now 1825USD for a controller for my motorbike. Now to get the data info out of the system I now need a CAN set of gauges... not sure how much.

So are they THAT much better? If I get one can a not get the programming interface?

 

[dangerzone]

considered the big sevcon brushless controllers? expensive and reasonably large but good for 30kW+ continuously

eg http://www.electricmotorsport.com/store/ems_ev_parts_controllers_sevcon_gen4_36-48_650.php

apologies if this has been considered/discussed before

I was also wondering whether the Colossus could use all the power of a G8055, but that seems a bit too much... Aren't there any cheaper controllers which use the same FETs and have similar specifications..?

Sevcon makes smaller controlers!
But I am very confident Collossus can take all the G8055 can put out.
The sevcon controlers mesure phase current so they limit the low end torque quite a bit and they use 150 volt fets so the rdson is high in comparison to other 100 volt fets!

Well, I am not worried about the voltage but am worried whether the current of 660 A might fry the little Collie even if it is for just 10 seconds..?

Cause such a controller as this G8055 with 220A continuous, 550 bursts for 60 seconds (just right for that tight corner 'twist of the wrist' fun ) and maaaybeee the 660A for 10 seconds might be just similar to the RG500 ICE output. There is a problem with RPM, does anyone know if the G8055 could handle the Collie spinning so fast at 80V..?

 

[dangerzone]

If I were to get a Sevcon controller then one of these would do
[pre]Model Nominal Operating Current Boost
G8 72-80 39.1-116
G8018 180 215
G8035 350 420
G8055 550 660[/pre]

Now the middle one would do me fine... BUT and it is a big BUT...
The middle one is priced at 925USD based on the link and they Highly recommend buying the programming device.. 900USD....

So that is now 1825USD for a controller for my motorbike. Now to get the data info out of the system I now need a CAN set of gauges... not sure how much.

So are they THAT much better? If I get one can a not get the programming interface?

Yeah, that is quite pricy for any controller. But on the other hand I think that with those specs they are way better than any 36FET conversion we could ever make. I would love to see a 36FET monster handle some 220A continuos and 550 peak but that seems just a bit out of its league. Didn't someone from ES consider the idea of making a controller using the same FETs as the Sevcon controllers..?

Splinter, great work, respect! Am really happy seeing you working on this idea...

 

[voicecoils]

If I were to get a Sevcon controller then one of these would do
<snip>
Now the middle one would do me fine... BUT and it is a big BUT...
The middle one is priced at 925USD based on the link and they Highly recommend buying the programming device.. 900USD....

So that is now 1825USD for a controller for my motorbike. Now to get the data info out of the system I now need a CAN set of gauges... not sure how much.

So are they THAT much better? If I get one can a not get the programming interface?

I would call some of the retailers and see if any rent the programming interface. It would not surprise me if some would.

The appeal to me of these controllers is that they have high power and should be very reliable. If they did fail, they could be replaced under warranty. I prefer an oversized controller to an undersized one. Especially if they can be programmed.

People seem to have good luck with the Kelly brushless controllers too and they're probably cheaper but you do hear about people popping them from time to time.

 

[Arlo1]

Talk to luke (liveforphysics)
He has tested one of these controlers.
He might have a interface for you.
If I was you I wouldnt worry about Collossus because luke said the phase limit makes these controlers slow Its 7 150v TO220c FETs in parrellel, 42 FETs total.
But 150v FETs are about half s 100v FET for current flow, so its not super good at Phase current!
The cheep China controlers we use just moniter battery current so off the line they have huge phase current with a 100 volt fet!!
Im going to order the biggest they have the G8055 and its for a smaller bike!

 

[dangerzone]

Talk to luke (liveforphysics)
He has tested one of these controlers.
He might have a interface for you.
If I was you I wouldnt worry about Collossus because luke said the phase limit makes these controlers slow Its 7 150v TO220c FETs in parrellel, 42 FETs total.
But 150v FETs are about half s 100v FET for current flow, so its not super good at Phase current!
The cheep China controlers we use just moniter battery current so off the line they have huge phase current with a 100 volt fet!!
Im going to order the biggest they have the G8055 and its for a smaller bike!

You know, don't call me crazy but I am also considering the idea of getting the G8055 if the Collie can stand the power... Can their interface control the motor at 6000rpm, are there any rpm limits or other limits..?

 

[SplinterOz]

Ok working on the 12volt setup this weekend (waiting bearings and a good bearing puller).
No photos as it does not look too interesting, however I am looking at battery sizes for the 12 volt and would like some opinions.

In Australia the Australian Design Rules (ADR) it states:
"An independent auxiliary (ELV) must be supplied to run safety equipement... in the event of a shutdown of the main battery system.
...
The auxiliary supply must be capable of operating the hazard lights (four-way flashers) as a normal duty cycle, for a minimum period of 20 continuous minutes."

Now I have 6 led indicators (front, rear and in the mirrors), LED tail light, LED running light at the front. Based on that I picked up a small sealed Lead Acid Battery...
[pre]Century PS1212 12V 1.2Ah Sealed Lead-Acid Battery Dimensions: L97 x W47 x H51 mm Weight: 550g[/pre]
I have just run my 4 indicators, tail and brake lights (don't have the front running light up yet) for one hour and the battery had dropped to about 35% capacity left.

So this will meet the ADR's... however will this battery cope during normal operation when I have the DC-DC converter driving 5 amps across it to run the headlights etc.

 

[jonescg]

If I recall, DC/DC converters aren't very efficient. In fact, a giant 50 W resistor would probably do the same job at similar efficiency. I need to find a decent 12 V axillary for my bike too. Was thinking of a 4S2P setup with A123 cells and charge it separately with my balance charger.

 

[SplinterOz]

I thought about that... but was concerned I might run out of auxiliary battery then not have the contractor function. I am getting one of these
http://www.current-logic.com/shop/index.php?main_page=product_info&cPath=10_20&products_id=176
as my convertor, it is meant to be 80% efficient and is fully isolated (required by ADR).

Really I only need the battery as the emergency lights... if I have a traction battery failure. I was thinking about two 9 volt batteries for a while

 

[Arlo1]

Methods tells everyone to get laptop power supplies!

 

[jonescg]

They take AC input. But will they take 100 V DC input?

 

[Arlo1]

They take AC input. But will they take 100 V DC input?

Yeh he claims they just run all the power through one half the diodes!
I have a dead laptop with a great powersupply I will try is as soon as I can but Im sure its fine!