ELECTRIC TRANSMISSION DELTA /WYE and SERIE/PAPALLEL stator

[rkosiorek]

here is a quick video showing the no load speed of a 2810 high-torque motor.

in WYE mode it is 41kph. in Delta the speed is 81kph but look at the current draw on the CA.

[youtube]kgqPw2sHDi0[/youtube]

rick

 

[rkosiorek]

[youtube]zLe5vtkB5CE[/youtube]

in this video i have the throttle set and i am repeatedly switching between Delta and Wye mode. i mounted 3 X 50A SPDT 12V relays inside of the motor. the controller is a stock infineon 72V 12FET that i bought from Keywin.

Delta uses a lot of current. to fully appreciate this conversion you'll want to use a high current controller.

Lucky37 - good job. nice and neat relay wiring.

rick

 

[F&P]

Hi Guys,

After already posting a thread about my Epitbike conversion with DELTA only, I then found this thread and thought someone might like my picture and video.
9C type motor with a no name 48V / 12Fet controller beefed up from 26A to 55A and with no issues, even in delta with continous high current draw. Had to also beef up cable size also.
Will do the star / delta relay switch thing soon to increase torque, but even as is with a smaller 20" wheel it really goes well.

Pitbike was free, motor and controller cost me $110AU (import 1000W motors myself for friends) and batts about $120, total = $240 all up.



[youtube]ncQCdolZ0VU[/youtube]

Matt

 

[Doctorbass]

80A Automotive Relay temperature endurance and torture test

Today for my Giant DH 24'' ( Stealth bpmber style) project i did some test with those great 80A automotive relay.

Me and many others are driving our X5 motor to extreme temperature around 140 celsius!! and more so i decided to test if those great relay can endure that ambiant temperature when they are installed inside the motor.

Since X5 can only be connected i Serie-Parallel and not Delta-Wye and that mean the relay can only be installed INDIDE THE MOTOR cause we need 10 to 12 wires from the stator to the relay.

At my job we have great thermotron to simulate diverse ambient temp for R&D projects, i used one that i decided to set to 140 celsius.

Since these relay dissipate around 1.5Watts of heat with their coil ( 12V 0.15A) I also connected them while they was insid ethe thermotron oven.

Then i connected a temperature prove to the coil to watch the coil temperature with that ambiant to 140 celsius...

and... I measured 180 celsius at the coil!!!.. I know that these coil insulation are made for up to 125 celsius ambiant so probably 150 at the coil... but they toke it no prob.

I've noted that the resistance of the coil increased while the temp increased. they started to ambiant ( 140 celsius) to 180 celsius!!! after i coonnected the coils. and when they colded down back to 25 celsius their coil resistance goes back to the nominal value :wink:

Something interesting is that the relay plastic cover had no problem hilding 140 celsius and it stayed hard enough for still holding the connector in place and not melting.

I know that my 5305 often reach 135 celsius but the motor is still not air cooled.

Probably that if i convert my 5303 to serie parallel ( 03 or 06) and that i make some nicely located holes on the cover, that will probably decrease the inside ambiant temp of the motor for probably 10 to 20 celsius .. i hope !!


140 celsius was done during 8 hours

On the pictures you can see the copper connector that had changed color (probably oxydised) and that the coil also had changed color. The plastic case also become a bit lighter than the original black color.

Nothing melted!!

Overall that's a good new that the relay was able to endure such high temp and torture test!!

Steveo might be interested by that i guess!

Doc

 

[Doctorbass]

continued...

 

[EVan]

Remember there is heat from the contact resistance as well Dr B!
However I think the part is probably up to the job. I've used and abused auto relays in extreme environments and they keep on working even when fairly melted!

And this is an extreme mod anyway..

I am just ordering my 9c today

 

[rkosiorek]

the contact current handling ability also decreases as the temp goes up. this may not be an issue for mere mortals running only a fraction of the rated current, but it may become an issue for those running the limits.

rick

 

[Doctorbass]

Yes, the contact resistance is also something i thought but i had not any high current cource for driving these contact while doing the test.

we still can aproximate what it would give because we know that the coil that is near 1.5W is increasing the inside temp of the relay of 40 celsius ( 140 to 180) and that the contact resisatnce is around 1 miliohm ( i measured that) so 1 miliohm at 30A average is around 900mW of heat...

1.5W gaved 40 aditional degree
so 0.9W would give probably 25 more degree celsius for a max of 205 celsius on the relay coil...

BUT THAT' S WITH MAXIMUM ABUSE.

Doc

 

[Doctorbass]

I will see if aircoling help decreasing the temp.

I also thought about making some holes on the relay plastic box to cool them but that could also do the opposite by let entering dust and particules and humidity in the relay and creating problem with contact resistance and making chain reaction about the degradation of the relay...

If i make holes on the motor that will probably help.. but i'm still hesitating on moding the motor i 5306 only or making that serie parallel mod...

Doc

 

[Doctorbass]

Here is a couple of pics before the entire serie parallel 03 or 06 mod!!

Please note that the holes in both side covers are in the wrong direction!!!.. some corrections will be acheived soon.

Doc

 

[Arlo1]

So does this meen I can make my x5304 a x5304/x5302 or a x5304/x5308?

 

[rkosiorek]

5304/5308

 

[Arlo1]

5304/5308

Hmmmmmmm SO double the launch off the line? I am thinking about racing it against some cars in the 100 foot dash next weekend at a big car show so....

How hard is it to dig out the series parellel wires and do this?

 

[Ariane]

Hi Guys,

Doc said that is 5303/5306 mods not 5304/5308 but my motor is X5304 and see allot of potentionel for a electric dirt bike in the future with this 5303/5306 setup. :wink:

Doc did you have cut a little bit the stator with your Dremell to fit the relay like it's seems on the picture ?

Good day!
Black Arrow

 

[Doctorbass]

Hi Guys,

Doc said that is 5303/5306 mods not 5304/5308 but my motor is X5304 and see allot of potentionel for a electric dirt bike in the future with this 5303/5306 setup. :wink:

Doc did you have cut a little bit the stator with your Dremell to fit the relay like it's seems on the picture ?

Good day!
Black Arrow

Yes i had to cut the stator aluminum armature.. but it will not affect the structure stability.

now i'm desining a PCB to hold and solder and link every relay tabs in serie parallel connection.. no teflon wire .. just big beffy trace!!

Less risk of melting wire skin, and the relay tab structure will stay in place in case where overtemp could occur.. remember i tested these relay to up to 140 celsius and the only thing i noted i sthat the tab just oxydise a bit but the plastic structure of the relay stay in place...

Here is a picture i did to help me finding out the relay tab holes on the PCM:


( Now i'm wondering if i connect al 6 relay coil in serie to get 72V needed at low current or if i connect all them in parallel for 12V needed and high current demand from the coil...

I guess that serie is better cause all coil would have the exact same current and probably will energize all at the same time making the 6 contacts synchronization better, and also helping in case where one relay would malfunction.. all other would cut and avoiding erratic combinaison of relay ON and OFF making a big BOOM on the controller with some smoke!...

Doc

 

[steveo]

Hi Guys,

Doc said that is 5303/5306 mods not 5304/5308 but my motor is X5304 and see allot of potentionel for a electric dirt bike in the future with this 5303/5306 setup. :wink:

Doc did you have cut a little bit the stator with your Dremell to fit the relay like it's seems on the picture ?

Good day!
Black Arrow

Yes i had to cut the stator aluminum armature.. but it will not affect the structure stability.

now i'm desining a PCB to hold and solder and link every relay tabs in serie parallel connection.. no teflon wire .. just big beffy trace!!

Less risk of melting wire skin, and the relay tab structure will stay in place in case where overtemp could occur.. remember i tested these relay to up to 140 celsius and the only thing i noted i sthat the tab just oxydise a bit but the plastic structure of the relay stay in place...

Here is a picture i did to help me finding out the relay tab holes on the PCM:


( Now i'm wondering if i connect al 6 relay coil in serie to get 72V needed at low current or if i connect all them in parallel for 12V needed and high current demand from the coil...

I guess that serie is better cause all coil would have the exact same current and probably will energize all at the same time making the 6 contacts synchronization better, and also helping in case where one relay would malfunction.. all other would cut and avoiding erratic combinaison of relay ON and OFF making a big BOOM on the controller with some smoke!...

Doc

Hey Doc,

Awsome job on the motor cover... can't wait to see that motor on the road!

I guess you will not have any disc brake on it eh ?

keep use posted how your gona ling the relays together.

-steveo

 

[Doctorbass]

SERIE PARALLEL DIAGRAM FOR EACH OF THE 3 PHASES

In this configuration,
When the relays are at rest (normally closed position) the phase winding are wired in parallel ( normal 5303 mode). When ground is applied to each coil (energizing the relay coils), the phases winding are wired in parallel.

 

[Doctorbass]

Here is some easy explainations about

HOW TO CONVERT:

your 5302 to a 5304
or your 5303 to a 5306
or your 5304 to a 5308
or your 5305 to a 5310
or your 5306 to a 5312

Each brushless hub motor like the crystalyte, 9 continents or golden etc have each 3 phases that are connected in STAR mode wich mean each of these 3 phases have one end that connect to the other 2 phases ( known as COMMON)

Each of these phase are made of not one big strand of cooper but many smaller size parallel strands . Ex: a 5303 have 8 strands for each of the 3 phases.

Now, if we decide to split each phase into two individual group of equal strands number, we will convert from one winding per phase, to two insulated and individual winding per phase.

And, if we connect those two winding per phase in serie, we will double the actual number of turn per pole and will by that way modify the Kv of the motor ( RPM per volt) or if you preffer the speed of the motor.

As we know a 5306 is slower but for the same current you get twice torque as a 5303.

YOU CAN NOT CONVERT ANY ORIGINAL 5304 into 5302 or 5306 into 5303 it's only one way and you can only increase the torque and decrease the speed per volt of your motor.

Power will remain the same.


You can do it permanantly or have the possibility to switch back to any mode by using relay to change the connections automaticly by energizing 6 relays that you need to fit into the motor. Unlike the Delta star, where the relay can be installed outside the motor, serie Parallel need to use 12 seperated connections and it's nearly impossible to cary all that thru the axel... so RELAY ARE INSIDE THE HUB.... IF YOU HAVE THE ROOM FOR THEM AS WELL !

Doc

 

[liveforphysics]

Don't forget two factors with delta-wye switching.


If the motor was neutrally timed when in WYE, the hall position will be either 30deg advanced, or 30deg retarded depending on the sensor location and direction of rotation.

For an equal power input, delta and wye both have equal torque potential, but the wye curve will reach it's max efficiency point sooner (at a lower speed).

 

[rkosiorek]

DoctorBass,

i like your idea of a circuit board to connect all of the relays. it will make it much easier to fit all of the wiring under the covers. worse problem i had was to tie down all of the interconnect wires flat against the hub so they would not rub against the covers.

i also wired the relay coils in series. this meant that i did not need to drop the voltage to operate the relays.

i saw a video on youtube of steveo's 5303/5306. GOOD JOB Steveo!!!
[youtube]wNiKJJa90jc[/youtube]

maybe later this year you and he should get together for a 5303/5306 shoot-out?

rick

 

[Doctorbass]

Hey Rick,

Probably not just this year.. but this more month! I just need to finalize the PCB. I might also make it availlable cause i will ask the PCB company to make more than one unit in case where some people will need to try that serie parallel :wink:

Curiously the room for the PCM is very similar to a DVD size!!.. near 121mm diameter. I cutted teh legs of these 80A relay to 3mm long and i think that will make psible to put that mod in the new gen X5 motor ( that are narrower)

The problem Steveo had is that all the wires inside the motor take too much room and some are hitting the side cover....

I have the chance to have a old gen 5303 that are wider so there is no problem.. but i preffer designing the PCB and relay assemble to fit on both motor!

These two speeds motor mod will for X5 and X400 will make a revolution as just like the 4110 fets did in our ebike community!!

I know that some people worry about buying low winding motor because they also want torque... so with that mod they will get both!

The only thing to pay attention is that WE MUST PAY ATTENTION TO THE SERIE MODE SWITCH BACK WHEN DRIVING HIGH SPEED.. THE BACK EMF OF THE MOTOR WILL BE TOO HIGH FOR THE CONTROLLER MAX VOLTAGE... SO WE MUST ALSO INCLUDE A CIRCUIT TO ABOUG SWITCHING BACK TO SERIE MODE DURING HIGH SPEED TO AVOID BLOWING THE CONTROLLER...

I remember xomeone here in this thread already built a circuit for that.. I will seach and post the link of the page...


Luke, Yes in Delta WYE the timing may vary from one mode to the next but i did not experienced any trouble in real life test. and probably it create a current spike but i did not blown any controller since. I know that in industrial DELTA WYE control box they also include a sync circuit for that 30 degree offset delay compensation.

Doc

 

[liveforphysics]

Luke, Yes in Delta WYE the timing may vary from one mode to the next but i did not experienced any trouble in real life test. and probably it create a current spike but i did not blown any controller since. I know that in industrial DELTA WYE control box they also include a sync circuit for that 30 degree offset delay compensation.

Doc

In industrial delta-wye, you're using induction motors running in slip at an RPM determined by frequency.

A 9C LOVES to be switched into delta because it happens to advance timing 30deg on top of the delta-wye switch, so you pay a tax on efficiency in exchange for the added high speed power potential.

 

[Miles]

These two speeds motor mod will for X5 and X400 will make a revolution as just like the 4110 fets did in our ebike community!!

I know that some people worry about buying low winding motor because they also want torque... so with that mod they will get both!

They won't necessarily get more torque.............. They certainly won't increase the continuous torque rating, significantly.

Now that we are seeing batteries with more than 50C discharge rates, there seems less justification for Delta/Wye switching?

What am I missing?

 

[Arlo1]

These two speeds motor mod will for X5 and X400 will make a revolution as just like the 4110 fets did in our ebike community!!

I know that some people worry about buying low winding motor because they also want torque... so with that mod they will get both!

They won't necessarily get more torque..............

Now that we are seeing batteries with more than 50C discharge rates, there seems less justification for Delta/Wye switching.

What am I missing?

The AMP limit of the controller is what determins the torque right? So 200 amps at 100 volts WYE or 200 amps at 100 volts delta would be about 1.73 x different from each other right?

 

[Arlo1]

Just though of something! The 2 speed switch meens regen will be ably to slow me at very low speeds as well!