Water cooled car alternator

stepus

1 W
Joined
Sep 3, 2019
Messages
62
Hi, i have bought two water cooled car alternators, for remake a rotor with magnets. But i measured parameters like no load current, revs and BEMF. And its not bad at all. I wonder if it's not worth using as is and not remake it. Because I will have a variable "magnet strength" = variable BEMF, and variable losses. BLDC and PMSM motors have a cogging effect, and stable magnetic field of magnets doing eddy currents when free spinning. In alternator i can turn off power to magnet winding and there are only bearing losses so vehicle will slow down very long => very low losses. So I thing acceleration will be bad to compare with PMSM - I tested with my second remaked alternator on tricycle motor can handle 3kg weight on 110mm diameter with 23W consuption with magnet rotor, i tested same load with same diameter wheel, and i must put almost 200W to magnet coil, and about 5W to stator coil - i dont have a controller yet so its only approximate test.
But at cruising speed i thing its better to have a variable magnetic field, so losses can be smaller. And Variable BEMF / RPMS are also better, you dont need shifting gears, at standstill you can send 200W to magnet coil, and with speed decreasing that magnet coil power which will decrease losses in iron. BEMF is sine shaped. So what do you thing ? its better to modify it and not have any wasted energy in magnetic field or leave it as its and try it in real conditions? I think this water cooled alternator can handle 20kW of power like nothing :D
 

Attachments

  • DS4_QuickPrint4.png
    DS4_QuickPrint4.png
    25.8 KB · Views: 2,070
  • DS4_QuickPrint5.png
    DS4_QuickPrint5.png
    11.4 KB · Views: 2,070
  • IMG_1295.JPG
    IMG_1295.JPG
    316.3 KB · Views: 2,070
  • IMG_1301.JPG
    IMG_1301.JPG
    479.6 KB · Views: 2,070
  • IMG_1308.JPG
    IMG_1308.JPG
    378.5 KB · Views: 2,070
  • IMG_1309.JPG
    IMG_1309.JPG
    356.3 KB · Views: 2,070
  • IMG_1356.JPG
    IMG_1356.JPG
    329.1 KB · Views: 2,070
  • chlazeny altik.PNG
    chlazeny altik.PNG
    14.2 KB · Views: 2,070
Alternators are only around 70% efficient when used as motors due to losses from thick laminations and being optimized for low cost, not high efficiency.
You'll definitely need a radiator and coolant to keep the motor from overheating at all but the smallest power levels. A temperature sensor is a must-have as well. To power the motor, a standard e-bike controller should work if you provide the rotor current yourself. A buck converter with a constant-current output would be a good way to do it.
 
yes, i ran it with RC ESC for that measurments i done, but i thought to mount hall sensors inside alternator, and use my own ESC with HALL feedback - there i can do same test as on my tricycle where i add a little PWM, motor lifts load to air, and then decreasing PWM to find spot, where motor is still able to hold load, but draws minimum power before load will fall to ground. A small efficiency is a cost for a low price of alternator, i bought these two as damaged (both have rectifier short) for a 18 dolars, so one "motor" costs me about 9 dollar :D And efficiency isnt that bad, when i riding tricycle about 2-3kw continous power, 7kW peak, alternator is slightly warm and this is forced air cooled one.
 
stepus said:
Hi, i have bought two water cooled car alternators, for remake a rotor with magnets. But i measured parameters like no load current, revs and BEMF. And its not bad at all. I wonder if it's not worth using as is and not remake it. Because I will have a variable "magnet strength" = variable BEMF, and variable losses. BLDC and PMSM motors have a cogging effect, and stable magnetic field of magnets doing eddy currents when free spinning. In alternator i can turn off power to magnet winding and there are only bearing losses so vehicle will slow down very long => very low losses. So I thing acceleration will be bad to compare with PMSM - I tested with my second remaked alternator on tricycle motor can handle 3kg weight on 110mm diameter with 23W consuption with magnet rotor, i tested same load with same diameter wheel, and i must put almost 200W to magnet coil, and about 5W to stator coil - i dont have a controller yet so its only approximate test.
But at cruising speed i thing its better to have a variable magnetic field, so losses can be smaller. And Variable BEMF / RPMS are also better, you dont need shifting gears, at standstill you can send 200W to magnet coil, and with speed decreasing that magnet coil power which will decrease losses in iron. BEMF is sine shaped. So what do you thing ? its better to modify it and not have any wasted energy in magnetic field or leave it as its and try it in real conditions? I think this water cooled alternator can handle 20kW of power like nothing :D

Hey Stepus, i am happy that i actually saw a post of yours in ES, i saw your ebike build (and then a trike build) in YT and loved it, one of the first youtubers that actually invested to make real use of alternators as motors, and even to tackle the idea of using permanent magnet rotors for the build, i still want to see others trying that too. About your new watercooled alternators, there's something that you probably didn't realized: Your new alternators are built like those mild hybrid GM BAS, with the rotor "claws/wedge poles" already with a smaller angle and closer to the rectangular shape of BLDC rotors' magnets, and also have small ferrite (i think) magnets between the "teeth", that help with the magnetic flux as well. From the picture that shows the stator from inside, the part that lost the coating shows that the laminations are probably thinner than regular alternators, helping more with eddy currents, and since you will be able to use watercooling will be better for more power.

thorlancaster328 said:
Alternators are only around 70% efficient when used as motors due to losses from thick laminations and being optimized for low cost, not high efficiency.
You'll definitely need a radiator and coolant to keep the motor from overheating at all but the smallest power levels. A temperature sensor is a must-have as well. To power the motor, a standard e-bike controller should work if you provide the rotor current yourself. A buck converter with a constant-current output would be a good way to do it.

You're right, alternators when converted to motors are generally ~70% efficient with thick laminations, but you need to know that those alternators that he got have thinner laminations, so will be more efficient and have less eddy currents for a same level of magnertic flux. The radiator and coolant will be a bit better than some build that i saw because they don't even use forced cooling, nor checking the internal fan's rotation to see how to at least keep some airflow.

stepus said:
yes, i ran it with RC ESC for that measurments i done, but i thought to mount hall sensors inside alternator, and use my own ESC with HALL feedback - there i can do same test as on my tricycle where i add a little PWM, motor lifts load to air, and then decreasing PWM to find spot, where motor is still able to hold load, but draws minimum power before load will fall to ground. A small efficiency is a cost for a low price of alternator, i bought these two as damaged (both have rectifier short) for a 18 dolars, so one "motor" costs me about 9 dollar :D And efficiency isnt that bad, when i riding tricycle about 2-3kw continous power, 7kW peak, alternator is slightly warm and this is forced air cooled one.

Installing the hall sensors will be a very good idea, and keeping voltages up to 72V (and even 96V since you will use water cooling), those alternators are promising at least 10kW peaks for sure, i will wait and follow!!
 
Yes its a ferrite between teeths, i can measure thick of lamination when i will reopen alternator. Efficiency will never by comparable to profesional design motors, they have some simulations of fluxes softwares etc. but as DIY project with plenty of power and little heat it worth for that price :D
I will see what i will do with my all electric stuff, now i bought Sur-ron l1e, so i will make modifications on it. In my country its very stricted,what can be on roads. So i bought street legal surron. Its definitly more street legal than my tricycle, fatbike and more :D Thats a big disadvantage for this DIY stuff, you build it only for riding on your garden, and snow removal in winter :D riding it every day on roads is very risky.
 
you can buy a Belt-driven Starter/Generator motor,It is a permanent magnet synchronous motor,It is very efficient,Maximum power is greater than 15Kw
 
KAUDIO said:
you can buy a Belt-driven Starter/Generator motor,It is a permanent magnet synchronous motor,It is very efficient,Maximum power is greater than 15Kw

Buying those would be interesting to use, but you didn't told us their brand, voltage, or where its sold.
There also are other starter/generators that people already have been using:
GM's eAssist: AC Induction, Hyundai's HSG: IPM PMAC, Honda's IMA: SPM PMAC

I wanted to see people using the GM's first BAS generation, but the closest that i will get is this thread, because it looks very similar to them
 
Hi, a little update with some info.
I have 3d printed adapter for hall sensor PCB - using them on all my alternator conversions, luckily they have same pole number, Its from PC CD/DVD drive motor, so i salvaged them, and i can easy adjust timing/angle with rotating Hall PCB. Usualy i adjust timing to draw low current as possible or can retard timing to get higher RPM/torque, or a big change of angle PCB cause motor to spin other way :D
So "motor" is spinning from zero RPMs very softly without cogging but with adjusting rotor field it can also have some torque. When i spin motor to some RPMs, and turn off field and controller motor spins forever :D it has no coging when field coil isnt powered and there is only bearing losses...
Motor can be also driven without delivering power to rotor for field, but it takes some time to spin up, and it have i think a small torque.

In video iam doing same test to hold 3kg load on 110mm wheel, (alternator with magnets do this for 23W) - i was adding amps to rotor to have a stronger field looking for lowest power need on field and controller.

first test was 1.12Amp to field, and 20V@5.41A to controller so its about 3W to field coil and 108W to controller total 111W
second test was 1.81Amp to field, and 20V@2.73A to controller so its about 7W to field coil and 52W to controller total 59W
last test was 4,54Amp to field, and 20V@1.26A to controller so its about 45W to field coil and 25W to controller total 70W
https://youtu.be/m9y6H9_qcxs
https://youtu.be/gNlJZD9k63M
https://youtu.be/YpmtNa92eNM

on another video is clearly seen how long it takes to stop rotor with no power on rotor and from controller, on second video is still powered rotor and it stops in seconds. - There can be benefit
https://youtu.be/sY5VQuA1C7s
https://youtu.be/MrpXB_g8zUs
 

Attachments

  • IMG_1535.jpg
    IMG_1535.jpg
    621.7 KB · Views: 1,669
  • IMG_1537.jpg
    IMG_1537.jpg
    242.5 KB · Views: 1,669
  • IMG_1536.jpg
    IMG_1536.jpg
    305.1 KB · Views: 1,669
If you're targeting 10s of kW, 45W to the field didn't seem so much?

If you have any plan to use field weakening, having a field that you can you know... weaken... by using less power rather than have the controller apply Id seems pretty useful. Using Id to reduce the field would probably use far more power.

But maybe with permanent magnets you could get far higher field strengths thus lower kV and more torque?
 
yes 45W of additional power for motor isnt much compare to 10kW of power to windings :D but in lower power its significant percentage of power input, that rotor with magnets do not have. So this "motor" is unefficient at power like <500W and as you mentioned there can be another benefit with that variable field - controller without FW - use less energy, and less heat... it would be worth a test to compare this with some BLDC/PMSM motor at high speeds.
for output torque compare i did tests above, with my alternator rebuild with rotor with neodyms, it needs 23W to hold in air 3kgs of load. I did same tests with this "motor" and second test was with minimum power input (7W field, 52W controller) so it draws about 60W compare to 23W - with same torque as my rebuilded alternator with neodyms.
tests was done on power supply that is not lieary scalable (this is why test currents was 1.12A,1.81A and 4.54A), so i cannot increase current to rotor lineary, so i thing i can find better point where will for example field draws 15W and controller 25W so i thing "motor" can hold 3kg load with even lower power input than 60W... This can be a little playing with it to find point of where is minimum current to field and minimum to controller - there will be best efficiency...
 
Hi who can help me finding the correct synchronization stator field to Rotor.

Greetings Robert
61e6dad9eaa8f8c138ec233767a00845.jpg
bf59606096d786d321e797822ec98972.jpg


Gesendet von meinem CPH1911 mit Tapatalk


28fd1f23d02b227deb3028036071efca.jpg
 
monette999 said:
Hi who can help me finding the correct synchronization stator field to Rotor.

Greetings Robert
61e6dad9eaa8f8c138ec233767a00845.jpg
bf59606096d786d321e797822ec98972.jpg


Gesendet von meinem CPH1911 mit Tapatalk


28fd1f23d02b227deb3028036071efca.jpg

Do your alternator rotor have 12, 14, or 16 teeth?
 
monette999 said:
My rotor is 8 polepair.
8cfefa6fdb7039f0e44f9a1a940e6e71.jpg


Gesendet von meinem CPH1911 mit Tapatalk

Alright.
From what i researched (not an expert, so bear with me please), light (the space between teeth on the disk) on the photosensor will give output 1, so the spaces are north poles. You need a cheaper phone with GPS and android, go to https://supermagnete.de and search for the Pole Detector app, follow its instructions about everything, and feed the rotor electromagnet with a small battery in the same polarity that you will use. Then the pole app will tell which teeth are N and which are S
 
Okay and what am I going to do when I know where is north and south?
3ecaf79d2859b7254b32248687822848.jpg
a6d0b8b3125d78fd373efac347262ae9.jpg


Gesendet von meinem CPH1911 mit Tapatalk

 
Hi, what is your problem with sensors ? on scope photo i can see a perfect timing phase-sensor, so what is problem ? controller cannot spinup motor ?
 
My motor does not starts smooth.
In some cases the first pulse is backwards then it runs forward.
Or it get stuck and I have to turn the shaft to make it run!

Here on the scope I am measuring Phase A And Sensor A Signal.
How do I find the correct sequence for my project.

Thanks for the support
Bob
c831fc39ed63e5d71cf7bd496cc5b1a1.jpg
2cfe0ff4297e91875c5ced5d46f3ab40.jpg
b8980abcfedbbbb3b67fc12b5f0a2eb2.jpg


Gesendet von meinem CPH1911 mit Tapatalk

 
how much amps do you delivering to rotor ? Did you tried to rotate a littlebit that pcb mounted on rotor shaft? like 2degree clockwise and anticlockwise above rotor position?
When i setting timing i find position when it runs, and then run it at constant PWM dutycycle, and measuring current to controller, and littlebit adjusting timing to have smallest current draw as posible - its not good for highest torque and speed, but its efficient.
 
Hi I am running 14V 6Amp through my rotor. I have to be careful. When the rotor locked up I just need a slight turn on the rotor and it starts running. So I really have to watch my fingers.

How can I setup the Sensor correct?
Like on the scetch.

Thanks for the support.
375dc035acd2daa1f71fad619fda4bac.jpg


Gesendet von meinem CPH1911 mit Tapatalk

 
can you supply to rotor less current ? like 1-2Amps ? it can be this, alternator have also cogging effect when you power too much rotor, so controller must deliver bigger current for startup to overcome cogging. If you have right pole number sensor and alternator, and right angle between that photocouplers then it must work, did you tried disconnecting halls from controller ? or tried to look at all three sensors signals - if they are inverted etc ? did you bought that sensor or made it yourself ?
 
https://youtube.com/shorts/zE6h52T6AMw?feature=share

I have this video where you see a fail start. The trigger wheel is going backwards for 1/4 Tooth.

Greeting.

Gesendet von meinem CPH1911 mit Tapatalk

 
in video motor sounds like its having bad timing, its running hard. Did you tried to adjust timing ? Can you do this test with sensors to we will be 100% sure they are good ? Put some LEDs with resistors on sensors output and ground, and leds should light in these order like in table Hall sensors "1" means led will be ON when you slowly rotate with rotor
CommutationTable_CCW-800x226.png
 
Hi Stepus

Here my link.
https://youtu.be/m4MHyDjYqFw

Sensors work as they should, but how do I find the correct stator pair for the sensors.

Gesendet von meinem CPH1911 mit Tapatalk



 
Back
Top