how to build a motor test stand?

[izeman]

summer is alonost over and testing motor and controller will be hard when the sun goes down early and it's cold and rainy.
but i still want to make my lebowski controller and mac work together.
i have 3 motors left.
so i thought of building a very simple test stand. one motor turning another one working as generator. theoratically with two identical motors i could run both full power. i don't need to do it for long, just for a few seconds to see if the controller can accelerate the mac up to full speed unter load.
mechanically this should be no big challenge, but how would i consume the energy produced by the generator (motor #2)? running regen alone will not be enough as regen is only a few amps.
any ideas? the mac should run at 50v and around 45a. so 2.5kw power produced/consumed.

 

[izeman]

ooops. sorry. i didn't see that an almost identical thread was started a week ago. i'm on my iphone here.

 

[Miles]

This might help: http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45869

 

[izeman]

Omg! Thx

 

[Burtie]

Another example of a hub motor test stand... uses array of halogen lamps to dump the electrical power:

http://endless-sphere.com/forums/viewtopic.php?f=30&t=44910&start=150#p966937

 

[major]

You can power the motor controller and generator controller from the same battery. That way you do not have to worry about dissipating the load energy from the generator. The DC from the generator inverter subtracts from the DC drawn by the motor inverter from the battery. The battery ends up supplying only the losses to the system and not the load power. I've done this several times using a battery source and also a bridge rectifier off the mains. If the motor and generator are identical, you can assume equal losses in the two and calculate shaft power and therefore shaft torque accurate within a few percent.

 

[izeman]

If i power the motor from the generator i would need two controllers right? Which controller can deliver the same amount of regen current as it can supply for driving a motor? Those infineons i have only have a few amps of regen.
If i use a car engine start motor those only give 12v which would need to be up converted to 48v. Possible but maybe expensive.
This motor stand has only ONE purpose: i want to test all settings of andy's 3kw lebowski controller on the bench.
I don't care about power usage or power loss (from a cost perspective). If it's cheaper to re-use parts of the energy I'll do it.

 

[major]

If i power the motor from the generator i would need two controllers right? Which controller can deliver the same amount of regen current as it can supply for driving a motor? Those infineons i have only have a few amps of regen.
If i use a car engine start motor those only give 12v which would need to be up converted to 48v. Possible but maybe expensive.
This motor stand has only ONE purpose: i want to test all settings of andy's 3kw lebowski controller on the bench.
I don't care about power usage or power loss (from a cost perspective). If it's cheaper to re-use parts of the energy I'll do it.

Yes, you need a controller for each; motor and generator. The controllers do not have to be the same. The weaker of the two will limit your power. You can use the motor as a generator load without a controller by putting resistors across the phase leads but that doesn't give you much control on the load and 3kW is a sizable resistor load, though not undoable. As we speak I am running a battery test into resistors at 50V, 50 to 100A. Those are pretty large and if I go very long I need to put a big fan blowing on them. The car cranking motor would be difficult to use. Some are PM nowadays. If it is the old series wound motor, forgetaboutit, they make unstable generators.

 

[izeman]

But i still don't understand how I can make the generator's controller regen with full amps, equivalent to the motors current.
Normally regen is only a few amps, but i need 50 amps.

 

[izeman]

Regarding load i could buy 30x 100w resistors with the correct resistance and wire them as needed between 100 and 1000w per phase and connect them between each phase? So 3x 1000w resistance should give the right load?!

ABWE!Gold Tone Wire Wound Aluminum Shell Resistor 100w Power 10 Ohm 5%
http://www.aliexpress.com/item/ABWE-Gold-Tone-Wire-Wound-Aluminum-Shell-Resistor-100w-Power-10-Ohm-5/32405866400.html

 

[major]

But i still don't understand how I can make the generator's controller regen with full amps, equivalent to the motors current.
Normally regen is only a few amps, but i need 50 amps.

The current out of the generator's controller will always be less than the current into the motor's controller. No free lunch. I don't know about the specific controllers you have, but the ones I've worked with for EV cars can generate at or in excess of the motoring power.

 

[major]

Regarding load i could buy 30x 100w resistors with the correct resistance and wire them as needed between 100 and 1000w per phase and connect them between each phase? So 3x 1000w resistance should give the right load?!

ABWE!Gold Tone Wire Wound Aluminum Shell Resistor 100w Power 10 Ohm 5%
http://www.aliexpress.com/item/ABWE-Gold-Tone-Wire-Wound-Aluminum-Shell-Resistor-100w-Power-10-Ohm-5/32405866400.html

Yep, that would be the idea. Or you could use a 3 phase rectifier bridge and a single bank of resistors. That might be easier to adjust than 3 separate banks.

 

[izeman]

Something like that?

2015 Hot A black 3 phase 90 SQL90A a1200v diode rectifier bridge
http://www.aliexpress.com/item/2015-Hot-A-black-3-phase-90-SQL90A-a1200v-diode-rectifier-bridge/32263464630.html

 

[major]

Something like that?

2015 Hot A black 3 phase 90 SQL90A a1200v diode rectifier bridge
http://www.aliexpress.com/item/2015-Hot-A-black-3-phase-90-SQL90A-a1200v-diode-rectifier-bridge/32263464630.html

That should work.

 

[izeman]

so i did some math. i'm gonna buy 30pcs 1ohm 100watt resistors for a total of 3kw load.
to have 500w load @50v this is 10a or 5ohm resistance. 5 resistors of 1ohm in series have 5ohm. correct?
if i take two of those 5ohm strings and connect them in parallel it get 2.5ohm and 1000w for 20a load.
two of them in series makes it 1.25ohm and 2000w and 40a.
to reach 3000w i need another 1000w string in parallel to the 2000w network making it 0.83ohm and 60a.
did i calculate right? will all resistors see the same load and will it be capable of burning 3000w?

 

[major]

so i did some math. i'm gonna buy 30pcs 1ohm 100watt resistors for a total of 3kw load.
to have 500w load @50v this is 10a or 5ohm resistance. 5 resistors of 1ohm in series have 5ohm. correct?
if i take two of those 5ohm strings and connect them in parallel it get 2.5ohm and 1000w for 20a load.
two of them in series makes it 1.25ohm and 2000w and 40a.
to reach 3000w i need another 1000w string in parallel to the 2000w network making it 0.83ohm and 60a.
did i calculate right? will all resistors see the same load and will it be capable of burning 3000w?

Looks good to me. Assuming you go with the rectifier. And you may have to adjust down the number of resistors if there are voltage drops in the generator motor, diode, etc. Also, often times the rated power is that from the battery and coming off the generator rectifier will be less. Leave yourself some adjustment flexibility when wiring those resistors. Good luck.

 

[izeman]

Sure. You're right. The inducted voltage of the generator raises with rpm and will be at 50v if both motor run at the same speed.
Would it be better to use light bulbs? It shouldn't matter i guess. It's both ohm' resistors.

 

[izeman]

i thought again about it: there will be no load at all at 0rpm. how would i simulate a start from a standstill? or even worse: a start on a steep road?

 

[Lebowski]

i thought again about it: there will be no load at all at 0rpm. how would i simulate a start from a standstill? or even worse: a start on a steep road?

Maybe add some mechanical brakes for this ?

 

[teklektik]

Justin has a few posts about his dynos and measurement techniques:

• Stall measurement using load cell. This is more of a motor test, but the braking technique might be modified/applied for controller testing.
• Using bearing in place of dropout slots to measure torque directly
• Using a CA V3 to measure both motor electrical and mechanical power.
• Using load cell from a bathroom scale for torque measurement
• He has a post about feeding the regen power back into the battery but I can't seem to find it just now...

There are typically some interesting posts surrounding these, so a bit of 'nearby reading' may be helpful.

A somewhat less elegant load might be made using a car/truck alternator dumping into a load. The load current can be varied by controlling the field current with a bench supply in place of the usual regulator.

 

[izeman]

i thought again about it: there will be no load at all at 0rpm. how would i simulate a start from a standstill? or even worse: a start on a steep road?

Maybe add some mechanical brakes for this ?

thought of that as well, but the load is very hard to reproduce and i'd like to do some repeatable testing.

 

[izeman]

A somewhat less elegant load might be made using a car/truck alternator dumping into a load. The load current can be varied by controlling the field current with a bench supply in place of the usual regulator.

but wouldn't the car alternator have the same 'problem'? zero volt output at zero rpm leading to zero current?

 

[teklektik]

Ya - that's why I gave the link above to J's stall measurement technique. He mentions in there a difficulty with controller's shutting down w/o some detectable commutation, so the scenario may be a little more complicated to stage as a static measurement. You may simply need to approach zero rpm on a pull and then measure the stall torque via a second test if that particular data is important to your study...

I wasn't trying to answer the one specific question, just doing a dump of dyno-related stuff I knew about from the guy who has built more than one bench dyno...

The alternator was a means to get a continuously variable load which is problematic with the simple resistor approach... and junkyards are full of alternators...

 

[izeman]

Thanks teklektik. I appreciate your help.
To make it clearer : I'm not interested in any measurement at all.
I have problems to find settings of my controller to make the motor turn at all if there is some load.
So i need to simulate the start of the bike. Exactly that situation that has zero resistance at the moment

 

[izeman]

double post deleted