Workshop Dynamometer using BLDC motor

Irving

Joined Jan 30, 2016
3,843
The more I think on this, given the situation, the more I think @LowQCab is right. It is a simpler approach. The question is now how to instrument it to give you the output needed. On at least 1 of the stator windings you need to measure current and generated voltage; ideally do it on all 3 since they wont be perfectly identical. The easiest way to measure the current is with a 75mV shunt, eg https://www.ebay.co.uk/itm/164828952645 inserted in one of the shorting link

1637831822414.png
You also need to measure the voltage generated, from the centre join to the shorted connection, like this...

alternator.png

Then you can measure current & voltage across the cycle and calculate Power Out. From that and revs you can calculate torque load.

But you will need to calibrate it,. One way is with a pulley and dropped weight.
 

Irving

Joined Jan 30, 2016
3,843
Ah, yes of course, I'd forgotten that option. I once built an eddy current dyno for small to medium electric motors, using a 20cm (8") spinning aluminium rotor and 2 large Chinese 50kg pull electromagnets mounted on a concentrically pivoted 0.5m arm and a force gauge.

Come to think on it, that's probably an option here 'cos the motors I was testing then were 5 - 8Nm stepper and servo motors.
 

shortbus

Joined Sep 30, 2009
10,045
I want to eventually be able to do repeatable transient run
And that can be done with the prony if you use a little imagination. Using a stepper motor to control the application of the friction on the motor shaft.

Using an alternator like is being suggested is opening another whole can of worms for you.

The alternator in a car is over driven, meaning that you have a ~8" pulley on the crankshaft and ~2 1/2" on the alternator. This is there to get the alternator to give an output at lower engine speed, like around the town driving. You will need some sort of belt or other drive from your motor to the alternator, and have to figure any slippage and power consumption into your test numbers.

With that done you have to be aware that alternator output isn't linear. Quote: " alternators have greater output at higher speeds. This increase in output is not linear but is actually a curve. This means that as rpm changes, the alternator will increase and decrease its output at different rates. " From - https://www.motortrend.com/how-to/0206sr-understanding-your-alternator/ So you will also need to factor that curve into your results.

Where as a prony brake doesn't have any of those constraints. I know I keep bringing up the prony brake, but that is because it's still the simplest way of doing what you want. Its worked for 200 years, this 2021 is it's 200th birthday after all. just modernize it with a little electronics on the scale and power application and make life easy.
 

LowQCab

Joined Nov 6, 2012
4,023
""
Using an alternator like is being suggested is opening another whole can of worms for you.

The alternator in a car is over driven, meaning that you have a ~8" pulley on the crankshaft and ~2 1/2" on the alternator. This is there to get the alternator to give an output at lower engine speed, like around the town driving. You will need some sort of belt or other drive from your motor to the alternator, and have to figure any slippage and power consumption into your test numbers.

With that done you have to be aware that alternator output isn't linear. Quote: " alternators have greater output at higher speeds. This increase in output is not linear but is actually a curve. This means that as rpm changes, the alternator will increase and decrease its output at different rates. " From - https://www.motortrend.com/how-to/0206sr-understanding-your-alternator/ So you will also need to factor that curve into your results. ""

You evidently are missing a variety of factors .........
When using a Mechanical-Friction-Brake, You still have to dissipate the same ~1640-Watts of Heat.
This could probably be done with a Disc-Brake off of a small Motorcycle, along with a Cooling-Fan.

Most Fiction-Brakes will require more or less pressure to generate the same Torque-Value
as the Temperatures increase with time, and the amount of Power absorbed.
A Water-Pump-Brake would be substantially more consistent than a Friction-Brake,
That's why they use them in full-scale, Commercial-Dynomometers,
although, some of the best "High-End" Dyno manufacturers currently use
Alternators instead of a Water-Brake.

https://dynomitedyno.com/tech-corner/absorber-comparison/

Nobody uses a Leather-Belt, wrapped around a steel drum, which is what a "Prony-Brake" is.

Your quote regarding "Alternator-Output" flies out the window when running the Alternator in
a "shorted" condition, rather than the Alternator trying to maintain a fixed Output-Voltage.
The shorted Windings make the Alternator much more linear, and in any case,
the desired outcome is NOT Voltage-Generation, but Torque-Generation.

An Alternator with shorted Windings will generate its maximum Torque at speeds as low as ~1000-RPM.
The Engine in question here generates its maximum Torque at around ~4000-RPM,
and generates its maximum Horsepower at around ~6000-RPM,
and makes useful Power up to around ~8000-RPM.
The Engine Idles between ~1000 and 1500-RPM,
at that speed the Alternator will easily produce enough Torque to stall the Engine,
so no increase in RPM with Pulleys or Gears is necessary at all.
The average Alternator can comfortably operate at ~20,000-RPM without damage.

No separate Brake actuating devices and associated Linkages, Motors, Gearbox, etc. are required.
just a standard Bench-Power-Supply, and an accurate Electronic-Scale, or Electronic-Pressure-Transducer.

The Torque generated responds to the Voltage applied to the Rotor-Winding fast enough for
an automatic, "Closed-Loop", feed-back scheme to be used for real-time-RPM-control.

The Voltage or Current that is generated is completely irrelevant, and does not need to be measured.
The only factor that should be monitored is Temperature,
which is the case no matter what means of generating Braking-Torque is employed.

1640-Watts is not a trivial amount of Heat to be handled,
it requires a large mass,
and a large surface-area for Heat-Dissipation,
and a Cooling-Fan to remove the excess Heat.
.
.
.
 

Irving

Joined Jan 30, 2016
3,843
1640-Watts is not a trivial amount of Heat to be handled,
it requires a large mass,
and a large surface-area for Heat-Dissipation,
and a Cooling-Fan to remove the excess Heat.
That's one reason I like the eddy current dyno. The spinning ally disc is it's own heatsink.

Friction brakes always have problems as the friction coefficient varies with temperature and it's hard to maintain a constant load.

As you say, electromagnetic brakes have an almost linear load to excitation relationship and a fast enough response to be part of a closed loop setup.

My only concern with the alternator approach is the stability of the rotor winding at high rpm. As I have no experience of that I'll take your word on it. The eddy current brake is easy to make and a spinning disc has few issues with high rotation speeds (and it's easy to calculate size to be sure).

This is a good read:
https://mechanicalengineering.blog/dynamometer-introduction-types/
 
Last edited:

LowQCab

Joined Nov 6, 2012
4,023
Eddy-Current-Brakes are the superior solution, no doubt,
but will require a substantial outlay of Cash, and or, Precision-Machine-Work-Construction.

Small Eddy-Current-Brakes are readily available commercially, but are quite expensive,
a Car-Alternator can be obtained from an Automotive-Junk-Yard for less than ~$20.oo, and will do the job.
.
.
.
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
Hi all thanks for all the help, looks like I've got a few options to try. I'll try source a car alternator and see if it works. I've also managed to find an old 2.2kW 3phase AC induction motor in the university basement so will try putting some DC across that for some braking as well. Hopefully one of the ideas works!
 

shortbus

Joined Sep 30, 2009
10,045
The Engine Idles between ~1000 and 1500-RPM,
at that speed the Alternator will easily produce enough Torque to stall the Engine,
So your saying that a ~1600W load will stall a 8KW motor?

a Car-Alternator can be obtained from an Automotive-Junk-Yard for less than ~$20.oo, and will do the job.
.
The only alternator you will get at a junk yard, or at least one around here, is one that is burned out, and even they are worth more than that as rebuild cores. I have a friend that has a business buying andselling core and they are worth to him much more than $20.
,
,
,
,
,
,
,
 

shortbus

Joined Sep 30, 2009
10,045
Your quote regarding "Alternator-Output" flies out the window when running the Alternator in
a "shorted" condition, rather than the Alternator trying to maintain a fixed Output-Voltage.
The shorted Windings make the Alternator much more linear, and in any case,
the desired outcome is NOT Voltage-Generation, but Torque-Generation.
You still need a load cell or scale to measure that torque. So that in effect still makes it a Prony brake.
,
,
,
,
,
,
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
So your saying that a ~1600W load will stall a 8KW motor?



The only alternator you will get at a junk yard, or at least one around here, is one that is burned out, and even they are worth more than that as rebuild cores. I have a friend that has a business buying andselling core and they are worth to him much more than $20.
,
,
,
,
,
,
,
It doesnt need to stall 8kW my engine only goes up to 1.6kW.

I've already found a few alternators for around £30-£50 here in the UK which is within my budget so should be fine
 

shortbus

Joined Sep 30, 2009
10,045
I've already found a few alternators for around £30-£50 here in the UK which is within my budget so should be fine
All of the talk of shorting alternator windings may work out for you, but from my years of working on cars, when I've found an alternator with a shorted winding you know that it is shorted because the insulation on that winding is burned off, due to the heat in it.

Exactly how do you plan on measuring the torque from the shorted alternator? Or for that matter how did you plan to do it if you still used the original BLDC motor?
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
All of the talk of shorting alternator windings may work out for you, but from my years of working on cars, when I've found an alternator with a shorted winding you know that it is shorted because the insulation on that winding is burned off, due to the heat in it.

Exactly how do you plan on measuring the torque from the shorted alternator? Or for that matter how did you plan to do it if you still used the original BLDC motor?
I wont be doing long tests and I have a powerful fan to cool it. I was also thinking of shorting it to some external load that could be more easily cooled if necessary.

Measuring the torque should be simple, I will just use a lever arm and a set of scales and some form of vibration damping.
 

LowQCab

Joined Nov 6, 2012
4,023
"" I wont be doing long tests and I have a powerful fan to cool it.
I was also thinking of shorting it to some external load that could be more easily cooled if necessary. ""


Most Alternators have a built-in Fan that works pretty well.

Using an external Electrical-Load will reduce the maximum Torque that the Alternator can generate.

Starting from room-temperature,
and using the Engine that You described, running at wide-open Throttle,
You will probably have a good ~5-minutes of heavy-Load run-time before You
need to start being concerned about over-heating a large-frame Alternator with its original Fan.

An Alternator that has developed a damaged Winding,
which eventually became an internal-Short-Circuit between individual turns of Wire in one of the Coils,
is an entirely different scenario than purposefully connecting the Output-Wires of all 3-Windings.

Usually, a burnt-Winding in an Alternator starts out as a failed Diode in the Diode-Pack,
( "3-Phase-Bridge-Rectifier" is the proper description or term ),
and then continuing to be run for days or months,
with no one realizing that there was actually a problem with the Alternator,
until one day the Battery is finally so low on Charge that the Engine won't start.

The Windings can sometimes get so HOT that it will melt a Soldered-Wire-Spice, ( ~750F ),
and still not suffer any debilitating degradation of the Coil's-Insulation.
For this reason, mechanical "Crimp-Splices" should be used in an Alternator, not Soldered-Splices.
.
.
.
 

shortbus

Joined Sep 30, 2009
10,045
I wont be doing long tests and I have a powerful fan to cool it. I was also thinking of shorting it to some external load that could be more easily cooled if necessary.

Measuring the torque should be simple, I will just use a lever arm and a set of scales and some form of vibration damping.
I was also going to ask if you have ever seen them do a dyno test? It only lasts long enough to get the different torque levels on the RPM points they decide on. Usually less than around 5 seconds.

If you are still going to use a lever and scale you still have in effect a Prony brake, sorry but that's the fact. They don't normally use anything but a friction prony on small engines like this. Bigger high HP engines need the hydraulic or eddy current brakes because they would over come a friction brake.

I worked with a couple of guys that often enter something called the "Engine Masters Challenge" and have gone to a couple as a observer.
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
I was also going to ask if you have ever seen them do a dyno test? It only lasts long enough to get the different torque levels on the RPM points they decide on. Usually less than around 5 seconds.

If you are still going to use a lever and scale you still have in effect a Prony brake, sorry but that's the fact. They don't normally use anything but a friction prony on small engines like this. Bigger high HP engines need the hydraulic or eddy current brakes because they would over come a friction brake.

I worked with a couple of guys that often enter something called the "Engine Masters Challenge" and have gone to a couple as a observer.
Yeah ive worked with full scale dynos for a while now and nearly every dynamometer uses a load cell to measure torque, i knew what a prony brake was and thats not the hard bit. I want to do more than just measure torque, i want to map the ECU which will require the engine to be accurately held at a desired rpm for at least a few minutes. I also then want to simulate how the engine would accelerate through the rev ranges, with varying load to simualte how drag increases with speed and measure fuel consumption . All of this would be much easier with an electrically controlled dynamometer and not a friction brake. I may resort to a friction brake if i cant this to work but it wont be ideal
 

LowQCab

Joined Nov 6, 2012
4,023
The Carburetor that comes on most small 2-Strokes is usually not well suited to "part-throttle" operation.
A Micro-Squirt Fuel-Injection-Computer will do an excellent job though.
You may want to stabilize things by using Timing-Retard for the Idle through mid-throttle ranges,
much the same as the arrangement that older 2-Stroke Out-Board-Motors used to use.
This will alleviate the need to run in "4-Stroking-Mode" in the lowest RPM ranges.
But this will also cause Engine-Vacuum to be very low, reducing the Mapping-resolution and
possibly necessitating the use of a mixture of "Speed-Density" and "TPS/RPM" Load-Mapping.
.
.
.
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
The Carburetor that comes on most small 2-Strokes is usually not well suited to "part-throttle" operation.
A Micro-Squirt Fuel-Injection-Computer will do an excellent job though.
You may want to stabilize things by using Timing-Retard for the Idle through mid-throttle ranges,
much the same as the arrangement that older 2-Stroke Out-Board-Motors used to use.
This will alleviate the need to run in "4-Stroking-Mode" in the lowest RPM ranges.
But this will also cause Engine-Vacuum to be very low, reducing the Mapping-resolution and
possibly necessitating the use of a mixture of "Speed-Density" and "TPS/RPM" Load-Mapping.
.
.
.
Its ok it is a 4 stroke, and ive already converted it to use fuel injection
 

shortbus

Joined Sep 30, 2009
10,045
I also then want to simulate how the engine would accelerate through the rev ranges, with varying load to simualte how drag increases with speed
That type of information should have been presented much earlier in the thread. so then if you go with a modified alternator how are you going to know what rotor voltages present what amount of load on the alternator? You do understand that rotor voltage is what will control the load? LowQ suggested that the rotor will take the full 12V, but in reality and in a vehicle they never see that high a voltage, so watch to see if there is a heat problem on the rotor.
 

Thread Starter

ElNormo123

Joined Nov 17, 2021
24
That type of information should have been presented much earlier in the thread. so then if you go with a modified alternator how are you going to know what rotor voltages present what amount of load on the alternator? You do understand that rotor voltage is what will control the load? LowQ suggested that the rotor will take the full 12V, but in reality and in a vehicle they never see that high a voltage, so watch to see if there is a heat problem on the rotor.
I said i need a motor that can hold the engine at all speeds. The control side wasnt really what i needed answering Ive already got a PID system that can control the voltage accordingly depending on the measured speed. Ive got it to work on a similar project with lower power but the principle should be the same.
Thanks for the help but like i said, a friction brake isnt what im looking for. Might be a good backup though
 
Top