PDA

View Full Version : Dyno's for mini Z?


ScoobyPete
2011.02.03, 12:44 PM
I have been looking into a Dyno for our club to use with mini-z's and it seems that there is only really DIY projects or $2000 rolling roads. Basicly we want to see the diff between motors.

All the links included download sites for the software.

The 1st one I found was this one.
http://www.sci-spot.com/Mechanical/dyno.htm

Its made the PCB from a Mouse it uses the ball movement sensors to work out the specs. It's limited to 3000rpm though so would need a gearbox to reduce the RPM.

http://www.sci-spot.com/images/Dyno/interruptclose.jpg

Another one I have found is a tradtional rolling road type, this uses a coil and magnet to wrok out the specs.

(These are all the same dyno just built by 3 diff people.)

http://sites.google.com/site/simpledyno/the-dyno

http://wn.com/Build_Your_Own_RC_Car_Dyno

http://www.msuk-forum.co.uk/topic/81...built-rc-dyno/

http://www.rccarclub.co.uk/images/car.jpg


We've had some advice in regards to the 1st mouse based Dynometer from Reaper about a system to work out Torque.
with your set-up RPM will probably be achievable
torque is going to be a problem
for an inertia type dyno you will want a motor that you know the ratings of , you use this one as a slave
connect your motor that you want to test directly to your "slave" motor and spin it up wile measuring the output of the slave, you may also need to run this setup with a flywheel to determine power-up and run-down speeds and torque
add all this to some heavy maths and your somewhere close

Anyone able to offer advice or even a details of one you have made?

arch2b
2011.02.03, 12:48 PM
the tamiya speed checker (http://mini-zracer.com/forums/showthread.php?t=11729&highlight=tamiya+rpm) is the only thing i know of and i'ts not really accurate as it only determines speed, not speed + torque. it may be a good base for a custom dyno though.

ScoobyPete
2011.02.03, 12:53 PM
Thanks Arch, I should have mentioned we already have the Tamyia Speed Cheecker :) and as you say its not really accurate as it only determines speed and that isnt all that accurate either. We want a bit more detail than just speed :(

DamoRC
2011.02.05, 03:45 PM
Hi,

I tried to build a rolling road type dyno for my Radline some time back. Initially I tried using the mouse based approach using the site you have linked to above. I could never get it to work just the way I wanted and the mouse I was using couldn't handle rpm over about 3000.

So I decided to build my own detection system (initially the coil magnet approach) and write my own software (which is the SimpleDyno application you also referred to above)

I should point out that I have seen a couple of projects described where the mouse based approach was successful, one of which was a system for testing slot car motors - so it does work and may be worth a try.

In terms of advice, it depends on what you want to do. Do you want to test the motors "in car" (like a rolling road) or "out of car"?

You should not use a CD based flywheel if you are building a motor dyno - they shatter typically at around 23,000 - 25,000 rpm (although a few glued together might handle higher rpm)

There is no issue with calculating torque using an inertia type dyno. You don't need a slave motor setup to get these numbers. A slave motor approach could be used if you want to build an absorption type dyno. However, these can be more complicated and, using a brushed motor as a slave, you will face problems with different outputs based on brush wear and motor temp.

If you are building a rolling road type dyno I would advise using a "one roller per axle" approach. They are simpler to build and easier to operate because you don't have to ensure that both rollers sitting under a single axle are being powered equally (which can have a big impact on the numbers you get).

For good reference numbers, AtomicMods have a nice "motor comparison" section at their site.

Hope this helps and good luck.

DamoRC

ScoobyPete
2011.02.07, 07:01 PM
Thanks very much for your reply. We haven't had a chance to try the software yet. Great work in writing it though. Well beyond any of my skills.

Thanks for the pointer to the comparison chart, never knew that was there.
Thats pretty much exactly what we want to see. We need to build some kind of rig to do that. If we built a robust rig with the mouse option using 2 or 3 cd's, and with 98% of the testing PN70 HFAY and similar spec motors. Do you think the PN70T would shatter the cd's? And what do you think the Atomic would have to use?

DamoRC
2011.02.07, 11:19 PM
I am not sure that I can answer your question. The AtomicMods numbers suggest that most of these types of motors will go to 25-30k RPM under their test conditions, beyond the shatter point of a single CD. With 2 or 3 appropriately stuck together they may tolerate higher RPM. You could just try it and make sure that you cover the CDs with a safety cage or box during runs.

ScoobyPete
2011.02.08, 04:33 AM
We have decided to try and gear the rig down by using a spur gear on the shaft and a pinion on the motor to drive the rig. This will reduce the rpm on the CD. I guess we'll just have do it by trial and lots of errors :)

Thanks Again :)

DamoRC
2011.02.08, 10:03 AM
An amusing "study" (http://www.paintbug.com/cdexplode/) on the rpm limits of cd's

Team Nissmo
2011.02.25, 11:18 AM
Finished a proto type:cool:
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/dyno%20small.JPG

Test results of 3 PN70`s
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/pn%2070%20rpm.JPG
Motor 1 True peak RPM: 25125.5
Motor 2 True peak RPM: 25148.75
Motor 3 True peak RPM: 24590.75
Power:
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/dyno%20power.JPG
Torque:
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/dyno%20torque.JPG

ScoobyPete
2011.02.26, 08:40 AM
Please bear in mind this is just a Protype and the results of the 1st time we got it running in any form.

First of we would like to thank DamoRC very a very useful piece of Software.

As we have used a different set up for the rig we have guessed at the settings for the sensor set up. So the Results will not be accurate at this stage. Our thinking is that with the same settings used for all tests the results will still be consistent and at this stage that is enough untill we can refine the rig and work out the proper maths for the sensor settings.

.
The RPM figure is scaled down as the sensor only has a 7500 limit so we put the resistor across the motor to slow it down. We then worked out the ration of the pinion and Spur gear (3.875)and multiplied the result by that to give a true RPM figure.

The power and torque settings are the least accurate as we don't know what values the driven motor provides as a substitute flywheel. We put some values in tweaking them till we got constant power results for a know motor like Atomic Stock R's.

Now to work out how work excel :D

DamoRC
2011.02.26, 02:37 PM
Sorry Scooby Pete, but to clarify, I'm not responsible for the sci-spot dynamomter software (the one that uses the mouse scroll wheel). My application is the SimpleDyno software which will not work with the mouse scroll wheel setup.

Thanks anyway.

DamoRC

ScoobyPete
2011.02.27, 02:40 PM
No worries we'll thank you for some inspiration then :)

DamoRC
2011.03.01, 12:06 AM
Okay - I'll take some marks for inspiration.

I have a few comments / observations on the data presented.

First, I think that the system that you have will be useful for motor-to-motor comparison (or pre-tweak-post-tweak modifications). How precise are the data? (precision being the ability to get the same answer time after time). To truly compare the motors, each one should be tested a minimum of 2 times, ideally three so you can see what the variability run-to-run looks like. Then you will have an idea as to what kind of differences you should be able to detect motor-to-motor.

I am not sure that you are ever going to be able to tweak the input settings on the software to ever get accurate (accuracy being how close you are to the truth) results for power and torque. The reason I say this is that the calculations that an inertial dyno software application will perform are based on assumptions of a constant moment of inertia flywheel.

In your applications, a slave motor with a resistor across the terminals represents a variable load so the response will not be linear (like having a flywheel that gets heavier as it spins faster). This is not a problem though as precision is more important than accuracy for your application (although now that you have decided to use a geared down approach, you are safely back in CD flywheel territory if you wanted to try it).

Alternatively, you could build a calibration curve for your slave motor (run at different RPM, measure the voltage coming out of the slave at each RPM, calculate the power dissipated through the resistor using ohms law at each RPM, build a curve that fits these data, test your driven motor, use the curve to calculate the power at each RPM, assume 100% efficiency therefore power dissipated by the resistor = mechanical power of the drive motor = torque by RPM (in radians per second) of the motor under test) or something like that.

Anyway, it's nice to see the data - I look forward to seeing more.

DamoRC

ScoobyPete
2011.03.01, 04:36 AM
Yeah these are things we have thought about, as we are trying to fool the software into thinking the motor represesnts a Flywheel with a set mass.
The main fault with this is that the nmotor reaches it maximum Torque an power almost straight way so all the power and Torque curves are much narrower than they should be.

As I said before though if all the motors are using the same rig with the same variables then all the results will have the erros in them.

This more a case of comparing the motors than getting a true reading, we can atleast see what the curves look like for each motor.

I think we need to build a rig with the CD Flywheel with the pionoin and Spur gear to gear it down for accurate results.

Here are some of the results we have produced using the rig, it's still in development though.

We are getting pretty consistant results but the PN 70 motors look slower than the kyosho stock:confused:
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/dyno%20tests.JPG

http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/rpm.JPG

I have just rebuilt an Atomic stock R that was known to of lost power
http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/re%20build.JPG

DamoRC
2011.03.01, 07:00 PM
The main fault with this is that the nmotor reaches it maximum Torque an power almost straight way so all the power and Torque curves are much narrower than they should be.

Don't worry about the torque readings maxing out straight away, they are supposed to, brushed motors have maximum torque at 0 RPM.

In the second of the three latest graphs is the x axis seconds? If so it looks like your spooling up time is fine, if not a little slow.

DamoRC

ScoobyPete
2011.03.02, 05:01 AM
Don't worry about the torque readings maxing out straight away, they are supposed to, brushed motors have maximum torque at 0 RPM.

In the second of the three latest graphs is the x axis seconds? If so it looks like your spooling up time is fine, if not a little slow.

DamoRC

Yeah the 2nd Graph is RPM v Time. And you need to multiuply the results by 3.875 for the true RPM. So we are getting resluts close to manufactures spec. Although the vast majority of Motors we have are betweeen 2 years and 18 months old, so we would expect them to vary a bit.

The results we are getting are very consistant, very little differences between results for the moptors with only a few percent diff between runs.

We have now had an offer for someone to do independant tests for RPM and Torque so then we will have a bench mark to work from to dial the rig in a bit.


Here are the results of 4 PN70T HFAY motors from 3 Diff owners all with diff usage times and some have been run with com drops.

As you can see there is very little Diff between them, although I think my HFAY2 motor (HFAY2PB) may have been binding or loose whem we ran the test as this gave results closer to the other last time we ran it through the rig.

http://memberfiles.freewebs.com/77/47/38974777/photos/Motor-Dyno/untitled.bmp