Thread: VSD power tests

Originally Posted by Michael G

. . . . Stuart, can you manage a waterbath or even just a pond pump running water over the brake wheel?

I'd be worried about the water being flicked everywhere. I still reckon a fan will do the job, even a jet of expanding compressed air

Were to start lol

Hi Phil,

Thanks for the picture. Funnily enough I'm been thinking about replacing Telstra rope with hemp, was a little worried it might catch fire, so water had crossed my mind(seems I'm not completely crazy lol). If I can find some real rope I might just dip it in a bucket of water first and see what happens. It doesn't need to last long

Hi Chris,
I did think about other loads. An alternator was likely in 3rd spot after a more "normal" brake setup. I picked this setup as the alignment of the brake shoe(lol or rope in this case) with the drum is easy on a drill press. Anything else involved alignment of two shafts, which I would have to be able to repeat after the motor was removed for rewiring, which really meant I would need to make some all together more complicated. With this brake its just the weight on the scales, nothing else comes into it.
Also with an alternator some of the load is missing(though granted I dont think its much) even if we assume I can measure the output. Turns out with my gear I cant even measure the output of a variac with any degree of accuracy.

Of course if I cant make this simple brake work it doesnt matter so much that is was simple lol

Stuart

I had a crack at making a de Prony brake setup.

The motor is bolted to an 8 mm steel plate to which is welded a pair of angle iron tabs.
Bolted to each upright part of each tab are 500 mm lengths of HD Unistrut.
Between the tops of the Unistrut is a piece of angle that can be bolted/clamped at any height between the Unistrut.

Attached to this frame is a bike brake handle (4) and suspended from the angle will be two force/weight sensors (only one shown (7)) as per Stuar'ts arrangement.
A steel drum (6) consisting of a 12mm inner disc, and a 115 mm diam, 25 mm wide x 6 mm thick rim, is attached to the motor.
A 16 mm x 1 mm spring steel strap wraps around the drum and is attached to a fixed position sensor (7) and one floating sensor (not shown but will be located at 5).
The floating sensor is attached to the bike brake handle via a standard brake cable. All other connections are 2 or 3 mm steel cable.



Heres a close up of the drum - a 0.5 mm recess wide enough for the steel strap is turned into the drum to help the steel strap ride true.
Because it is only going to be used intermittently I made no special effort to balance the drum but it runs reasonably true and there's not much vibe added to what is already a smooth motor.




The vertical position of the fixed sensor can be adjusted by the all thread rod.
The spring is the return spring for the brake.



Brake cable pathway.


Here is the brake in action and a meaningless torque reading.


I've tried it out on 50 Hz and the noise is not that bad, certainly nowhere near as bad as I thought it would be The brake arrangement enables the spring steel band to be rapidly engaged and disengaged and I had plans for extensive cooling but is does not appear necessary.
The band and drum rim do get quite hot but they also cool down quite quickly from the spinning drum so half a dozen brake applications can be made in a couple of minutes without any problems

I cannot make any real measurements until I get the second sensor but by engaging the brake I can make the 1HP motor briefly draw more than 5.5A from the single phase supply which would normally be ~2 HP so I assume from that that the motor is well and truly doing its business.

For bigger motors I will make a bigger drum.

Things get done in your shed a little faster than mine!

Originally Posted by Stustoys

Things get done in your shed a little faster than mine!

Well although I have many distractions (two Border Collies constantly shoving a soggy ball in my face) I was able to clear yesterday and today to solely work on that project. Plus I spent all of monday morning driving around picking up bits of scrap steel and whatnot from a couple of stores. It took me all opt yesterday to make the drum/pulley but after that things wen't pretty quickly.

Plus I had the benefit of your prototyping !!

I got the second weight/force sensor (digital luggage scale) my DeProny brake some time ago but have been working (yes for money) so have had no time to work on these motor testing side projects.
The paying job finished last week so managed to get some time on it this morning to see if I can get any sensible HP measurements with this setup.

Here is the set up,
The difference between T1 and T2 times the radius of the drum brake give the applied torque (back Torque is applied via a bike brake handle on the side of the rig - see posts above)
Input Amps/volts to the VFD are given by the panel meter
3 Phase current is measured by a cable clamp and shown on the digital multimeter
An Oscilloscope is used to check the output of the VFD wave form to check it is roughly sinusoidal
The tacho of course measures RPM

By setting everything up like this and applying the hand brake incrementally and taking a series of photo I can collect all the data needed very quickly.
On a couple of photos the shot must have been taken when some of the LCD digits were changing but that only happened a couple of times




Now here is the sort of data one can get.
All I did here was kept the VFD at the same frequency and varies the applied torque.
As you can see as torque is applied the RPMs drop as the motor starts to slip.
The nominally 1HP motor at 1440 RPM (fee revs art 1507 RPM) for this motor seems too be almost bang on, I get 1445 RPM
The motor is able to generate about 1.4HP before it starts to slip significantly although I would probably not want to drive it even that hard.
So all in all, the DeProny Brake seems to be working nicely, the data can be collected very quickly and I think I am ready to test motor Power/Torque Versus frequency

Hi Bob,

The question that arises most often is the case where a 415V motor can't be converted to 240 delta ( for a multitude of reasons ) the question is what happens to the power output at half voltage.. The answer is either half power or quarter power.. I'm in the camp that says half power.. but we need to design an experiment to prove it one way or another.

I say half power because as the voltage drops the field weakens and the slip increases which increases the current in direct proportion to the rotor slippage, so by the time you get to half voltage the current hasn't decreased.. the limiting case being when the current increases to the locked rotor current that voltage is the lowest you can go.. below that the motor doesn't turn, it's stalled. AC induction motors are self balancing systems as the load increases the rotor slip increases and the current increases.

The argument for quarter power says that because the voltage has dropped the current also drops and at half voltage the current is halved. So 0.5*0.5==0.25

Your setup should be able to answer the question..

Ray

Ok here is the procedure for the test... take a motor wired for 415V operation ( star ) start it up on 240V 3 phase, run it up to 50Hz and then load it up with the dyno
until you reach nameplate current. At that point you should have half the rated nameplate power.. If you then reduce the voltage, while keeping the frequency at 50Hz the motor current should increase to try and maintain the load.

Since the rotor current will be higher, the I²R losses in the rotor will be a little higher and efficiency will be a little less...

Hi Bob,

The photo idea is pretty clever.
I'm glad someone is getting roundtoit
Mine is still set up on the drill but hasn't advanced at all.

Hi Ray,

Could these tests be done by changing the base freq on a 415V vsd?

Stuart

hi every one. fee free to tell me to Pi++ off. and before i read the start of the thread.
what about a motor driven hydralic pump and som pfm turning the pressure reading to
a torque value?
aaron

Originally Posted by azzrock

hi every one. fee free to tell me to Pi++ off. and before i read the start of the thread.
what about a motor driven hydralic pump and som pfm turning the pressure reading to
a torque value?
aaron

Sure it you like and have some of the parts already.
My next favorite dyno was a second hand torque converter and a strain gauge. But it was a little on the big side, involves extra bearings and alignments.

What I think is the outstanding feature of this set up is it costs about $6 for the parts you don't already have laying around.

Stuart

Cheers Guys,

Ray, Thanks for reminding me to test a motor configured for 415V operation. I should have added that the motor I tested above was rated at 1HP for 240V 3 phase ∆ config.

PM
azzrock, I wonder how accurately a pressure reading could be converted into a true torque value given the mechanical losses that would occur in such a setup ?
As Stu says the DeProny setup is very simple. The main problem with the brake is heat as the steel brake band on my setup does get hot, but using the camera means that about a reading per second can be made so a few seconds are all that is needed to ensure one has a reliable reading, and a series/sequence of readings can be made in ~10 seconds. The drum is pretty chunky and so starting from RT has enough thermal coolth inertia so I can touch the band with my hand about 30s after the 10 seconds with of brake application.

The drum I made has a radius of 61 mm and the luggage balances have a max weight range of 50 kg. If a motor can generate this amount of force on that pulley this translates into about 10 HP rating but realistically with only single phase in the shed I'm not going to be going over about 5HP. I expect the steel brake band might be getting pretty hot at that power rating so I have rigged up a fan for extra cooling of the drum - I will take some more photos of the setup later tonight.

Any more ideas for experiments are welcome
So far I propose
- Rays suggestion for testing the power output of a 415V wired motor, ∆ and Y
- The one we all want to know of HP/torque output as a function of frequency
- testing different powered VFDs with the same motor (e.g. 1HP motor with 2, 3 and 4HP VFDs)
Anything else?

Originally Posted by Stustoys

Sure it you like and have some of the parts already.
My next favorite dyno was a second hand torque converter and a strain gauge. But it was a little on the big side, involves extra bearings and alignments.

What I think is the outstanding feature of this set up is it costs about $6 for the parts you don't already have laying around.

Stuart

hi stuart, good point mate. i thought every one had tonnes of that stuff holding up
the corner of there welding bench

Originally Posted by azzrock

hi every one. fee free to tell me to Pi++ off. and before i read the start of the thread.
what about a motor driven hydralic pump and som pfm turning the pressure reading to
a torque value?
aaron

yes... a hydraulic pump would do this, maybe a car power steering pump would do.
Put a needle valve or even a ball valve on the pressure pipe to adjust the pressure/flow and return pipe from the valve straight back into the tank. if you could make a mounting bracket for the pump with a brg on the shaft and a brg at the back of the pump so that the pump could pivot on its axis, then have an arm extending 1 ft from the centre of axis sideways from the pump with a foot on the end that sits on a set of scales, bolt it to the bench and use the existing pump pulley to drive it. tiz likely that the front bearing could be mounted on the front part of the pulley boss, might need a bush?

the more you restrict the flow the greater the pressure applied to the scales

shed

ps. if'n you want to have a crack at the hydraulic version Stu I can give you a pump

A few more photos of the setup.

General View

One thing you might notice compared to the photos I took back in April when I first made the rig is that I slapped a coat of paint on the rig as there was a fair bit of bare metal showing.

Then I upgraded the mounting mechanism and made a new handbrake arrangement (a much more angle handle bar) as the old one did not have allow for sufficient throw to disengage the steel band at "no load"


Hi tech tacho mounting mechanism


Something else I learned was that the circular polariser does not play well with the polarisers used on the DVM display panel.
Sometimes the DVM display numbers cannot be seen and other times a rainbow spray can mask the readings.

Originally Posted by shedhappens

yes... a hydraulic pump would do this, maybe a car power steering pump would do.
Put a needle valve or even a ball valve on the pressure pipe to adjust the pressure/flow and return pipe from the valve straight back into the tank. if you could make a mounting bracket for the pump with a brg on the shaft and a brg at the back of the pump so that the pump could pivot on its axis, then have an arm extending 1 ft from the centre of axis sideways from the pump with a foot on the end that sits on a set of scales, bolt it to the bench and use the existing pump pulley to drive it. tiz likely that the front bearing could be mounted on the front part of the pulley boss, might need a bush?

the more you restrict the flow the greater the pressure applied to the scales

shed

ps. if'n you want to have a crack at the hydraulic version Stu I can give you a pump

Nice one. But is it possible to get the torque reading from measuring the pressure or flow from the pump instead of from reading the scales.
i was just thinking a flex coupling straight from motor to pump.

bob love your set up. that tacho reads from way back there. luve your lab jack where can i get one.
aaron