top

A new test rig

Post new topic Reply to topic FizzX.org Forum Index | Steorn Technology Two: Permanent Magnet Motor (Only Permanent Magnets) Goto page Previous  1, 2   Page 2 of 2

Sat Feb 24, 2007 12:40 am PostPost subject:
TylerD1
Contributor
Contributor


Joined: 28 Oct 2006
Posts: 15

Reply with quote

avid_engineer wrote:
I thought CPR stood for Cycles per Revolution, where 1 Cycle = 4 Pulses for a quadrature encoder. So therefore 5000 Pulses per Revolution would equal 1250 CPR? (just to clarify).

So far, the best explanation of the terminology that I saw while going through the RoboteQ forum was from this thread:

"Pulses are the number of pulse that you could physically see with a scope on a single channel output."

"Counts are the number of transitions on the combined channels. Since you have two channels and each channel transitions twice per pulse (low to high and high to low), the number of CPRs is 4 x the numbers of PPRs."

And this is a modified/labeled version of one of the optical encoder pictures from RoboteQ's website.

But then, looking over an optical encoder manufacturer's website, it lists:

"Resolution is measured in cycles per revolution. Each cycle can provide 1, 2 or 4 counts.
Example 1000 CPR provides 1000, 2000 or 4000 counts or pulses per revolution."

Which the last part of the quote doesn't make any sense compared to the previous quote. So I'm assuming the last part of the quote from the manufacturer's website should be something like:

"Example 1000 CPR provides 1000, 2000 or 4000 counts per revolution."

And, therefore, a cycle would be the part in the picture between the blue lines, bounding a single pulse on channel A and channel B. So, with a 1250 CPR providing 5000 counts per revolution, if we assume the frequency refers to the quadrature signal we would presumably have:
Code:
(Frequency in Hz) = (RPM/60) * 1250 * 4

RPM = 60 * (Frequency in Hz)/(1250 * 4)

Max Frequency = 250000 Hz

Max RPM = 60 * (250000)/(1250 * 4) = 3000

Assuming the frequency is referring to the single channel signal, the max RPM would be higher than 3000.

avid_engineer wrote:
With regards to the roboteq software being buggy, i appreciate the ground work on the forums! i will have a play tomorrow and see what conclusions i can draw.

No problem, I look forward to the results. Very Happy
Back to top
View user's profile Send private message
 
Sun Feb 25, 2007 12:57 pm PostPost subject:
avid_engineer
Regular Contributor
Regular Contributor


Joined: 26 Oct 2006
Posts: 162
Location: Dorset, United Kingdom

Reply with quote

Well i've had a poke around with roborun, and as far as i can see it is sending the correct time base values to the controller. Range is between 1 and 64.

All the problems i am seeing are because of a bad implementation of the RPM formulas in the roborun software. If you ignore all the RPM readings, and take note of the time base and measured relative speed values you can do the math in excel and it works out fine!

Time base and measured rel speed values are taken directly from the controller, no nonsense. So i've no idea how they managed to mess up the maths so bad for RPM. lol

It would appear that the max RPM calculation multiplies the PPR by 4 before using it is the equation. Therefore requiring 1250 as an appropriate value for a 5000PPR sensor. The RPM calculated appears to also use a PPR of 1250 for the input to its formula, but with time base increments of 256us.... absolutely crazy!
Back to top
View user's profile Send private message
 
Wed Feb 28, 2007 8:16 pm PostPost subject:
verticalvoid
Contributor
Contributor


Joined: 31 Oct 2006
Posts: 19

Reply with quote

So what's the word on the latest test rig, a_e?

~V
Back to top
View user's profile Send private message
 
Wed Feb 28, 2007 10:42 pm PostPost subject:
avid_engineer
Regular Contributor
Regular Contributor


Joined: 26 Oct 2006
Posts: 162
Location: Dorset, United Kingdom

Reply with quote

Had to go away on a work related trip for a couple of days and havn't had much time yet this week to play. Also not been feeling so good, bring on the summer!

I know now how to get the RPM readings correctly, and can poll the data logger in 20ms intervals for power stats so just need to bring it all together with the software (again).

Before i went away i did a bit of work on the shaft and rotor to balance it properly, its running well now.

The next stage is to get the software working which i will try and get on with in the evenings this week. Then start work on mounting the magents on the rotor, and finally begin the stator mount which requires some thought and deliberation. I have a large block of transparent acryllic which i managed to scavenge from work so just need to find time to move forward with the project. I'm sure i will make time this coming weekend so should at least have some spreadsheet data and an up to date pic to post by sunday. Lets hope so Smile

Thinking about getting one of these...

http://www.velleman.be/ot/en/product/view/?id=362986

I thought it would be interesting to rig a special stator mount with a ghetto piezo sensor to observe variance in force at different frequencies of magentic interaction to see if there was a 'sweet' RPM. One of the problems with the previous piezo tests by Photon and others on Steorn.com forums was that the piezo sensor only produces a voltage upon change in force, not relative to applied force over time. So, i had an idea that i could apply the same princple to the piezo sensor as i did to the pulse width modulation voltage measurement (suggested by Tyler) whereby a series of resistors and capacitors give an average voltage reading over time varying with pulse width. I'm quite excited to see what results i get with this, but need the scope first.
Back to top
View user's profile Send private message
 
Sun Mar 04, 2007 7:58 pm PostPost subject:
avid_engineer
Regular Contributor
Regular Contributor


Joined: 26 Oct 2006
Posts: 162
Location: Dorset, United Kingdom

Reply with quote

A quick update as promised...

Been building the stator mount, was going well till i tore off part of my thumbnail with the milling machine (dumbass Rolling Eyes).

Some pics of the rig so far...

http://www.fictionary.co.uk/images/rig/new/smoothing_circuit.jpg
http://www.fictionary.co.uk/images/rig/new/rig_progress.jpg
http://www.fictionary.co.uk/images/rig/new/stator_rotor_progress.jpg

As you can see, i have mounted the rotor magnets now. They need sealing in to stop from flying out at high RPM (though they are a tight fit).

The stator mount will be adjustable to slide toward or away from the rotor for different tests and the L shaped block to the left is the unfinished mount for the stator magnets.

Thought i would post a closeup of the circuitry for smoothing the PWM voltage as discussed previously. I intend to try a similar ciruit for measuring current and have ordered some parts to build a smoothed current sensing circuit. Also ordered the oscilloscope so it should be an interesting week ahead Very Happy
Back to top
View user's profile Send private message
 
Fri Mar 09, 2007 9:08 pm PostPost subject:
avid_engineer
Regular Contributor
Regular Contributor


Joined: 26 Oct 2006
Posts: 162
Location: Dorset, United Kingdom

Reply with quote

I'll be spending some more time on the stator mount this weekend, i've an idea on how to build in a piezo element to record force on the stator magnets which should be interesting. My DSO turned up this week so i've been playing with the scope software trying to familiarise myself with how the readout works. Had a few initial problems but seems to be working ok. Very cool reading the waveform from a piezo element, and also tested the TX line of a serial adapter Very Happy

The plan is to use the scope to record data from piezo sensors near real time, and use the Roboteq controller and Data Harvest logger to read periodic RPM, Current, and Voltage (possibly updated at 50ms intervals).

It's proving very difficult indeed to get an accurate reading of power consumption from the motor. This is due to the PWM, and although we managed to work around the voltage measurement issue it is still very difficult to get current readings because they fluctuate as rapidly as the PWM voltage did.

Unless anyone can suggest a possible solution i am seriously considering running the motor straight from a 12V battery source to keep things simple. This means sacrificing speed control but the experiment would be worthless without accurate power consumption data Sad
Back to top
View user's profile Send private message
 
Wed May 09, 2007 6:30 am PostPost subject:
alsetalokin
Site Admin


Joined: 14 Apr 2007
Posts: 640
Location: Sol III

Reply with quote

Hey avid!
What kind (brand) of DSO did you get? Depending on its features, it may be possible to have it display the output current and voltage waveforms of your PWM controller and do the math on those waveforms automatically to give you pretty accurate power measurements going into your device. I have used a LeCroy DSO pretty extensively and I know it can be programmed for this.
Read the voltage directly on one channel, use a 1-ohm high-wattage resistor in series with the motor and view the voltage drop across that resistor for the current waveform on the other channel, and if you got a good scope, it should be able to do the appropriate maths.
Another possible solution would be to include the PWM electronics in the "box"--that is, take the "power in" measurements from the (presumably smooth DC) input to the PWM controller, rather than its output pulses.
An estimated efficiency correction factor could then be applied--or not, as necessary.
If you have a lightmeter, like a photographer's accessory, or a more sophisticated one, you can use the PWM output to illuminate a tungsten filiament bulb to a specific brightness, then illuminate the same bulb to the same brightness with straight DC. The power dissipation will be equal, or very nearly so, in both cases, and this will allow you to compute an efficiency correction factor for your PWM system, and also roughly calibrate its controls.
I think. Therefore I drink.
In the old days with analog scopes we would trace the traces on tracing paper, cut them out with scissors, and weigh the paper cutouts on an analytical balance to do the integration.

Anyway, keep us posted.
_________________
"Abandon hope, all ye who enter here..."
Back to top
View user's profile Send private message
 
Sun May 13, 2007 12:44 pm PostPost subject:
avid_engineer
Regular Contributor
Regular Contributor


Joined: 26 Oct 2006
Posts: 162
Location: Dorset, United Kingdom

Reply with quote

Hi alsetalokin,

Thanks for the suggestion, i ended up buying a Velleman PCSU1000.

http://www.velleman.be/ot/en/product/view/?id=362986

It's my first scope so can't really comment on its quality compared to the rest of the market, but in fact i did have some teething problems with it.

Firstly i had problems getting the thing to run out of the box, not sure if it was a driver issue but the software kept hanging when it loaded. When i finally got that sorted i further investigated the software development support and found it to be a little backward to say the least... in order to develop software to extract data from the scope you have to run the primary application and then use a dll to communicate with that, not the scope drivers / libraries directly. Most irritating.

Finally, after a couple of weeks one of the probes stopped working. It seems that it had a faulty connection where the cable joined the probe. Rather than spend an eternity dealing with my supplier, and possibly having to send the whole unit back i bought a new set of probes on eBay for a very good price.

Having said all that, the software that comes with the scope is very straight forward, and Velleman has an excellent forum on which the actual software developers are active. They even respond to requests for features in the software!

When it works properly, great. But if buying again i would look elsewhere.

As for the test rig, and the purpose of this thread... i've kinda moved on to other tests right now as the PWM current measurements just made the whole experiment over complicated in an area which is already complicated enough. lol.

I do appreciate your contribution though, good ideas for future projects Smile
Back to top
View user's profile Send private message
 
Mon May 14, 2007 1:01 am PostPost subject:
alsetalokin
Site Admin


Joined: 14 Apr 2007
Posts: 640
Location: Sol III

Reply with quote

Hi Avid--
I've built a bunch of Velleman kits, they seem like a pretty good company overall, but I've never used one of their 'scopes, so I can't comment there. The LeCroy that I use mostly has gotten me pretty well spoiled--it will do everything but pour the beer, it seems.
But I do use Velleman's PWM controllers in a couple of projects around the lab here, and they seem to work pretty well.
Good luck!
_________________
"Abandon hope, all ye who enter here..."
Back to top
View user's profile Send private message
 
Post new topic Reply to topic FizzX.org Forum Index | Steorn Technology Two: Permanent Magnet Motor (Only Permanent Magnets)
View previous topic
View next topic
Display posts from previous:   




You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum