electrodynamometer

[The Electrician]

Here's a movie I made showing the experiment suggested by JimB:

https://tinypic.com/r/34japs7/6

I started out at a frequency of 1/10 Hz and ending at 10 Hz. I didn't increase the frequency smoothly, but rather in arbitrary steps. This meter is underdamped, and you can see the swing of the needle increases at one point, but then decreases from then on.

The movie file is large.

Edit:

I should mention that what I have done is not exactly what JimB suggested. I didn't use a moving coil meter; I used a moving iron meter. A moving iron meter responds in the same way as an electrodynamometer meter--the deflection force is proportional to the square of the applied voltage. Thus, there is an upward deflection no matter what the polarity of the applied voltage. The meter even responds to applied DC.

 

[MrAl]

Ha ha, that's cool. At one point i think we can see the slight overshoot too, which is something i wanted to see firsthand.

 

[JimB]

Excellent!

That is just the sort of thing I had in mind.
I should have made an effort to do that myself but I am not au fait with movie techniques and uploading to suitable sites.
Maybe time for me to learn?

JimB

 

[MrAl]

Here's a movie I made showing the experiment suggested by JimB:

https://tinypic.com/r/34japs7/6

I started out at a frequency of 1/10 Hz and ending at 10 Hz. I didn't increase the frequency smoothly, but rather in arbitrary steps. This meter is underdamped, and you can see the swing of the needle increases at one point, but then decreases from then on.

The movie file is large.

Edit:

I should mention that what I have done is not exactly what JimB suggested. I didn't use a moving coil meter; I used a moving iron meter. A moving iron meter responds in the same way as an electrodynamometer meter--the deflection force is proportional to the square of the applied voltage. Thus, there is an upward deflection no matter what the polarity of the applied voltage. The meter even responds to applied DC.

Hi again,


So the frequency was 0.1Hz to 10 Hz. Was this using a pure sine wave or some other shape, and what was the peak voltage ?

is there any way to download the movie file ?

Thanks

 

[The Electrician]

Hi again,


So the frequency was 0.1Hz to 10 Hz. Was this using a pure sine wave or some other shape, and what was the peak voltage ?

is there any way to download the movie file ?

Thanks

The meter responds to, and the calibration is correct for, DC. The waveform was a sine wave from a function generator, with a peak value of 7 volts (and, of course, an RMS value of 4.949 volts). This meter indicates the true RMS (AC+DC) value of a waveform. The iron vanes of the movement can retain some residual magnetism when the meter has been used on DC, and then the indication will have a small bias, but it's not bad for a simple panel meter.

For the purposes of this thread, it does a good job of showing how things work.

I don't think the file can be downloaded from Tinypic.

I used a still camera's movie mode, so the movie isn't great, but it was more convenient than getting out the camcorder.

 

[MrAl]

Hi again,

Oh i thought the quality wasnt that bad really as i could clearly see the meter face and the movement too pretty well.

The only question i had was that if the meter reads 7v peak for a 7v peak slow varying signal then it if was a square law meter it should read 0.35v for a fast varying sine signal. That's because the average of the square of the sine is 0.5 while the average of just the abs of the sine is 0.64 approximately. So a meter that reads 7v for the peak of a slow signal and 5v for a fast sine means it must be just taking the absolute value of the sine and averaging that.

So i was wondering how did you find out it was the type of meter that works by the square law?

 

[The Electrician]

Hi again,

Oh i thought the quality wasnt that bad really as i could clearly see the meter face and the movement too pretty well.

The only question i had was that if the meter reads 7v peak for a 7v peak slow varying signal then it if was a square law meter it should read 0.35v for a fast varying sine signal. That's because the average of the square of the sine is 0.5 while the average of just the abs of the sine is 0.64 approximately. So a meter that reads 7v for the peak of a slow signal and 5v for a fast sine means it must be just taking the absolute value of the sine and averaging that.

You discussed in post #8:

I think what is happening here is that once the meter face is calibrated, then it indicates RMS. After all if we calibrate the meter face as a square root function, we'll see the exact RMS value appear on the meter face.

If the scale on the meter is calibrated to read RMS (as I think every AC voltmeter is), then I would expect a sine with a 7 volt peak to read 7*.707 = ~5 volts. If an AC voltmeter reads right on AC waveforms (reads the RMS value), then to determine if the basic meter movement is average or RMS responding, determine what it reads with DC (assuming it responds to DC at all).

My old Triplett analog VOM, which is a rectifier type instrument, is an average responding meter, with the scale calibrated to read RMS on the AC scale. I applied 7V DC and on the AC scale it reads 7.8 volts; theoretically it should read 7*1.1107 = 7.775 volts. It has a heavily damped taut band movement (slow to respond). I applied a .01 Hz sine with a 7 volt peak voltage and watched to see what the peak measured on the same AC scale; it read 7.8 volts, the same thing it read with 7V DC applied. This confirms that the movement is an average responding movement.

The meter in the movie reads 7.0 (given the readability and accuracy of a panel meter) with 7V DC applied, and ~5V with a 7 volt peak 10 Hz AC waveform applied. This shows that the movement is responding to the square of the applied voltage (with the scale taking the square root, so to speak). The RMS value of DC is just the value of the DC. An RMS responding movement doesn't have the 1/1.1107 factor applied to the reading, that an average responding movement does.

So i was wondering how did you find out it was the type of meter that works by the square law?

I have an old text that shows various meter types; it describes the moving iron movement and says that the torque is proportional to the square of the current. It also shows two styles of construction for moving iron movements; see the first attached image.

The second attached image shows the internal construction of the meter in the movie. The two vanes are clearly visible; this meter uses the radial vane type of construction.

There's a web page with some discussion of moving iron movements, mentioning that the torque is proportional to the square of the current:

https://electrical-engineering-portal.com/moving-iron-instruments-voltmeter-and-ammeter

 

[MrAl]

Hello again,


Ok that's good, i forgot about the meter face that could be taking the square root, thanks for reminding me. This seems to work out then.

If we apply 7vdc we get 49 and the square root of that is 7 so we see 7 on the face.
If we apply 7vacpk we get 49 peak and the average of that is 24.5 (the average of sin^2 is 1/2 the peak) and taking the square root of that we get just a hair under 5 volts so we read 5 volts on the face.

Works out nicely, so that tells us that they really do calibrate the face to take the square root. Looking at the face, it is definitely non linear, and the distance in the vicinity of 10 is larger than the distance around 5 and it looks like it's roughly in the right proportion.

So i guess we proved not only is the moving vane meter taking the average of the square, the face is taking the square root so that works out to Vpeak/sqrt(2) as it should.

All is well in meterland

I have an old Simpson 160 analog multimeter. Dont know if they make the batteries for them anymore, but apparently they still sell the meters (on the web). I've had that meter since the 70's.
I also have an old Western Electric ammeter with a face that is bigger than most large wall clocks! It's so old it was made with a custom face plate. I'll have to dig it out one of these days and see what kind of movement it has.

 

[The Electrician]

Panel meters of the moving iron type are very handy for experimenting with power supply designs of the type that hobby builders often ask about on the forums.

They can be had for very low cost on eBay. Panel meters that are designed to measure AC amps are almost always moving iron type. Sometimes thermocouple types are available, but they are much rarer. Typically they are standardized to have a 5 amp full scale movement; if the scale reads higher, they are probably intended to be used with a current transformer. However, sometimes you can get one which measures high currents directly, without needing a current transformer. I have gotten moving iron panel AC ammeters that could measure 50 amps full scale directly. It can measure DC as well as AC, and that's a handy meter to have around.

Moving iron panel ammeters have several advantages. They are inexpensive, so you can buy several and measure several places in a power supply at once. They respond as true RMS AC+DC. They are built with a coil of relatively heavy wire and they are fairly immune to damage from momentary overload. The heavy wire coil has low resistance so they don't drop much extra voltage in the circuit (you can put them in series with a filter cap to measure ripple current in the cap and leave them in place without much change to circuit operation).

The main disadvantage is that they aren't especially accurate, but if you have one DVM, you can use that to check the calibration of the panel meter(s) by comparison. Extreme accuracy is probably not necessary anyway to learn about power supply operation.

Here are some examples of meters available now on eBay:

This one measures 5 amps directly, and is quite inexpensive:

**broken link removed**

Here are two ammeters in one auction. They don't measure 800 amps directly. If you look at the lower left part of the image, you can see the legend "use 160/1 C.T." That means that you would need to use a current transformer to get 800 amps full scale range. The meter itself is a 5 amp meter. You could mark the face plate to indicate 0-5 amps. But, two meters for 99 cents? Wow.

**broken link removed**

Here's a meter that could need a current transformer, or it could measure 50 amps directly; I can't tell from the image. If you bought it, and it turns out that it needs a C.T. to measure 50 amps, then you can use it to measure 5 amps full scale directly. The price is right.

**broken link removed**

If a hobbyist needs to measure currents for a power supply project, you can't go wrong with these moving iron meters.

 

[MrAl]

Hello again,

OH yes some nice meters there for very good prices. Shipping adds to the price, but that's life i guess.

I see the trend now in the face plate of these meters. They have a slight compression between digits as the numerical value gets lower. The far left end is very compressed due to the required scaling. By contrast, most of the meters i use have fairly linear scales which means they cant be square law. I dont use them that much anymore though as i use digital for almost everything, but i do use one panel meter for monitoring the line, which also has a linear face and i think i can see a small coil inside although i'd have to look better. So it's probably moving coil with stationary magnet. It's not too bad really though, with a slightly wider face than normal.

 

[PG1995]

Thank you very much for the movie, The Electrician.

is there any way to download the movie file ?

@MrAl: I have already captured the movie for you and can upload it to some file sharing website. But I have just realized that there was no need for the capture because I was already getting "Download This Video" option on the top right corner of the video. This option is available if you have Real Player installed on your machine. Nonetheless, I can still upload it if you like. Please let me know. Thanks.

Best wishes
PG

 

[MrAl]

Hi,

I cant seem to download it, but what format is it in?
I can read mpg, avi, etc.

 

[PG1995]

Hi MrAl

I have attached the video file. Let me know if there is a problem.

Regards
PG

 

[MrAl]

Hi again,

Nice. I was able to download this and view it without going to the web site. Thanks much PG, and thanks again to Electrician for taking the time to do this experiment, which illustrates how it really works.

It would be interesting to disassemble the meter and make some physical measurements (lengths, widths, masses) and model the whole meter. We could use that data in the model that i presented earlier in this thread. I had to guess at the measurements for that simulation but really it was just sticking some numbers in. Having all the numbers would tell the whole story
Would be quite a lot of work to do though and we'd probably have to test the metals too to find out their magnetic characteristics.