Page 4 of 4

PWM Motors.

PostPosted: Sat Jun 30, 2007 2:03 pm
by Steve Anderson
Klaas Robers wrote:Steve, be aware that a Permanent Magnet DC motor cannot be driven by pulse width "hard" voltages. Then the motor current goes to very high values. I assume then that outside the pulse the motor is shorted, i.e. voltage = zero.

I have to admit I have done very little work with PWM driven motors so I'll take your advice.

As I mentioned before I don't know who the author of that PWM motor-drive circuit is, I don't recall how I came to have it or where it came from. It looks to be the schematic output of a circuit simulation program, but I don't recognize the format. I have no idea if the author built it and got it working.

I agree that with PVM drive to a DC brushed motor that the drive should be push-pull which might(?) help with dynamic braking of the motor. Obviously that circuit is single-ended.

Klaas Robers wrote:You can use pulse width modulation, but then the pulse frequency should be so high, that the self inductance of the coil of the motor defines the current. I think of a frequency of 20 kHz.

It seems that most use frequencies of that order, if nothing else so there is no audable buzz or whine from the windings, but most likely for the reasons you mention.

Klaas Robers wrote:But on the other hand I think that the dissipation of the transistor is not a problem. We are working in the watts region. Don't make things more complicated than they are....

As you say, in this application the linear versions dissipation is not a lot, so the advantage is minimal unless you're using a really large motor. (Unlikely for NBTV).

As for complication, it is a complicated diagram, and not how I would at least start out. It was a follow-on to Andrews' mention of state diagrams and Karnaugh maps. This would produce logic out, so it seemed that a pulse driven system was an ideal way to control the motor. It was just a suggestion for discussion.

Steve A.

A confession..

PostPosted: Mon Jul 02, 2007 4:48 pm
by Steve Anderson
Now, returning to the original topic of this thread, I use a different frame sync system here as it's easier to detect than the missing sync method.

The way I have done it here is the normal line pulses are eight pixels out of 120 at a sample rate of 48kHz = 167µs (6.7% of line time), for the frame sync I use a narrow pulse of four pixels, which equals 83µs. The remaining four samples are simply set to black.

It's quite easy to differentiate between the two and you have a continuous stream of pulses at 400Hz for line and motor sync purposes. This does require a higher bandwidth than the original 10kHz, but it does still fit into a 20kHz channel.

At first I used a broad pulse of 1.25ms, half a line, but abandoned it as I was reluctant to give up 3% of the resolution.

The problem here is that it is not compatable with the 'missing sync' system, so I'm still looking for ideas and suggestions....unless the missing sync sytem is trained to ignore pulses of less than about 125µs.

Steve A.

PostPosted: Mon Jul 02, 2007 6:34 pm
by Steve Anderson
Now, how about this Guys and Gals...

We reverse the method that I'm using above and make the line sync pulse 83µs wide, and the frame pulse 167µs wide, or thereabouts.

What benefit does this confer? It gives us time to establish a back porch where correct black-level clamping can be done, thereby negating using the use of bottom of sync as a reference.

Just an idea I thought I'd throw in...

Steve A.

PostPosted: Mon Jul 02, 2007 10:57 pm
by AncientBrit

How about 31 standard sync pulses and a super white pulse for frame sync?

Bandwidth requirements are normal, you might need to wind the gain down to cope with a slightly larger signal.

And you would probably need to fire a frame mono to hide the pulse on screen if it was distracting.

It's also compatible with standard 31 syncs using a standard sync sep.


PostPosted: Tue Jul 03, 2007 12:25 am
by Stephen
AncientBrit wrote:How about 31 standard sync pulses and a super white pulse for frame sync?
I like this, Graham! It would embed minimum and maximum reference signal levels right in the video signal and allow those who follow a more traditional opto-mechanical means of synchronising to have automatic frame synchronisation. For instance, one could use a phonic motor for line synchronisation and a variation of one of John Logie Baird's early synchronisation schemes for with this for frame synchronisation.

The frame synchronisation would comprise two side-by-side photocells aimed on each side of the super-white pulse hooked up to a difference amplifier. Drifting of the white pulse to either side would generate an error signal that would advance or retard the drive motor slightly. See Mr. Baird's British Patent 269,834, filed 21 October 1925, and 275,318, filed 3 May 1926, in the Patents and Articles section.

PostPosted: Tue Jul 03, 2007 10:35 pm
by AncientBrit

I used a variant of this on my interlaced experiments.

Both odd and even frames had only 31 line syncs.

On the odd frame on line 0 (I count from zero) for 4 pixels where the line sync was missing I added a mid grey pulse of about 50% amplitude to act as a flag for Odd/Even synch

Worked well.