Synch woes

[Panrock]

80 lines? Forget it! I am back to basics now. Have replaced the synch circuitry I can't even get the thing to lock properly now.

I have had it synching for a while but it's 'delicate'. Am I right in thinking that all the resistors after pin 13 of the 4046 really should be reduced by a factor of... say 5? After all, my motor isn't the usual little 'club' cassette motor but is a great big Lada heater fan motor consuming 3 to 4 amps. I'm worried that the gate of the power FET isn't being fed with enough "oomph".

(My version of the club circuit attached - wires that 'go nowhere' actually go to deck.)

Steve

[Steve Anderson]

.....but is a great big Lada heater fan motor..... Steve

Lada?.....Lada?.....'com boy, go down the scrappie and get a part with class, say a Jaguar version....

Joking aside, the size and mass of the motor and what it drives will dictate the values of the components within the PPL loop-filter. But first....

Driving the MOSFET with shown arrangement is fine, they're voltage-input devices when used under these conditions, much like a valve/tube. DC wise there's only the leakage current to consider which will be well under 10nA, = 500 Megohms.

They have far more input capacitance than a 'regular' transistor, but this is swamped by the 100n from the gate to 0V in this circuit. So the device is not the problem here.

Assuming the pulses at pins 3 & 14 of the 4046 are clean and stable (within themselves) we can say that all to the left of pin 13 of the 4046 in the diagram is operating correctly. So the problem is with the handful of passive components remaining around the MOSFET.

I would disconnect the 100k resistor from pin 13 of the 4046 (or pull it out of its socket if used) and apply +4.5V to its free end (two 1k resistors in series from +9V to 0V), and run the motor. The pot should give a range of speed control which includes 750RPM. Make a note of min. and max. RPM (if you dare), and the voltage across the motor for each of these three settings. The RPM can be calculated by measing the frequency of the pulses at pin 14 of the 4046, fitted or not.

What I'm trying to determine here is the approximate DC gain with in the loop-filter, when it comes to the AC gain into the equation comes the mechanical inertia of the moving bits. We'll deal with that later.

This circuit has been optimised for small cassette-type motors with far less inertia and generally appears to 'work' with most of them.

Steve, I'll need those numbers before we can go any further.

Steve A.

[Viewmaster]

I don't know if this will help, but Klaas changed the 2 100k resistors to a damping pot in this thread.....
http://www.taswegian.com/NBTV/forum/vie ... c&start=90

Albert.

[Panrock]

Great response chaps - helpful as ever - many thanks!

Last night I tried changing the 100K resistors for 22K, resulting in a much firmer action. However, still the same see-sawing and going faster when it should have been going slower etc... that's why Viewmaster's link has proved such interesting reading.

Then lying in bed this morning it struck me... why not try pins 3 and 14 on the 4046 round the other way? I've just done this and blow me - it works perfectly! No hunting, frame phasing automatically right. Hah hah ha ... on to 80 lines now!!! (Phew)

[Steve Anderson]

No hunting, frame phasing automatically right. Hah hah ha ... on to 80 lines now!!! (Phew)

Yep, you're correct, pin 3 should be in the feedback from the loop, not pin 14, something I missed until you just pointed it out and I had a quick look at the datasheet for the 4046.

I wish you had posted your sync arrangements earlier as I could have saved you a lot of trouble in this regard. There is no need for the 555s and all the other stuff, the following will do what is needed. Feed point 'A' into the appropriate pin of the 4046. The PC2 in the 4046, used here, is edge-sensitive and doesn't care about the duration of the pulses.

This is a current-driven Schmitt trigger which converts the current from the 'opto-fork' (or whatever arrangement you have) into logic-level signals for the 4046. However a quick experiment needs to be done to measure the current produced by the phototransistor as shown in the 'test circuit'. This needs to be done with the final set-up, emitter and detector in thier final places and with the stationary disc between them. Do not use a photodarlington (too slow) or a photodiode (current too low).

Also don't use BCxyz transistor types, also too slow, use proper switching transistors.

There is nothing 'magic' about the value of R1, 680 Ohms is the most common, but it could be anything from 150 to over 1000, depends on the LED used. But it's a good starting value.

There should be almost zero current when obscured, make a note of the current when a sync hole is placed between them. Select resistors from the table and to suit supply volts.

Steve A.

[Panrock]

Thanks for this Steve A. Yes, so much more elegant a way of doing things and well worth keeping on file along with all your other circuits which have so improved my apparatus. In effect you have designed a complete photomultiplier-based camera & monitor project.

However, it is all now working so well that I am going to give into my natural laziness and leave it in bloated form for now, including the 555s. After all, the apparatus is 'bloated' in its physical presence already. And looking at nice fat pulses locked in perfect synchrony on a scope is so satisfying somehow. How sad can you get....?

Viewers at this year's convention should certainly notice an improvement on what was shown before. Brighter, rock solid, pictures with more subtle detail . Even the 'over-vivid' color gives way when the controls are adjusted correctly and whatever you throw at it, the colours are pretty accurate.

So I've now got a really good basic colour system. You can tell I'm delighted - chuffed - but this has only been made possible by the help of Steve A and other members of this board !

Next I shall wait to see how the help from other kind members (Volker and Andrew) pans out regarding obtaining some workable 80-line discs. In the meantime I shall next address myself to the vexed question of the stereo/3D...

Steve

[Steve Anderson]

.....I am going to give into my natural laziness and leave it in bloated form for now.....Steve

..."If it ain't broke, don't fix it." Agreed, the circuits won't impress anyone at the convention, but the results you demonstrate surely will. Especially if you can 'upgrade' to 80 lines.

Steve A.

An afterthought, at 80 lines at 12.5Hz = 1000Hz line-rate. It might be worthwhile considering reducing the values of the 10n caps in the 555 circuits to 4n7 or the 68k Rs to 33k. As shown the pulse length is about 750us...getting a little close to the line duration IMHO. Should still work OK for 32 lines.

[Panrock]

Agreed, the circuits won't impress anyone at the convention, but the results you demonstrate surely will.

"Electronics" has never been my strong point and my abilities here are only rudimentary. I regard electronics as a necessary evil to deal with whenever I want to try out new ideas; otherwise I don't have the brain for it. That's why it's so useful that we have the electronic whizzes in the club (such as yourself).

I think it's important we maintain a balance between the innovative side of NBTV and the purely electronic discussions. There do seem to be two different types of member in this regard.

An afterthought, at 80 lines at 12.5Hz = 1000Hz line-rate. It might be worthwhile considering reducing the values of the 10n caps in the 555 circuits to 4n7 or the 68k Rs to 33k. As shown the pulse length is about 750us...getting a little close to the line duration IMHO. Should still work OK for 32 lines.

Yes I'm prepared for this. The answer is to include 50K presets in the time constant circuits and tweak 'em when necessary.

Cheers,

Steve O

[Steve Anderson]

...my abilities here are only rudimentary...Steve O

Rudimentary? How many others could honestly hold up their hands and say 'Yes, me too'. From just a schematic you breathed life into it, and now we can see it works. There were no easy-to-assemble PCBs, no layout diagrams, you just got on and did it!

Electronics, a sub-set of physics and to a large degree chemistry is worth at least a cursory study as to why electrons do what they do and how they assist us in all walks of life.

.. There do seem to be two different types of member in this regard.

And those in-between, I'm sure anyone who has been a member here for more than a few weeks knows which camp I come from, but I admire and am entranced by those that have the mechanical aptitute I wish I had.

Please never view electronics as a neccessary evil, more a complimentary friend who obays the laws of physics just as Newton and Einstein proposed.

Steve A.

[Panrock]

Please never view electronics as a neccessary evil, more a complimentary friend who obays the laws of physics just as Newton and Einstein proposed.

Perhaps my words were a little harsh. Don't worry. I like electronics really; it's just that I'm no bl**dy good at it!

I'll finish with a shot taken this evening of my now-happy (thanks to you) monitor and camera in the position they'll be in for the next few weeks - out of the way of my main working area and hopefully with enough space in front for the stereo apparatus. I'll start another thread about this when I get going...

Steve