Narrow-Bandwidth Television Association

I've noticed this problem show up now that I have a very bright image which requires me to turn down the contrast. Below about 1/2 setting on the contrast knob, I lose synchronisation on the motor control circuit.

I have measured the incoming synch pulses to the motor control circuit, and they are rock-solid 12V at 61ms wide. They do not vary even slightly when the contrast is adjusted.

Yet, my synchronisation fails based on the setting of contrast. I don't understand this at all, as AFAIK, there is no connection between the contrast of the image, and the synchronisation! I can only surmise some weird transient power fluctuation?

The whole assembly (circuits and the mechanical bits) forms a PPL (Phase Locked Loop). This requires two 'inputs' to function.

I have measured the incoming synch pulses to the motor control circuit, and they are rock-solid 12V at 61ms wide.

This I assume is from the disc sensor, the IR pair. The other input, the reference, is the signal that comes from the signal source, the CD player. If this is fed into the motor control circuit after a pot it will lose lock at some point as you turn down the gain.

Without having complete details of your arrangement it's hard to offer much more help. But I would check with a scope both inputs to the 4046 on pins 3 and 14. Quoting the STM datasheet for the device...

The phase-comparator signal input (terminal 14) can be direct-coupled provided the signal swing is within CMOS logic levels.......for smaller swings the signal must be capacitively coupled......

So pin 14 should have a logic level input signal, 0V to +12V (I assume you're using 12V). OR have a capacitor on its input for smaller signals, but here they still need to be more than 1Vp/p for reliable operation.

Pin 3 is a logic input and the signal here should go from 0V to +12V, no ifs or buts.

Steve A.

I attach two datasheets for the device for those that don't have them.

I'm also a little concerned as to your measurement of 61mS...do you really mean milliseconds (mS) or microseconds (uS)? Can this board support Greek characters so often used in electronics and mathematics?

Steve Anderson wrote:The whole assembly (circuits and the mechanical bits) forms a PPL (Phase Locked Loop). This requires two 'inputs' to function.

The phase-comparator signal input (terminal 14) can be direct-coupled provided the signal swing is within CMOS logic levels.......for smaller swings the signal must be capacitively coupled......

So pin 14 should have a logic level input signal, 0V to +12V (I assume you're using 12V). OR have a capacitor on its input for smaller signals, but here they still need to be more than 1Vp/p for reliable operation.

Pin 3 is a logic input and the signal here should go from 0V to +12V, no ifs or buts.

I'm also a little concerned as to your measurement of 61mS...do you really mean milliseconds (mS) or microseconds (uS)? Can this board support Greek characters so often used in electronics and mathematics?

61µs it is. Me bad. This is the width of the synch pulse coming in to pin 14 of the 4046. It is rock-solid 11.9V and does not vary when I adjust the conrast knob. Symbols such as 'µ' can be cut/pasted to the board or inserted via your keyboard depending on your O/S. In Windows you can hold down the Alt key whilst typing the 4-digit ASCII code of the symbol you want on the numeric keypad. 'Micro' is Alt-0-1-8-1

The IR LED signal into the 4046 is about 9V, and it too does not vary when I adjust the contrast knob.

You say pin 3 is a logic input, but pin 3 is the IR input on the circuit diagram. This is definitely not on/off 12V -- it varies, of course, based on the speed of the disc and the strength of IR signal received.

I wonder if pin 3 and 14 are 'the wrong way around', and that pin 3 should be receiving the sync signal from the NBTV waveform, and pin 14 should be receiving the IR signal?

The IR LED signal into the 4046 is about 9V, and it too does not vary when I adjust the contrast knob.

You say pin 3 is a logic input, but pin 3 is the IR input on the circuit diagram. This is definitely not on/off 12V -- it varies, of course, based on the speed of the disc and the strength of IR signal received.

Well, it's amplitude shouldn't vary, whatever the speed of the disc, Klaas's additional gates will see to that. Ideally it should snap from 0V to 12V in a very short period of time (less than 1uS) even if you turn the disc very slowly by hand. The hysteresis provided by these extra gates will also ensure there is no 'bouncing' or ragged edges on the IR signal.

Steve A.

P.S. I tried inserting the Greek 'mu', but because this PC is bisexual...er, I mean bilingual, it tried to insert a Thai character. Just have a look at the keyboard....

Steve Anderson wrote:

The IR LED signal into the 4046 is about 9V, and it too does not vary when I adjust the contrast knob.

You say pin 3 is a logic input, but pin 3 is the IR input on the circuit diagram. This is definitely not on/off 12V -- it varies, of course, based on the speed of the disc and the strength of IR signal received.

Well, it's amplitude shouldn't vary, whatever the speed of the disc, Klaas's additional gates will see to that. Ideally it should snap from 0V to 12V in a very short period of time (less than 1uS) even if you turn the disc very slowly by hand. The hysteresis provided by these extra gates will also ensure there is no 'bouncing' or ragged edges on the IR signal.

Steve A.

OK, well I haven't installed these. I guess I will do that ASAP and see how it goes. At the moment, I'm using the original circuit.

I have a theory. Since the LED and the motor share the same power supply -- the 21V supply -- then when the power usage of the LED changes, the power to the motor changes. Since the synch circuit is incredibly sensitive to the motor adjustment, it is simply a matter of the contrast changing the characteristics of the power usage by the LED, which in turn affects the power usage of the motor... and at some point the LED current requirement becomes so low that the motor gets more, and it speeds up too much for the sync circuit to dampen.

This will all be corrected if/when I eventually figure out how to adjust things such that the synch circuit can either slow down the motor, or there is sufficient dampening of the disk rotation to do that in a more effective manner than is happening now.

I have a theory. Since the LED and the motor share the same power supply -- the 21V supply -- then when the power usage of the LED changes, the power to the motor changes.

This is entirely plausable. You might have noticed my references to clean and stable power supplies, it seems that it's come home to roost.

There is an inate problem in the circuit you are using that there is some negative feedback from the drain to the gate of the MOSFET. So what that device is trying to do is keep its drain Voltage the same whatever. This means that if your supply voltage increases due to you turning down the current through the LEDs there'll be more across the motor speeding it up and the PPL not being able to cope.

Solution? Simple. Use a stable regulated supply for everything. I cannot stress this enough.

The alternative is to rearrange the motor drive circuit whereby it isn't susceptable to power supply variations. Quite easy, and actually far more 'hygienic'.

Steve A.

Buy yourself a second transformer for the motor, or even better for the LEDs. Then you can have a higher voltage for the LEDs in less series and the voltage for the motor is independant from the LED current.

However, if you change the setting of the contrast pot the setting of the clipping level of the sunc separator is no more correct. Then you will see that the line sync pulses coming from the sync separator are no more correct. This is especially if you lower the contrast. The sync separator is optimal when the video coming out of the separator board, going to the LED-driver board, is 1.4 volt from bottom sync to peak white.