Getting a frame pulse early !

[Viewmaster]

Steve, I have started building the new cct and had another
thought (fatal that!)

To demonstrate to myself and others how a NBTV picture changes as the framing is moved slowly one could add this ganged switch to the cct.

This would switch in a larger C and put a larger R pot in series with the existing one, in the timing RCcct of the 2-3ms 4538 as I show, giving a new range of timing from 3ms to 80ms.

This would enable the frame pulse to be shifted anywhere in the picture at the flick of a switch and a turn of the new pot, and hence roll the picture until giddy

I do not know what the highest/lowest values for the R and C in the 4538
timing can be , although I have seen as low as 10k for the R and up to 30n for the C.
Do you know what the limiting values are please?

[Steve Anderson]

Albert, yep, I don't see any problem with that...if it goes over 80ms weird things might happen but no harm would be done.

The minimum value of Rt is 5k, no maximum except board leakage, say 10M in practice.

There is no limit to Ct at either extreme. At the lower end you'll have stray capacitance to contend with, the larger values above 10µF will require a discharge diode as per Fig. 11 in the datasheet I sent you.

The 4538 is specified for pulse lengths from 10µs to 10s (seconds) but performs best between 100µs and 1s.

With all these HEF/CD monos be prepared for wide differences between the calculated time period and those observed in practice. Although the formula is Pw = Rt x Ct be prepared for a ±50% variation. (at least). This is due to the CMOS process used which is better suited to logic applications rather than linear. The datasheets often say they're much better...don't believe them.

Steve A.

[Viewmaster]

Thanks Steve, for info, which I didn't see on the data sheet.
As the pot size can be at least 2meg I don't need to change the C at all .......only need a switch across the 2 meg.

As I have a dual trace scope so shall put a 400 kHz line pulse on one trace as a standard time check and look at the o/p from the vibs to be adjusted timewise as a comparison on t'other.

Now Steve, I shall get back to soldering up your cct. and leave you in peace to get on with your own projects

[Steve Anderson]

Oh heck! I didn't realise you were actually building it! As I said, it was just an idea in my head, I've never built or experimented with it myself.

I'm fairly certain it will work, the only proviso is that there are 31 sync pulses and no glitches or other rubbish mixed in with them. But that's generally true of most frame sync detection arrangements.

I wish you good luck.

Steve A.

[Viewmaster]

Oh heck! I didn't realise you were actually building it! As I said, it was just an idea in my head, I've never built or experimented with it myself.
.

Not to worry Steve, I shan't try to sue you if it doesn't work.
It's your idea, but I shall do the experimenting.

I'm fairly certain it will work, the only proviso is that there are 31 sync pulses and no glitches or other rubbish mixed in with them. But that's generally true of most frame sync detection arrangements.
.

I shall have C's across the power lines to the critical IC's.
Of course with this Niptrix display, it should work OK with only the 31 pulses coming off the sync sep. Just that the 1st line of LEDs would not be clocked and so missing.

[Steve Anderson]

Of course with this Niptrix display, it should work OK with only the 31 pulses coming off the sync sep. Just that the 1st line of LEDs would not be clocked and so missing.

I don't quite grasp that, I can't see any reason why line 1 (presumably) would be omitted. The whole raison d'être for this sort of arrangement is to ensure that the framing is precise which includes displaying line 1 in its entirety and correctly positioned.

But then again I could be missing something...not just line 1.

Steve A.

[Viewmaster]

Of course with this Niptrix display, it should work OK with only the 31 pulses coming off the sync sep. Just that the 1st line of LEDs would not be clocked and so missing.

I don't quite grasp that, I can't see any reason why line 1 (presumably) would be omitted. The whole raison d'être for this sort of arrangement is to ensure that the framing is precise which includes displaying line 1 in its entirety and correctly positioned.

But then again I could be missing something...not just line 1.

Steve A.

O Steve, I'm more likely to be missing something, rather than you.
But if I never used an 32 line pulse inserter, then it could well be the last line that was missing. First line, last line, not sure which, but a line would be missing on the display.

BTW, if your cct doesn't work, (if it doesn't, it's probably a wiring fault on my part! see below), I can always fall back on my own cruder idea using just a 2.6 and 80ms vibs to put the frame/line pulse in one frame late as I mentioned. As you said at the time non accurate 400 line pulses would be a trouble etc.

I now have it all mainly finished.........I inserted the first vib and the counter and switched on.........big smoke from the vib and the 12v fuse blew So will try again today after cct checking.

[Steve Anderson]

.........I inserted the first vib and the counter and switched on.........big smoke from the vib and the 12v fuse blew

There is something about that smell of frying chips which is never forgotten...and has nothing to do with fish...

Steve A.

[Viewmaster]

All sorted out now Steve, the complete cct sems to be ticking away OK.
The smoke was an uncut track to the 12v line which a pin on a 4538
didn't like.
No sign of any noise or nasties on the scope at the various stages so it all augurs well methinks.

That's the 4017 cascade switching board and your cct now up and running, so we progess. We progress.

[Steve Anderson]

Good to hear that you got the problem fixed and that the counter/detector seems to be OK. I perhaps should draw it up properly and submit it to Jeremy. I don't know if this method has been published in the past. If anyone does know, please let me know which issue of the newsletter it's in.

On second thoughts maybe I won't. As it forms part of the NIPTRIX perhaps it's better included in that. That's assuming Albert will be submitting this sometime.

Steve A.

[Steve Anderson]

Another way to this is with a modification of the 'Eureka' device. This assumes you've already extracted the syncs from the video. It works exactly as the above with separately adjustable frame and line 1 triggers as well as continuous syncs.

If the individual adjustment isn't required just wind VR102 to the ground end or simply omit it and permanently ground pin 6. At start-up the micro checks the voltage at pin 6 and if less than 0.5V the frame and line 1 pulses are the same and only VR101 has any effect. Pins 6 & 7 could probably do with having a 100n cap to ground just to mop-up any crud. That's it...one 8-pin chip...in its entirety!

I've not written the code for this, it should be a simple hack of the 'Eureka' code. But if anyone's interested and would like a sample pre-programmed chip just let me know...no catch, it's free...as long as 2,654 people don't apply all at once! The postage is likely to be more than the cost of the chip...which is almost insignificant anyway.

Serious builders only please...and regular contributors to this forum and/or the newsletter....I do want to see it used. The polarity of the output pulses can be inverted if desired, so too the input, just let me know.

The down-side of this device is it's 5V only (or 3.3V) so it won't directly interface with 12V HEF/CD logic.

Steve A.

[Viewmaster]

On second thoughts maybe I won't. As it forms part of the NIPTRIX perhaps it's better included in that. That's assuming Albert will be submitting this sometime.
Steve A.

Well it's your cct Steve, but I do intend to submit the NIPTRIX write up when complete and will show your cct with your permission.

Two things still at present though....it has not been tested in combat yet, so any publication may be premature.

Also I am still using my OR gate to add the 31 line pulses to the new 32 pulse, not you final version. I find that it is easier for me to understand doing it that way. I haven't yet finished that part but all the rest gives pulses OK....reset MR pulse/counter o/p/last stages. The OR gate should be tested today and I see no problems there.........
......provided that I don't put 12v on the wrong pin again !

The last duplicate 2-3ms stage that we added for 32 line pulse generation will require twigging with the pot I would think in order to put that new 32 pulse exactly between line 31 and line 1, hopefully within one pixel or less. I believe that I am thinking right here, but you never know with me.

[Steve Anderson]

...it has not been tested in combat yet, so any publication may be premature

Yes, agreed. being the positive-minded individual I am I'm thinking that this will in due course work and that a submission would be made for publication. So I'll leave it up to your good self to include said arrangement.

I've not had any contact from Jeremy in a while so I don't know what his status is regarding enough material for the newsletter. But it's clear that this is not going to be ready for the usual December issue. In conjunction with a demonstration at the convention next April would be ideal...but things in life don't generally work like that...

Steve A.

[Viewmaster]

...it has not been tested in combat yet, so any publication may be premature

Yes, agreed. being the positive-minded individual I am I'm thinking that this will in due course work and that a submission would be made for publication. So I'll leave it up to your good self to include said arrangement.
.

Of course it will work, Steve.....you designed the timing cct. and I built it
I tried the OR gate this morning but missing pulse still not there.......you see, I built it.

Then the penny dropped.............Your cct shows neg going input pulses, whilst mine are pos from the club sync seperator. So just a few changes to be made, which I had wrong.
I am new to these vibs with rise/falling triggered inputs and pos/neg outputs. But I have sorted it out now and retrieved the penny !
Just 6 input pins to be changed, that's all.

[
I've not had any contact from Jeremy in a while so I don't know what his status is regarding enough material for the newsletter. But it's clear that this is not going to be ready for the usual December issue. In conjunction with a demonstration at the convention next April would be ideal...but things in life don't generally work like that...
Steve A.

I would not write it up before it works, so Dec is out as you say, Steve.

Over 3,000 more joints to be done so we shall see for convention in time.
The demonstration, if any, may be of a NIPTRIX that doesn't work properly yet...........
"Roll up, roll up, see the jerky moving abstract pictures

[gary]

Over 3,000 more joints to be done so we shall see for convention in time.

If I was doing that I would be more worried about MY joints

Re newsletter, there are, of course, many examples of "work in progress" which are interesting in their own right (as this thread attests!). However, I must admit, unlike many brave souls in the newsletters and on this forum, I would prefer to have something working before sticking my neck out. The again I'm a known coward.