EEVblog #574 - NEC Analog TV IF Modulator Teardown

[mikeselectricstuff]

Maybe the reason for using a TCXO rather than an OCXO is that there is no warm-up time - for broadcast gear you want to be able to swap stuff out quickly in case of a fault - you don't want to have to wait for an oven to stabilise.

[Galaxyrise]

Getting all that scanned in looks like it could be quite a project! I'm definitely interested in the oven controller schematic, though.

[brabus]

Ladies, just one question, why everybody is like: "Oh, Dave, you should have said that and that", "Dave, you should have done more", blah blah blah...

You know what? Next time YOU go to the TV transmitter station, provide something interesting to teardown, and say whatever you want!
Sorry but I can't stand futile criticizing.



On my side, I can only say: GOSH.  AWESOME piece of engineering, wish I could smell that '80s essence...
You lucky 'astard, Dave! Go ahead with the other two puppies! 

[firewalker]

You should really call them and ask about the dust thing.

Alexander.

[SeanB]

SAW filter heating is simple, it is a crystal, and with temperature changes the dimensions change and the passband and rejection drifts, and the ripples in response vary up and down. Keeping at a constant temperature means the output filtering can be adjusted to handle the ripple and attenuate it. At the receive end the drift can be a lot worse as it has little visible effect, and the AFC will tend to compensate a lot for the drift.  At the transmitter end if it drifts it will drop out of the SAW filter ranges on some receivers and give rise to vision on sound or sound on vision as the filter drifts and the transmit frequency varies. Your transmitted signal has to be good enough so that a broadcast quality picture can be received on any good set. Most sets are far from that, but you will get a good number of very good receivers and the viewers of those will be the most critical of the broadcast quality. The whole chain is only as good as the worst part, and you really want the customer end to be that.

Dust is easy, you have HEPA filtering of the air coming in, and positive pressure inside the whole facility. Note the doors coming in are at least 3 sets of double doors with sealing strips on the sides, and this allows the area to be kept dust free, along with having dirt trapping mats on all entrances. Major source of dust is then the people.

Like the 4 trim cuts on that power heater resistor, all done so as to make the trim even out to a constant heat per segment. Nice work on that one. Also remember using those TO66 heatsink adaptors for TO39 packages as well. They are pretty expensive, costing a lot more than the transistor itself.

[jippie]

What is a "pedestal clamp"? Eg. @ 0:52:33.

[vk6zgo]

What is a "pedestal clamp"? Eg. @ 0:52:33.

The short intervals of blanking level (PAL) or black level (NTSC) before & after the Sync Pulse in the  Composite Video signal are collectively known as the "Pedestal".

"Keyed Clamping" is a DC restoration technique which  produces a narrow pulse from the sync pulse.
This is then delayed in time until it is coincident with the "back porch",where a switch (classically a diode bridge,or,as in this case,a FET) is turned on by the pulse,returning the signal line to zero volts,or whatever voltage you wish to clamp the signal blanking (PAL),or black level (NTSC) to.

This can be used to remove such signal perturbations as hum,loss of DC component following  an AC  coupled Amplifier,etc.

With colour TV,care must be taken that the clamp does not distort the colour burst.

[ram]

dave scanned documentation please it really helps in my under graduation programme studies

[jippie]

What is a "pedestal clamp"? Eg. @ 0:52:33.

The short intervals of blanking level (PAL) or black level (NTSC) before & after the Sync Pulse in the  Composite Video signal are collectively known as the "Pedestal".

For a moment I thought you meant the front and back porch until I read:

This is then delayed in time until it is coincident with the "back porch",

But I get the essence of what you wrote, thnx!

[jnissen]

30 minutes of reading from a manual! If YouTube has specs on how many folks fast forwarded through your first 30 minutes I bet it would be high! 

[JoannaK]

30 minutes of reading from a manual! If YouTube has specs on how many folks fast forwarded through your first 30 minutes I bet it would be high! 

Well.. I didn't. I don't know what it tells about me, but I watched it entirely. 

Of course it would be nicer to read the copy of the manual at the same time as the Dave talks about it, but it's not mandatory.

[N2IXK]

I watched the whole damned thing, as well. 

As far as what else could have been shown, how about firing the system up, applying audio and video signals, and showing the output on a spectrum analyzer and a TV set?

Signal tracing the video through all the processing and correction stages (and tweaking the trimpots) would be neat, too, but probably a bit too much for most viewers.

[EvilGeniusSkis]

Dave, @1:00:00 I think the caps are hexagonal so that they tessellate(fit together better).

[FrankBuss]

30 minutes of reading from a manual! If YouTube has specs on how many folks fast forwarded through your first 30 minutes I bet it would be high! 

Youtube has such a statistic, it is called "Audience retention", but only the uploader has access to it (for one of my videos:

). I admit, I skipped the manual reading as well. Might be interesting to see the curves for this video

[jippie]

30 minutes of reading from a manual! If YouTube has specs on how many folks fast forwarded through your first 30 minutes I bet it would be high! 

Dave might have hidden a secret give-away somewhere half way the manual reading for all you know I know I carefully viewed the entire video

[david77]

I also watched Dave reading from the manual, I love this kind of technical documentation. The manuals for analogue broadcast equipment are especially interesting, it's fascinating to see how the engineers went to town on stuff where money was no object.

[Rasz]

Dave, I'm not sure the content fulfilled the expectations given by the "teaser". It's a good teardown, but certainly not your best, with respect, and I felt a bit deflated... after 3 weeks of expecting something more.

Like what?

30 minutes of reading from a manual! If YouTube has specs on how many folks fast forwarded

as a matter of a fact they do have this data I dont know if they share it with content creators, but they do collect it.
DASH
http://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
lets them know which parts of the video you request + YT player pings back which parts you actually watched.

edit:

Youtube has such a statistic, it is called "Audience retention", but only the uploader has access to it (for one of my videos:

)

ha

[Towger]

It really needs to be fired up one last time.

[vk6zgo]

What is a "pedestal clamp"? Eg. @ 0:52:33.

The short intervals of blanking level (PAL) or black level (NTSC) before & after the Sync Pulse in the  Composite Video signal are collectively known as the "Pedestal".

For a moment I thought you meant the front and back porch until I read:

This is then delayed in time until it is coincident with the "back porch",

But I get the essence of what you wrote, thnx!

Yes,the "Pedestal" is the  front & back porches,so clamping on the back porch is "Pedestal" clamping.

"Pedestal" is an old term,which is apparently more used in Japan than elsewhere.

I think it originated in the UK or USA---looking at an inverted sync pulse it looks a bit like a monument mounted on a Pedestal,so  that's probably where the name came from.

It is easier to generate the clamp pulse & delay it a few microseconds to clamp on the back porch,than to clamp sync tips (which is done),as that requires either a very fast clemp pulse generator to clamp on the same sync pulse.,or a very long delay(64uS),so as to clamp on the next pulse.
The front porch duration is quite short,& I have never seen any attempt to clamp there.

[vk6zgo]

I watched the whole damned thing, as well. 

As far as what else could have been shown, how about firing the system up, applying audio and video signals, and showing the output on a spectrum analyzer and a TV set?

Signal tracing the video through all the processing and correction stages (and tweaking the trimpots) would be neat, too, but probably a bit too much for most viewers.

Dave doesn't have the mixer (upconverter) so nix on the TV set,but he could do the rest.
But---Guys,we have to remember Dave's got a life outside letting us "perve" on interesting Electronics!

[dave_k]

It really needs to be fired up one last time.

Gee, I'm tempted to fire up the old Ch 7 transmitter one last time into a test load. It would be interesting to see if it all works properly after sitting for 8 weeks with no power!

[vk6zgo]

It really needs to be fired up one last time.

Gee, I'm tempted to fire up the old Ch 7 transmitter one last time into a test load. It would be interesting to see if it all works properly after sitting for 8 weeks with no power!

It's an NEC,Dave,------of course it will!

What I liked when we installed the second 13kW Tx at TVW7,was that our test results were exactly the same as those achieved by the last test at the factory before they pulled it apart for delivery.

With most other Manufacturers you have to adjust them on arrival to get them into spec!

[aroby]

Great teardown.

How would the manufacturer go about testing something like this?   The debugging must have been horrendous.  Today you can use  computer simulations, back 35 years ago, how would this be done?

[coppice]

1) As a guess, the SAW filter is temperature stabilized to control differential phase shift.  You really do not want the relative phase across the passband to change in color TV system as it will screw up the displayed colors. (This may be more an issue with NTSC  than PAL, however. All my limited experience is with NTSC systems.)

2) The only TV-locked frequency reference systems I'm familiar with tied to the color burst.  In the USA, network generated programming used a color reference tied to a rubidium standard, but locally originated programming used crystal based color burst. (Talking about the 1970's when rubidium standards were extremely expensive.) There were kits available that would tap into the video of a TV set and bring out the color subcarrier reference. When live (not tape delayed) network programming was being aired, this provided a stable and accurate 3.58... MHz reference tied back to a Rb standard sitting in a rack in New York or Los Angeles.  However, as technology advanced, the locally transmitted color reference was no longer tied to the network Rb standard - the exact details of the change escape me at the moment. Might have been frame buffering or the change to greater video tape use and satellite distribution instead of coaxial cable leased from the telephone companies or something else. Way too many years ago and it was not an area I worked in.

1. Piezo materials for saw filters are not the most temperature stable in the world. They are often temperature controlled, typically by peltier devices.

2. Rhubidium locking for TV signals was really important in the 1960s, 70s, and into the early 80s, as changing the frame rate or frame phase of a signal was really hard. Large studio complexes had a room full of coax cables of different lengths, and other forms of delay line. Those gave you the various delays needed to perform frame phase adjustments. The atomic clock ensured the frame phase slewed so slowly it didn't matter. As people like Quantel massively brought down the cost of frame stores, and frame rate adjustment, the need to atomic locking went away.

There was never a need for an accurate carrier frequency for TV transmitters. TCXOs were normal for setting the carrier frequency. They can give you something like 1in 10^7 without any warmup time, which is goo d enough. The ones in this unit must have been made just prior to the switch to digitally controlled TCXOs.

[vaualbus]

I'd have liked to see a tear down of the power output section but ...

In case it wasn't clear enough, teardowns on the other two bits will follow...
The intention was to show all 3 briefly, but you know who takes always works out...

So they other module will come.... Almost two yaers after no more video on any of them.......
Also you never scanned the schematics!