Multi-Channel VHF Modulator Design - Page 2 - UK Vintage Radio Repair and Restoration Discussion Forum

oldeurope · 25th Feb 2005, 9:58 pm

Hi David,
I have a few ideas to use the AM sound section of the IC.

One way is to change the 4MHz XTAL.

But there is an other independent way I tested.

This IC uses an intercarrier sound mixing system.
You can see on the Analyzer two sound carriers one below
and one over the picture carrier.

This means the sound VCO works on 3,5 MHz not at the
high output frequency like the vision carrier does.

So you can use the 3,5 MHz sound osc without a PLL.
You have to put a DC voltage from a voltage divider
to the input SPLLFLT. I used a 10uF block and a 1K variable
resistor and 1K2 to VCCA. (ca.0,9V)
The frequency stability is much better than I expectet!

Kind regards
Darius

David_Robinson · 28th Feb 2005, 4:47 pm

I have now drawn the schematic in OrCAD, saved as PDF, which is a lot easier to read than the previous photos. There is one more thing outstanding namely the 3.5MHz notch filter on the video input. I am just waiting for parts to arrive, then I'll update the drawing again.

oldeurope · 28th Feb 2005, 5:37 pm

Hi David,
thanks for posting your schematic. Looks very easy. The most work
is to programm the PIC. I am not able to do this.
Congratulationes

Please look at the basis of the PNP audio emitterfollower.
The input resistance is ca. 100K Ohms, so a 0,1uF cer. cap
is better than the 10uF electrolytic. The leak current of this cap
can change the bias! (picture thread 8)
It is better to place the electrolytic from the bias pot to ground
to block noise from the 5V.
The bias poti is not a level adjust.

I am sure your PIC will work in the axing modulator too.
Did you switch off the audio carrier or not?
Please note that the audio carrier level should be 7dB under peack
white of the vision carrier.

Kind regards
Darius

flyback · 28th Feb 2005, 5:57 pm

Hi all,

I agree with Darius about the Electrolytic as a coupler and the Bias de-coupling, I use Tantalums in my circuits.

Darius, the Bias pot will set the "resting" level of the sound carrier(which is really coming from a vision modulator with it's clamp overidden), with it you can set the ratio of the sound to vision carriers correctly.

On Domino I did not provide two L/R inputs for sound, I assumed only mono vintage material would be played, most Vintage DVD TV stuff is mono!.

I have never seen/heard the need for a 3.5MHz filter on the video feed, your video should be band limited to ~3.0MHz, is this somthing that you have found is a problem, maybe with 405 line VHS playback?.

Malcolm

oldeurope · 28th Feb 2005, 6:16 pm

Hi Malcolm,
not from VHS of course, but from the camera. My camera makes
vision components at 3,5MHz.
The converter too if you have no filter at the output of the DA converter.

The bias adj. must be set 0,5V higher than the sync tips would be.
(Hope you can understand my english)
Than you can use the full dynamic of the modulator and the carrier
is 6 dB under peak white. Of course the bias changes the audio level
too. But is it not the audio level adjustment!!!

Kind regards
Darius

David_Robinson · 28th Feb 2005, 6:19 pm

Quote:

Originally Posted by oldeurope

... programm the PIC. I am not able to do this.

If you want to build this design I can send you the programming file. You just need a PIC programmer (mine cost about £50).

Quote:

The leak current of this cap can change the bias!

The cap I used is a ceramic chip (surface mount) not an electrolytic. The leakage is very low.

Quote:

... place the electrolytic from the bias pot to ground to block noise from the 5V.

Good point although I haven't noticed a problem.

Quote:

Originally Posted by flyback

... your video should be band limited to ~3.0MHz, is this somthing that you have found is a problem?

Well, keep this to yourself as it is very hush-hush, but I am developing a standards converter too, and this produces the input video bandwidth times 405/625 which is a bit more than 3MHz...

oldeurope · 28th Feb 2005, 6:34 pm

Quote:

The cap I used is a ceramic chip (surface mount) not an electrolytic. The leakage is very low.

10uF

?

Darius

flyback · 28th Feb 2005, 7:15 pm

Hi all in Telly land..

10µF in ceramic is get-able but at £1.20 each (1210 size, Farnell 578-289) thats a costly coupling cap! and you show three of them.

David_Robinson said..
"Well, keep this to yourself as it is very hush-hush, but I am developing a standards converter too, and this produces the input video bandwidth times 405/625 which is a bit more than 3MHz... "

Ooh er!, do tell, will this be a design you publish or something you will make to sell?.

I may have to add a "Go Faster Stripe" to Domino

or make a ++Version!

Malcolm

oldeurope · 28th Feb 2005, 7:15 pm

http://www.voti.nl/swp/

Good evening David,

thanks for your offer.

First I have to learn something about PIC's
I found some fundamental infos on this website.

Kind regards
Darius

David_Robinson · 28th Feb 2005, 7:44 pm

Quote:

Originally Posted by flyback

... 10µF in ceramic is get-able but at £1.20 each

Blimey! You're right, that is expensive. I used Farnell 471-6711 at £0.12 each.

Quote:

Ooh er!, do tell

To be honest I haven't really thought that far ahead and I'm some way from completion at the moment.

flyback · 1st Mar 2005, 1:08 am

Hi all..

Yes Dave you’re right, upon further digging in the catalogue (Tiny print to go with Tiny components!) I see the range of SM Caps, it says there intended as an alternative to Electrolytic and Tantalum capacitors and cheaper too.

Very impressive, I didn’t know there was the capacity to make capacitors with that capacitance in cases with such a small capacity!!.

Apart from the ADC and Frame FIFO’s the only SMD in Domino is a BB149A varicap diode, it’s ‘the’ most fiddly part to fit!, I’ve often twanged one off into the land of the lost never to be seen again.

Dave said,
'To be honest I haven't really thought that far ahead and I'm some way from completion at the moment.'

I’ll hold off on the ‘go faster stripe and UV under lights’ then, please tell me your not using the dreaded ‘Xilinx’ or some such device in your converter..
(quiet Jeff B will hear!

).

Malcolm

David_Robinson · 1st Mar 2005, 8:55 am

Hi Malcolm, have you tried the search facilities on the Farnell website? They are excellent - you can search for a 10µF capacitor and then sort the results by price. No tiny print here... As I'm sure you know, prices are typically 20% below RS.

Xilinx (spit)? No no, I'm using the market leader, ALTERA! (Sorry Jeffrey). 100 pin chips with 0.5mm pin spacing - steady hand required.

I'm all too aware though, that it's one thing to build a working prototype but it's an order of magnitude more impressive to produce a marketable product with a proper PCB and all that. I leave that kind of thing to others.

David_Robinson · 3rd Mar 2005, 11:53 am

OK, attached is the latest version of the modulator schematic. This includes the 3.5MHz notch filter, if anyone wants it.

Quote:

Originally Posted by flyback

On Domino I did not provide two L/R inputs for sound, I assumed only mono vintage material would be played

Not at all - have a look at http://homepage.ntlworld.com/david.r...20Majestic.JPG . A 10" loudspeaker and 1 watt "undistorted" output. What more could you want?

oldeurope · 3rd Mar 2005, 5:54 pm

Hi David,

why did you use an NPN in the video buffer?

Kind regards
Darius

David_Robinson · 4th Mar 2005, 9:16 am

Hi Darius

The video buffer is there so that the input cable sees a more accurate 75Ω termination. This avoids any reflections in the cable.

Quote:

Originally Posted by oldeurope

Did you switch off the audio carrier or not?

Sorry I did not answer this question before. Yes, the sound carrier on both modulator chips is switched off.

oldeurope · 4th Mar 2005, 9:21 am

Good morning David,

no question that you need a buffer.
But why an NPN and not a PNP Transistor?

Kind regards
Darius

David_Robinson · 4th Mar 2005, 10:37 am

I don't think it makes much difference either way. In theory an NPN will give better rejection of noise on the +5v rail, but I doubt you could measure the difference. The PIC doesn't run continuously, so there's really nothing to make any noise...

ppppenguin · 4th Mar 2005, 10:59 am

It really doesn't make much difference for an emitter follower (unlike a common emitter stage where the input and output are referenced to different power rails). A small amount of muck on the +ve rail can be injected via the bias chain for PNP or NPN. For NPN there is virtually no sensitivity to the +ve rail. For PNP it's a very slight sensitivity due to the finite output impedance.

oldeurope · 4th Mar 2005, 11:24 am

Hi,

it is not the noise from the DC that makes an NPN a better choise.
With your biasing the output resistance is max 7 Ohms.

There is an other problem.
After this stage the DC restorer followes.
There goes a high current through the 10nF at the sync tips.
So you must make sure that there is a strong source to pull
down!
A NPN emitter follower is a strong source to pull up and
a PNP emitter follower is a strong source to pull down.
This is a very importand aspect.

In your case the 470 Ohms makes the pull down.
So if you have a DC restorer as a load and a negative sync
it is better to take a PNP !!!

The output resistance of the emitter follower should be as low
as possible to make sure that the sync will not be damaged
or gets a level reduce.

You placed a 180 Ohms at the output of the emitter follower
and this will be added to the output resistance.

In front of the DC restorer this is a bad mistake.

It is a DC restorer
not a sync separator!!!

There are a lot of commercial modulators that makes trouble at the top of
the picture because of sync level changes and other mistakes
caused by a wrong driven DC restorer.

I hope I was able to show you that there is a big difference in
the use of a PNP or NPN emitter follower.
In this case the PNP is the better one.

Kind regards
Darius

ppppenguin · 4th Mar 2005, 12:17 pm

I was looking at the audio emitter followers. For video you need to take account of the load as Darius correctly says. If the load is a DC restorer, this can take high-ish current pulses at the sync tips so for -ve syncs a PNP follower is better. If the input of the modulator chip is a keyed clamp the current pulses could be of either polarity. The series resistance definitely does not help a DC restorer but won't make much difference to a keyed clamp unless you are trying to make a very hard clamp. You do not want a hard clamp at the input to a modulator as it will distort the colour burst.

25th Feb 2005, 9:58 pm	#21
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Hi David, I have a few ideas to use the AM sound section of the IC. One way is to change the 4MHz XTAL. But there is an other independent way I tested. This IC uses an intercarrier sound mixing system. You can see on the Analyzer two sound carriers one below and one over the picture carrier. This means the sound VCO works on 3,5 MHz not at the high output frequency like the vision carrier does. So you can use the 3,5 MHz sound osc without a PLL. You have to put a DC voltage from a voltage divider to the input SPLLFLT. I used a 10uF block and a 1K variable resistor and 1K2 to VCCA. (ca.0,9V) The frequency stability is much better than I expectet! Kind regards Darius Attached Thumbnails

28th Feb 2005, 5:37 pm	#23
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Hi David, thanks for posting your schematic. Looks very easy. The most work is to programm the PIC. I am not able to do this. Congratulationes Please look at the basis of the PNP audio emitterfollower. The input resistance is ca. 100K Ohms, so a 0,1uF cer. cap is better than the 10uF electrolytic. The leak current of this cap can change the bias! (picture thread 8) It is better to place the electrolytic from the bias pot to ground to block noise from the 5V. The bias poti is not a level adjust. I am sure your PIC will work in the axing modulator too. Did you switch off the audio carrier or not? Please note that the audio carrier level should be 7dB under peack white of the vision carrier. Kind regards Darius Last edited by oldeurope; 28th Feb 2005 at 5:50 pm. Reason: picture thread 8

28th Feb 2005, 5:57 pm	#24
flyback Retired Dormant Member Join Date: Apr 2003 Location: Swindon Wiltshire Posts: 26	Re: Multi-Channel VHF Modulator Design Hi all, I agree with Darius about the Electrolytic as a coupler and the Bias de-coupling, I use Tantalums in my circuits. Darius, the Bias pot will set the "resting" level of the sound carrier(which is really coming from a vision modulator with it's clamp overidden), with it you can set the ratio of the sound to vision carriers correctly. On Domino I did not provide two L/R inputs for sound, I assumed only mono vintage material would be played, most Vintage DVD TV stuff is mono!. I have never seen/heard the need for a 3.5MHz filter on the video feed, your video should be band limited to ~3.0MHz, is this somthing that you have found is a problem, maybe with 405 line VHS playback?. Malcolm

28th Feb 2005, 6:16 pm	#25
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Hi Malcolm, not from VHS of course, but from the camera. My camera makes vision components at 3,5MHz. The converter too if you have no filter at the output of the DA converter. The bias adj. must be set 0,5V higher than the sync tips would be. (Hope you can understand my english) Than you can use the full dynamic of the modulator and the carrier is 6 dB under peak white. Of course the bias changes the audio level too. But is it not the audio level adjustment!!! Kind regards Darius

28th Feb 2005, 7:15 pm	#28
flyback Retired Dormant Member Join Date: Apr 2003 Location: Swindon Wiltshire Posts: 26	Re: Multi-Channel VHF Modulator Design Hi all in Telly land.. 10µF in ceramic is get-able but at £1.20 each (1210 size, Farnell 578-289) thats a costly coupling cap! and you show three of them. David_Robinson said.. "Well, keep this to yourself as it is very hush-hush, but I am developing a standards converter too, and this produces the input video bandwidth times 405/625 which is a bit more than 3MHz... " Ooh er!, do tell, will this be a design you publish or something you will make to sell?. I may have to add a "Go Faster Stripe" to Domino or make a ++Version! Malcolm

1st Mar 2005, 1:08 am	#31
flyback Retired Dormant Member Join Date: Apr 2003 Location: Swindon Wiltshire Posts: 26	Re: Multi-Channel VHF Modulator Design Hi all.. Yes Dave you’re right, upon further digging in the catalogue (Tiny print to go with Tiny components!) I see the range of SM Caps, it says there intended as an alternative to Electrolytic and Tantalum capacitors and cheaper too. Very impressive, I didn’t know there was the capacity to make capacitors with that capacitance in cases with such a small capacity!!. Apart from the ADC and Frame FIFO’s the only SMD in Domino is a BB149A varicap diode, it’s ‘the’ most fiddly part to fit!, I’ve often twanged one off into the land of the lost never to be seen again. Dave said, 'To be honest I haven't really thought that far ahead and I'm some way from completion at the moment.' I’ll hold off on the ‘go faster stripe and UV under lights’ then, please tell me your not using the dreaded ‘Xilinx’ or some such device in your converter.. (quiet Jeff B will hear! ). Malcolm

1st Mar 2005, 8:55 am	#32
David_Robinson Retired Dormant Member Join Date: Feb 2004 Location: Nottingham, UK Posts: 94	Re: Multi-Channel VHF Modulator Design Hi Malcolm, have you tried the search facilities on the Farnell website? They are excellent - you can search for a 10µF capacitor and then sort the results by price. No tiny print here... As I'm sure you know, prices are typically 20% below RS. Xilinx (spit)? No no, I'm using the market leader, ALTERA! (Sorry Jeffrey). 100 pin chips with 0.5mm pin spacing - steady hand required. I'm all too aware though, that it's one thing to build a working prototype but it's an order of magnitude more impressive to produce a marketable product with a proper PCB and all that. I leave that kind of thing to others.

3rd Mar 2005, 5:54 pm	#34
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Hi David, why did you use an NPN in the video buffer? Kind regards Darius

4th Mar 2005, 9:21 am	#36
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Good morning David, no question that you need a buffer. But why an NPN and not a PNP Transistor? Kind regards Darius

4th Mar 2005, 10:37 am	#37
David_Robinson Retired Dormant Member Join Date: Feb 2004 Location: Nottingham, UK Posts: 94	Re: Multi-Channel VHF Modulator Design I don't think it makes much difference either way. In theory an NPN will give better rejection of noise on the +5v rail, but I doubt you could measure the difference. The PIC doesn't run continuously, so there's really nothing to make any noise...

4th Mar 2005, 10:59 am	#38
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Multi-Channel VHF Modulator Design It really doesn't make much difference for an emitter follower (unlike a common emitter stage where the input and output are referenced to different power rails). A small amount of muck on the +ve rail can be injected via the bias chain for PNP or NPN. For NPN there is virtually no sensitivity to the +ve rail. For PNP it's a very slight sensitivity due to the finite output impedance.

4th Mar 2005, 11:24 am	#39
oldeurope Retired Dormant Member Join Date: Dec 2003 Location: Solingen, Germany Posts: 727	Re: Multi-Channel VHF Modulator Design Hi, it is not the noise from the DC that makes an NPN a better choise. With your biasing the output resistance is max 7 Ohms. There is an other problem. After this stage the DC restorer followes. There goes a high current through the 10nF at the sync tips. So you must make sure that there is a strong source to pull down! A NPN emitter follower is a strong source to pull up and a PNP emitter follower is a strong source to pull down. This is a very importand aspect. In your case the 470 Ohms makes the pull down. So if you have a DC restorer as a load and a negative sync it is better to take a PNP !!! The output resistance of the emitter follower should be as low as possible to make sure that the sync will not be damaged or gets a level reduce. You placed a 180 Ohms at the output of the emitter follower and this will be added to the output resistance. In front of the DC restorer this is a bad mistake. It is a DC restorer not a sync separator!!! There are a lot of commercial modulators that makes trouble at the top of the picture because of sync level changes and other mistakes caused by a wrong driven DC restorer. I hope I was able to show you that there is a big difference in the use of a PNP or NPN emitter follower. In this case the PNP is the better one. Kind regards Darius Last edited by oldeurope; 4th Mar 2005 at 11:33 am. Reason: grammar

4th Mar 2005, 12:17 pm	#40
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Multi-Channel VHF Modulator Design I was looking at the audio emitter followers. For video you need to take account of the load as Darius correctly says. If the load is a DC restorer, this can take high-ish current pulses at the sync tips so for -ve syncs a PNP follower is better. If the input of the modulator chip is a keyed clamp the current pulses could be of either polarity. The series resistance definitely does not help a DC restorer but won't make much difference to a keyed clamp unless you are trying to make a very hard clamp. You do not want a hard clamp at the input to a modulator as it will distort the colour burst.