UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (https://www.vintage-radio.net/forum/index.php)
-   Television Standards Converters, Modulators etc (https://www.vintage-radio.net/forum/forumdisplay.php?f=95)
-   -   My 819 line French TV project (https://www.vintage-radio.net/forum/showthread.php?t=167674)

Semir_DE 7th Jun 2020 8:55 pm

My 819 line French TV project
 
5 Attachment(s)
Dear fellow vintage TV enthusiasts,

Last year I was able to get my hands on a couple of TV sets that support the French 819 Line standard, so naturally I wanted to generate some signals for those sets. Among the sets I have are a Sony TV9-90UM the French version of the TV 9-90UB and a Radiola (Built by Philips) TV originally built for the Belgian market. This latter set is in mint condition and the CRT is like new.

I ruled out using an up-conversion from 625, because I wanted to get an idea about the true resolution of this vintage HD format. So the best approach in my mind was to use a PC to generate the required signals from 720p and 1080p footage.

Luckily I had come across Kat Mantons threads regarding this subject matter which helped me a lot in getting my project up and running. I would like to thank Kat for sharing this information regarding the mode lines for 819, this is much appreciated!

Modern Linux systems can be made to generate odd resolutions very easily nowadays. Here are the steps I took to get my Linux Mint system running on an HP 8440p laptop computer to output 738i i.e. the French 819 line system.:

Step #1) First we need to find out the internal designation of the VGA port. In order to do that we need to open the console and input this command:

xrandr | grep -e " connected [^(]" | sed -e "s/\([A-Z0-9]\+\) connected.*/\1/"

This will yield some output like:

LVDS1
VGA-0

Hint: You can copy the above by using the usual "control+c" keys. In the console you need, however, to use "Shift+Control+v" to paste.

The first is the internal LCD panel of my laptop, the second (VGA-0) is the VGA port. Please take note that the VGA port will only show up in the list if something like a monitor or adapter is connected to it. Here "VGA-0" is just an example, depending on the PC system used this may have other numbers.

Step #2) Now we need to generate the new resolution using Kat Mantons mode line. We key in this command in the console:

xrandr --newmode "819_TV" 29.3203 1152 1168 1240 1432 738 744 745 819 -hsync -vsync interlace

and make it known to the VGA port by using this command:

xrandr --addmode VGA-0 819_TV

As can be seen we use the internal designation of the VGA port "VGA-0" obtained in step #1 for this second command. If everything went well the VGA output should now have an 819 line signal at its output. This will show up as a second monitor to the right of the current display. In order to play a video through the VGA port simply drag the the player window to the "monitor" on the right and put it in full screen mode. I usually use VLC and set its aspect ratio to 2,21:1 to get the correct format of 16:9 footage on a 4:3 screen.

Please take note that the new resolution is not permanent and will have to be re-configured after a re-boot. In order to make this change permanent the easiest way is to edit a file named ".profile" that can be found i the home folder. Just add the two lines from step two at the bottom of the file.

Hint for the Linux novice: Please take note that this is a hidden file and you will only be able to see it if you hit Control+h on the keyboard. Any file or directory in Linux that begins with a "." will be hidden.

In order to obtain a good quality 1Vpp video signal from the VGA port of my PC I built a circuit that will entirely run off the 5V supplied by the VGA port on pin #9. The circuit itself runs on 3,3V and provides 2 outputs with 1Vpp into a 75Ω load. Since not all PCs and graphics adapters provide a 5V output powerful enough to run the circuit which needs around 16mA I have made provisions for an extra power input via a micro USB port. A blue LED lights up if the supplied voltage is high enough to power the circuit i.e. above 4.2V.

I use the THS7314 video amplifier to drive the outputs and amplify the signal to the required 1Vpp. A simple matrix generates the luminance signal from the red, green and blue signals. The H and V sync signals are fed into an XOR gate to turn them into a form of composite sync. This is not fully system compliant but my TVs work very well with this signal. In order for this to work with only a single 74HC1G86 XOR gate the vertical sync from the VGA port has to be inverted. This can be achieved by simply changing the mode line at the end. instead of having "-vsync" there it should be changed to "+vsync" that's it. Here is the complete mode line:

"819_TV" 29.3203 1152 1168 1240 1432 738 744 745 819 -hsync +vsync interlace

Here is the circuit diagram and a few images of the board I have designed in KiCAD. It mostly uses SMD components.

Attachment 207978

Attachment 207979

Attachment 207980

The project data for this board is available in KiCAD format. Please send me a PM or mail and I will send it to you.

In order to get the signal onto an RF carrier I used a modified version of my system A modulator that I had introduced here in the past. This uses only one crystal to generate the vision carrier and a simple LC circuit for generating the difference frequency between vision and sound carriers which in this case is 11,15Mc. I was able to use an original FM IF filter for the oscillator that generates the 11,15Mc intermediate carrier. This intermediate carrier is then mixed with the vision carrier signal and depending on the French channel to be generated the difference or sum frequency needs to be filtered out giving the sound carrier. The sound carrier generated by this method then goes to an AM modulator that is fed with the audio signal. I shall be posting the final circuit of my system A modulator with the French modification here in an upcoming post.

Attachment 207982

I was lucky to find a 56,57Mc crystal in my junk box which is close enough to the French channel F3 at 56,15Mc. Since channel F3 is an odd numbered channel the sound carrier in this case is above the vision carrier by 11,15Mc i.e. at 67,30Mc or in my case due to the slightly off crystal at 67,72Mc. It is worth noting that even numbered channels have the sound carrier below the vision carrier in the French system, while odd numbered ones have it above the vision carrier. So I tuned the sound carrier filter in my modulator to 67,7Mc and was able to get sound and vision in glorious 819 lines on my TVs...

Attachment 207981

To be continued...

Richard_FM 7th Jun 2020 10:39 pm

Re: My 819 line French TV project
 
Sounds interesting, I've seen two 819 line sets at different times running off an 819 line Aurora, both of which were producing a very sharp & clear picture.

Jac 8th Jun 2020 6:46 am

Re: My 819 line French TV project
 
2 Attachment(s)
Hi Semir,

Good results!

I recently restored my Philips TF651A 819 line set.
Pictures, see below.

Jac

FRANK.C 8th Jun 2020 8:23 am

Re: My 819 line French TV project
 
Hi Semir
Great results. 819 lines looks really good.

Frank

FRANK.C 8th Jun 2020 9:04 am

Re: My 819 line French TV project
 
Hi again Semir
Just wondering how the THS7314 video amplifier preforms with 819 lines.
Does it's 8.5 MHz low pass filter have any noticeable effect on the video? I suspect not.

Frank

Richard_FM 8th Jun 2020 11:08 am

Re: My 819 line French TV project
 
2 Attachment(s)
Here are the two I've seen at events, the first is the Radiola on the left, and the next year I managed to get a close-up of the Philips.

Semir_DE 8th Jun 2020 7:13 pm

Re: My 819 line French TV project
 
Quote:

Originally Posted by FRANK.C (Post 1257295)
Hi again Semir
Just wondering how the THS7314 video amplifier preforms with 819 lines.
Does it's 8.5 MHz low pass filter have any noticeable effect on the video? I suspect not.

Frank

Hi Frank,

Before using the THS7314 I had built a simple adder network to get the luminance signal. I can not see any difference between this solution and the simple one. Since my TVs are of later build their IF bandwidth is limited at around 6Mc which poses the real bottleneck in the signal chain. So no the THS 7314 works fine in this application. The only effect I could see was in the H-sync this had some ringing which I had not noticed in the simple solution. I added the 1kΩ/33pF LPF to reduce high frequency content in the sync and then it passed through the THS7314 without exhibiting the ringing.

I do have a friend with an older French TV that may have the full 10Mc bandwidth. When he gets it repaired I will check it out. I do need to get some test cards though to be able to better judge the resolution.

Cheers

Semir

rambo1152 8th Jun 2020 9:33 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Hi Semir,

This post documents my "Eureka" moment when I got Kat's idea to work with a Windows PC, simply by using a compatible graphics card and Jeroni Paul's "WinModelines"
I use it for 405 lines, Peter Scott uses it for his pre war 405/240 line set.

I use this modulator
and this very simple interface,Attachment 208059 as described by Jeroni Paul.

Here's the system in action.

https://youtu.be/ahzdx4KgWBg

peter_scott 8th Jun 2020 10:15 pm

Re: My 819 line French TV project
 
I had a similar problem with the simple XOR of the H and V syncs on 405 lines and built a crude monostable arrangement to get a truer pulse stream.

Peter

peter_scott 9th Jun 2020 10:01 am

Re: My 819 line French TV project
 
1 Attachment(s)
I don't have an 819 line set but looking back I see that I used a different modeline to you not as high resolution but on VHF 405 my Sony 9-90 couldn't deal with higher bandwidth signals:

Modeline "600x778_25 20.5kHz 50.0Hz" 15.070 600 608 683 736 778 786 800 819 interlace -hsync -vsync

but could only display it on 405 lines.

Peter

Semir_DE 9th Jun 2020 4:54 pm

Re: My 819 line French TV project
 
Quote:

Originally Posted by rambo1152 (Post 1257514)

and this very simple interface,Attachment 208059 as described by Jeroni Paul.

Hi Graham,

this is the circuit I used initially before making the more sophisticated one. I had also come across the site with all the mode line info a few years ago but had this project on the back burner. Now that I am on the German version of Furlough (Kurzarbeit) thanks to corona I have some extra free time to work on many neglected projects...;D

Thanks for posting the video, your TV looks great! which reminds me that I still have to finish restoring my Bush TV62...

Semir_DE 9th Jun 2020 5:12 pm

Re: My 819 line French TV project
 
Quote:

Originally Posted by peter_scott (Post 1257641)
I don't have an 819 line set but looking back I see that I used a different modeline to you not as high resolution but on VHF 405 my Sony 9-90 couldn't deal with higher bandwidth signals

Peter

Hi Peter, That is interesting that the 9-90UB can display the 819 signals in the way you show. Well the H frequency of 819 is roughly twice that of the 405 signal so I guess it syncs on every second H-pulse...

I have also thought about using some additional logic to get more compliant V pulses, but my TVs so far are quite happy with the current offering of XORed pulse trains. The French V-sync is basically very simple on odd fields it's just a single 20s pulse starting from where the H-sync would've started. In the even field the 20s V-sync starts halfway through the line and ends just before the next H-Sync. Since the TVs are happy I will leave things as they are for now, however.

peter_scott 9th Jun 2020 7:29 pm

Re: My 819 line French TV project
 
Hi Semir,

I hadn't looked at the 819 line waveform before. As you say it's very simple and I understand now why you are staying with the XOR.

Thanks,

Peter

Semir_DE 9th Jun 2020 7:46 pm

Re: My 819 line French TV project
 
2 Attachment(s)
Hi Peter,

here is the spec for the original waveform of the French system:

Attachment 208135

Attachment 208136

I think it is quite surprising that the 20s pulses are sufficient for v-sync...

Since the granularity of the mode line configuration for vertical timings is one full line, the PC generated sync which is a full line of 49s in each field is more than twice as long as the original one, but it works...

peter_scott 9th Jun 2020 9:02 pm

Re: My 819 line French TV project
 
Thanks Semir, I've copied the two docs to my standards conversion file. Your screen picture sync looks good.

Peter

Karen O 10th Jun 2020 12:54 am

Re: My 819 line French TV project
 
A very impressive accomplishment!

Maximum respect! :)

Semir_DE 11th Jun 2020 11:02 pm

Re: My 819 line French TV project
 
2 Attachment(s)
Dear all,

here is an update of this project. After mulling over the modulator design I decided to build both the VGA to composite video converter and the A/V modulator onto a single PCB. The design is SMD-based an will fit in an aluminium box roughly the size of a cigarette pack. Power options will be either USB or VGA, although I believe the latter will not work on most PCs. Here is the circuit diagram some values might still change slightly when I build the circuit:

Attachment 208294

I have included a low pass filter at the output in order to prevent hamonics from spilling into Band III where I have all my 625 channels. Band I will be used for exotic signals such as this one.

I have found a company that still offers to make custom crystals in small quantities, at 30 each not a cheap option though. I have ordered two vision and two sound carrier crystals each for the French F3 Channel i.e. 56.15Mc vision, and 67.3Mc sound. They are scheduled to arrive in early July.

Today I finished the first revision of the PCB. This will work as a single layer, but I will use a double layer one where the top layer will be reserved for ground. Since this is an RF circuit it is always a good idea to have a large ground plane:

Attachment 208295

Weather permitting i.e. if it rains :) over the weekend I will assemble a prototype and post results here.

Cheers

Semir

Jac 13th Jun 2020 7:23 am

Re: My 819 line French TV project
 
Impressive work Semir!

I have a very high resolution (ORTF) 819 line test card. Just send me a pm with your email address.

Jac

Semir_DE 13th Jun 2020 11:15 am

Re: My 819 line French TV project
 
Quote:

Originally Posted by Jac (Post 1258984)
Impressive work Semir!

I have a very high resolution (ORTF) 819 line test card. Just send me a pm with your email address.

Jac

Hi Jac,

Thank you very much for offering the test card, that was exactly what I was looking for as an authentic add-on to this project! PM is on its way ;D

Semir_DE 13th Jun 2020 11:38 am

Re: My 819 line French TV project
 
3 Attachment(s)
Hi again,

yesterday, or rather early this morning ;D I assembled the first unit. Since I am still waiting for the crystals I had ordered last week, I tried it with what I had at hand 65.53Mc and 76.64Mc which is not really a French channel but the vision carrier is close enough. Sound is above not below vision but I tried anyway and it worked on my Sony TV9-90UM. Here are some images of the assembled unit:

Attachment 208381

Attachment 208382

Attachment 208383

Once the crystals arrive I will finish the project. probably in early July. The difficult part will be to make a nice looking cut out for the VGA connector in the back of the aluminium case. I have a friend with a laser cutting machine though and I am planning to ask his help...

I will also test this with 405 once I get the mode line working on a PC here. I have crystals for channel A1 so that should not pose a problem. If anyone here is interested in the KiCAD design data I can send it by mail. Just send me a message.

Cheers

Semir

rambo1152 13th Jun 2020 7:17 pm

Re: My 819 line French TV project
 
1 Attachment(s)
No crystals or custom PCB in mine Semir I'm afraid.
The PVC tape join hides the RGB sync matrix!

Actually, drift of the 51.75 & 48.25 MHz "Holme Moss" free-running carriers is not a problem.

peter_scott 13th Jun 2020 8:19 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Gosh Graham, and I thought that my construction was bad!

Peter ;)

rambo1152 13th Jun 2020 9:32 pm

Re: My 819 line French TV project
 
2 Attachment(s)
That's the problem when the initial proof-of-concept breadboard edition works so well, there is little impetus for improvement.

In my defence I did make a start on a slightly neater iteration over two years ago, but my only 405 line set has been poorly until recently, so I really should complete it now.

Attachment 208427

I also need to write up grafting the LOPTX that David (FERNSEH) kindly sent me into the Cossor.

It's working fine; I just want to optimise the value of a tuning capacitor I found necessary, and mount the Tx in the original tin can.

Attachment 208428

Semir_DE 14th Jun 2020 3:08 pm

Re: My 819 line French TV project
 
2 Attachment(s)
Dear all,

here I would like to share some images of my TVs with the ORTF test card that Jac kindly sent to me.

Attachment 208458

This one needs some alignment, I know...

Attachment 208459

In high resolution these images would show that the horizontal resolution of the TVs is limited to about 400 TVL while the vertical resolution is over 500 TVL. This is probably due to the fact that more recent 819 sets operated with a limited IF bandwidth e.g. the Radiola has roughly 6Mc and the small TV9-90 most likely even less. It would be interesting to see how the test card would look on an older French TV with the full IF bandwidth. On the other hand older sets may have worn out CRTs so resolution might be limited due to lack of focus in the CRT itself, but this is just an assumption on my side.

Cheers

Semir

Jac 14th Jun 2020 4:00 pm

Re: My 819 line French TV project
 
2 Attachment(s)
Hi Semir,

My Philips TF651A indeed displays a somewhat better definition.
The modulator I used is MC44BS373 based and thus has a limited bandwidth.

Nevertheless, with this the TF651A can display 650 lines horizontal (819 vertical).
To the naked eye it looks better than I managed to catch in a photo.

Attachment 208460

Attachment 208461

Due to influence from the magnetic field of the mains transformer, the picture wobbles a tiny bit, and this does not help for a sharp photo (and combine that with my very limited photographic capabilities :) )

Jac

peter_scott 14th Jun 2020 5:10 pm

Re: My 819 line French TV project
 
Are these resolution bars square wave or sine?

Peter

Jac 14th Jun 2020 5:35 pm

Re: My 819 line French TV project
 
Hi Peter,

The bars are square waves, but since the higher harmonics (of the narrower bars) will not pass through the system, this might not make much difference?

Jac

peter_scott 14th Jun 2020 7:34 pm

Re: My 819 line French TV project
 
Hello Jac,

On your detail (second) picture if you look at the top where 350 is just truncated the lines look sharper than they do for the 350 in the centre left.

Peter

Jac 14th Jun 2020 7:43 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Hi Peter,

Yes they do. I think it has a lot to do with the contrast of the original:

Attachment 208477

I must try the set with a better test pattern (like I sent to Semir).

Jac

peter_scott 14th Jun 2020 8:05 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Well that looks different but perhaps not in the screen shot with brightness reduced in the lower block.

Peter

Jac 15th Jun 2020 5:29 pm

Re: My 819 line French TV project
 
3 Attachment(s)
Hi Peter,

I had another go at photographing the test card on the Philips TF651A.
(Due to my inadequate photographic skills and the wobble in the picture, it looks a bit better to the naked eye though.)

Attachment 208545

Attachment 208546

Attachment 208547

Jac

Semir_DE 15th Jun 2020 6:39 pm

Re: My 819 line French TV project
 
Hi Jac,

looks pretty good. Much better resolution than my Radiola TV. On your TV I'm sure you can probably read menus e.g. of VLC very clearly. That is a very nice vintage TV you have I just found this web site where yours is actually pictured.

My dream TV would be to get an E1 441 Line "Einheitsfernseher" but as only about 50 of them were ever built they are like gold dust and priced accordingly...

Cheers

Semir

Jac 15th Jun 2020 6:56 pm

Re: My 819 line French TV project
 
Hi Semir,

Thanks.
The set I hope to add to my small collection would be a (any) pre-war Philips.
I must confess that an E1 is also on my wish list, but no chance of ever finding one in this country.

Best regards,
Jac

peter_scott 15th Jun 2020 9:11 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Hello Jac,

That does look good, especially in the close up.

Thanks,

Peter

red-duck 16th Jun 2020 12:31 am

Re: My 819 line French TV project
 
I'm curious that the test card image is for ORTF 2.
Did the second channel ever transmit in 819 lines ? I thought it was 625 only.
I'd be interested in knowing the early history of ORTF 2.

peter_scott 16th Jun 2020 8:40 am

Re: My 819 line French TV project
 
1 Attachment(s)
In my previous post I was trying to relate Jac's 819 line bandwidth to a 405 line benchmark. The lower half of the image shows a crop of the TRF test card displayed on a 405 line receiver that has 3MHz bandwidth. (The 3MHz grating on Test Card C equates to something like 425 lines on the RTF test card.)

Jac's 819 receiver has better resolution than the 405 line receiver and the active part of the 405 line is 80.8uS whereas the active part of the 819 line is 41uS so Jac's receiver has bandwidth greater than twice the 3MHz of the 405 line receiver. ~7MHz.

Peter

Richard_FM 16th Jun 2020 8:56 am

Re: My 819 line French TV project
 
Quote:

Originally Posted by red-duck (Post 1259833)
I'm curious that the test card image is for ORTF 2.
Did the second channel ever transmit in 819 lines ? I thought it was 625 only.
I'd be interested in knowing the early history of ORTF 2.

As far as I know it was 625 lines from the start, which I presume was system L without the SECAM encoding at first.

Semir_DE 20th Jun 2020 9:34 pm

Re: My 819 line French TV project
 
5 Attachment(s)
Dear all,

I would like to share some more findings here regarding the implementation of an 819/576/405 signal generation system on a Linux PC in the hope some of you might find this useful. My first experiments were with two laptop computers an HP 8440p and an even older Dell Latitude 6400

Both employ the Intel integrated GMA graphics solution which seems to be limited when it comes to clock rates below 20Mc. While this is not a limitation for higher resolution output formats it will limit possibilities for outputting low res formats like PAL (576i) or 405i. Sine the french 819i format requires a clock close to 30Mc this format could be generated with ease on both laptop PCs. I was unable, however, to get a PAL or 405 output to work on those PCs.

Since I was aiming for a solution that could be switched to 405i, 576i and 819i I decided to conduct some more experiments. A few considerations had to be made for the whole system to work well with all desired formats:

1) The output of the TV signal should be over a VGA connector and this should be a second head on the PC. This meant a graphics card with both DVI/HDMI and ananlog VGA outputs needed to be found.

2) The main or primary monitor should have a decent resolution e.g. 1280x1024 or higher, so all OS menus, VLC player etc. could be run in a normal fashion

3) It should be easy to switch resolutions for the secondary monitor without a reboot and without interrupting playback of a video. This is to be able to demonstrate the different formats while playing the same video file.

4) The PC should be powerful enough to play HD content since only HD content could be down scaled to 819i while SD content would be up scaled thus not using the full potential of the 819i format. For a 405i only system a much less powerful PC would suffice as only SD content would need to be playable.

The problem, or so I thought, was to find a suitable graphics adapter that could be configured to output all desired formats while at the same time having an interface that was compatible with a modern PC motherboard. As I had read here that old GeForce 4 graphics cards were a suitable choice I tried an old GeForce 6600 with AGP interface. I still had this as well as the quite dated PC with a Pentium 4 processor. The results with the GeForce 6600 were disappointing. It worked with 405i but on 819i it produced a second vertical sync pulse halfway across the picture. Since the old Pentium 4 PC was not an ideal choice for playing HD content anyway I decided to investigate a more modern PC architecture.

I also had an old small PC that I had built a few years ago with an Asus N4L-VM Motherboard and an Intel core2duo this has a PCI-Express 16 interface which makes the choice of graphics card much easier. Also not the most recent technology but just powerful enough to play HD content as long as a good graphics adapter was used.

Originally this had an ATi Radeon 4000 (EAH4350) in it which worked well under Windows 8 as an HTPC. When I installed Linux Mint XFCE 32 Bit on this machine it also worked well for the OS and I could get modelines to work for 405i. It would not work with 819i, however, in this case the sync of every second field was inverted and no matter which setting I chose in the modeline this issue remained. Back to the drawing board...

As a next test I extracted an Nvidia N210 card made by MSI from another machine I had and put it into the one I was testing. This was the answer! All modes were supported immediately and I could switch between formats even while playing a video in full screen on the second "monitor" i.e. VGA output to my modulator. All this worked until I decided as suggested by the Linux OS driver manager to install the proprietary Nvidia drivers. After doing so the modelines no longer worked, so be warned only use the integrated Xorg drivers...

As a third choice I had ordered or more recent Nvidia N710 card made by Zotac. This costs around 35 and is very sleek and does not need a noisy fan. This card also works perfectly with all modes. Here is an image of the card which can still be bought online:

Attachment 209075

Attachment 209076

So as a conclusion from this non representative test it seems that Nvidia cards even of newer design work with all modes we need for generating 405i, 576i and 819i. Summary:

Intel integrated graphics: good for 819i not good for 405i (576i was not tested)
ATi Radeon EAH4350 (Radeon 4000 series): good for 405i, not Good for 819i (576i was not tested)
Nvidia N210 made by MSI: good for all modes 405i, 576i and 819i
Nvidia N710 made by Zotac: good for all modes 405i, 576i and 819i

Here are a few images of the settings:

Attachment 209077
for a 405i output

Attachment 209078
for a 819i output



A few hints for setting up VLC:

- I used the detached video interface setting, so the window containing the play list will remain on the PC screen while only the full screen video can be moved to the secondary monitor. This will allow adding new clips to the play list while playback is ongoing

- It is a good idea to select a fixed video output renderer instead of "auto" in the VLC video settings. If this is not done the VLC interface will initialize every time it starts playing a clip and this will move the full screen video back to the PC screen.

Finally an image of the PC in its case:

Attachment 209079

I hope this information will be useful to others here while experimenting with modeline signal generation.

Cheers

Semir

peter_scott 21st Jun 2020 5:33 pm

Re: My 819 line French TV project
 
I'm using an ATI Radeon HD 2400 and the fan is rather noisy so it's interesting that the Nvidia N710 works well without a fan. I thought there was some incompatibility between Nvidia cards and Jeroni Paul's Modeline program?

Peter

Semir_DE 22nd Jun 2020 8:27 pm

Re: My 819 line French TV project
 
Hi Peter,

I did not use the "Winmodelines" program as such, just the posted modelines and the ones made by Kat Manton. These worked without any issues. I also use Linux not Windows, so the incompatibility described on Jeroni Pauls page plays no role.

I'll post the modelines I used here later when I'm on my other pc...

Semir_DE 22nd Jun 2020 9:56 pm

Re: My 819 line French TV project
 
1 Attachment(s)
Dear all,

Here are the mode lines I have currently in my ".profile" file on my Mint Linux system.

one more I am planning to test is a version for 441 lines.

This is all running very nicely on a fan-less Nvidia N710 PCI Express 16 graphics card.

xrandr --newmode "819" 29.3203 1152 1168 1240 1432 738 744 745 819 -hsync +vsync interlace
xrandr --addmode VGA-1 819

xrandr --newmode "405" 8.10 664 680 752 800 377 378 385 405 -hsync +vsync interlace
xrandr --addmode VGA-1 405

xrandr --newmode "PAL" 13.875 720 741 806 888 576 581 586 625 interlace -hsync +vsync
xrandr --addmode VGA-1 PAL

xrandr --newmode "PAL-SQP" 14.750 768 790 859 944 576 581 586 625 interlace -hsync +vsync
xrandr --addmode VGA-1 PAL-SQP

xrandr --newmode "NTSC" 13.846 720 744 809 880 480 488 494 525 interlace -hsync +vsync
xrandr --addmode VGA-1 NTSC

Please take note that this configuration uses a positive vsync and a negative hsync. This is owed to the fact that I am using a single XOR gate to generate a somewhat compliant vertical sync with h-serrations. If you intend to use the simple resistor adder network for generating the composite signal please change the "+vsync" in each modeline to "-vsync"

you can find the .profile file I'm using in the attached zip file.

Cheers

Semir

Semir_DE 26th Jun 2020 10:44 am

Re: My 819 line French TV project
 
Dear All,

just quick update on this subject matter.

As I am also experimenting with the old German 441 line standard I decided to tinker a little bit with the modeline parameters.

I found a very useful web site that allows you to tweak an existing modeline and will show you the timings of h-sync, v rate etc.

I started out with the 405 modeline I had and changed the line count and other items until I had the required timing. Here is a modeline that will generate the 441 line format:

Modeline "656x383" 8.82 656 665 744 800 383 383 391 441 -hsync +vsync Interlace

Again due to way my adapter processes the syncs the vsync is positive, for a simple adapter this should be changed to -vsync.

This modeline generates a vertical timing that is spot on, only h frequencies are slightly higher at 11.030kc instead of the 11,025 as defined by the standard.

Unfortunately I do not have a real 441 line TV (yet) but I have modified some older surveillance monitors to run on 441 and they work very nicely with my modeline signals. Also UK 405 sets can be used with this signal if the h-hold is adjusted to accommodate the higher horizontal frequency.

If anybody here has a 441 set I would be interested to know if this works on yours. As a modulator any 405 line modulator should work, especially the simple designs with LC tanks should be easy to modify to output the required frequencies.

Cheers

Semir

Semir_DE 13th Oct 2020 2:26 pm

Re: My 819 line French TV project
 
3 Attachment(s)
Dear All,

Here is a short update to this project. A nice friend of mine who has a CNC machine kindly made me some really fancy front and back plates for my VGA 2 RF converter.

Attachment 217787 Attachment 217788 Attachment 217789

I am planning to make one for the 405 line system as well. This one is for the French system. It uses custom made crystals for channel F3. Since I have crystals for UK channel 1 I will make a VGA 2 RF for that too.

Another update is related to the 819 line modeline. I have created one that has a 4:3 aspect ratio. This is easier to set up on more modern 4.3 TVs an VLC:

Modeline "984x738" 25.066 984 998 1060 1224 738 744 744 819 -hsync +vsync Interlace

It has a slightly lower clock rate, so I am not sure if it will work with Intel GMA graphics. It is working very well on my Nvidia card.

I have also found out that the Nvidia 710 card can handle triple outputs, so I now have the following setup:

- SVGA PC monitor for control on DVI
- HDMI to Video converter on HDMI this provides a PAL output
- 405/441/819 (selectable in Linux graphics management) monochrome output on VGA

The "monitors" are lined up horizontally, the arrangement is from left to right:

PC monitor - HDMI (PAL) "monitor" - VGA (405/819) "monitor"

So I can just drag the player window to the desired screen.

Audio output can be selected in VLC to be either HDMI (PAL) or Analog (405/819) modulator.

If I run two instances of VLC I can now output PAL and 405 (or 819) at the same time. Unfortunately there does not seem to be a way to synchronize two VLC players, so I have to do that by hand which is a little cumbersome.

I am also thinking of a possibility to create a 3D file with identical left and right images and then play them in one VLC. If I could drag one image to each monitor this might work for simultaneous output. I would like to have this for demonstration purposes to show the differences of the standards side by side.

I use this setup sometimes to watch old movies. Yesterday I watched "Sabrina" a classic with Audrey Hepburn and Humphrey Bogart in glorious 819. This black and white film was shot in 4:3 so it is ideal for watching on an old set.

The original is from a Blu Ray and I converted it to 1440x1080 25p using "Handbrake" a very neat program that can shrink the very large BD file of ~30GB to about 3GB without any visible impact on image quality. "Handbrake" does not decrypt any discs it merely converts video formats and can change the resolution in the process as well. It uses all available CPU cores almost to 100% and thus slows down the PC notably this is probably why the program is called "handbrake" :)

Cheers

Semir

AdrianH 14th Oct 2020 11:28 am

Re: My 819 line French TV project
 
1 Attachment(s)
Hello Semir,

I think I need to pick your brain and ask for help. I use Linux, generally Ubuntu 18.04 and ubuntu 20. I am having absolutely no success, (I think) getting your method to work. So I need to check with you if I am doing things correctly

I am at present only interested in 405 line, I have tried several Video cards from NVIDIA 510, HD8350 and now today a AMD/ATI HD2400 Pro XT.
In shell this is what I am getting for the video card.

lspci
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] RV610 [Radeon HD 2400 PRO/XT]

On the card I have a DMS-59 cable to two VGA connectors.

when I try xrandr in shell I get:-

Screen 0: minimum 320 x 200, current 2624 x 915, maximum 8192 x 8192
DIN disconnected (normal left inverted right x axis y axis)
DVI-0 connected primary 1600x900+0+15 (normal left inverted right x axis y axis) 434mm x 236mm
1600x900 60.00*+
1280x1024 75.02 60.02
1152x864 75.00
1280x720 60.00
1024x768 75.03 70.07 60.00
832x624 74.55
800x600 75.00 60.32
640x480 75.00 59.94
720x400 70.08
DVI-1 connected 1024x768+1600+0 (normal left inverted right x axis y axis) 0mm x 0mm
1024x768 60.00*
800x600 60.32 56.25
848x480 60.00
640x480 59.94

So the first thing that seems different is that it is saying DVI-0, DVI-1 rather than VGA.

I try the following:-
xrandr --newmode "405" 8.10 664 680 752 800 377 378 385 405 -hsync -vsync interlace
xrandr --addmode DVI-1 405

Then xrandr in shell reports for DVI-1 as follows:-

DVI-1 connected 1024x768+1600+0 (normal left inverted right x axis y axis) 0mm x 0mm
1024x768 60.00*
800x600 60.32 56.25
848x480 60.00
640x480 59.94
405 50.00

So it is in there, but if I now go into Ubuntu display settings I can not select the 405 setting for the second display, it is not there for me to select?

Attachment 217876

Please can you assist?

Adrian


p.s.

I have just used another line in shell of :-
xrandr --output DVI-1 --mode 405

This seems to force the mode to work, but any changes even to the other screen in video settings will revert the display away from 405 lines, so not sure if that is the correct method.

Semir_DE 15th Oct 2020 8:16 am

Re: My 819 line French TV project
 
Hello Adrian,

I have conducted a few tests yesterday. The short version is: Use Linux Mint XFCE version.

I will post my findings in more detail later. As I had to find out there are many variables in this equation e.g. graphics adapter, Linux version used and driver used.

Cheers

Semir

AdrianH 15th Oct 2020 10:10 am

Re: My 819 line French TV project
 
Using Linux Mint is not an option for me as I am firmly embedded into Ubuntu for other reasons, but I have asked on the Ubuntu forums to see if there is another way, or to why it is not seen in the display settings GUI. I can work around it using the xrandr line, and I could find more video cards will work if I try them again.

I will see if I get any responses from the other forum.

Adrian

Semir_DE 15th Oct 2020 11:52 am

Re: My 819 line French TV project
 
Hi Adrian,

Too bad you are stuck with Ubuntu... Mint by the way is a derivative of Ubuntu and uses mostly the same repositories. For some reason graphics management front ends in Ubuntu and Mint cinnamon version seem to hide graphics modes that have a low resolution. I have searched online but thus far have not been able to find out why this is happening. I am pretty sure a simple tweak of a configuration file might be all that is needed to show all resolutions. But as most users out there want more rather than less resolution I believe chances of find an answer in the forums etc are slim. Out of the box only the xfce windows environment seems to list all modes. For what its worth here is what I found:

1) Nvidia proprietary (faster) drivers are not compatible with xrandr. You need to use the x-org drivers with Nvidia for xrandr to work.

2) on Ubuntu 18.4 I have been able to use an AMD (ATI) graphics card (RV710) successfully. You do need to use the "output" command line once, however, to get the correct output. After that it will continue to work even after a reboot.

I am far from being a Linux "guru" so I am also curious of anybody else here might jump in and suggest a solution.

Cheers

Semir

AdrianH 15th Oct 2020 12:29 pm

Re: My 819 line French TV project
 
I have just plugged in a previous video card
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Cedar [Radeon HD 7350/8350 / R5 220]

and tried that with the added line of xrandr --output DVI-1 --mode 405

That card is also working for me providing composite video, it will be interesting to see how many of my other cards also work with this method.

Adrian

jhalphen 17th Nov 2020 2:21 pm

Re: My 819 line French TV project
 
Hello to all,
Hi Semir,

I'm a big admirer of your work on standards conversion.

There is a current thread on the French RadioFil TV forum on synthesizing B&W 819 line signals using Linux or other means.
I have flagged your topic for them as "must read" stuff.

Would you care to spend a few minutes reading about their approach...

Here's the link via Google French to English translation :
https://translate.google.com/transla...5%26t%3D256196

Best Regards
jhalphen
Paris/France

PS : as for me, i have a bunch of Auroras...

jhalphen 30th Dec 2020 7:28 pm

Re: My 819 line French TV project
 
Hello to all,

Still on the topic of the French 819 E standard, here is a full description of a DIY VHF Band III modulator. It operates on a fixed frequency, Channel 8A which was used in Paris on the Eiffel Tower from 1949 to 1983 (end of service). Channel 8A was also used in Lille, northern France.

The design by engineer JC Jardine has been around for quite some time and has been used by many 819 enthusiasts.

http://819lignes.free.fr/Realisation...r_VHF_F8A.html

Vision carrier is 185.25 MHz
Audio carrier is 174.10MHz

I've built a pair. It works well but serious shielding is necessary as the RF output is quite powerful.

Best Regards
jhalphen
Paris/France


All times are GMT. The time now is 10:11 am.

Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.