Dragon 32 RGB SECAM and video tests

For the discussion of all hardware related topics.
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Dragon 32 RGB SECAM and video tests

Post by Azerpy »

Hello.
I have got two Dragon RGB SECAM and I think I'm not the only one here.
The motherboard PN 41503 (issue 2 and issue 2A for me) and the SECAM PSU board PN 41504 are already in the hardware section. The quartz frequency is 14,2180 MHz.
This machine has two interesting properties :
- a RGB circuit on the PSU board that delivers a very good signal on a 7 pins DIN socket
- a real SECAM circuit on the motherboard that feeds a UHF converter.

As I had no DIN / Scart junction, I looked inside to understand and make one cable. I put here the schématics of this special video section of the machine.
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Re: Dragon 32 RGB SECAM and video tests

Post by Azerpy »

Dragon SECAM video on motherboard
Dragon SECAM video on motherboard
D32_RGB-SECAM1.jpg (1.44 MiB) Viewed 10627 times
Dragon SECAM video powerboard
Dragon SECAM video powerboard
D32_RGB-SECAM2.jpg (2.14 MiB) Viewed 10627 times
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Re: Dragon 32 RGB SECAM and video tests

Post by Azerpy »

RGB tests
I have tested the RGB on the Scart input of different LCD TVs. It gives a nice picture, but with vertical lines. However, there are 2 things to fix (see schematics for explainations) :
- the voltage at pin 16 of the Scart is too low to switch the RGB ; if not corrected, you could only get the black and white luminance signal of pin 20, whitch is unstable and of bad quality if not filtered.
- the picture is too bright and blue, so I added resistors on the outputs (60 ohms for Red and Green, 120 ohms for Blue.

SCART to HDMI box
I bought a SCART to HDMI box that works very well with the LCD TVs. I could'nt get a picture with my Viewsonic monitor until I add a HDMI to VGA converter. What a miracle ?

I also tested that HDMI box with a HDMI to USB capture device when I saw Mike Miller's PC screen with a Dragon window. Thank you very much Mike for the idea ! I have now a VLC Reader window with a Dragon on my PC screen, happy to be there with his brothers Xroar and Mame.

GBS 8220 device
I tested the RGB outputs with a GBS 8220 card, but it was impossible for me to have a stable image Because of 50 Hz not accepted by the card ?

The SECAM circuit
For fun, I looked at the SECAM circuit. I adjusted RV1, L6 and L7 for picture and L1 for sound to have a correct signal on the UHF input of a TV. Not very good quality.
To get a composite SECAM output for a RCA input of a TV, I replaced the UHF converter by a 680 ohms resistor and captured the signal with a little video amplifier (gain x -5). Not better picture than with UHF. One day, I will try to send a S Video signal to my TV. As Tormod Voden suggested here once ago. Luminance going out of TR4 is bad quality and must be filtered.
With the same amplifier, I also can get a picture of Y and LUM signals. All that 3 signals have peaks going up, and must be amplified and inverted. Y contains a composite synch H + V and LUM is better with added lines.

About the vertical lines "jail bars"
Both of my machines present vertical bars. I can mitigate the phenomenom with the settings of the screen or VLC Reader. But if anyone as ideas about the cause ... I have not tested the machines with other external PSU.
I think it is possible to get rid of them because I also have got a coco 2B RGB which is, by me, the ultimate reference of video quality.
sorchard
Posts: 530
Joined: Sat Jun 07, 2014 9:43 pm
Location: Norwich UK

Re: Dragon 32 RGB SECAM and video tests

Post by sorchard »

One thing that causes vertical bars, particularly bars that have the same spacing as text columns, is insufficient supply bypassing on the memory array.

This motherboard picture in the archive is a good example: https://archive.worldofdragon.org/index ... _8x64K.png

Look at all those missing capacitors around the memory chips in the upper right of the board. I've had machines like this with vertical bars and fitting the missing caps completely cured it. Perhaps double-check that the empty positions are actually supplies first. The boards with jumpers can put signals on these lines, depending on which type of memory ics are fitted.
Stew
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Re: Dragon 32 RGB SECAM and video tests

Post by Azerpy »

My motherboard is like this one.
https://archive.worldofdragon.org/image ... ue3%29.jpg

We can see on the picture that the 8 x OKI M3732H RAMs have the caps.
and also that caps have been added, by exemple near the 6809.

I don't understand what you mean by "The boards with jumpers can put signals on these lines, depending on which type of memory ics are fitted."
There is a jumper named LK2 near the RAMs. Besides, this strap disturbs my project of changing to 64 K RAM (74 LS 02 has no gate unused).
sorchard
Posts: 530
Joined: Sat Jun 07, 2014 9:43 pm
Location: Norwich UK

Re: Dragon 32 RGB SECAM and video tests

Post by sorchard »

I've not seen your board type before so I'm not sure. Judging by the large size of the caps that have been fitted, this might be a known problem with this board layout. It might be worth adding some more bypass caps (100 - 220nF) across the dram supply rails on the back of the board just to see if it makes a difference. I would also try running some heavy wires to take 0V and 5V directly from the power connector to the dram supplies.

Some boards (not yours) have five jumpers near the power connector. These decide how the dram ics are connected. I think one of them selects between +5V and Z7, so if it's in the Z7 position, then bypass caps should not be fitted on this line.

The LK2 jumper on your board could be an inverter for the Z7 line. If it is, then you should be able to reclaim the gate for your 64K upgrade. Definitely worth a closer look to see what it does.
Stew
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Re: Dragon 32 RGB SECAM and video tests

Post by Azerpy »

Thank you Stew for all thoses precious advices, I will look at the power supply.

I have taken a picture of the screen with the bars.
And also a view of the Y signal given by IC 6847 compared with the Video Clock (oscillo at 0,2 µS per division)

When I read the 6847 data sheet, in the case of a 14,2180 MHz quartz, the timings are :
- VClock (upper signal) is at 3,55 MHz, the cycle is 0,28 µS
- one line of 256 dots on the screen is 36 µS (64 µS for a total line)
- one dot is 0,14 µS = half a VClock cycle
We see on the picture and on Y (lower signal) an alternation of black and white bars at each dot
and if we zoom the picture, another alternation of black and white bars (3 or 4 per dot).
That does not explain the cause !
Gérard
Azerpy
Posts: 54
Joined: Tue Apr 09, 2019 1:16 pm
Location: France Pyrenées

Re: Dragon 32 RGB SECAM and video tests

Post by Azerpy »

vertical lines on a Dragon 32
vertical lines on a Dragon 32
vertical lines on dragon 32.jpg (7.03 MiB) Viewed 10587 times
Video clock and Y signal at 0,2 µS per division
Video clock and Y signal at 0,2 µS per division
video_clk and y.jpg (3.76 MiB) Viewed 10587 times
sorchard
Posts: 530
Joined: Sat Jun 07, 2014 9:43 pm
Location: Norwich UK

Re: Dragon 32 RGB SECAM and video tests

Post by sorchard »

Hmmm, those bars are at a higher frequency than what I had in mind, so probably not dram noise. As your photos show, that looks like the video clock coupling into the luma signal.

If the noise is on the supply rails, then some bypass caps might help.

I notice in the picture of your motherboard that the 6847 has no suitable bypass caps at all. It might be worth connecting 100nF directly between the supply pins (1 & 17) to see if that helps.

The video clock is routed to several ICs: IC14, IC22, IC23, IC29. These can all put noise on the supply rails.

If that doesn't help, then it may be more of a crosstalk issue resulting from the board layout. It would be interesting to find out if anyone else with that board type has the same pattern of noise in the display.

I can see the same pattern faintly on my MKII Dragon, but nowhere near as noticeable as in your photo.
Stew
dublevay
Posts: 173
Joined: Wed Jan 23, 2019 3:41 pm

Re: Dragon 32 RGB SECAM and video tests

Post by dublevay »

Hi,

I have one of these, and though fascinating in itself (I wanted the board because of the motherboard layout - I didn't realise beforehand that these were the SECAM machines), the RGB output simply was not as good as I expected. The main issue I found was the excessive blue on the screen. I have been meaning to trace the video circuitry, but just haven't got round to it.

I also tried to tune in the RF output (after telling my 2006 Sony TV that it lived in France), and managed to tune it in, but the display was awful.

So very interested in where this thread goes. Ought to get my machine running again to compare what we are seeing. Incidentally, I have never had an issue with the RGB switching voltage - but I think I will add the resistors as you suggest to modify the colours.
Post Reply