Harlequin video modifications

A place to discuss the ZX80's and ZX81's younger brother - the Sinclair ZX Spectrum
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Harlequin video modifications

Post by 1024MAK »

By the time the relevant signal has been processed by the TV/monitor, there may be some minor timing differences. But the horizontal and vertical sync signals (a composite or combined sync being both these combined) is the primary timing source for analogue video systems. This sync signal is combined with the picture information to form the LUMA signal (monochrome composite video).

A RGB input is in effect three separate baseband video signals. So there is no need for, and you don’t actually want any chrominance subcarrier signal anywhere near them. Or in the sync signal(s).

However, TV RGB SCART inputs should not have a problem with picture information or chrominance subcarrier signals on the ‘sync’ input, as this is used for composite video as well, so the TV should include all the required filtering.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

What do you think about normal colour / bright colour voltage levels?

Watching various clips and pictures online, it seems to me that the contrast between normal blue and bright blue is too low.
According to schematics, bright blue voltage across the 75 ohm resistor is about 0.700V, while normal blue voltage is about 0.510V, for a normal blue to bright blue ratio NB/BB=0.729.

In contrast, ZX Spectrum 128 (+2) testscreen looks good, with better normal blue / bright blue contrast. They use TEA2000 as PAL encoder, which use 2 bits per colour; normal blue is encoded as binary 10 and bright blue as binary 11 (decimal 2 vs decimal 3), so their ratio is NB/BB=2/3=0.667.

TEA2000 also applies a gamma correction of 1/2.4 (apparently) so that ratio becomes 0.667 ^ (1/2.4)=0.667^0.417=0.845. This new intensity ratio seems ok according with https://en.m.wikipedia.org/wiki/ZX_Spec ... phic_modes as D7 hexa /FF hexa =215/255=0.843

It seems that the Harlequin voltage ratio (0.729) sits between TEA2000 voltage ratio not gamma corrected (0.667) and gamma corrected (0.845). I found no information if AD724 also applies gamma correction, but I think it does since this is consistent with the reduced blue normal/bright contrast observed.

If this is true, some new values for the resistors are needed so that ratio of 0.667 is applied to AD724.
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

For example, changing resistors R34, R19, R20, R38, R21, R43 from 68 ohm to 75 ohm, and adding 3 x 56k resistors in paralel to 18k resistors R23, R26, R29 should give a bright voltage of 0.701V and a normal voltage of 0.467V, for a ratio of 0.667.

I will test this in near future and come back with results.
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Harlequin video modifications

Post by 1024MAK »

The trouble is, there is no set specification for what the standard ZX Spectrum colours or the ZX Spectrum bright colours are supposed to be other than the words black, blue, red, magenta etc.

The issue has been debated numerous times when emulator writers have asked in the past.

Back when the only video connection method for the 16K and 48K rubber key models was the TV RF/UHF output, the actual displayed colours on a CRT TV depended on the TV and the user controls for brightness, contrast and colour. So everyone set these at their preferred values. And it was common for bright blue to be almost indistinguishable from normal blue.

If you find a set of resistor combinations that produce a display that you are happy with, that’s fine :D. It would be good if you could report back what values you end up with.

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

It seems that three 47k resistors soldered in paralel with each of the 18k resistors (R23, R26, R29 Harlequin 128k) would do the trick! (and replacing three 68 ohm resistors with 75 ohm ones - R34, R38, R43)
Since my Harlequin is not complete yet, can anyone test this before I do, and provide some photos?

Without 47k resistors (and without replacing 68 ohm resistors), voltages across 75 ohm load resistors are 0.680 V (bright) and 0.502 V (non-bright), for a ratio of 0.738.
With 47k resistors added (and replacing 68 ohm with 75 ohm resistors), these voltages become 0.685 V (bright) and 0.459 V (non-bright), for a ratio of 0.670.

Voltages are measured with one 75 ohm resistor conected at SCART pins, thus simulating full load for the transistor (SCART and AD724).
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

Alternatively, one might try replacing those 18k resistors with 13k ones (instead of adding 47k in paralel).
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

Does anyone know the purpose of R48 (270 ohm)? It connects Q1 emitter to the ground and is in fact in parallel to R18 + 75 ohm (from TV).
It seems pointless to me...
User avatar
1024MAK
Posts: 5101
Joined: Mon Sep 26, 2011 10:56 am
Location: Looking forward to summer in Somerset, UK...

Re: Harlequin video modifications

Post by 1024MAK »

dac wrote: Wed Jul 14, 2021 7:55 pm Does anyone know the purpose of R48 (270 ohm)? It connects Q1 emitter to the ground and is in fact in parallel to R18 + 75 ohm (from TV).
It seems pointless to me...
Are you sure it goes to GND? The different schematics disagree if it goes to VCC or GND. Try leaving it out.

I’m a bit pressed for time this week, so not much hobby time to try any of this, or investigate any of it on read hardware :(

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Buffer Amp

:!: Standby alert :!:
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer later in the year.
dac
Posts: 10
Joined: Wed Jul 07, 2021 2:42 pm

Re: Harlequin video modifications

Post by dac »

It goes to GND on my PCB. All it does is insignificantly lowering the sync voltage (by almost doubling Q1 Ic current).
superfo
Posts: 74
Joined: Wed Jul 08, 2009 9:12 am

Re: Harlequin video modifications

Post by superfo »

Post Reply