Help with DTX1 needed urgently

[G3GJA]

I'm trying to get GB3EY on the air using the logic developed for the analogue version of EY. The logic is based on the BATC I2C project and uses a Maplin TEA2000 PAL encoder. Until the Repeater Group can qualify for a bursary to get a modern modulator, we are using my DTX1 to get started.

The output from the encoder is not recognised by the DTX1 and it just outputs a blue screen. The subcarrier frequency has been corrected as has the position of the burst that was caused by insufficient pull-up on the composite sync output from the I2C.

Levels are exactly right, 0.3v sync and 0.7v white. Burst amplitude is good too. All timings are better than 0.1uS of the nominal PAL spec. So why does the DTX1 ignore the output of the TEA2000 PAL coder yet it works perfectly OK with a composite signal from a Raspberry Pi or Richard Russell's test card generator?

The DTX1 works OK with composite video from the set-top box used as a receiver for digital signals. It's only when the video is switched to the logic generated PAL signal does the DTX1 go blue screen,

The video from the TEA2000 logic produces normal colour pictures when applied to two different PAL TVs. Has anyone seen similar behaviour from the DTX1?

Clive G3GJA

[G8GKQ]

Clive

I see that the TEA2000 accepts external Comp Sync. Is the comp sync source that you are driving it with true 625/25 or perhaps 312/50?

Dave

[G3GJA]

Dave

The comp sync for the TEA2000 is derived from the I2C VDU board's SAA5243 Enhanced Computer Controlled Teletext chip.

It's definitely capable of 625/50 interlaced. The data sheet shows a standard 625 line waveform complete with equalizing pulses either side of the broad pulses forming the field sync, but it also shows a 312/50 waveform. Presumably it can be switched. I haven't gone through the datasheet yet to see how interlace is enabled/disabled. That might be the cause. Update; the interlace on/off is set by bits in register 0, so not easy to change but the odd/even field output is active so would suggest interlaced?

I'm also looking at noise; the through video is much less noisy than the the logic generated PAL video.

Every other PAL video source I've tried works with the DTX1; from £13 eBay Chinese camera specials to PAL coders I built in the 90s such as the Project 100 SPG / ATV Handbook Electronic Testcard and TEA1002 based colour bar generators.

Thanks for the input; trying to get EY on before the NoV gets to its second birthday!

Clive

[G3GJA]

After a lot of pfaffing, I eventually found that the output from the BATC I2C VDU is NOT interlaced. A hardware solution has solved the problem and now the DTX1 takes the VDU's video.

I built an SPG around the ZNA134 and a 2.5625MHz crystal (thank you so much to whoever was selling £1 bags of 2 x ZNA134 and the crystal at CAT18). Fed the composite syncs to a new PAL encoder, the AD724 plus a 4fsc oscillator and buffer, and also sent them to the VDU board for genlocking after fitting the missing SAA5231teletext slicer chip.

Syncs coming out the SAA5243 graphics chip still seem to be non-interlaced but the RGB video encodes very nicely and produces a superb interlaced PAL output. Nice thing about the AD724 is that it doesn't need a delay line and the chroma filtering is built-in too.

Clive G3GJA

[G8GKQ]

Hi Clive

Really pleased that you found the solution and that the ZNA134 was useful. The ZNA134s and the crystals were a donation to the club, so I bagged them up for sale at CAT 18. That makes all the effort worthwhile!

Sadly, I can't recall who made the original donation - but thanks must go to them.

Dave