The Sinclair QL Forum

Don't do this unless you're using a Fluke or other quality multimeter. Some cheap multimeters put the whole battery voltage across the pins to measure resistance, and that might be 9V.

Very. Bad. Things.

Ahhhh, didn't think of that. Okiez, will hold off until the clock crystals arrive (had to order them from overseas, but they should be here soon(. Hoping it is a missing clock signal to blame.

In the meantime, I tried all the socketted chips from thus dead QL in my working QL and it runs fine - so definitely not one of the custom chips etc.

I noticed there is a third clock crystal on the mb near the microdrives - 4.xxx mhz - I assume this is only for the microdrives?

abraXXious wrote:I noticed there is a third clock crystal on the mb near the microdrives - 4.xxx mhz - I assume this is only for the microdrives?

That will be for the Colour encoder to make the PAL composite signal. It will be 4.433619MHz, standard for an analogue TV.

Martin, while you're here - I have your updated reset generator circuit here and it works reasonably well (2 clock corruptions in 100 resets compared to 44 clock losses in 100 resets with the unaltered board). Do you have any changes you'd like to make on it in the 27 years since you released it? Also, do you have any objections to it being incorporated into the Issue 8 QL board?

How interesting!

I attempted to build Martin's cct in one of my mods last year and, due to my rubbish discrete hardware soldering skills, never got it to work and, in the end, bought a cheap Reset IC/Battery switcher (ADM-692) which works well enough, though probably not as good as 2 in 100 resets...

(Not meaning to redirect the thread...) I've been wondering if the PCEN-L _also_ needs to be held high to the 8302 like it does with SRAM - or is it gated properly from the RESET-L internally and thus doesn't need the extra step? I saw in one of the SQB (I think?) alternate designs for 8302 reset which did seem to try to disable PCEN-L during a reset.

Anyways, looking forward to hearing from abraXXious whether the crystals fix his dead QL issue, when they arrive...

The whole point is to pull PCENL to Vcc when power fails and/or reset occurs, as this is how the clock gets corrupt (spurious writes to 8302 registers). It's not the Vbat switchover - this is easy to do using two small schottky diodes. An interestig circuit to look at is the original Qimi PCENL decoder and battery circuit. It is not directly usable with lithium batteries as it incorporates a trickle charger (big no no for lithium!), but it is a clever circuit. It pulls up PCENL as soon as the supply voltage drops below Vbat. Another way to implement (early) power fail detect would be on the 9V input voltage, prior to the power regulator.

Martin_Head wrote:

abraXXious wrote:I noticed there is a third clock crystal on the mb near the microdrives - 4.xxx mhz - I assume this is only for the microdrives?

That will be for the Colour encoder to make the PAL composite signal. It will be 4.433619MHz, standard for an analogue TV.

Ahhhh. Okiez, so if THIS clock was faulty then the RF would not work, but the monitor out would be unaffected?

RGB and monochrome would work, but composite would not. Not sure if it would be a blank or B/W picture.

Okiez. This cannot be my problem though as pressing F1 etc does nothing, so the entire system is failing to boot. I will wait for the clock crystals to arrive and go fro there.

Dave wrote:Martin, while you're here - I have your updated reset generator circuit here and it works reasonably well (2 clock corruptions in 100 resets compared to 44 clock losses in 100 resets with the unaltered board). Do you have any changes you'd like to make on it in the 27 years since you released it? Also, do you have any objections to it being incorporated into the Issue 8 QL board?

Cor! That's a rave from the grave... I've no objections to you using it. And, no I haven't made any changes in the last 27 years.