weekends projects #2
My Kicad project seem ok.
I had to check the original device…
My project (boxed):
Maplin SPO256-Al2 Kicad projet (temporary release, 1.8M):
http://zx81.ordi5.free.fr/brico/maplin/ ... SPO256.zip
There is a component, may be a quartz oscillator (capacitor?) near the pin 27-28 (in and out clock).
In the original feature, the PIN#28 (out clock) is not used, but it seen to be link on the PCB.
Yes, It looks very much like a quarts oscillator.
If you look at the back picture there is also another small capacitor connected between pin 27 and what looks like C5.
My guess is that this is replacing the clock line which probably had the trace cut under IC6 for the mod. I was actually surprised that it was using the zx81 clock instead of an oscillator when I first saw the schematic.
yes, I did spend an entire evening downloading and reading through the Corrigenda section of every issue for several years to see if they mentioned anything about the zx81 talkback project
If I plug it in with A4 connected the zx81 won't start.
I get no signal on my display.
However, if I unplug A4 until after everything has booted and I've got the inverted k.
Then connect the wire from A4 to IC2 pin 8 - everything is fine and it all works.
In the failure state the breadboad get's power and the SP0256 sometimes chirps.
This makes me believe it's a power overload problem.
Possibly pulling too much current all at once on start up.
But, I can't figure out why not connecting A4 would solve the problem?
I only found out this works because I originally thought there was a wiring issue started by pulling A4-A7 to diagnose it.
That turned out to be a lucky guess.
But I've triple checked and I'm wired according to the diagram.
Anyone have a theory as to what's going on?
"No signal", a black screen… or a white screen ?
In case of the White screen, the memory test in ROM seem to loop to the A4 address… A4 (from the 74LS28 IC2) may be set to 1 at start… you had to insert a diode (1n4148) to avoid it… or this case is out of order.
In case of black screen, the inboard 7805 from the Zx81 seem pull down and hag the ULA/display...
Multimeter shows good voltage all around.
Other address pin (A6) reads high.
But A4 reads around 1.7-1.8v
If I force it high or low I fail to boot.
If I let it float I boot fine.
If I add a diode
A4 - >|- IC2(p8)
It boots, but then I can't POKE 16417,197 to get a sound
The design calls for IC2 to be a 74LS28
Those aren't available to me.
So I substituted a 74LS02
According to the datasheet they look almost identical.
Pin for pin compatible. Both quad 2-input Positive-NOR gates.
The only difference I could find was max current of the 28 seems to be a bit higher.
Time to sleep to see if any ideas come while I'm away from the problem.
It's worth noting that the transistor is also acting as a buffer between the speech chip and what ever power amp you connect to so as not to over burden the out put of the chip, much in the same way as when doing a composite display mod it's not recommended to connect the ULA directly to the outside world with out some sort of buffer in between to do the heavy lifting even though it may seem to produce a display(issue three ULA)
The first thing I tried this morning was to boot with the SP0256 reset and sbyreset forced low.
It booted fine.
I have a 16k internal memory upgrade. The original article mentions that everything "should" work with up to 16k but that higher might require a larger cap at C13 to keep a crash from happening.
The article mentions that NEW will also cause the fault.
I just tested, and sure enough NEW causes the fault.
I've tried increasing up to a 10uf cap and it still faults.
But, at least now I feel like I'm making progress
IC4b, C13, and D1 are clearly not doing their job.
Increasing C13 is supposed to help it work more effectively.
I started think about the purpose of C13.
I did silly thing and moved C13 to the other side of D1
And now it works. I can even put C13 back to a lower value.
I did a quick check and this seems to make IC4b and D1 superfluous. I'm clearly missing something.
Still trying to wrap my brain around why.
And why the original circuit isn't working.
Now I'm REALLY confused.
Inconsistent data sheets show pins 1G and 2G as either same (active low) or opposite (1G active high)
At least my 240's test out to have both 1G and 2G active low