exxos's Atari, Amiga & retro forum

The alternative is something is shorting the RAS line so that the signal never shows up on the scope.

Pull all the RAM chips and check again, in case one of those is shorting RAS to ground.

Also note that it's not a 72LS139 but a 74S139, which is faster. There are many case in the Beeb where the S variant is required.

stephen_usher wrote: 19 May 2021 15:11 The alternative is something is shorting the RAS line so that the signal never shows up on the scope.

Pull all the RAM chips and check again, in case one of those is shorting RAS to ground.

Also note that it's not a 72LS139 but a 74S139, which is faster. There are many case in the Beeb where the S variant is required.

Sorry, yes, you're quite correct, I've just been looking at so many LS's of late I auto added it!!

Good point regarding the RAM shorting the RAS line, I will check that out before I break out the soldering iron.

@stephen_usher There's no change in the RAS with the RAM removed, so it doesn't look like a RAM short.

Could just continuity test it to ground also...

BennehBoy wrote: 19 May 2021 17:11 Could just continuity test it to ground also...

Well replacing the 74LS74 made no difference. Digging deeper into the service manual I found this:


This is what the SM shows as the CAS signal, and it's what I also see, more or less, on the Fisher-Price oscilloscope.

Now we come to RAS:


This I do not see at all, just a flat line that has a bit of a sine wave to it if I tweak the scope enough. I see this same signal all the way back to pin 12 of IC43, the inverter output, and more or less the same on the input at pin 13.

Since the Fisher-Price scope only has 1 channel, I can't tell if it is inverted, but IC43 is new, so I expect it is.

Since the output from pin 13 is taken from pin 6 of the 74LS74, and they have continuity between them, and the signal is the same at both points, that would lead me to believe that given CAS is looking ok, there is a problem with either the master clock, and the video IC is sending out the wrong divided clock signals, or the master clock is good and the video IC is not delivering a proper 8Mhz signal to drive IC44, the 74LS74.

The RAS circuit is really not a major feat of logic, so if continuity is good (I am going back to check - again), then it can only come down to the flip flop inputs not being able to drive the necessary outputs.

So I have rechecked all the continuity from end to end here:


So all along pins 1-6 of IC 43 (I have noway to check the caps) go to exactly where I'd expect them to go, and the resistors R91,R92, R102,R98 & R107 all checkout value wise. Pin 11 of IC 40 is correctly connected to pin 8 of IC6.

All of the 74LS74 pins are connected to exactly where they are supposed to go and the RAS inverter and R106 also all checkout.

Heres the problem I face now. I discovered that by touching pin 4 or 5 on IC43 with my scope probe, which forms part of the 16Mhz oscillator circuit, that the speaker stops making a tone and begins to make very low white noise.

In addition, I start to see what looks like normal RAS activity, so it's almost as though the extra capacitance of the probe kicks off the oscillator running, just not properly. If I boot whilst touching pin 4 or 5 (They are directly connected) I also see RAS, but I do need to plug into a TV to see if there is any signal in that state.

Things would be a lot clearer with a scope that could see the clocks properly. I'm now wondering about replacing the caps and crystal for the main clock circuit. I have the parts list and they should be simple parts to get.

@rubber_jonnie do you have the data sheet for the dram does it show the ras and cas timing that it needs

Tom

Tomswork wrote: 20 May 2021 06:16 @rubber_jonnie do you have the data sheet for the dram does it show the ras and cas timing that it needs

Tom

Yes, I have the datasheets.

The CAS has always been firing, and until I discovered the thing with the 16mhz clock last night the RAS wasn't.

If I boot with the probe touching the pin, RAS starts with the machine, though it still doesn't boot. If I don't touch the pin, RAS doesn't start on boot but will start if I subsequently touch the pin.

I've been testing at all points of the RAS circuit from all the RAM right back to the point at which all of the necessary signals come together to generate it, plus all the continuity of the signal paths. No shorts or continuity problems at all.

I also have RAS and CAS timings from the datasheet, but theproblem is no RAS at all, plus the scope I have just isn't up to snuff as it's only single channel.

I'm starting to suspect the 16MHz clock is borked.

The clock has a lot of capacitors in it they are non polarized ceramics do the show signs of cracking or black spots on them

Tom

Tomswork wrote: 20 May 2021 09:28 The clock has a lot of capacitors in it they are non polarized ceramics do the show signs of cracking or black spots on them

Tom

I'm going to order some replacement capacitors, regardless of condition, there is clearly something off with the clock. It's all kind of weird that the speaker starts making odd noises when I touch pin 4/5 of IC43 and suddenly there is a RAS signal when there was nothing.

The resistors are fine and within tolerance so they'll be left as they are.

At least one of the caps has a chipped corner, and I think I'll get a crystal, but I'll only change that should the replacement caps make no difference.