dragon 64 problem

[centrespot]

Hi,
I have a dragon 64 that displays nothing but @@@ every where, has this been seen before and is there a fix ?
I have checked the psu and all voltages are correct, I have swapped chips IC38 & IC39 with a dragon 32 that works (can't check IC34 & IC37).
hope someone can help,
cheers

[sixxie]

RAM is always a possibility... and while I don't have much experience with dead Dragon 64s (mine have held up rather well!), I'd reckon on that: there's a checkerboard pattern of $00 and $ff that, for some reason, seems to be the default state of RAM on power on - if you're seeing all $00 (inverse '@'), that implies to me either dead RAM, or the data isn't getting to the VDG.

Unlikely, but can you "type blind" and generate sound or MOTOR ON / MOTOR OFF and hear the relay? If so, perhaps give IC20 a look (octal buffer that clocks data from RAM data lines to VDG).

All guesswork though - there are others here with more experience actually fixing stuff.

[centrespot]

I tried what sixxie suggested

motor on :: motor off :: sound 1,1 but nothing happening.

I removed IC34 & IC37 (ram ?) and the checked board was displayed first in black/white then after a lot of screen distortion a brown/yellow checked board was displayed.

Excuse my lack of knowledge but are these the ram chips ? should this be happening ? if they are faulty where can I get replacements or will any 27128 chip do ?

[sixxie]

No those contain the BASIC ROM (and the alternate one that is copied into RAM in 64K BASIC mode) - so replacement is going to be tricky (and do make sure you got them back in the right sockets - they won't work swapped).

I don't have a board in front of me, but the schematic has it that the RAMs are IC15, 16, 21, 22, 26, 27, 32 & 33.

Perhaps first trigger off output 1/2 of the 74LS138 connected to SAM S0,1,2 and see if the ROM is serving up useful data. Hopefully someone else can offer some more obvious next things to check...

[Alastair]

In case you have missed them, there are some schematics at http://archive.worldofdragon.org/archiv ... chematics/ and the official Dragon Data Ltd - Dragon 32 Service Info at http://archive.worldofdragon.org/archiv ... r=Manuals/ . These may be of some use to you, but remember that you use them at your own risk.

[centrespot]

Again I am showing my ignorance, it seems that by getting lazy and using pc's/Microsoft I have forgotten most of my board level computer knowledge. It is making me determined to get this knowledge back, I might even take a college/university course, if they still do them on 8 bit machines.

Could you explain in lay-person terms what needs to be done to :

"first trigger off output 1/2 of the 74LS138 connected to SAM S0,1,2 and see if the ROM is serving up useful data."

Was that a massive groan I heard ?

Seriously, you would be helping me greatly if you could guide me through this, even if the outcome is fruitless. I would benefit from the experience.

[sixxie]

I was suggesting poking an oscilloscope at some bits to see if the ROM's ok:

The SAM (IC39) does address decoding and device select amongst other things, and it has three outputs, S0, S1 & S2 that form a binary number that breaks the address space down into large blocks. In the Dragon, these are fed to a 74138 device, a 3->8 demultiplexer, that converts that to 8 device select lines.

S=1 and S=2 both refer to areas in the BASIC ROM, so I imagined you could set a scope to trigger off a transition of those lines on the demultiplexer to see what data came off the ROM's data lines. The first addresses the CPU should access will be the RESET vector, which is in the S=2 region. I guess really you could just trigger off the RESET line...

I was about to suggest trying transplanting the Dragon 32 ROM, but I see it's split into two smaller chips - shame.

The biggest help is going to be the schematics and the data sheets for the various parts - and hopefully someone knowing of more common faults to check out first without having to go into all this detail!

[centrespot]

Hi Sixxie & Alistair

I have got myself another Dragon 64 that is working. I have done as Sixxie suggested and checked the following voltages :

Reset switch Working Dragon Faulty Dragon

S0 (pin 25) not pressed 0.18v 0.18v
S0 (pin 25) held in 4.62v 4.42v

S1 (pin 26) not 4.59v 1.3 to 2.4v fluctuating
S1 (pin 26) held in 4.62v 4.42v

S2 (pin 27) not pressed 0.18v 1.3 to 2.4v fluctuating
S2 (pin 27) held in 4.62v 4.42v

I have swapped the chips that I can and they all work properly in the working Dragon.
Am I right in suspecting a faulty IC40 (74LS138) or does the IC39 output to IC40 so it is somewhere else ?

Cheers

[centrespot]

The next test I did is :

Removed IC39 from both Dragons and compaired voltages on all pins of IC40 and they are the same/similar.
I now suspect that it is something else prior to IC39 that is affecting S1 and S2.
But first I will remove the IC34 & IC37 RAM chips and compair results.

[centrespot]

Original results re-done to read more easily



_______________Reset switch _______Working Dragon ____________________Faulty Dragon

S0 (pin 25) _____not pressed ___________0.18v______________________________ 0.18v
S0 (pin 25) _____held in _______________4.62v ______________________________4.42v

S1 (pin 26) _____not __________________4.59v ________________________1.3 to 2.4v fluctuating
S1 (pin 26) _____held in _______________4.62v ______________________________4.42v

S2 (pin 27) _____not pressed ___________0.18v ________________________1.3 to 2.4v fluctuating
S2 (pin 27) _____held in _______________4.62v ______________________________4.42v