Hello, before making the final casing for my Orwell computer I decided to try adding a 6581 SID to my machine. It is one pulled from a dead C64.

I am not having much luck though. I know these chips are notorious for failing but it would be good to check I am using it properly first.

The chip has only one /CS line. My machine is using Garth's address decoder from the primer. I am using an additional 74HC00 NAND for the extra decoding I need. I feed the /CS from the decoder into both inputs on one gate as an inverter. The output is feed into a second gate whose other input is the A8 address line. I think that should give me the correct /CS on the SID when POKEing to location $4100 (16640 decimal since my machine runs BASIC). If I check on the scope running a simple poke in a loop program I can see the correct negative pulse.

The SID though seems a bit suspect. It basically sits there making feeble noises all the time. But the noise is affected by whatever else is happening on the machine. So typing, running code, etc makes the noise change. It's a bit like all the computer in science fiction films who make noises when they are doing something.

I suspect the SID is dead (or badly wounded!) because even if I physically hold the /CS on it high it still makes the odd noises. The datasheet does say the reset needs to be 10 phi2 cycles to reset the chip registers to zero. I have a 100mS reset pulse and the machine runs at 1Mhz so that should be fine. That should make the chip silent on start-up.

It's as if the /CS isn't working so things are being written to it all the time making it do odd things and cause the strange noises I hear. Is that a common sort of failure with these chips? Or is there something amiss with my extra address decode I am missing?

I have ordered one of the Italian SwinSIDs off eBay to try (http://www.ebay.com/itm/Commodore-64-swinsid-replacement-for-6581-8580-SID-/171728731586?ssPageName=ADME:L:COSI:US:1123). I unfortunately don't have, or know anyone with, a good C64 to try the SID in.

Simon

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

Have you tried disconnecting and tying /CS high - that should test the reset-to-silence behaviour. (I have no idea if that behaviour is real - my experience is with the Beeb, where the software has to reset the sound to silence during initialisation, hence the double beep.)

Yes, I tried that. I figure with the correct reset and the /CS held high the chip should be silent. I am pretty sure it's broken as no matter what I do I see the same noisy signal coming from the output.

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

Oh, sorry, you did say you'd already tried that!

That's OK! I ask in case I am missing something obvious so I don't mind if people mention things I have already tried! Means I am thinking along the right lines

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

Your description convinces me that what you're hearing is induced noise on the ground wiring, not the actual output of the SID. I don't have enough information to suggest why the SID fails to produce any real output of its own. (Can you verify a DC level of about 6 volts on pin 27?) But I can describe how to reduce/eliminate the false signal you're hearing. Here (below) is a schematic representation of the SID and the pertinent connections.
In this case it matters a lot what the physical layout is. Specifically, pin 14 is is the place -- and no other -- where the various ground connections should physically attach together. I've illustrated that below.
This will reduce the noise. But most of these details won't affect the underlying problem, I'm afraid -- except maybe the bypass caps on 25 and 28, that is. Make sure both ends of both caps connect physically close to the respective IC pins.

When the underlying problem is fixed, I think you'll find the SID output is much louder than the noise. So reducing the noise is almost incidental. I wish I could be more helpful...

-- Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

I tidied the wiring as best as I could It's a bit tricky as it's bread boarded. It didn't seem to make much difference. I think you're right that is just noise. I had the gain right up on my amp too making it sound worse. Even with that the SID should produce something way more than that itself.

I get around 5.6 volts on pin 27. The noise on that seems to be about 150mV. Interestingly if I remove the 1k resistor to ground that drops to 50mV.

The other thing I notice is a little negative spike on the /CS line sometimes. About 40nS long, dropping to about 2 volts. This happens just as the R/W line is going high ans when I am not specifically addressing that chip. If I actually address the SID (say poking 0 to it in a loop) I see the correct /CS low pulse with the RW line neatly going low for the second half of that.

I am pretty sure the CS logic is ok. Not sure about that little glitch though.

Simon

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

It sounds right, although the four cascaded levels of '00 will give you more propagation delay than you'd get with a '139 doing the same thing: A15 goes to G\, A14 goes to input B, and A8 goes to input A. Then the Y3 output goes to CS\ of the SID.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Just a follow up to my SID issues. I went ahead and made up a better Veroboard daughter board for my machine. I took care over the earth wiring this time and that has got rid of the weird noise issues. Still no audio output from the SID though so I think it's duff. I am however able to read values on the POT x and Y pins so I think my addressing and chip select is all fine. Funny, I spent a lot of time designing and making a little IO board with ADCs so I could read pots for a joystick and now with the SID I get those for free!

My SwinSID is on it's way and it is a plug in replacement I am hoping that solves the issues. I'm going to go ahead and start making the housing for my machine now. Even if the SID doesn't work the way the board connect means it can easily be left out of the final machine.

Simon

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

I'll ask the obvious question. Is there a suitable clock signal going to it?

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

It uses the Phi2 signal straight from the 6502 so that should be fine I think.

I think the fact that I can read the pot pins correctly means all addressing/clocking etc is working fine. It's just the sound output not working and apparently that's a common failure on these old SID chips.

Simon

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer