CPC464 - sound is present but 'out of tune'

Started by gw0udm, 01:28, 07 July 21

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

gw0udm


Hi




I'm not having much luck with CPC464s at the moment. I have two and both seem to have sound issues. The first one I've already posted about and has some issue with the audio amplifier because the sound is there but basically inaudible. I'm still stuck on that, have replaced almost all the caps with no improvement so suspecting the amplifier now.




On my other one, the sound apparently works and is nice and loud (beeps etc fine). However when I try some games, although the music plays it sounds terrible - it sounds out of tune and out of time as well with all sorts of horrible squeaks and clicks. If you know what you are listening for you can basically hear the tune but it's not a pleasant experience.




Other than this though the machine works absolutely fine.




Is it as simple as a faulty AY chip? Ideally not - because I don't have any spares. Any thoughts?




James


Bryce

Without hearing it, it's very difficult to guess what it could be.

Bryce.

ajcasado

Hi,
I had a similar sound issue. The problem was that I had changed the Z80 with a CMOS one, and certain new games and demos that uses the Arkos player from Targhan had the sound scrambled as the Arkos code uses the undocumented instruction OUT (c),0. This instruction outs a 0 in a NMOS Z80 and a 255 in a CMOS Z80.
CPC 664

Empiezas a envejecer cuando dejas de aprender.
You start to get old when you stop learning.

gw0udm

Quote from: Bryce on 10:57, 07 July 21
Without hearing it, it's very difficult to guess what it could be.

Bryce.


Fair point! It is hard to describe. I'll post up a Youtube video with some examples.

gw0udm

Right here is a video with a demo of the issue:



https://youtu.be/vQxOdu9ZviI


I'm playing the intro music from 'Head Over Heels'


The first one is the music sounding correct - on the CPC with the quiet sound. The second one is the one with the music problem.


Listening to it again, it almost is like the main channel is OK but the other channels are either absent or playing all the wrong notes.


I'd be grateful for any thoughts. I've had a look at the main board and there's no apparent evidence of anything being changed but it need to have another looj.


Thanks!


James

Bryce

That's painful! Probably a dodgy AY.

Bryce.

gw0udm

Quote from: Bryce on 00:46, 08 July 21
That's painful! Probably a dodgy AY.

Bryce.


Thanks - rather what I feared. Although it's a relief that it's not meant to sound like that, I started off thinking "hmmm maybe the CPC sound isn't as good as everyone makes out".


is there any reliable source for replacement AYs? The plan now is to marry the best bits of the two together (basically put the cassette mech etc from the dodgy AY one into the one with the busted amplifier) so I should get one which works correctly and one with issues (although the mech needs attention).


I'd still like to fix the issues though...


James
James

Bryce

They are easy to find on ebay.

Bryce.

Shaun M. Neary

Oh man, that's aural sodomy at it's finest, right there!
Replace AY, should be okay.
Currently playing on: 2xCPC464, 1xCPC6128, 1x464Plus, 1x6128Plus, 2xGX4000. M4 board, ZMem 1MB and still forever playing Bruce Lee.
No cheats, snapshots or emulation. I play my games as they're intended to be played. What about you?

darkhalf

Don't get the A version from Microchip. I got a CPC464 with dead row of keyboard inputs. Was unsure about the China eBay ones so got one from Amigastore. The input voltages are slightly different to the GE version. Result is a bunch of random characters following key presses


I've got one from eBay USA seller with GE on its way now
https://www.ebay.com/itm/294034880267
CPC6128, 2xCPC464, C64, A500

Bryce

Quote from: darkhalf on 10:56, 14 July 21
Don't get the A version from Microchip. I got a CPC464 with dead row of keyboard inputs. Was unsure about the China eBay ones so got one from Amigastore. The input voltages are slightly different to the GE version. Result is a bunch of random characters following key presses


I've got one from eBay USA seller with GE on its way now
https://www.ebay.com/itm/294034880267

It's not the input voltages that are different, rather the fact that one version has internal pull-up resistors and the other one doesn't.

Bryce.

gerald

Quote from: Bryce on 11:03, 14 July 21
It's not the input voltages that are different, rather the fact that one version has internal pull-up resistors and the other one doesn't.

Bryce.
I beg to differ.
Both have pull up, but the A version ones are weaker and it takes more time to get the input high again than the scan loop period.


darkhalf

#12
I went and compared both data sheets after discovering this issue (as my second 464 still isn't working).


As per my previous post, the levels of the input voltages to the AY are different:
Looking at the I/O ports, the levels are slightly different with the A being lower, hence the problem
Low:  0-0.6 vs 0-0.5


Looking at the I/O ports, the levels are slightly different with the A being lower, hence the problem
Low:  0-0.6 vs 0-0.5
High: 2.4-VCC vs 2.2-VCC



So as you mentioned Bryce, this seems due to different internal pullups between these two versions, hence the change in levels


I also wrote to ikonsgr745 about this, and he mentioned that the CPC6128 has a spare footprint for a pullup resistor network, which I can see next to the AY


This was utilised in later 6128 which used the A variant. Unfortunately the 464 doesn't have this, so needs the original as OP required


Bryce - do you know what the values of the pullup resistor network for the AY inputs should be for those later boards?
CPC6128, 2xCPC464, C64, A500

Bryce

NR102 on the Plus is an 8x 2.2K array.

Bryce.

Arnold

LOL I say keep the faulty one for its very funny sound I am still laughing as I am writing this comment :)

GUNHED

Well, the AY can be the problem, but also a "dry" solder joint I guess.
However if you replace it you can add a socket anyway.
http://futureos.de --> Get the revolutionary FutureOS (Update: 2022.03.09)
http://futureos.cpc-live.com/files/LambdaSpeak_RSX_by_TFM.zip --> Get the RSX-ROM for LambdaSpeak :-) (Updated: 2021.12.26)

SerErris

I cannot see how solderjoints would result in that problem.

One channel of the three is playing correctly .. So that means all the inputs are correct (including clock address and data lines). Also the keyboard works, so the IO pins are also working correctly. The output lines are then the only ones in questions that might have solder joint issues ... 

I also would suspect that two of the three voices are broken in the chip.

I have created a small basic program that you can try to identify what is wrong.

It plays 
  • simple sounds test on each channel,
  • Noise test on each channel
  • Volume Envelope teston each channel
  • Tone Envelope test on each channel
Proud owner of 2 Schneider CPC 464, 1 Schneider CPC 6128, GT65 and lots of books
Still learning all the details on how things work.

gw0udm

Quote from: SerErris on 19:57, 24 March 22I cannot see how solderjoints would result in that problem.

One channel of the three is playing correctly .. So that means all the inputs are correct (including clock address and data lines). Also the keyboard works, so the IO pins are also working correctly. The output lines are then the only ones in questions that might have solder joint issues ...

I also would suspect that two of the three voices are broken in the chip.

I have created a small basic program that you can try to identify what is wrong.

It plays
  • simple sounds test on each channel,
  • Noise test on each channel
  • Volume Envelope teston each channel
  • Tone Envelope test on each channel

Ahh thanks so much for this. Sorry for the late response, have not been active on here for a while and have had to put the CPC projects away for a spell. However I am working on getting a new retro area established at home so when that's ready I'll look into this further. Sounds like it's definitely worth reflowing the joints on the AY as a starting point so I will do that.

However what I have now is probably a fairly unique 'circuit bent' CPC 464 although definitely not to my taste!

I'll report back when I've tried your listing out.

James

Powered by SMFPacks Menu Editor Mod