Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

C64 Sound Too Quiet

642 views
Skip to first unread message

Justin the Almighty

unread,
Apr 18, 2001, 8:41:13 PM4/18/01
to
I have a Commodore 64, and the sound is very, very quiet, to the point that
I
have to crank the volume on my 1702 monitor to maximum to hear anything at
all, and of course you get this annoying buzz at that point, and there's
this
really loud *POP* every once in a while (I think it's when the sound is
turning on and off internally). I don't think the problem is with the
monitor
or cable, because I had basically the same problem using an RF switch on a
television (without the pop, and much statickier). I don't think the problem
is software, either, because I've tried it with several different games and
some SID playing software.

My question: Is there some easy fix for this, like a well-hidden volume
setting or a misaligned cable, or do I have to replace something? If so,
what?
Or do you know if there's any kind of (cheap) equipment I could use to
filter
out the static and amplify the sound?

Also, the system isn't grounded--could this be doing it?

Thanks in advance.

------------------------------------------------------------
Get your FREE web-based e-mail and newsgroup access at:
http://MailAndNews.com

Create a new mailbox, or access your existing IMAP4 or
POP3 mailbox from anywhere with just a web browser.
------------------------------------------------------------

Matthew Montchalin

unread,
Apr 18, 2001, 9:55:50 PM4/18/01
to
On Wed, 18 Apr 2001, Justin the Almighty wrote:
|I have a Commodore 64, and the sound is very, very quiet, to
|the point that I have to crank the volume on my 1702 monitor
|to maximum to hear anything at all, and of course you get this
|annoying buzz at that point, and there's this really loud *POP*
|every once in a while (I think it's when the sound is turning
|on and off internally). I don't think the problem is with the
|monitor or cable, because I had basically the same problem
|using an RF switch on a television (without the pop, and much
|statickier).

Okay.

|I don't think the problem is software, either, because I've
|tried it with several different games and some SID playing
|software.

Hmmm. With my 1701 monitor, if I plugged the audio jack into
the wrong plug, I would get really crappy audio. Have you tried
plugging the things into different sockets? Or how about using
different cables?

|My question: Is there some easy fix for this, like a well-
|hidden volume setting or a misaligned cable, or do I have to
|replace something? If so, what?

If your C-64 is used, there is a chance that the SID is partially
blown.

|Or do you know if there's any kind of (cheap) equipment I could
|use to filter out the static and amplify the sound?

How I wish the Super SID Expander board were a reality! You'd
just unplug the old SID, and plug in the new thing, and run a
cable out the back (after drilling a small hole in the case)...

|Also, the system isn't grounded--could this be doing it?

No, that shouldn't be it.

Cameron Kaiser

unread,
Apr 18, 2001, 11:27:17 PM4/18/01
to
Matthew Montchalin <mmon...@OregonVOS.net> writes:

>>I have a Commodore 64, and the sound is very, very quiet, to
>>the point that I have to crank the volume on my 1702 monitor
>>to maximum to hear anything at all,

:
:


>If your C-64 is used, there is a chance that the SID is partially
>blown.

Yes, I concur. This was the same symptom when I blew the first of two SIDs
on my SX-64. Replacing the SID cured the problem, until I totally walloped
it a second time. :-P

--
Cameron Kaiser * cka...@stockholm.ptloma.edu * posting with a Commodore 128
personal page: http://www.armory.com/~spectre/
** Computer Workshops: games, productivity software and more for C64/128! **
** http://www.armory.com/~spectre/cwi/ **

Matthew Montchalin

unread,
Apr 19, 2001, 3:19:55 AM4/19/01
to
On 18 Apr 2001, Cameron Kaiser wrote:
|>If your C-64 is used, there is a chance that the SID is partially
|>blown.
|
|Yes, I concur. This was the same symptom when I blew the first of
|two SIDs on my SX-64. Replacing the SID cured the problem, until I
|totally walloped it a second time. :-P

And each time a person does that, the world has one fewer SID chip-

If only there were a SID expander board, a lot of C-64's could be
saved from being parted out. (Sigh!)

Kevin Horton

unread,
Apr 19, 2001, 4:34:22 AM4/19/01
to
In article <Pine.SUN.3.96.101041...@compass.oregonvos.net>, Matthew Montchalin <mmon...@OregonVOS.net> wrote:
>On 18 Apr 2001, Cameron Kaiser wrote:
>|>If your C-64 is used, there is a chance that the SID is partially
>|>blown.
>|
>|Yes, I concur. This was the same symptom when I blew the first of
>|two SIDs on my SX-64. Replacing the SID cured the problem, until I
>|totally walloped it a second time. :-P
>
>And each time a person does that, the world has one fewer SID chip-

I've gutted 6 or 7 C64's so far for their SIDs <ducks>. Jameco used to sell
them new, but they got past $18 a pop before disappearing all together. (At
that price I can buy several C64's from a thrift, but I digress)

>If only there were a SID expander board, a lot of C-64's could be
>saved from being parted out. (Sigh!)

I've been reading with interest some of the discussions about replacements
for the SID chip. I've been thinking about this for awhile and
I think it is doable.

For those wondering, the SID's internal waveform generators use 12 bit DACs,
run through an 8 bit multiplying DAC to get envelope control, and finally to
the mixer which directs it nowhere, to the filter, or to the output bypassing
the filter. The output passes through a final 4 bit multiplying DAC which
serves as the master volume control.

All of the above could be reproduced in hardware simply by using a larger word
(say, 16 bits) and then using some adding/multiplication to emulate the
multiplying DACs. You'd need two DACs- one for audio that is run through the
filter, and one for audio that is not run through the filter. The two outputs
can then be combined into the final audio.

It would require an FPGA of some sort to hold all the logic stuff, then 2 16
bit DACs and a reproduction of the SID's all-important filter- fortunately I
have info on the filter used in it. Specifically an equivelant circuit that
uses a couple op-amps and DACs along with the usual 2 capacitors.

Eventually, I'd like to make a small portable SID player that has a SID
"clone" which does not draw nearly 1 watt while operating.

While I'm waiting for that ship to come in, I came up with this:

http://tripoint.org/kevtris/sidplay.html

Juergen Oppermann

unread,
Apr 22, 2001, 1:32:02 PM4/22/01
to

Kevin Horton wrote:
>
> It would require an FPGA of some sort to hold all the logic stuff, then 2 16
> bit DACs and a reproduction of the SID's all-important filter- fortunately I
> have info on the filter used in it. Specifically an equivelant circuit that
> uses a couple op-amps and DACs along with the usual 2 capacitors.
sounds like finally sb is knowing sth (;
i'd be intrested to learn about the how-it-was-done concening the
filters as well, please get in touch.

--
* Jürgen Oppermann
* http://www.joogn.de
* Deep Thought Systems Present "The Answer to all Questions" of the Day:
* 41,9999999999 (Powered by InTeL Pentium)

Christer Palm

unread,
Apr 22, 2001, 2:45:25 PM4/22/01
to
Kevin Horton wrote:
>
> All of the above could be reproduced in hardware simply by using a larger word
> (say, 16 bits) and then using some adding/multiplication to emulate the
> multiplying DACs. You'd need two DACs- one for audio that is run through the
> filter, and one for audio that is not run through the filter. The two outputs
> can then be combined into the final audio.
>
> It would require an FPGA of some sort to hold all the logic stuff, then 2 16
> bit DACs and a reproduction of the SID's all-important filter- fortunately I
> have info on the filter used in it. Specifically an equivelant circuit that
> uses a couple op-amps and DACs along with the usual 2 capacitors.
>

You can do the filter digitally as well.
Me and some friends had a project going a while back where we designed a
very simple 4-bit CPU to emulate the SID's state variable filter
circuit.

palm.vcf
0 new messages