Re: 6567 revisions (was Re: 264/TED/Plus4 Story)

From: Hársfalvi Levente <>
Date: Thu, 25 Aug 2011 22:58:31 +0200
Message-ID: <>
Yeah; in the 6581, something around the envelope DACs put a constant and
pretty large DC offset to the output, which is then summed by the mixer
(...occassionally inverted, if the respective channel is routed through
the multimode filter first), which still remains when the mixed signal
reaches the 4-bit multiplying DAC of the master volume.

With the 8580, the envelope generators have been fixed. They don't put
any DC load to the mixer anymore.

The extin trick does something similar in principle (...but different in
details) - as the external in signal is simply fed to the mixer,
similarly to the SID channels (...either filtered or not), a DC offset
put here biases the mixer, ie. adds a DC to the signal fed to the
"master volume" DAC, similarly as seen before.

It might be important to note that 1.) biasing extin slightly increases
overall distortion, due to the signal path being biased ie. the
"amplitude" (to be dealt with the SID inners) being larger; that's
clearly audible in such modified setups. 2.) The trick mostly does its
job, but is effectively not fully similar to the original one (where
filtering some channels would invert the DC level of the respective
channel, thus change the amount of DC put onto the mixer - that one
won't happen here, since the DC won't come from those sources).


On 2011-08-25 22:07, Rainer Buchty wrote:
> On Thu, 25 Aug 2011, Gerrit Heitsch wrote:
>> The trick to 'fix' the 8580 was to pull the EXT_IN pin (!) to GND via
>> a 330 KOhm or larger resistor... Makes me wonder what that does inside
>> the chip.
> It adds the DC level which was (IIRC by mistake) present in the 6581,
> hence you again hear the "crackling noise" when changing the volume
> setting.
> Rainer
>       Message was sent through the cbm-hackers mailing list

       Message was sent through the cbm-hackers mailing list
Received on 2011-08-25 21:00:29

Archive generated by hypermail 2.2.0.