hi all,
I made some sounds with it for people to hear. Just incase your
wondering, I didn't try to make it musical at all. Infact I'm trying to
show the amount of distortion I get for a few different settings. mostly
justing hitting different Cs.
These are unfiltered inputs from a sawtooth oscillator. I didn't have
the compander hooked up when I recorded. I cant' say that the noise you
hear is from the delay or not because it seems everything I record is as
noisy.
This one uses the switched cap filters clocked by the delay clock
divided by 4. The PT2395 is in short mode/256k. this one doesn't sound
that great.
http://www.sdiy.org/destrukto/mp3/pt2395_div_by_4.mp3
same with divide by 8 for filter clock. This is much better (not
perfect), and I think i'm going to build it as a module like this. max
delay is about 1.5 or 1.6 seconds.
http://www.sdiy.org/destrukto/mp3/pt2395_div_by_8.mp3
This one shows the minimum delay I'm getting. well almost, I had it just
a tad shorter on the breadboard. I still need to adjust the 74hc4046 RC
values a bit. This is 64k/short mode.
http://www.sdiy.org/destrukto/mp3/pt2395_64k_short.mp3
They all sound much nicer with a filtered input, and of course playing a
nice sequence or maybe if someone who could play music was controlling
it. I've been reading a little about delta modulation (the ADC and DAC
it uses) and it seems that the 5th order filters I'm using shouldn't be
needed. One of the advantages for this type of ADC is supposed to be
easy filtering. The data sheet claims that at 4mhz clock, the
samplerate is 500khz. this would mean the sample rate is clock/8, which
is what i'm using for the filter clock. the filter cutoff is clock/100,
so the cutoff should be delay_clock/100. that seems like it should be
more than enough room for a 5th order filter even without an
oversampling converter. anyone know about these things?
about really short delays with cut address line. I decided there was no
reason why I shouldn't test this now. it's only soldering in 1 more
resistor... so I cut A7 and put a 10k in Ra7. this just made the output
very distorted (I suspected it would), but I wonder if that is what was
intended, or if anyone has some other idea about doing this?
,Ryan
I made some sounds with it for people to hear. Just incase your
wondering, I didn't try to make it musical at all. Infact I'm trying to
show the amount of distortion I get for a few different settings. mostly
justing hitting different Cs.
These are unfiltered inputs from a sawtooth oscillator. I didn't have
the compander hooked up when I recorded. I cant' say that the noise you
hear is from the delay or not because it seems everything I record is as
noisy.
This one uses the switched cap filters clocked by the delay clock
divided by 4. The PT2395 is in short mode/256k. this one doesn't sound
that great.
http://www.sdiy.org/destrukto/mp3/pt2395_div_by_4.mp3
same with divide by 8 for filter clock. This is much better (not
perfect), and I think i'm going to build it as a module like this. max
delay is about 1.5 or 1.6 seconds.
http://www.sdiy.org/destrukto/mp3/pt2395_div_by_8.mp3
This one shows the minimum delay I'm getting. well almost, I had it just
a tad shorter on the breadboard. I still need to adjust the 74hc4046 RC
values a bit. This is 64k/short mode.
http://www.sdiy.org/destrukto/mp3/pt2395_64k_short.mp3
They all sound much nicer with a filtered input, and of course playing a
nice sequence or maybe if someone who could play music was controlling
it. I've been reading a little about delta modulation (the ADC and DAC
it uses) and it seems that the 5th order filters I'm using shouldn't be
needed. One of the advantages for this type of ADC is supposed to be
easy filtering. The data sheet claims that at 4mhz clock, the
samplerate is 500khz. this would mean the sample rate is clock/8, which
is what i'm using for the filter clock. the filter cutoff is clock/100,
so the cutoff should be delay_clock/100. that seems like it should be
more than enough room for a 5th order filter even without an
oversampling converter. anyone know about these things?
about really short delays with cut address line. I decided there was no
reason why I shouldn't test this now. it's only soldering in 1 more
resistor... so I cut A7 and put a 10k in Ra7. this just made the output
very distorted (I suspected it would), but I wonder if that is what was
intended, or if anyone has some other idea about doing this?
,Ryan