Errors in PT-80 delay schematic?

[geertjacobs]

In this thread was mentioned that when the PT-80 ((schematic))is engaged the high frequencies are boosted a bit compared to true bypass. In that same thread Mark Hammer was wondering about the filtering at the input buffer.
Normally I would expect that the input buffer is more or less acting as a transparent bandpass filter that lets the frequencies of a guitar pass to the companding and delay circuit. Same for the output.
So after the remark of Mark I started calculating:
At the input buffer a 15n cap and a 1K resistor in series form a high-pass filter at 10.6kHz.
A 10K resistor and 470p cap in parallel form a low-pass filter at 33.9 kHz.

At the output buffer we have a slightly different filter setup because the opamp is used to add the wet and dry signal. The bandpass filter is made in the feedback loop and has the same high-pass filter of a 15n cap and a 1k resistor in series, so cut-off is again at 10.6kHz. The low pass filter is 10K in parallel to 47pF, which sets the cut-off frequency at 338kHz?

These seem some very strange values for a bandpass filter in a guitar effect pedal, but of course I could have missed something...
Can anyone please verify my calculations or point me to the error I'm making?

[geertjacobs]

 
When comparing the PT-80 with the AD-80 i saw that the input and output is exactly the same on both schematics.
Apparenty it serves as pre-emphasis and de-emphasis.
I don't get the frequency ranges that have been chosen though...
How does the de-emphasis to compensate the pre-emphasis?

[bioroids]

How does the de-emphasis to compensate the pre-emphasis?

You boost the highs before the delay circuit, and cut the exact amount of highs after it. So in the end you get the flat frequency response, but you also get to cut the hiss of the BBD (or the noise generated by the delay circuit).

But the frequencies should be in the guitar range, so may be an error on your calculations...

Luck

Miguel

[geertjacobs]

You boost the highs before the delay circuit, and cut the exact amount of highs after it.

I understand that, but I think that this isn't happening.
The input filters are 15n - 1K and 470p - 10k.
The output filters are 15n - 1K and 47p -10K.

Even though i don't seem to use the right formula to calculate the cutoff frequencies, I still would expect pre-emphasis and de-emphasis to use the exact same cutoff frequencies, right?

Since the values are literally copied from the Ibanez AD-80, maybe the filtering is changed on purpose to get a small high-frequency boost?

[shoegazed]

glad i looked this up.

would a worthwhile experiment be to change the 47n to 470, or vice versa?

[geertjacobs]

Although it shouldn't matter which one you change as long as the filters are identical, I would think it makes more sense to change the 47p to a 470p since this would provide a cutoff frequency of 33.6kHz.
Wonder if i have cap like that in my junkbox...

[shoegazed]

i'll try it out, and see what happens...

thanks...andy

[geertjacobs]

Just tried putting a 470p cap in parallel to the 47p cap using two crocodile clamps.
I couldn't hear any difference. Probably makes sense, since the cutoff frequency is still out of the audible range...
Would it have been a kind of copy-protection from Ibanez?

Still don't know where the high boost is coming from...