DOD FX-25 filter-type mod confirmed

[chemosis]

im having issues getting a noticeable result from the decay mod also. it seems so sutle if working at all. ive had the same problem with a decay mod on my bassballs. I built 3 bassballs over the course of 2 years all with decay mod but only one worked effectively like in the video. I tried marks suggestion. maybe a 2m pot with a 22k resistor this time

[Mark Hammer]

The decay time depends on how much current is stored by the cap, how quickly it charges up, and how quickly it discharges.

I'm wondering if the 1M resistor between V+ and the 22uf cap is the better path to pursue.  It is reminiscent of what happens in the Dynacomp with the 10uf cap and 150k resistor to V+.  That resistance sets how log it takes the storage cap to "recover".  Try throwing a 470k in parallel with the 1M and see if that shortens the decay time.

[daeg]

Hope you guys don't mind me reviving this thread. I'm gearing up to build one of these and have some curiosities.

What would happen if you mismatch the resistance (and thereby current) going into each 'Amp Bias Input'?

On this schematic, R12 & R13 are both 10k. What should we expect if 5k and the other 15k?

[PRR]

> What would happen if you mismatch the resistance (and thereby current) going into each 'Amp Bias Input'?

It becomes even more unpredictable than it is. Especially at very low current (low frequency).

I don't know why you would want to offset the two filters, but surely a safer technique would be to offset values of C9 C10.

[Rob Strand]

Changing the ratio of the currents will make the gains of the integrators different.   Off hand, that's one way to change the Q of the state-variable band-pass filter.

There's some stuff here on the analysis which you could modify,
https://www.diystompboxes.com/smfforum/index.php?topic=125170.0

However you could poke around of the web for state-variable filters for a web page showing the effect of changing the gain terms, the k's.


EDIT:
Here's the idea,
https://www.electronics-tutorials.ws/filter/state-variable-filter.html

- circuit about 10% down,
- then at about half way down, there's an expression for Q.

If you imagine increasing the gain of the last integrator, is similar to decreasing R4, and that shows Q increasing (I think).
(The difference between the integrator change and the R4 change is it predicts a different gain at Vout.)

[daeg]

I'm going to try this 'Balance Pot' idea in the next week or so. I'm thinking of trying a 10k pot with series with a 10k fixed resistor.

                                        [Env Follower]
                                                  ↓
                                        [10k Resistor]
                                                  ↓
Amp Bias Input A (Pin1)  ← [b10k Pot] → Amp Bias Input B (Pin16)

Do you guys think that will have a noticeable effect? Can we think of the 'Amp Bias Input' as having an input impedance?
My understanding of OTAs is weak and I can't find that info on the LM1700 data sheet.

[PRR]

The Iabc pin (called ABI here) is a "zero impedance".

Why do you need to know?

[daeg]

The Iabc pin (called ABI here) is a "zero impedance".

Why do you need to know?

Just making sure I understand it correctly and I'm trying a mod that will actually do something.

[daeg]

Based on what PRR's last post, the scheme above would probably not work. Going to try this instead.

                            [Env Follower]
                                      ↓
     [10k Resistor] ← [b10k Pot] → [10k Resistor]
               ↓                                           ↓
          [ABI-A]                                 [ABI-B]
           (Pin1)                                  (Pin16)