PT2399 Delay with true bypass and unbuffered dry path?

[swever]

It that even possible? Any projects?

I've breadboarded the small time delay (http://www.diystompboxes.com/smfforum/index.php?topic=95632.0e) yesterday. It works well, but ruins the tone and dynamics. There is a huge level drop with it (bypassed) in the chain after a distortion that can go above unity as compared to without it there at all.

Another thing I still can't understand, that is closely related to the questions above, is why is it not possible to have tails and true bypass together? Isn't it possible to mix unbuffered dry with buffered wet delay?
I do have a clue that it has to do with input&output impedances, loading, etc - the concepts that I can't wrap my head around because thats just beyond my mental abilities

[Phoenix]

It that even possible? Any projects?

Well it's possible-ish, you could use a passive mixer, but results would be unpredictable with unknown source and load impedances, and the signal would ALWAYS be attenuated with the effect on, and then of course if it's true bypass, your bypass sound is no longer going to be attenuated, so your dry signal would always be less than unity when the effect is engaged - your volume would jump up when you bypassed it (the wet signal could be amplified, but it'd be weird to have a delay with repeats louder than the dry signal).

I've breadboarded the small time delay (http://www.diystompboxes.com/smfforum/index.php?topic=95632.0e) yesterday. It works well, but ruins the tone and dynamics. There is a huge level drop with it (bypassed) in the chain after a distortion that can go above unity as compared to without it there at all.

I'm not sure I'm following exactly what you're saying, but it would be helpful if you could tell us what distortion box you're using, as that may explain what you're experiencing.

Another thing I still can't understand, that is closely related to the questions above, is why is it not possible to have tails and true bypass together? Isn't it possible to mix unbuffered dry with buffered wet delay?

Well true bypass by its very definition completely disconnects the circuit from the signal path when bypassed, you can't very well inject tails into the signal if it's disconnected!

I didn't notice your white text until I went to reply, here it is for those who want to be able to read it.

I do have a clue that it has to do with input&output impedances, loading, etc - the concepts that I can't wrap my head around because thats just beyond my mental abilities

Yes, you're right, it's all about input and output impedances. It becomes easier to understand when you realise that impedance is just AC resistance. Then it's just all voltage dividers, something which most people grasp much more easily. Sure, impedance changes with frequency if there's any inductance or capacitance, but as a first-order approximation you just don't need to worry about that - until you do.

[swever]

Well it's possible-ish, you could use a passive mixer, but results would be unpredictable with unknown source and load impedances, and the signal would ALWAYS be attenuated with the effect on, and then of course if it's true bypass, your bypass sound is no longer going to be attenuated, so your dry signal would always be less than unity when the effect is engaged - your volume would jump up when you bypassed it (the wet signal could be amplified, but it'd be weird to have a delay with repeats louder than the dry signal).

Well true bypass by its very definition completely disconnects the circuit from the signal path when bypassed, you can't very well inject tails into the signal if it's disconnected!

Yea, that makes sence. I somehow forgot that you have to split the signal into dry and wet, which is impossibe without any buffers/active ciruitry. Guess I'm having too many beers over the weekend!

I'm not sure I'm following exactly what you're saying, but it would be helpful if you could tell us what distortion box you're using, as that may explain what you're experiencing.

It's a dod250/mxr dist+ clone i've built. What I'm trying to say is that when the dist is bypassed (or is not present in the chain), there is not so much difference in output level between the small delay being present in the signal chain (in it's bypass mode) and no small dealy in the chain. Once I engage the dist (that can boost the peak level significantly at it's max volume setting), the difference becomes apparent. It's much louder (and better sounding) without the small delay in the chain. All of that applies to the dry part when the small delay is on.

[samhay]

> It works well
That's my experience

> but ruins the tone and dynamics.
This is not my experience, nor is it working well. I suggest you take another look at the breadboard as I suspect you have made an error somewhere

[ElectricDruid]

I suggest you take another look at the breadboard as I suspect you have made an error somewhere

+1 agree. I think he might be onto something. The Small Time Delay circuit looks solid to me - I wouldn't expect it to produce any undue loading or cause any problems. It's all properly buffered in and out, and should act reliably the same pretty much whatever you connect it to. If that's not happening, perhaps you've got a breadboard error somewhere?

HTH,
Tom

[anotherjim]

I would say you need to check your components around the input op-amp of your Small Time, particularly the resistors -  are they the values you think?

There's nothing in the ST that would limit levels any more than any other 9v pedal could. As with any "Tails" delay, you can have an additional footswitch that give full bypass if you want, although it may need a bigger box.

Dist+ has a 10k resistor in it's output (before the clipping diodes), so too low input impedance in the next thing you plug it in to will load down it's output level. As you also built that, is it 10k?

It's no sin to confirm resistor values with your DMM. I do it all the time.

[blackieNYC]

If you want TB, skip the tails. And having one with and one without tails, skip the tails.  It's really not all that exciting IMO. The excitement was in coming up with a way to do it.

[ElectricDruid]

It's no sin to confirm resistor values with your DMM. I do it all the time.

Not only is it no sin, but I gave up actually reading the damn stripes ages ago. Much easier to dob a meter on them and check the value. There are still plenty of values I recognise, but it's far too easy to mix up a 10K and a 100K, for example.

T.

[blackieNYC]

+1 - banana plugs on the meter's input, and I hold the resistor leads in there.

[bluebunny]

. . . but it's far too easy to mix up a 10K and a 100K, for example.

Particularly when the orange looks kinda brown.  And the red looks kinda brown.  And the violet...   

[swever]

Thanks for all the suggestions. I will double check if I have any errors or wrong values. I'm usually quite careful, go slowly and don't make too many mistakes on breadboard. But then this time it looks like this, so it there might be a mistake somewhere. 



The other suspect is the opamp. I use really cheap stuff from ebay. Maybe it's not working properly.  I think I'm going to try 4558 is going to make any difference.

[swever]

Breadboard error confirmed! Had the R14 (10k) going from pin3 to pin6 of the opamp, instead of pin1 to pin6. Now it sounds great - no audible difference between buffered bypass and true bypass, at least with a distortion in front of it.

Thanks peeps!

[samhay]

excellent - another win for team DIYSB