Recreating the Binson Echorec with PT2399s

Started by mth5044, July 14, 2009, 09:48:22 PM

Previous topic - Next topic

mth5044

There is something about the sound of the echorec you can't get from a regular delay. I need to get this delay sound  ;D I've been pondering it for a bit now, and based on my limited knowledge, the best way I can come up with is using 8 pt2399s  :-\

I based the partial design around Hiwatt Bob's explanation in this video:

http://www.youtube.com/watch?v=ApEdnkiEGOY&feature=PlayList&p=8D6C53BABD49D7B9&index=0&playnext=1

So what I was thinking is each head selector button on there could be a separate chip, set for a different delay time on a trimmer (since you can't change the delay time on the echorec). The first four would just be a single, equal volume repeat at equal levels of time up to 310ms or so, the second 4 would be the same delay time settings as the first four, but with decay and feedback control. From what I get, the feedback is from the signal going around and coming back through the playback heads, but I can't seem to figure out a way to simulate that with the pt2399's, so I thought using 8 would work just as well.

From what I read, this is the delay part of the pt2399

http://i23.photobucket.com/albums/b379/sjaltenb/rebotecircuit.jpg x8  :-\

I was thinking, to make this project even cooler, was to use tube input buffers, instead of the tl072 (IIRC) that the echo base, rebote, pt80, etc use. Haven't done any research on that yet, but now is a good time to start.

I guess besides that, I'd have to have a mixer to combine all 8 chips and some elaborate switching system to incorporate the selector switch... so..

Any feedback or ideas? This project is likely to take a while to get all the ideas massed together and layouts drawn, so I have plenty of time for feedback and incorporation. Thanks dudes  8)

soggybag

Seems like it would be easier and more satisfying to make something new that gave a nod to the Binson. Imagine a box with 3 or 4 PT2399 chips. Rather than setting these up with a fixed delay time, it's easier to just add a knob. With an elaborate switching mechanism, you could set each delay in parallel or series.

Add a couple LFOs. That could be switched between or shared by various delays. Then you could add an envelope that could be applied to one or more delays. Of course by this point you have small modular delay synth that is best programmed by patch chords.

The BYOC delay does this with 2 PT2399 chips. I think they have a schematic on their site. The Echo Base is also good for some ideas. Oh yea, and Doug Deeper posted a schematic for his Clari(not). This was a PT2399 with an envelope tied to the delay time.

Too many possibilities. I think Binson made the choice they made to fit the technology they had. If they had the choice they probably would have mad different decisions.

mth5044

I completely understand where you are coming from, but with working with the technology they had, they created a sound that is unique to other delays. Sure, the DOD250 is smaller and easier to build then a tubescreamer, both are overdrives, but it's hard to get a tubescreamer sound from a DOD250.

And sure, the echo base modulation is cool (I have built it and it is my main delay) and the clari(not) is unique sounding, but it's not something I would use all the time. I have built the BYOC ping pong delay, and I have to say I am not impressed with the ping pong function.

When you say it is easier to just add a knob, yes, that is true, but then it won't be as easy to synchronise the delay times and whatnot. A trimmer is just as easy.

Maybe I'm just under a binson echorec spell, I dont know. I've been in love with Floyd's 'Live at Pompeii' cuts, and the video I posted.. I dont know.. it's just magical. I wouldn't mind adding in extras, i.e. modulation and whatnot, but those are all add-ons, once the basic layout is put together.

Thanks for your input  :)

Mark Hammer

Looking at the video, there are certainly a lot of sonic possibilities in he unit, some of which I've thought about before, not realizing they already existed.

Seems to me your best bet is really to score yourself a wad of VL3205 chips (Coolaudio replacement for MN3205), and have them all driven by the same master clock (with buffered clock lines to feed each 3205).  Conceivably, there is a way to sync a bunch of 2399 chips to the same master clock; I just don't know what it is.  Whichever path is chosen, there will be a need to have:
a) a master clock so that there is ONE hf-clock signal to contend with and filter out, and not heterodyning from multiple clocks,
b) buffering between consecutive SS "playback heads" so that a regeneration signal can be tapped OR an input signal fed to that tap.

I had been working on something distantly related to that using a 6-tap MN3011 a few years back, but it is unfinished and gathering dust.  The basis for it can be seen here: http://www.adadepot.com/adagear/gearpages/effects/ADA-STD-1.htm  You may find some applicable ideas in the documentation.  Strikes me as a cousin (albeit at a shorter delay range) to the Binson.

skiraly017

#4
I may be in over my head here but could you get away with 2 PT2399's? After watching the video you could set up the first PT2399 to function as the 1 through 4 one repeat only buttons via toggles or a rotary switch. The second PT2399 could be wired up as a regular delay...time, repeats and volume and you should be able to mimic the second bank of four knobs, yes? The idea is in my head and I know what I want to say but I can't get the words out!

Edit - The first PT2399 is wired for fixed delay times, repeats and volume which should be easy to do. I'm curious though, how would a Binson 2399 be different than running two separate delays?
"Why do things that happen to stupid people keep happening to me?" - Homer Simpson

slacker

There was some discussion a while ago that might be useful. http://www.diystompboxes.com/smfforum/index.php?topic=73717.0 I don't think anything ever came of it though.

skiraly017

Quote from: slacker on July 15, 2009, 01:03:18 PM
There was some discussion a while ago that might be useful. http://www.diystompboxes.com/smfforum/index.php?topic=73717.0 I don't think anything ever came of it though.

After reading that thread I have a better understanding of what's going on but wonder if the same (or close to) couldn't be achieved with two PT2399's?
"Why do things that happen to stupid people keep happening to me?" - Homer Simpson

slacker

Yeah you can get similar effects using just 2 stages by putting them in series, especially if you then feed the output back into both of the stages.

mth5044

Slacker - thanks for the thread link, I didn't think about actually feeding the signal back through the pt2399.. or atleast I don't think I did  ???  The trouble there is how do you control the feedback then? How the signal degrade? Maybe when you feed the signal back through the pt2399, put a 'volume control' that reduces the volume everytime it comes back around.. kind of like feedback I guess.

Skiraly - The first chip could be wired for fixed delay, and I can understand how to set each switch to have a different resistance to set the delay level, but how would it work if I wanted to active at heads 3 and 4? That would mess up the delay time if it's just one chip, I think. I can see how 2 chips would work if it were only to have one head activated at one time, but I think the echorec can have multiple heads active at once...

Mark - I've never got into (what I assume to be) an analogue or similar delay chips. I'll have a search around the forum and see if I can make some sense of your post   :P

JKowalski

#9
Quote from: mth5044 on July 15, 2009, 09:37:50 PM
Slacker - thanks for the thread link, I didn't think about actually feeding the signal back through the pt2399.. or atleast I don't think I did  ???  The trouble there is how do you control the feedback then? How the signal degrade? Maybe when you feed the signal back through the pt2399, put a 'volume control' that reduces the volume everytime it comes back around.. kind of like feedback I guess.

Look at every standard delay application of the PT2399 (rebote!) and you can see how the feedback system works. You just take the output delay signal, and mix it back in (rebote uses a standard op-amp mixer, see schematic - PT2399 makes it easy by incorporating two op-amp stages in the chip itself for this very purpose) with the input of the PT2399, and you have your multiple repeats. THe value of the resistor in the feedback loop changes the decay of each repeat, since the resistance of that feedback loop goes into a op-amp mixer.

Quote from: mth5044 on July 15, 2009, 09:37:50 PM
Skiraly - The first chip could be wired for fixed delay, and I can understand how to set each switch to have a different resistance to set the delay level, but how would it work if I wanted to active at heads 3 and 4? That would mess up the delay time if it's just one chip, I think. I can see how 2 chips would work if it were only to have one head activated at one time, but I think the echorec can have multiple heads active at once...

Okay - this entire application is EXTREMELY simple - I think you guys are all overthinking it.

You get 4 PT2399s. You set the first one with a trimpot for a certain repeat speed, set the second one to twice that, set the third to three times that, set the fourth to four times that.

Construt the full PT2399 delay stage from a rebote (you can easily see how they make this part in the schematic for the rebote) for each chip. Put a DPST/DPDT for each chip that either cuts it out of the circuit or puts it in parallel.

You now have the exact layout that you see with the first four buttons on the top of the echorec. Now, for the four bottom buttons, you make it connect the feedback loop to the corresponding PT2399. The value of the feedback resistor may be tricky - as with the normal echorec you change the "regeneration" for all four at the same time - you would need a quad set of voltage controlled resistors, or a pot that has four sections in it that change at the same time.


Ill make you a block diagram of the circuit. Its really easy.





JKowalski

#10


There. ECHO 1,2,3 &4 ON/OFF corresponds to the 1,2,3 &4 buttons on top, REG ON/OFF 1,2,3 & 4 corresponds to the buttons on the bottom. The four PT2399s are set with multi-turn trimpots on pin 6 each to get the exact delay times as in the BINSON ECHOREC's 1,2,3 and 4 heads. The REGENERATION needs to be controlled by a master regeneration knob - so look into voltage controlled resistors or something/circuit that will change four different resistances equally at the same time with one knob. The guy in the video was very unclear about the workings of the switching system so Iamgoing to have to look further into that to incorporate that into the diagram as well.

JKowalski

#11
Alright so I just watched the video a couple more times and this is what I gathered:



ECHO mode lets you ONLY use the four hard repeats (one repeat for each head)

REPEAT MODE lets you choose any of the four hard repeats and lets you add a regeneration to each head at your discretion.

SWELL mode engages all four heads and all four regenerations and lets you have no control over anything except for the swell knob.



SO - Seeing that a switching system in a conventional sense would be a little complicated, it might be best (easier, cleaner and smaller) to set up a simple digital switching system:






If you haven't already guessed, the main reason I am doing all of this is because I am very interested in the idea and may build this someday (not a clone as you want, but more of a extended concept kind of thing).

Mark Hammer

Again, I have to say that have a little reticence over so many disparate clocks co-existing on one board.  Individually, yes, they are high-frequency and unlikely to be audible, but their difference products may well be audible.  This is why I feel more comfortable with the idea of a couple of 4096-stage BBDs slaved to a single clock.

Gus

#13
I have no experience with the PT2399.  I just looked at the spec sheet I see an external clock out at pin 5.  I guess as a way to set the delay with a counter.

It is not shown how the resistor at pin 6 sets the internal clock except a VCO block.

If it had an external clock input you could do some fun things

  Has anyone tried clocking pin 6 with a pulse or squarewave generator to see what happens(I don't know if this will harm the chip)*?  If one could external clock that would allow a divider circuit or micro controller to change delay times.

If you can external clock this chip MAYBE use frequency setting resistors at pin 6 of one of them and divide the output from pin 5 to drive the other chips.  When you change the frequency of the one the other three track.

*I would try gen cap coupled to pin 6
Cap coupled with a resistor pull up to +5VDC or pull down to ground at pin 6 maybe 10K
Gen direct coupled to pin 6 and maybe a series resistor between the gen and pin 6
direct coupled to pin6
Limit the signal to between 0 and 5VDC.


JKowalski

#14
Quote from: Mark Hammer on July 16, 2009, 09:02:04 AM
Again, I have to say that have a little reticence over so many disparate clocks co-existing on one board.  Individually, yes, they are high-frequency and unlikely to be audible, but their difference products may well be audible.  This is why I feel more comfortable with the idea of a couple of 4096-stage BBDs slaved to a single clock.

I don't think this will be too big of a problem. The ping pong delay from BYOC has two chips, and I have not heard any reports of noticable clock noise. The Belton reverb module is supposedly a bunch of PT2399s in a box, and likewise I have not heard of excessive clock noise in those either. I don't think it will be a problem.

I guess it's just up to the builder - each way (BBD/PT2399) would probably work fine. In any case, the same layouts I posted would work either way.




By the way I totally screwed up drawing that switching thing. It was lateee and I wasnt thinking straight. Here's a new one:


mth5044

I'm glad got back to this thread before my weekend started. Now I can sit and ponder instead of sit and vegetate  :P

JKowalksi - I didn't know there was such a feedback part of the PT2399 circuits... that definitly makes this easier.

Thanks everybody  :)

JKowalski

#16
It's not really "part of the circuit". It's just a property of delays. Think of it.

You have a device that outputs a signal at a certain time after the original signal (once). Now, if the output of that device goes back into the input, that delayed signal will go through the delay device again, and comes out an equally spaced time period afterwards - and so on, and so on..... If the delayed signal is made smaller every time it goes back to the input, the overall feedback becomes less and less on every repeat. If it is exactly equal, it goes on infinitely. That's just how delays in general work. So, basically, to make a delay you just need one device that plays it's input signal a certain time at the output after a certain amount of time and you are set!

mth5044

Alright, so I think I'm coming to the conclusion that having the ability to switch between the echo, swell and whatnot modes are just novelty and were probably used because they were easy to implement in the binsons design. For this sound recreation, I think it is just going to be easier to have all 8 'head' switches avaliable. It might not be too hard, but the amount of work it would take for having the ability to either have the repeat heads switchable or all on is not worth it, when all I would have to do is use the 4 avaliable switches to turn them on  :)

So, I've dumbed down the design for eight switches, just 4 repeat on/off, then the regeneration for each head (using head because typing pt2339 circuit is longer :p). That still leaves all the sound avaliable, I suppose, but just a little easier to make  :)

I'm guessing the feedback switching could just be switching the fixed single repeat resistor out for the feedback voltage controlled circuit. Off to see what I can find about voltage controlled resistors :D

I want to also look into 4096-stage BBD's, but the ones I found at smallbear were $16 each  :-\

mth5044

Ugh, having trouble finding a way to split the input to 4 different circuits in parallel then having them mixed back together at the end.

I think I could just use JD Sleeps mini mixer for the end:

http://www.generalguitargadgets.com/diagrams/mixer_sc.gif

but I'm not sure how to blend the clean signal in with it too.

I can't seem to find anything about the inputs in parallel, as I doubt it would be easy as just sending the signals to the inputs  :icon_lol:

JKowalski

#19
Quote from: mth5044 on August 01, 2009, 03:08:11 PM
Ugh, having trouble finding a way to split the input to 4 different circuits in parallel then having them mixed back together at the end.

I think I could just use JD Sleeps mini mixer for the end:

http://www.generalguitargadgets.com/diagrams/mixer_sc.gif

but I'm not sure how to blend the clean signal in with it too.

I can't seem to find anything about the inputs in parallel, as I doubt it would be easy as just sending the signals to the inputs  :icon_lol:

Alright - it's actually really easy. Take a look again at the tonepad rebote 2.5 delay schematic. As I said before, the PT2399 has two op-amp circuits in it's design, you can see those at the bottom of the schematic, shown as normal op-amps but with the pin numbers of the PT2399 chip.

Now  - since you are using the feedback loop, that means the PT2399 input mixer op amp (the way it is used in the tonepad schematic) is being used as the input signal/feedback mixer. If you simply wired the pedal input to four of the input mixer signals in parallel, you would get crosstalk, as each feedback path would send a little bit of itself into the other feedback/input mixers. So you need an additional buffer in front of each PT2399, to prevent the feedback signal of each PT2399 going to the other ones.

As for the mixer for the output, again look at how it is done on the rebote 2.5 - you see the op-amp at the end of the schematic, before the output? You see how one resistor comes fromthe first op-amp in the circuit, and one comes from the pt2399 output? That's your mixer. It's just a simple inverting amplifier mixer, you don't need anything complicated. If you want to mix more inputs, you just connect the additional pt2399 stages in the same way the single one is connected in the rebote, with that resistor to the (-) input.

That's exactly what JD's mixer does, though I have no idea why he added an extra inverting amplifier stage. In any case, you can copy the rebote almost exactly, and just add more stages. It's a simple project, now that you gave up the switching system  :icon_biggrin:



Good choice with the simpified switching, by the way.


The feedback control circuit would be alot simpler if you just used seperate pots for each PT2399... Although I can see the appeal in changing them all at the same time.