Echo Base - a new PT2399 delay

[jlullo]

yeah, ian, i'm with everyone else here-  this sounds great!  awesome job

[Ben N]

Great looking project, Slacker, can't wait for the schematic. BTW, I had to do a bit of a double take, but you have the terms "open" and "closed" with regards to switching switched around. "Closed" is generally understood to mean that the circuit is closed, i.e. completed. "Open" means the circuit is broken. Maybe counter-intuitive, and the opposite of my mother saying to "open the light", but there it is. So your CMOS switches are both closed when the effect is on and open when the effect is off, Boss-style.

BTW, the old-fashioed way to do delay tails is with a volume pedal in front of the delay. Of course, if you want to play over the tail, the VP has to be a panning type, with the second output going to the amp or a mixer. Your solution is obviously superior.

[BubbaKahuna]

I bet with the settings right on this it does a really good imitation of a Magnatone Tremelo.
I have a couple spare PT2399 chips laying around here, I might build one of these.
Sounds great!

Cheers,
- JJ

[Jaicen_solo]

Looks like a great project, I think i'll be getting into once I clear my current backlog!
Just a though, but have you considered adding the plug detector as a voltage input for the delay time??
I don't know if you're familiar with it, but it's basically a voltage output that increases with speed of picking. I'm thinking that would be a neat effect for the delay to track, though i've no idea how you'd calibrate something like that!

[slacker]

Thanks for all the feedback, some good ideas for mods. A wet feedback loop should just be as simple as breaking the connection between lug 2 of the feedback pot and the 20k resistor and splicing your sent/return jacks in there. So long as whatever pedal you put in the loop has a high input impedance I don't think you'll need a buffer.

BTW, I had to do a bit of a double take, but you have the terms "open" and "closed" with regards to switching switched around. "Closed" is generally understood to mean that the circuit is closed, i.e. completed. "Open" means the circuit is broken.

Slaps head! yeah you're right I got it the wrong way round, unfortunately I can't edit the post, hope it doesn't confuse too many folks.

I have been thinking about PT2399 delays a lot lately. A couple questions keep coming up.

First, the PT2399 chip runs off 5V so the signal fed into it should be kept in a range of 0 to 5V. The PT 80 delay uses a NE570 to compand the signal. But the Rebote and others usually use an op-amp with a gain of about 1 as the input buffer. My Rebote sounds pretty good. Is there a chance of clipping? Would it be better to use an input buffer with a gain of something like 1/2 and recover gain at the output buffer?

I think if you reduced the signal going in and then boosted it afterwards there would be a risk of increased noise. I've never heard anyone complaining about distorting a PT2399 based circuit so perhaps it's not a problem. I guess if the signal was loud enough to cause distortion it would already be quite dirty.

Second, most of the examples I see use inverting stages for the input and output buffer. An inverting stage at the input is more prone to noise. This is due to the need for large input and feedback resistors. Larger resistors create more noise but are required to keep the input impedance high. Then again my Rebote sounds good. Maybe the metal film resistors produce less noise, or any noise they create adds "mojo" to the sound.

I wondered about that, because I've read about noise caused by high value resistors, but never really understood what it meant.
The reason for using an inverting input buffer in these sort of things is that the easiest way to mix the wet and dry signals is with an inverting opamp mixer. To keep the dry signal at the output in phase relative to the input you then need an inverting input buffer. I did use a non inverting buffer originally but I figured as the dry signal would always be going through the pedal it was best to keep it in phase, so I switched to an inverting input buffer.
I can't hear any noise caused by the pedal, and it doesn't appear to colour the sound when bypassed so I'm happy with it.
Hopefully someone with a bit more knowledge can explain the noise issue and whether it's something to be worried about. Or maybe there's a way to bootstrap an inverting stage or something so you can get a high impedance without big resistors?

Just a though, but have you considered adding the plug detector as a voltage input for the delay time??

Interesting, I think you could do this or an envelope follower. Basically if you get the voltages right you could feed whatever CV signal you wanted into the base of the PNP transistor to control the delay time. I've thought about adding an expression pedal/CV input but I don't think I've got room in the current box.

I'll try and do some soundclips of the more subtle modulation effects, I think I went into "Demo mode" a bit with the original ones. Hopefully the schematic should be ready tomorrow.

[Ben N]

Slaps head!

Which is exactly what I just did when I realized the schematic was linked in your first post. Looks fantastic, and the samples are awesome.

Is there a way, do you think to do a momentary-on switch using the 4066 (i.e. effect on only when holding the switch down), or something like the SIB Echo 'Slam" switch, with this circuit?

Ben

[slacker]

The bypass switch is just a switch to ground, when it's open the effect is bypassed, when it's closed it's on, got the terminology right that time If you used a momentary press to make switch instead of a latching one then it would stay on whilst you held the switch down. You could even wire a momentary in parallel with a latching one to get both options.

I think you could emulate the Slam switch, I had a look at the SIB website and it basically selects the shortest delay time and increases the feedback to maximum.
If you short out the delay time pot that gives you shortest delay time, then to increase the feedback you could short out the 20k resistor. That should work, I'll try it and let you know.

[Ucho]

Great work slaker, thanks for sharing it.

For those "true-bypass-or-die" guys, since the tail will only last a few seconds (or less), i think you could add a classic 3pdt switch to bypass all the circuit.
If you place the two switches close enough you can play with the delay on (both switches 'on') then turn off the original switch to have the tail, when the tail ends you can push both switches so the circuit is fully bypassed with the original bypass set 'on' so next time you just have to switch the 3pdt (or you can 'turn off' the 3pdt and later turn on buth switches togheter).
I don't know if it's clear, and maybe is overkill, and not worth the extra work.

Also, now I'm thinking that, since delays are usually placed at the end of chain (o near it) the "always on" buffer could be useful for driving the signal through the cable to the amp.

Another idea to make use of hte tail. If it would be possibile to have infinite repeats (as in Rebote Delay, IIRC) it would be possible to have something like a pedal note, or a chord, or a very small riff, playing "forever" while you play over it...

Just some thoughts.

BTW, again, great project slacker, and thanks for it.
Actually Im completing my PT80 (I have the box already drilled and painted, and the PCB ready), but I'm sure I'll need an other delay in the future... 

[caress]

i noticed that you don't have the same LPF'ing as in the rebote and i don't see any caps to ground on the echo base.  the soundclips are pretty similar, though...how does your LPF'ing differ?
also, is it possible to splice in a send/return earlier in the circuit?  i've heard people complain about effects loops on delays because there is always one clean delay, then the effected delays.  is this a problem, here?  i was thinking about trying to put one in after the 1uf but before pin16 of the rebote2.5, so i guess a similar spot would be before pin16 on the echo base? 

[tommy.genes]

Brian:

The trick is to place the external loop so that it affects the signal going to both the LEVEL and FEEDBACK knobs. That way you will get the effect on the very first repeat, with increasing effect on each subsequent repeat. If you place the send/return so that it only affects the FEEDBACK knob, the first repeat will be clean and the effect will appear on the 2nd repeat and build after that. Placing the send/receive so that it only affects the LEVEL knob will provide effect on the first repeat as well as the subsequent repeats, but the effect will not "build" over time due to multiple passes through the loop. I think all three options could be used musically, and it might even be desirable to have a switch that selects from the three modes.

-- T. G. --

[slacker]

i noticed that you don't have the same LPF'ing as in the rebote and i don't see any caps to ground on the echo base.  the soundclips are pretty similar, though...how does your LPF'ing differ?

The first lowpass filter is the 20k resistor and 4n7 cap to ground in the top right of the schematic, and the components attached to pins 15 and 16 of the PT2399. The second is the components attached to pins 11 and 12 of the PT2399. These are basically the same as on the Rebote2.5 except that has an extra RC lowpass filter in front of those stages. It looks different because the Rebote2.5 schematic shows those pins as individual  opamps which makes more sense to look at.

also, is it possible to splice in a send/return earlier in the circuit?  i've heard people complain about effects loops on delays because there is always one clean delay, then the effected delays.  is this a problem, here?  i was thinking about trying to put one in after the 1uf but before pin16 of the rebote2.5, so i guess a similar spot would be before pin16 on the echo base? 

Like T.G said if you put the loop before pin 16 then everything will get processed by what you put in the loop, but you won't get the layering effect where each repeat goes back through the effect.
On the Rebote2.5 if you splice the FXloop in between the 1uF cap off pin 14 and the top of the repeats/level pots, then the FXloop is in the feedback loop so you get the layering effect, and the first repeat will be processed.

[swt]

excellent work slacker!! as i'm the only one who built the slackfilter...will have to give this a try  . Maybe i'll make a pcb for this...so any mods...are still on time. If you have a perfboard design, that i can take a look at, that would be great to have an initial kick, and maybe convert it to pcb. Thanks for sharing!!

[Solidhex]

Yo

  That thing sounds great!

--Brad

[slacker]

excellent work slacker!! as i'm the only one who built the slackfilter...will have to give this a try  .

Yours is probably the only one in existence then, mines in bits 

Maybe i'll make a pcb for this...so any mods...are still on time. If you have a perfboard design, that i can take a look at, that would be great to have an initial kick, and maybe convert it to pcb. Thanks for sharing!!

If you or anyone else could do a PCB layout that would be great, I don't have any experience in that department.

[Skreddy]

Sounds really good!  You did a great job setting all your LPFs to get that amount of delay time without noise!

I've been playing around with the same basic idea in my spare time (my resistor and cap values aren't all the same as yours but not surprisingly similar given the common origins).  For the modulation, I decided to go with a very primitive 1-transistor LFO and use an LED/LDR just to keep the parts count low, and intentionally bias it into an asymmetrical waveform to give it a clunky, mechanical sound like a tape echo dragging.  I'm trying really hard to get the "voicing" as close to my old tube Echoplex as possible.

I was actually thinking about using diodes to ground too, mainly just to prevent the signal from getting too loud when in self-oscillation mode.  I take it yours doesn't distort from those diodes but just clips in case of runaway echoes, to keep the signal from going too high?  The PT2399 really sounds awful when allowed to overdrive itself.  Here's a clip of mine being played with by a friend... http://skreddypedals.com/circuits/delay/elctmst_echo.mp3  Apparently self oscillation is the main thing folks are interested in when they try out an echo.

[Gila_Crisis]

wonderfull!!! is it capable of infinite delay/auto-oscillation???

btw i just wait for the vero layout!

greatings

[caress]

i noticed that you don't have the same LPF'ing as in the rebote and i don't see any caps to ground on the echo base.  the soundclips are pretty similar, though...how does your LPF'ing differ?

The first lowpass filter is the 20k resistor and 4n7 cap to ground in the top right of the schematic, and the components attached to pins 15 and 16 of the PT2399. The second is the components attached to pins 11 and 12 of the PT2399. These are basically the same as on the Rebote2.5 except that has an extra RC lowpass filter in front of those stages. It looks different because the Rebote2.5 schematic shows those pins as individual  opamps which makes more sense to look at.

also, is it possible to splice in a send/return earlier in the circuit?  i've heard people complain about effects loops on delays because there is always one clean delay, then the effected delays.  is this a problem, here?  i was thinking about trying to put one in after the 1uf but before pin16 of the rebote2.5, so i guess a similar spot would be before pin16 on the echo base? 

Like T.G said if you put the loop before pin 16 then everything will get processed by what you put in the loop, but you won't get the layering effect where each repeat goes back through the effect.
On the Rebote2.5 if you splice the FXloop in between the 1uF cap off pin 14 and the top of the repeats/level pots, then the FXloop is in the feedback loop so you get the layering effect, and the first repeat will be processed.

ahhh yes.  thanks!

[slacker]

As promised, here's the vero layout, I haven't posted it directly because it's pretty big.
http://www.aronnelson.com/gallery/v/slackers-stuff/echobasefinal.jpg.html?g2_imageViewsIndex=0

Sounds really good!  You did a great job setting all your LPFs to get that amount of delay time without noise!

Cheers, I just started with the datasheet values and then increased the size of the caps until I was happy. The filter before the feedback pot seems to make the most difference as far as controlling the background noise goes.

I was actually thinking about using diodes to ground too, mainly just to prevent the signal from getting too loud when in self-oscillation mode.  I take it yours doesn't distort from those diodes but just clips in case of runaway echoes, to keep the signal from going too high? 

Yeah the diodes only clip the signal when it's self oscillating just to stop it getting too loud. Means you can play around making weird noises without going deaf. Nice soundclip by the way.

wonderfull!!! is it capable of infinite delay/auto-oscillation???

Yes it can do self oscillation. You can sort of get infinite repeats but because of the the filtering the sound eventually becomes muddy and turns to mush. You can't do a DD3 style hold that keeps going forever.

[boyersdad]

This is so cool! I wish I didn't just build a PT-80! Time to order another 2399!

[Branimir]

Excellent sounding pedal!

I've built the rebote delay, but it always sounded kinda flat to me since it just repeats, there's no lush echo like in memory man for instance, but this one sounds it could do it.

Chorus sample is wow! Do you have any other samples? Can it delay with lots of echo going on? Good idea for delay tails!!