This Week on the Breadboard: Stereo Chorus

Chuck D. Bones

Circuit Wizard
This is my 2nd entry in Buddy's Circuit Design Contest. It's a stereo chorus based on a pair of PT2399s. The idea is a spin-off from the Stereo Digital Reverb project. You can think of it as a stereo version of the Sea Machine, minus three knobs, with a better LFO and all of the filters corrected.

Sheet 1 shows the two delay stages. They are identical except for the resistors that set the delay time. I wanted their delays and sweeps to be slightly different for a deeper stereo image. The delays on both are in the 44ms range. Kinda long for a chorus, but it works. There is some feedback around each delay for a more lavish tone. The delay times are modulated by two LFO signals.

Sheet 2 shows the buffers, mixers, LFO and power supply. U4B is the input buffer. It supplies the DRY signal to the two delays and the mixers. U4A & U4D are the output buffers. VR1A & VR1B form a dual mixer. At 7:00, the signal is 100% dry, at noon it's 50/50 and at 5:00 it's 100% wet for a pure vibrato effect. The PT2399s and the LFO all run on +5V. U6 generates the LFO waveform. Switch S2 selects between two waveforms: smooth triangle and random ramps. The smooth triangle is only available in the CB mod of the STOMPLFO. You can also run a stock STOMPLFO chip and change R60 to 12K so you get the regular triangle waveform, or 7.5K to get the sine waveform. Apart from the random ramps, the other waveforms are not too useful in a chorus. U5 buffers and inverts the LFO signal so that the two delays are driven by LFO signals that are mirror images of each other. In other words, while one delay is sweeping up, the other delay is sweeping down. R53-R55 & C44 scale and filter the LFO signal. The cutoff freq was set low to limit how fast the random ramps can sweep. The SPEED & DEPTH controls have more than enough range. One could hook up the TAP pin on U6 to a momentary stompswitch.

Stereo Chorus v0.5 sheet 1.png

Stereo Chorus v0.5 sheet 2.png

Stereo Chorus v0.5 breadboard 02.jpg
 
Last edited:
Love me something stereo chorus just barely blended in at the end of a chain. Ala Gilmour with his use of the Yamaha RA-200s. Pretty sure he even blended CE-2s end of chain but can’t remember if went into the Yamahas with them.
 
@Chuck D. Bones

Whatever became of this circuit? Anyone build one? Seems like it could do a descent faux Leslie with the different speeds.

Could one simulate a Leslie by have one of the lines, higher having a high pass filter set to fast and the second line set with a low pass filter set to slow?

I’m thinking something similar to what Gilmour used. Doppolas or Yamaha RA200
 
The downside to using PT2399 delay for flanger or chorus is the min delay is a bit long.
FV-1 is a better solution. BBDs would also work.
What is the shortest delay time you’ve been reliably seen out of a PT2399? I know too low and they overheat or lockup.
 
I was thinking the PT2399 wasn’t quite quick enough for chorus or flanger. But saw that you did this project. Does almost have a pseudo slapback sound to the chorus? Wonder if a clean blend would help with that any. I wouldn’t want to be running something like this 100% wet.

The BBE Mind Bender is a great sounding analog chorus using bbds. But even though it manages to get a descent Leslie sound there too much of a boost when engaged as well as that pseudo slap back thing. Sounds like a mix between small clone and electric mistress to me but way more prominent when engaged.
 
If the wet side is loud enough to hear, then you are gonna hear the slap-back. No way around it if you use a PT2399 delay chip for chorus. My stereo chorus circuit above is an experiment. I don't plan on building any.

Have you tried the Tri-Vibe?
 
Sorry if I’m beating a dead horse… but what are the sonic differences between the the tri-vibe and the ol Ez Vibe by Joe Gagan I believe it is?
 
The Easy-Vibe is by John Hollis and is a UniVibe clone using opamps and LEDs in place of transistors and an incandescent lamp. Moen copied it for the Shaky Jimi. I have one, it works pretty well. Like the UniVibe, the Easy-Vibe is a 4-stage, stagger-tuned phase shifter. It makes two notches. LEDs respond faster than an incandescent lamp, so the Easy-Vibe's sweep is different from the UniVibe.

The ROG Tri-Vibe is a 2-stage, stagger-tuned phase shifter. It uses OTAs instead of LDRs & LEDs. It was primarily designed to make vibrato, but also has the "chorus" mode of the UniVibe. It makes one notch.

The two actually sound pretty similar.
 
The Easy-Vibe is by John Hollis and is a UniVibe clone using opamps and LEDs in place of transistors and an incandescent lamp. Moen copied it for the Shaky Jimi. I have one, it works pretty well. Like the UniVibe, the Easy-Vibe is a 4-stage, stagger-tuned phase shifter. It makes two notches. LEDs respond faster than an incandescent lamp, so the Easy-Vibe's sweep is different from the UniVibe.

The ROG Tri-Vibe is a 2-stage, stagger-tuned phase shifter. It uses OTAs instead of LDRs & LEDs. It was primarily designed to make vibrato, but also has the "chorus" mode of the UniVibe. It makes one notch.

The two actually sound pretty similar.
I also have a EZ Vibe. As for the seasick wobble we think of with a Vibe, it’s close, but the ElectroVibe is much closer.
What I did find the EZ vibe does better than some truer Vibes, is turn off the dry signal (vibrato mode), set for med depth and fast speed does one of the better “Leslie on high” simulations compared to some other vibes
 
@ChuckBackTooth Not sure why I was thinking Gagan did the ez vibe. I may end up getting one of the Moens eventually. Pretty hard to beat for the price. I’ve owned expensive vibes but can never justify keeping it for such a high price.

Never really cared for vibrato lol…it makes me disoriented or something. I’m referring to deep vibrato though.

I just got a good deal on one of the Alabs NovaDrift pedals to try the rotary. I did have the Flamma yellow mod pedals and the rotary was suprisingly good on it but it didn’t have analog dry through. The Alab has analog dry thru, true stereo, tap tempo or parameter “tweak” when using the switch as a momentary switch. I haven’t had enough time to give and honest review yet but so far the rotary’s pretty good. Keep in mind I mainly use Rotary end of chain and blended in ala Gilmour.

Haven’t tried any of the other effects. Has univibe, chorus, flanger, tri chorus, phase, rotary, tremolo, envelope filter and ring mod for some reason.
 
I found a better way to accomplish a PT2399 chorus. Barry over at Guitar PCB has a PT2399 chorus which reminded me that the PT2399's clock can be sped up enough to get the delay down to around 25ms, perfect for a chorus. It requires a bit of trickery that I had seen elsewhere but not explored. Until now.

The Ibanez BC9 is a stereo chorus with a separate LFO for each BBD delay channel.

I combined the two ideas and made a 2-channel PT2399 chorus, each with its own LFO. The breadboard contains both channels. Other than a common BLEND pot, common WAVEFORM pot and power supply, the two channels are completely independent.

Knobs (L-R): LEFT SPEED - LEFT DEPTH - RIGHT SPEED - RIGHT DEPTH - WAVEFORM - BLEND
Stereo Chorus v0.9 breadboard 02.jpg

This schematic shows a single channel. For a stereo version, duplicate everything except the power supply and the start delay (Q1, R13-R15, C35 & D3).
At the moment, the WAVEFORM select is a pot rather than a switch. The useful waveforms are Triangle, Rounded Triangle*, Sine and Random Ramps. The 5V supply rail (Vdd) is bumped up to 5.6V by D2. This facilitates higher clock frequencies. The PT2399 will not work properly if the VCO resistor (R12) is less that 1K when power is first applied. Q1 and surrounding circuitry provide a start delay that presents a 2.4K resistance for the first several hundred milliseconds. After that, Q1 turns on and kicks the clock freq up to around 25MHz. With a VCO resistor that small, modulating the clock via the VCO pin does not work well. Instead, we modulate the clock via the REF pin (pin 2). I was skeptical at first, but it works. R10, R11, C2 & C3 filter the LFO signal feeding the REF pin. R5 & C9 provide a little feedback around the delay for some increased depth. The STOMPLFO chip (U3) from Electric Druid provides the LFO waveforms.

Running the two channels at different LFO freq and depth provides a glorious 3-D swirling effect to the chorus tone. The PT2399 has more bandwidth and at these short delays, less noise than a BBD. The parts count is significantly lower as well. The PT2399s must be hand selected because not all of them work well with such a small VCO resistor. I tried four of them and 2 worked well. It is easy to tell by monitoring the DC voltage on the VCO pin. Before Q1 turns on, the DC voltage is around 2.8V (1/2 of the 5.6V rail). After Q1 turns on, the voltage drops to about 1.4V on the "good" ones, and below 1V on the "not good" ones.


Single Channel Chorus v0.1 sheet 1.png

Single Channel Chorus v0.1 sheet 2.png

* The Rounded Triangle waveform is only available on the CB mod STOMPLFO.
 
Last edited:
Thanks for sharing this one, Chuck!
Everytime I've seen people use a transistor connected to the VCO in order to have small resistances work, it's been a BJT.
Any particular reason you've gone for a MOSFET?
 
Three, actually,
1. Lower Vsat.
2. Higher input impedance.
3. Higher turn-on threshold.

In the case of a MOSFET, Vsat = Rds,on * Id.

The only advantage a BJT has is cost, and we're talking pennies here.
 
Back
Top