Tonepad CE-2 Chorus: quick build report and a question

[Bucksears]

I'm very impressed. I was a little hesistant about trying this one again, due to problems I had the first time, but this was worth it. After reading all of the debugging threads, I realized some of the pitfalls of my first attempt.
I put the input/output jacks near the bottom of the case, put the DC jack at the top, used a very short wire from the input jack to the footswitch and used a shielded wire from the output jack to the footswitch.
No noise. Period. I'm shocked because the first one was all noise I'm only getting that pop from turning the effect on/off, but that will eventually go away.

Back to back with my Tonepad Small Clone build, I can definitely hear a difference. The Small Clone seems more prominent, but also more 'linear' (or 'focused') where the CE-2 is more 'watery'. I'm going to socket that 47pF (intensity) cap so I can try others in there and settle on one I like.

So here's my question: is there a cap in the circuit somewhere that affects the brightness/tone of the chorus? I'd like to bump up the treble just a little if at all possible. If not, that's ok.

[BAARON]

Removing C16 might make it brighter, and making C15 might also make it brighter.

Changing C22 has a big effect on the sound.  I set mine up with an on/off/on toggle switch to go between 33p, 33p+22p, and 33p+68p for three different sounds.  The switch is worth it: the 33p never gets deep enough for a "thick chorus" kind of sound, but it does mild flangey chorusing really well, while the 33+68p is so heavy that it's way too heavy when you turn it all the way up, but it sounds nice when you turn the intensity lower.

[Bucksears]

Thanks BAARON - have you tried modifying either of these caps? (do I raise or lower C15? you didn't mention)

I've also read somewhere that using a TL072 for IC4 (instead of the 4558) helps with clarity; an NE5532 even more so, but pulls more current.

Thanks,
- Buck

[Mark Hammer]

I have a half dozen or so choruses, and almost always include a "range" switch for the C22 equivalent in each.  It is a simple thing to wire up a 3-position on-off-on toggle to get yourself the stock setting, a "thicker" setting, and a "lighter" setting.

Example:  C22=47pf.  47pf in series with 100pf produces an effective capacitance of 32pf.  So, take a SPDT switch and solder a 47pf cap between the centre lug and one of the outside lugs.  Now solder a 100pf cap between the centre lug and the other outside lug.  Now run wires from the two outside lugs to the pads on the board where C22 used to be.  Ta-da!!  You now have 3 modes.  When the switch is in the middle position, the delay range will be shifted down by about 1/3.  So if the stock unit swept between 3msec and 12msec, the new range would be between 2 and 8msec for a different swirl.  Kinda like the shift from ice cream to gelato.  Flick the switch in one direction to shunt the 47pf cap, and you'll have a much thicker doubling that moves more into the Pat Metheny end of the spectrum (assuming our hypothetical example, it would shift from 3-12msec to 6-24msec.  Flick the switch in the opposite direction so that the 100pf cap is shunted and you get the normal stock range.

Caveats: The wires leading to the switch should be short and kept away from the audio path or else you can experience clock whine.  The longer delay setting might risk a bit of audible whine when used in tandem with treble-enhancing changes to the circuit.

C16 and C15 are there to help manage clock noise.  You really don't want to  monkey with them.

Although not the same thing as treble boosting, inserting a cap just ahead of R22 will attenuate the bass in the wet signal.  This has the effect of making the pitch wobble less noticeable by attenuating the fundamentals in the wet signal, while keeping the swirl of the harmonics.  A cap of .0068uf inserted in series with R22 will give a rolloff beginning at around 500hz, and a 4700pf cap gets you a rolloff around 720hz.  To defeat the bass cut, just wire up a SPST toggle with the cap soldered to the toggle lugs, and leads from each lug going to the spot on the board where it would be inserted.

[Bucksears]

Thanks everybody - I think I'll socket C22 and try some different values in there, find one I like.
There's nothing WRONG with the 47pF, just might find something I like better; I'll start with 100pF and go from there.

If I find I need a 'brighter' chorus, I might have to look at a different circuit. This one does its thing well and don't want to mess it up trying to make it do something outside of its own range.

[Mark Hammer]

C3/R5 and C15/R23 provide pre-emphasis and de-emphasis to goose the treble a bit on input, and take away from it on output.  They provide symmetrical filtering action, but because the wet path adds hiss, it ends up taking that hiss away.

If one was to change the values of C3/R5, such that there was more pre-emphasis than de-emphasis, you could conceivably push the top end a bit more and increase brightness, however you might pay a price for that in terms of clarity or hiss-management.  You could try it in any event.

One way is to add the same amount of gain for the treble but move the starting point downward by changing C3 to .01uf (treble boost begins at 1.6khz instead of 2.4khz).  Another way is to keep the same starting point but increase the amount of treble advantage.  You could do this by means of C3=8200pf and R5=8k2.

[BAARON]

Good call on C15 + C16.  I only glanced at the schematic, so it doesn't surprise me that I was wrong.

C22: Mark's suggestion for how to implement the switch would work well.  Personally, I used an on/off/on switch with three caps instead of two: it requires the use of an extra cap (and thus some extra space and a few extra pennies spent), but it means you don't have to be quite as clever with your math when you're choosing the caps you're going to use.  I soldered my 33pF cap in place on the board, and then used the on/off/on toggle to control whether or not there were other caps in parallel with it (because caps in parallel add instead of doing strange math), so with my 22pF on one side and the 68pF on the other, it effectively gave me 33pF, 55pF, and 101pF settings overall.


And now that I've belittled clever math, I'm going to do some.  Mark, a question about your suggestion for the wet bass cut cap:
Isn't R20 setting the bias on IC1b?  If so, wouldn't a cap in series with R22 prevent it from doing so?
Also, a 6n8 cap wouldn't begin rolling frequencies off at 500Hz - that's the -6dB cutoff point!  In other words, frequencies at 500Hz would be attenuated by 50% already, and the attenuation would actually start somewhere above 5kHz.  I find a capacitor reactance chart pretty handy on occasions such as these (http://analogguru.an.ohost.de/techstuff/Reactance.htm for a quick reference, or http://www.electronics2000.co.uk/calc/reactance-calculator.php for a big beefy calculator).
R22 is forming an inverting op-amp circuit with IC1b, so the gain is set by the impedance of R24 (and its buddies) divided by R22 (and its potential buddies).  Normally it's a gain of ~1, give or take a bit depending on frequency (thanks to C15 and C16).  If one were to put a 6n8 cap in series with R22, it'll add way too much impedance (check reactance chart/calculator) and most of the chorused signal will vanish entirely.  (Plus, it potentially interferes with R20's bias-setting job?)

Instead, a value like 33n would probably work better.  It would provide an impedance of 60k at the bottom of the guitar's range (which, in conjunction with the R24/R22 inverting op-amp gain calculation, would cut the wet signal by about 60% on a low E), with a mild 3.8k at the highest fretted note on the guitar (a 7% wet signal gain reduction, hardly noticeable).  I have a pot in series with R22 right now as an "effect level" pot, and I'd say that by the time the pot gets up to ~50k, the wet signal is completely obscured by the dry, so that's all the cut you'd need.
In order to allow R20 to keep doing its job, putting a big resistor in parallel with the cap would allow the bias to still be set, but without significantly affecting the impedance of the 33n cap - say, 270k?  (270k||60k = 49k, or 49% wet signal reduction at 80Hz, and 270k||3.8k = 3.7k, essentially no change at the top of the instrument's range.)  To go back to the stock sound, you'd just bypass the 33nF||270k combo with a toggle switch.  (The 270k wouldn't affect the bias any more than the "Effect Level Mod" found in the Tonepad CE-2 PDF.)

Does my logic seem reasonable?

[Mark Hammer]

Your logic is entirely reasonable, and that's a good catch with respect to R20.  IN which case the cap to reduce the value of is C14.  If I had the habit of keeping notes, I could probably tell you whether that's what I changed in my own modded CE-2, and what the value is.

Suffice to say you can monkey around with smaller values until you find what you like.

[BAARON]

I thought about changing C14, but R20 is so big that it would have to be about a 1n-2n2 coupling cap to bring the bass cut up into the proper range, and I wasn't sure I wanted to use something so small as a coupling cap.  I'll give it a shot with the 33n/270k first because that'll be easier to wire in (with the way by board is set up right now) and if I don't like the sound, I'll pull the whole thing out and try changing C14 to something teeny tiny.

[Mark Hammer]

There are plenty of other places in the wet signal path that can serve to control bass content: C5 and C10 are every bit as valid as C14.  Heck, for that matter, by cutting out bass content at C5, one could actually boost the input level feeding that path (since the bass accounts for much of the amplitude), and get a better S/N ratio.

[DimebuGG]

I recently comparing the schematics of CE-2 and CE-3 as I'm about to build another one adapting the board with MN3207/MN3201 chips. My main concern is just
this R14/R39 in CE-2 and CE-3 respectively which are both 56K.

In CE-2's schematic, this resistor was connected to the main 9V but in the layout(from the service manual) and from pictures from this site http://www.pedalarea.com/chorus_ensemble.htm , it is actually connected right after R52(33 ohms) that shares the same power line with the BBD chips. Does this makes any difference?

[MetalGuy]

Hi,

Recently I was asked to build a chorus and I decided to build one for myself as well. It was time for experiments. I decided to go for Tonepad's CE-2 and to try the CE-3 as well. Then I started to look at different choruses schematics and a they all looked similar that's why I did several experiments on mine to see the difference if any.
In CE-2 and CE-3 the BBD chip is biased by the transistor while in other units the chip has separate bias trimmer and the opamp bias is derived from a simple 10k/10k resistor voltage divider as in CE-5 and CH-1 for example. That's why finally I redrew the PCB to make it look more like CE-5/CH-1 but without the filter/tone circuit. This is how my CE-2 ended - like a CE-5. Finally I modified the CE-3 BBD bias circuit to look like CE-5.
I had trouble making the MN3207 to work but finally I realized that 56k was still connected to the (+) supply. Don't forget that compared to MN3007 MN3207 has the power supply reversed including the resistor in question!
I didn't have 8n2 caps that's why I used 4n7+3n3 in parallel.
Below are the redrawn PCBs for CE-3 and for CE-5/CH-1  chorus only circuit. For input FET I used BF245B all other transistors are BC550C. I couldn't find space for that 100k LFO resistor that's why I had to solder it on the bottom. The pictures below are at 300dpi. Enjoy!

[Dr.maligno]

Hey, good job!!, are this circuits confirmed or revised?thanks a lot.

[MetalGuy]

I have the CE-3 working but you can always check the layouts against the corresponding schematics.