Mix Control for LA Chorus

[anotherjim]

You should be careful with that 10uF LFO coupling cap. A Capacitor in series can generate higher voltages than the supply at the pin it feeds.
Here's how...
The chip end is +2.5v the feed is at 0v. The cap charges to 2.5V with + at the chip pin.
The feed end swings up to +9v, the chip end voltage lifts up to 9+2.5V. The chip supply is 5v. If it survives, the chip pin asserts it's 2.5v and according to resistor values the cap discharges/recharges to 9-2.5V with + at the feed end. In principle, this is how a voltage boost charge-pump works.
In practice, the LFO can't swing 0 to 9v because the opamp used isn't rail-to-rail so can't. Also, the oscillator has a Schmitt trigger action (the 220k network to the +input) which positions the swing to a smaller range centred about 1/2 supply volts.
...and that's how to kill or at least lock-up a PT2399. The circuit avoids the bad condition when everything is ok, but make a mistake or mess about on the breadboard with power on, and you can accidentally duplicate the worst condition -  but the super-paranoid clamping diode will prevent the overvoltage.

As to why Ben's builds don't sound right, I still have
No
Eye
Deer.

[Ben Lyman]

You should be careful with that 10uF LFO coupling cap. A Capacitor in series can generate higher voltages than the supply at the pin it feeds.
Here's how...
The chip end is +2.5v the feed is at 0v. The cap charges to 2.5V with + at the chip pin.
The feed end swings up to +9v, the chip end voltage lifts up to 9+2.5V. The chip supply is 5v. If it survives, the chip pin asserts it's 2.5v and according to resistor values the cap discharges/recharges to 9-2.5V with + at the feed end. In principle, this is how a voltage boost charge-pump works.
In practice, the LFO can't swing 0 to 9v because the opamp used isn't rail-to-rail so can't. Also, the oscillator has a Schmitt trigger action (the 220k network to the +input) which positions the swing to a smaller range centred about 1/2 supply volts.
...and that's how to kill or at least lock-up a PT2399. The circuit avoids the bad condition when everything is ok, but make a mistake or mess about on the breadboard with power on, and you can accidentally duplicate the worst condition -  but the super-paranoid clamping diode will prevent the overvoltage.

As to why Ben's builds don't sound right, I still have
No
Eye
Deer.

Wow! Ok, so the Little Angel Chorus is not an example of perfect engineering and it is a circuit that is living on the edge, at best. Perhaps we can improve it and make it the finest affordable chorus to build. The PT2399 is still affordable, right?
I need to really examine the bread layout and make sure my schematic is right, but I'm fairly certain it is.
Difference in bread vs. build sound? I can only guess it's from different choice of capacitor types, although the values are the same.

[Ben Lyman]

FOUND IT!!!

Sloppy breadboarding work, I forgot to include that 10nF cap between pins 13 and 14, but I like the sound better this way.

My schematic, layout, and both final builds had this cap in there because it's supposed to be there, but it made the builds sound muddy.

So my concern is this: Does that 10nF do anything besides filtering? Will I be safe to omit it from my build?

Thanks!

[anotherjim]

Yes, it is just a low-pass filter formed by the 10nF. In theory, it's removing switching noise from the digital-analogue conversion process but in this circuit, the digital parts are worked at top speed and the noise frequencies are well above our guitar amp speakers range.

You can take another look at those filters on pins16,15 and 13,14. At least 2 resistors can be taken out.
Maybe it's because the schematics make the PT2399 look like a box with special functions on the pins or because the chip datasheet examples show certain components, but behind these pins are only opamps with the +input pins internally connected to the pin2 reference. These opamps can have any use, but pin15 is an amp output that is internally connected to the chips analogue-digital converter. Pins 16,14,13 have no other connections.

This is the datasheet usage of the opamps. Multiple feedback filters. You can calculate them here...
http://sim.okawa-denshi.jp/en/OPtazyuLowkeisan.htm
In many PT2399 designs out there, these filters are used, but to save parts, C1 is deleted. It still filters, but less deeply.
However, if you only need a simple filter, then...

A tool to work it out, and many other configs here...
http://sim.okawa-denshi.jp/en/opampkeisan.htm
But, in the simple filter case, it's no different than for a single RC filter. All I do is enter the C and Rf values in this tool...
http://sim.okawa-denshi.jp/en/CRlowkeisan.htm

If you recall the Low-cost chorus, the pin13,14 opamp is used as LFO, so the output is taken directly from the output converter pin12 and, if that works, you could leave this opamp out, and simply wire pins 13,14 together to stop it causing random noises.

[Ben Lyman]

Thanks Jim! Big help, I have made some changes to the bread and it is sounding really good.

Is this what you meant about taking the out from pin 12 and eliminating some resistors?

I changed the pins 11/12 filter cap to 22n to brighten it up more, it sounds great right now.

reversed the 10uF cap and added a diode to catch the over voltage. Did I do that correctly?

You'll also notice a 5k trim pot to set the delay time, the noise level in the build is nothing like the bread so I thought I would try to achieve minimum time. I will probably go back to the anti-latch circuit for best results.

[anotherjim]

Thanks Jim! Big help, I have made some changes to the bread and it is sounding really good.

Is this what you meant about taking the out from pin 12 and eliminating some resistors?

I changed the pins 11/12 filter cap to 22n to brighten it up more, it sounds great right now.

reversed the 10uF cap and added a diode to catch the over voltage. Did I do that correctly?

You'll also notice a 5k trim pot to set the delay time, the noise level in the build is nothing like the bread so I thought I would try to achieve minimum time. I will probably go back to the anti-latch circuit for best results.

Yep, that all as I was suggesting.
The pin6 resistor is a problem. Just fitting a low-value R there can lead to a situation where it works for a bit then locks up and needs power-off/on to get it going again - however, someone here said they never had a lockup with a fixed 510R on pin 6 and if it sounds ok at that, it might be a better solution.

[Ben Lyman]

Cool. I've been playing it for about an hour now and I really like it. I'm on the fence about the pin 11/12 filter cap, going back and forth between 22n, 47n, and 100n. Probably next build will have a socket to start with.

So, a reverse 1n4148 diode is sufficient on pin 2 to +5v, and keep the 10uF oriented like this?

About pin 6, I got it to latch up with trimmer wide open, then turned it a tiny bit, probably is somewhere between 500r to 1k, no latching yet. I think I might go back to the anti latch fix (if it doesn't add any noise)

Here's that anti latch circuit from the low cost chorus, with an LED. and another one from Electrosmash without the LED. Either look ok to use?

[Ben Lyman]

Going back through this thread, I've tried to gather all the info and include it in a new schematic. I have it like this on the bread and it sounds great.

Does it look like everything is protected correctly by all those diodes?
Did I forget something?
Anything that is redundant?

[anotherjim]

Looks good to me Ben.
If you want to post some sounds, perhaps start a new thread for it? The current title makes it seem like an average modding discussion when you've actually gone far enough to make it a project of its own.