PT2399 defeated by.. a few LEDs?

[HeavyFog]

So i set out to build a stereo delay a while back for my modular synth based around 2 rebote 2.5 PCBs i got from tonepad. Basically just 2 of these in parallel, with relay bypass switching, a mode for ping pong, and an added VU meter, all powered from +15vDC from the synth's psu. I also made the delay time voltage controlled so both delays could be properly controlled with one pot (i used the control circuit from the rebote 3). Got it wired up and it worked great for a while, until last night.

The relays switch on with +5vDC, which they were getting from the regulated +5v line on one of the delays. This was causing some minor issues when bypassing so I instead switched it to a separate regulated +5v line i took from the +15v rail. The problem is that now as soon as i power it up the delays put out this loud hum. I've noticed this before, but it would only happen if delay time got too short and also on powerup, and i could get rid of it by not having the delay time too short and switching it to bypassed and back, which eliminated the hum for some reason, but now its always there as long as the pedal isn't bypassed. After some digging i found that if i removed the bank of 5 warm white LEDs i'm using to backlight the VU meter it works fine without any hum, but as soon as i put the LEDs back in the hum starts immediately. I can also see that the LEDs are faintly flashing very quickly, kinda like incandescent bulbs, but only when the pedal is on (not bypassed). If i remove the LEDs and cycle the bypass it stops and works as normal!

I spent a few hours experimenting with it and cutting out anything unnecessary in the circuitry, but so far nothing seems to work. I know my wiring is right and i didn't fry anything since it works perfectly fine without the LEDs, and my power supply has more than enough current to supply everything. I even used higher rated electro caps throughout, so i cant seem to figure out why it works fine without, but not with the LEDs. Whatever whitchcraft is making it hum when i add a few LEDs in is kinda beyond me

[HeavyFog]

Also the actual delay circuit is more or less stock with the exception of the voltage controlled delay time (which also seems to work perfectly fine). I suspect that the relay bypassing might be involved, since the problem is definitely worse when the relays are powered, so it may be a power issue? Either way the only drop in voltage i measured was a drop of 0.1v on the main +15v power line when the LEDs are put in. I even tried powering the LEDs form the synths -15vDC line but the exact same thing happened It seems like if i can solve the issue of the hum appearing at power up (happens with and without the LEDs in) then this problem will resolve more easily, but so far no luck.

Schematic: http://www.tonepad.com/getFileInfo.asp?id=98

[antonis]

Hum is induced due to not enough ripple rejection..

The above might be caused either by low unregulated voltage or by heavy loading (usually by both of them)..
For full-wave rectification, a good rule of thumb is 4700μF for each Ampere of current draw for a ripple of  1 - 2 V_p-p..
The fact that hum decreases with delay time increased indicates high ripple due to current draw (Max PT2399 current draw is 30mA)..
Tonepad schematic calls for 47μF cap for 2V_p affordable ripple (9V - 5V - 2V_IN-OUT drop) which, IMHO, is marginally OK for a single circuit..

Could you plz post a full schematic of your power supply..??
(including relays, LEDs, etc..)

P.S.
Relays don't need a regulated supply 'cause they can work in a rather wide variety of voltage around their nominal rating..

[merlinb]

Sounds you are demanding too much current from the 5V regulator causing it to drop out, allowing hum from the raw supply to get in. Let the 5V run the PT23999s alone, and power all your non-critical stuff (LEDs / relays) from the main power source with resistors to drop the voltage where necessary, instead.

[duck_arse]

what are these warm white leds, anything special? and how are you driving the voltage that ".... also made the delay time voltage controlled ....."?

[ElectricDruid]

what are these warm white leds, anything special? and how are you driving the voltage that ".... also made the delay time voltage controlled ....."?

Yes, I'm curious about the voltage-controlled delay time too.

[HeavyFog]

Sounds you are demanding too much current from the 5V regulator causing it to drop out, allowing hum from the raw supply to get in. Let the 5V run the PT23999s alone, and power all your non-critical stuff (LEDs / relays) from the main power source with resistors to drop the voltage where necessary, instead.

Right now the PT2399s are the only thing power from their own regulators. I gave the relays their own 5v regulated line, which is also used as a source for the voltage controlled delay time (was previously tapping off one of the 5v lines from the delay board). Everything receives power from the +15v line, and from there is regulated on the delay boards for each pt2399 (as per the schematic) and separately on the relay power board.

For the voltage control, i just used the control circuit from the rebote 3 down to the component (http://www.tonepad.com/getFileInfo.asp?id=151). On the 3 it feeds 3 PT2399s and in my case 2. It works just fine with the exception of the humming when delay time get too fast.

I should also note that the hum is only in the wet path, and goes away as i turn down the level knob

[HeavyFog]

what are these warm white leds, anything special? and how are you driving the voltage that ".... also made the delay time voltage controlled ....."?

Here's the mouser page for the specific LEDs that i used. Come to think of it i haven't tried it with other LEDs yet. Maybe il try out a few generic reds to see if it acts up.

https://www.mouser.ca/ProductDetail/941-C513AMSSCW0Z0131

[duck_arse]

I can also see that the LEDs are faintly flashing very quickly, kinda like incandescent bulbs, but only when the pedal is on (not bypassed).

the rebote modulates those delay time transistors, are you also? if not, you might try a cap to ground across C31, like that C35. how fast, how deep is the flashing?

[anotherjim]

I wonder if the start up hum is all the current the caps on the 5v regs take to charge up robbing the main +15v caps of charge.
Once the 5v is happy, the +15v can fully charge and the hum stops.
And I wouldn't worry about LED type without first specifying the series resistor used with them - which I don't see mentioned.

TBH, the thought of dropping from +15v for a bunch of 5v loads would stop me. Too much power to waste in the drop and un-balancing the power transformer.
I'd prefer to add a 9v transformer in it that's running off the same mains supply. Can power 9v circuits and supply 5v regs happily without loading/unbalancing the existing synth supply.

[HeavyFog]

I can also see that the LEDs are faintly flashing very quickly, kinda like incandescent bulbs, but only when the pedal is on (not bypassed).

the rebote modulates those delay time transistors, are you also? if not, you might try a cap to ground across C31, like that C35. how fast, how deep is the flashing?

In my case the delay time isn't modulated since the core circuit is a rebote 2.5, only the 3 i believe has that. The modulation is pretty fast, and not very deep, but is just noticeable. Might be flashing at the same frequency as the hum, but i can't confirm this

[HeavyFog]

I wonder if the start up hum is all the current the caps on the 5v regs take to charge up robbing the main +15v caps of charge.
Once the 5v is happy, the +15v can fully charge and the hum stops.
And I wouldn't worry about LED type without first specifying the series resistor used with them - which I don't see mentioned.

TBH, the thought of dropping from +15v for a bunch of 5v loads would stop me. Too much power to waste in the drop and un-balancing the power transformer.
I'd prefer to add a 9v transformer in it that's running off the same mains supply. Can power 9v circuits and supply 5v regs happily without loading/unbalancing the existing synth supply.

That sounds like a possibility. Only thing is that the hum is constant on power up until i cycle the bypass, basically momentarily cutting power to the relays and turning it back on, which has always been an issue before i made the earlier change when the problem with the LEDs started happening.

Should also mention that before power goes to anything it goes through a separate 10ohm/22uf filter with a 0.1uf cap across the 15v/-15v lines. I did this since it's standard in most of the circuits in my modular. The same power filtering is seen in this schematic https://www.yusynth.net/Modular/Commun/MOOGVCF/Moogfilter-sch.jpg. The 5v live is also filtered like this post regulator. Starting to think maybe all that filtering could be contributing somehow to the problem

I did think about giving it it's own psu, but i didn't want to complicate an already complicated and wiring heavy circuit, plus i wanted to keep mains power away from the noise sensitive moog 905 spring reverb mounted in the modular. Powering it from the modular's DC supply made the most sense to me.

[HeavyFog]

Ok, so something i forgot to mention was that if i turn the delay time too short it starts humming, but when you turn delay time up again the hum doesn't go away until you cycle the bypass. I had just read that the current draw of the PT2399 goes up when delay time is shortened, so what i think may be playing a part here is that the delay time control circuit I lifted from the rebote 3 is actually letting delay time get faster than its supposed to, increasing current draw, and stressing the 78L05s to produce the hum, like Merlin mentioned. The delay time is probably starting very fast at power up because of the heavy filtering because lower voltage on the time pot = shorter delay time, which would explain why the hum is always there at power up. I'm gonna try replacing the regulators with sturdier LM7805s and see if that makes any difference.

[PRR]

The regulators don't hum. They do need some excess voltage, then they can remove hum.

If ripple-dips fall to 6V, the 5V regulator can't "fill in" the dips.

Your RAW power supply needs more capacitance. Maybe show us what it is?

[anotherjim]

The RC filter you added in the 15v feed should be useful, rather than causing a problem.
If you measure the voltage dropped across the 10R, Ohms Law V/R tells you the current draw from the 15v.
If you measure the voltage to 0v after the 10R (should be the 15v minus the volt drop) you can get an idea of what is going on.

[HeavyFog]

Was toying with it a bit tonight and testing some other pt2399s to see if the chip was at fault. Turns out the hum isn't coming from the pt2399 at all since i stil get the hum with no chip in place at all! Is obviously a lot quieter but it's still there and i can turn it down with the level control. I'm gonna keep running some tests and see if i can narrow down the issue further

[HeavyFog]

Ok, I also just ran another experiment where I wired one of the relays so the delay is always in the signal path, regardless of if the relay is powered or not. This way I can test to see if powering the relays has any effect on the hum. When the relays are powered, i get the hum, but when I disconnect the power to them the hum stops! Seems like the source of all this has something to do with the relays when they're powered, because with them off the delays are nice and clean.

Right now the relays get their power directly from the regulated 5v source, which passes through a 10R/10uF filter first. The power goes directly to both coils (in parallel with each other) and the grounds are tied directly to each other. This single ground line goes to a switch which cuts the ground to bypass the delay (with the ground cut the hum stops). The current draw is only 28mA per relay, 56mA total since they're in parallel, so it's not like the 78l05 doing the regulating is that close to it's limit. Still not sure exactly what's going on here, but i'm getting closer

At the moment i don't have a flywheel diode in place across the relay coils, but I'm going to try adding one tomorrow.

Here's the mouser page for the relays I'm using https://www.mouser.ca/ProductDetail/769-TQ2H-5V

[anotherjim]

The current isn't the power. It's power that causes heating in a device. What is the voltage drop across the regulator? Power is that voltage drop x the current.
It starts at 15v, drops some across the 10R feed resistor. 56mA x 10R = 560mV. That means 15-0.56v = 14.44v into the regulator. The output of the regulator is 5v so 14.44-5v=9.44 across the regulator. Power is 9.44v x 0.056A = 0.528W.
The next bit I always have trouble with. There is no maximum power spec because what matters is how hot the chip in the package gets so it's so many degrees C per Watt and I can't work out what the figure to use is in the datasheet. But either way, it shouldn't reach the thermal shutdown point at about 0.5W. But its early and I haven't finished my first cup of coffee yet.

[antonis]

But its early and I haven't finished my first cup of coffee yet.

After 2^nd cup of coffee here, your calculations seem to me alright..

Now, for a 78L09 (TO-92) regulator of 200^oC/W junction to ambient thermal resistance, half a watt power dissipation could result into internal limitation triggering..  

[HeavyFog]

My bad, the proper term here would be that the relays are "energized". Wasn't referring to power as in W, but the math is valid and i probably should replace the regulator with a proper 7805 when the ones i ordered from mouser show up. The delay time control is also powered from the same regulator, although the draw is pretty small (around 0.2mA). It would probably be better off with it's own regulated supply or better yet taping off one of the 5v lines on the delay board because when the relays are energized i do notice that the delay time gets a bit shorter. Also the combination of the time control feeding off that output and the lack of a flywheel diode on the relays could be playing a role here. Better safe than sorry.

Going to take some more measurements though of the relays since something is definitely up here