MAX1044 / LT1054 noise

[DUCKFACE]

Hello
i made a simple charge pump with LT1054 but the result is kind of bad. There is a screeping sound on the output of the stompbox. very high freq sound.
1. is there a specific way to wire the ground? (star topology?)
2. should i have to lift the LT ground from the stombox ground with 10k resistor ?
3. should i have to filter the power supply .... again ?

here is the schematic i used

[Kevin Mitchell]

Pin 1 must be not connected with an LT1054 - I see a jumper in the schematic but don't know your configuration.
This is the first time I've seen "C4" as 100uF and not 10uF. Not sure if that could cause your problem?
More filtering may help but try the above first.
You're going to want a regulator if your goal is +15v.

Could there be errors in your layout?

[DUCKFACE]

there is a jumper for boosting the MAX1044. For LT1054 pin 1 is not used.
the result is the same with 100uf and 10uf. iv tryed to use less than 10uf and its getting worse.
im not gonna use a regulator. +15 is the lowest voltage im gonna use. Right now the voltage is +18

[Processaurus]

Layout matters, star grounding the charge pumps helps with mystery whines.

Does the app note circuit in figure 22 sound like what you want? I see they do a little different thing with the diodes on the Vout.

https://www.ti.com/lit/ds/symlink/lt1054.pdf

[Rob Strand]

You need a 100uF cap across the input supply located right at the chip.
D7 should be fed from that cap as well.


https://www.ti.com/lit/wp/slpy005/slpy005.pdf
page 8,
"Passive-component effects on the
charge-pump application
CIN
The input capacitor, CIN, acts as charge reservoir
that aids a quick transfer of charge from the supply
to the switching capacitor during the charge phase
of operation. The input capacitor helps keep the
input voltage from drooping at the start of the
charge phase when the switching capacitor is
connected to the input. It also filters noise on the
input pin, keeping this noise out of sensitive internal
analog circuitry biased off the input line. Input
capacitance has a dominant and first-order effect
on input-ripple magnitude. Increasing the input-
capacitance value causes a proportional decrease in
input-voltage ripple."

It also reduces the level of current pulses on the input supply.  These cause noise which gets into the analog circuits.  The ground is particularly problematic.

[Clint Eastwood]

You need a 100uF cap across the input supply located right at the chip.

10 uF might be enough, if space is an issue. Just try.

[booza]

I been having a high pitch sqeal also. fixed it by grounding properly. forgot didn't have one half of the bread board grounded. 

[DUCKFACE]

changed a little bit. added more filtering capacitors. the result is the same. Still highfreq noises. Really need help

[Kevin Mitchell]

Hey man, save some schottkies for the rest of us!
I'm not quite sure what's going on with your circuit. It looks like some strange stuff going on with the negative output. Like what's up with C6? Are all those diodes necessary?

[anotherjim]

It's the datasheet dual power circuit figure 22.
https://www.ti.com/lit/ds/symlink/lt1054.pdf
I still don't see any bypass capacitors tried on pin 8. Datasheet circuits usually assume that the input power supply filtering is adequate or non-critical. Look at the layout example in figure 31.
It seems you can raise the switching frequency with a 5pF to 20pF between pins 2 and 7. Small values for us but you might have 10pF ceramic to try.
Shottky diodes may improve efficiency, but could also make fast noise current pulses stronger?

[Kevin Mitchell]

It's the datasheet dual power circuit figure 22.

Ah I overlooked that one. A negative doubler? That's a new one to me! I've done this with a bunch of series diodes & capacitors (forgot what the format is called).
Carry on...

[DUCKFACE]

Hey man, save some schottkies for the rest of us!
I'm not quite sure what's going on with your circuit. It looks like some strange stuff going on with the negative output. Like what's up with C6? Are all those diodes necessary?

im using the diagram from the producer pdf. https://www.ti.com/lit/ds/symlink/lt1054.pdf figure 22

[DUCKFACE]

It's the datasheet dual power circuit figure 22.
https://www.ti.com/lit/ds/symlink/lt1054.pdf
I still don't see any bypass capacitors tried on pin 8. Datasheet circuits usually assume that the input power supply filtering is adequate or non-critical. Look at the layout example in figure 31.
Shottky diodes may improve efficiency, but could also make fast noise current pulses stronger?

here is the hole idea. input supply is filtered and stabilized. Bypass capacitor is C3 1000uf
where to remove the shotkey and where to keep them ?
another interesting part is that when the relay is engaged by flipflop genrator the noises gets more noticeable

[PRR]

> when the relay is engaged by flipflop genrator the noises gets more noticeable

Power supply crap often increases with load.

Since your +12V seems(?) well filtered, I am looking at your LT1054 scheme. Is it at all possible to get two 9V batteries and run without the LT? (Yes the 12V relay may fail to clack; wire around it.)

[Rob Strand]

here is the hole idea. input supply is filtered and stabilized. Bypass capacitor is C3 1000uf
where to remove the shotkey and where to keep them ?
another interesting part is that when the relay is engaged by flipflop genrator the noises gets more noticeable

Technically the linear regulator IC2 should have a 220nF cap located right on IC2 input pins 1 and 2.

The view from outside might be the noise is increased when the Relay is active but the true cause might not be the load or relay itself.  If the relay is switching signal you are hearing different audio in the switched case.   If that audio path is noisy you will hear it.    The next step is to work out why that audio path is more noisy.   A high gain circuit like a fuzz or distortion pedal is going to be susceptible to noise.   The root cause of the noise could still be the charge pump or layout.

Causes are very specific things and can be very obscure.  The way you debug them is to isolate the problem to smaller regions of the circuit.   That's not so easy for a newbie, and it's not so easy when to-ing and fro-ing posts.

[anotherjim]

I don't know how relevant it might be, but the 555 state LEDs might be pushed to their maximum 20mA with 470R resistors. I would go higher possibly even to 4.7k.

I suggested the cap to change the LT1054 frequency to see if the noise is related to it. The stock frequency should be ultrasonic.

[DUCKFACE]

I suggested the cap to change the LT1054 frequency to see if the noise is related to it. The stock frequency should be ultrasonic.

iv tryed 5p1 8p2 15p. Te squeez is still there. only thing that was changed was the freq of the squeez. From hi 5p1 to mid 15p. The caps are connected 'on the fly'. I noticed that when i short pin7 and pin2 the squeez dissapear but the voltage start droping down.
btw voltage is +17.8v and -17.8v 

[anotherjim]

So the frequency of squeal changes with the change in clock frequency -  and the clock frequency should be ultrasonic, about 25Khz although that depends on supply voltage and I'm guessing that from the 12kHz - 35Khz datasheet range. You should not hear 25Khz!
I wonder if you have something else producing ultrasonic digital noise and the audible noise is the intermodulation result producing an audio frequency from 2 different ultrasonic frequencies. Perhaps it's your 12v supply?

[DUCKFACE]

You should not hear 25Khz!

maybe im kind of twisted
im seriously i can hear that squizz sound and its a nightmare.

I wonder if you have something else producing ultrasonic digital noise and the audible noise is the intermodulation result producing an audio frequency from 2 different ultrasonic frequencies. Perhaps it's your 12v supply?

iv changed the amp from transistor quad 405 to tube fender and the result is the same - hi sqizzing noise.
iv tryed with different power supply too.
opiton 1 casio 12v DC 1000mA transformer
option 2 panasonic KX-A11bs1 12v dc 500mA transformer

any other suggestions ?
im just curious about the output voltage? is it suppose to be +- 17.8v if in the input is +9v
according to pdf +VO ≈ 2 VIN − (VL + 2 VDiode)

here is the layout of the pcb