power supplies in parallel

[brownt]

hat happens on the wiper when two power supplies are put in parallel across a pot.

The scenario is two guitar effects units each with an expression jack. The jack outputs 5volts and is supposed to go across a potentiometer and the wiper voltage is fed back to the effects unit.

Assuming one potentiometer controlling two effects units the power supplies would be in parallel.
I have built a simulation circuit and things seem to work ok, and I guess I can actually build the circuit and see what happens.

but, what do you think would happen?

[homebrew]

Stop! I would highly recommend not putting arbitrary power supplies in parallel. Here are at lest four reasons why:

1) Depending on the topology, the supplies cannot handle reverse supply at all.
2) As the supplies will not have the EXACT same voltage, one supply will power both units which could overload the supplies easily.
3) Reverse supplying devices through I/O-Ports could override protection circuits.
4) In case there are several supply rails you would override all sequencing efforts. Maybe this voltage rail is not allowed if the rest of the unit is powered down?

So without the knowledge of the specific circuits I would not recommend any of this ...

!!! MAYBE !!! You could just feed the analog control voltage (from the wiper) to both units. But then you must ensure the integrity of the ground - that must be the same on both devices. Some sort of extra protection (i.e. adding a 1K resistor if the supply voltage is 5V) would be a good move. But again be careful ...

[brownt]

Stop! I would highly recommend not putting arbitrary power supplies in parallel. Here are at lest four reasons why:

1) Depending on the topology, the supplies cannot handle reverse supply at all.
2) As the supplies will not have the EXACT same voltage, one supply will power both units which could overload the supplies easily.
3) Reverse supplying devices through I/O-Ports could override protection circuits.
4) In case there are several supply rails you would override all sequencing efforts. Maybe this voltage rail is not allowed if the rest of the unit is powered down?

So without the knowledge of the specific circuits I would not recommend any of this ...

!!! MAYBE !!! You could just feed the analog control voltage (from the wiper) to both units. But then you must ensure the integrity of the ground - that must be the same on both devices. Some sort of extra protection (i.e. adding a 1K resistor if the supply voltage is 5V) would be a good move. But again be careful ...

thanks.
1. Why would there be a reverse supply?
2. Why does the exact same voltage matter, when you can put batteries in parallel and things work ok?
3. Where does the reverse supply come from?
4. I see.

[homebrew]

thanks.
1. Why would there be a reverse supply?
2. Why does the exact same voltage matter, when you can put batteries in parallel and things work ok?
3. Where does the reverse supply come from?
4. I see.

1) Imagine one device is turned off but still receives power through your contraption. Then the internal power supply 'sees' a higher output than input voltage. That means that IF current flows across its regulator, THEN it flows in REVERSE. That may (or may not) cause a problem - all depending on the specific design.
2) With accumulators this can be ok, as the cell voltage will average out over time. With fixed supplies, it is not. Imagine one supply delivering 5.2 volts while the other is putting out 5.1 volts. Then all current will flow from the 5.2V one. No averaging here.
3) When you connect your power supplies as described, then you would power on effect unit through the input port (See 1). If the device has any fuses in it, they will most likely be overridden as they are not in the current path anymore.

[james_s]

So if I understand this right, you have two voltage sources and you want a single variable voltage coming from the pot wiper? Can you just connect the grounds together and only use one of the V+ outputs so you're only using one power supply?

[brownt]

I don't want to modify the effect units. I want to drive two effect units with a single expression pedal.

It seems to work OK using the wiper voltage, only on the second effect unit, as suggested

Here is a schematic diagram of one of them.

I wonder what I need to be careful of.

[homebrew]

I don't want to modify the effect units. I want to drive two effect units with a single expression pedal.

It seems to work OK using the wiper voltage, only on the second effect unit, as suggested

Here is a schematic diagram of one of them.

I wonder what I need to be careful of.

Well, the threat started with the topic of "power supplies in parallel" and you explicitly stated the "Assuming one potentiometer controlling two effects units the power supplies would be in parallel."

That is where my concerns came from.

The schematic looks ok though. If this is equal to both effect units, I would follow james_s advice which you apparently did already. The current seems limited through the 1K resistor to AVDD. The only thing I would test is if the control voltages are in the same range so you don't feed a too large voltage into one of the effect units. In the particular schematic D6 and D7 (if installed) would protect against that scenario.

[brownt]

There is the potential for damage as some effects units use 5 volts on the control jack, but most use 3.3 volts.

I guess the units would all have the same diode protection, but maybe not. So, if I put in my own schottky diodes between ground and 5 volts would that work?

[obiwanjacobi]

Couple of series resistors and diodes and cross your fingers...

 

[brownt]

something like this circuit?

So, my diodes in parallel with some possible existing diodes in the effects units, is ok?

I guess a single limiting resistor will do?

Limiting will cause a voltage drop that will affect the voltage fed back to the A/D and so full modulation of an effect parameter will not occur, and for effect parameters like pitch, that is a problem.

All units I have found seem to have their own limiting, so perhaps that is not needed. A manufacturer would have to be crazy to not at least provide limiting?

[brownt]

that does not work reliably.

I thought the diodes would prevent the wiper rising above 3.3volts on the 3.3volt effect unit, but the wiper voltage comes out at around 4 volts. How can I limit that? The current is in microAmps so maybe it is not a problem anyway, or maybe the low current is the reason the clipping is not working as expected.

There is another problem also. Most of the dozens of effects units I have here all respond as expected when an isolated 3.3 volt supply is applied between the control jack tip (wiper) and ground. This is a simulation (minus the load) of the voltage fed back on the wiper from an expression pedal (10k pot).

But there is one effect unit that does not work with the simulated voltage, it is as if it needs a load to function correctly. If put a maximum of about 27K across the 3.3v of its control jack, it works ok.

Perhaps it is ok to do that. I don't know.

[obiwanjacobi]

Hmm, no that is not what I meant. I thought you could put the diodes in series with the supply of the two effect units in order to prevent power of one effect unit feeding back into the other unit. The series resistor was to limit current when a voltage difference exists between the supplies of these units...

Could you not mod the expression pedal by replacing the mono pot with a stereo pot? (or do you want more than two effect units connected?)
This would completely remove the problem and keep the units isolated.

[2c]

[brownt]

I see. A limiting resistor or diodes in series will cause a voltage drop, and that is no good as full modulation won't occur.

Stereo pot, no. I am trying to make a universal unit that any expression pedal can plug into. Things are working ok now, except for 4v where I think it should be 3.3 volts.

Is that the way those clipping diodes should work?