Taking the 'hummm' out of a circuit

Im working on a power supply for some musical bits. Unfortunately its
producing a great deal of "hum" in the circuit. Is there any way to filter
out residual AC in the DC line?

What I have:

+18V (wall wart) ---->3500uF
(electrolytic) ---->.22uF--->7809--->.1uF--->out
|
| |
ground
ground ground

the ground is common between the wall wart and all components. There are 4
7809s in parallel at the output.

 

If youve got any humm in the dc output still then something is wrong as it
sugests the regulator is dropping out. ie the input voltage is diping too
loo for the regulator to compensate.

Whats also likely is that you have ground or signal loops that are picking
up mains humm, particularly if you have a few bits of equipment conected
together or long signal leads earthed at both ends etc ..

Colin =^.^=

 

Subject: Taking the 'hummm' out of a circuit

You didn't mention if your wall wart is a switcher or a linear supply. If it's
relatively light, it's probably a switcher. Could be a lot of problems there.
But if it feels like it's got some iron in it, it's probably a linear. That
would narrow down your possibe problems.

Even if you're drawing an amp from the unregulated linear, you shouldn't have
more than 2V peak-to-peak of AC. If it's much greater than that, you've
probably got a bad 3500uF cap. Try replacing that first.

If you're still looking at problems with 2V p.p. of AC, your stuff may be very
sensitive to ripple in the power supply (the 7809s will probably cut that by at
least 60dB, which would give you less than 20mV ripple on the 9V output). If
you're measuring more than 20mV of ripple at the output of a 7809, you might
want to look at replacing them.

The noise on the power supply we've been talking about is 100Hz or 120Hz from
the wall wart. If it's higher frequency stuff, the output cap might be bad,
and you could try replacing that. If not, you might be getting crosstalk
between power supplies. A 7809 is very good at cutting down 100/120 Hz ripple,
but it's not very good at audio frequency. If that's an issue, you might want
to try putting an isolating resistor on each input to the 7809 like this (view
in fixed font or M$ Notepad):


Multiple 7809s ____
___ | |
o------o---o--.o-|___|-------o-----|7809||--o------o9V
| | | 10 ohm | |____| |
| | --- | ---
3500uF --- | --- | ---
--- | | === |
| | | GND |
o------o | === ____ ===
| | ___ GND | | GND
| o---o-|___|-------o-----|7809|----------o9V
=== | 10 ohm | |____| |
GND | --- | ---
: --- ---
: | |
| | |
: : === === ===
GND GND GND
:
:
____
| ___ | |
'---o-|___|-------o-----|7809|---o------o9V
10 ohm | |____| |
--- | ---
--- ---
| |
| | |
=== === ===
GND GND GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

These are the usual suspects.

Good luck
Chris

 

Is this to power guitar effects by any chance? If your components are
working properly, you might want to look into the possibility of a ground
loop somewhere. For instance, if you have your amp plugged in one outlet,
and your effects power supply in another one on the other side of the room,
you're almost guaranteed to create a ground loop. Try plugging all the gear
your plugging into into one outlet or power bar and see if the hum goes
away. Alternately, try unplugging (from the wall) each piece of equipment
one at a time to see if the hum goes away. If it does, you've found the
problem.

 

Looks like you are trying to get too much from the wall wart.
What is wall wart current rating at 18 volts and how much is
the load current?

-Bill

 

yes these are guitar effects. they are all plugged into the same power
strip. they work properly (normally).

 

The wall wart puts out 400 mA. For my test I was drawing ~5mA.

 

So your amp is also plugged into the same power strip?

 

One other question: Why do you have 4 7809s at the output? You could get
away with just one and run parallel lines to each effect from the one
ouptut. I wonder if you've got some goofy ground loop with all these
regulators.
FWIW I power 5 effects and a tuner with one wall wart connected to a single
7809 with about the same filtering you have here with no problems. Your plan
should work.
Also, have you tried the obvious culprits (bad cables, connectors, etc)?
Have you tried plugging in one effect at a time to see if any single effect
is the problem? Have you double checked your solder connections and wiring?

 

Weeeell... Ill try removing 3 of them. I didn't want to mess with a daisy
chain.

When I plug in my Boss DS-1 (which should draw 5mA max) I get tons of hum
and not much sound. Plugging in the echo box doesn't work at all. Both work
fine from the boss wall wart.

This circuit shouldn't be giving me fits like this. I think its a rite of
passage.

 

yes.

 

FWIW - the regulators share the ground line. I was hoping to spread out any
heat load by having multiple regulators. Im a long shot from an EE - which
is probably why Im having so many issues with this.

 

simple
as far as I see the output of the regulator is totally under-buffered
try putting an extra elco at the output of the reg, say 2200u something



Im working on a power supply for some musical bits. Unfortunately its
producing a great deal of "hum" in the circuit. Is there any way to filter
out residual AC in the DC line?

What I have:

+18V (wall wart) ---->3500uF
(electrolytic) ---->.22uF--->7809--->.1uF--->out
|
| |
ground
ground ground

the ground is common between the wall wart and all components. There are 4
7809s in parallel at the output.

 

"under buffered"? can you be more specific?

 

Seeing your schematic, I might say forget those textbook examples for
regulators, DON'T use them without an elco at the output since usually it's
a no-go.

I only see .1uF at the output of the regulator.
Regs are rather sensitive to input ripple and output current variations.
Input caps' size looks OK to me.
Try compensating by buffering the output with an elco of say 2200u instead
of larger input cap, so that any possible fed-through ripple gets caught by
the output cap.
By "under buffered" I meant "not having enough of energy reservoir
(capacity) at the output to handle voltage/current variations".

Above I assume your wall wart is capable of delivering enough current, but
as a side assumption :
As far as I seem to read, all 4 individual regs take appprox 400mA, thus
1.6A together
Is your wall wart capable of handling this ?
If not, input voltage will drop until a level where input ripple (lower
level) drops below 12V , thus making it impossible of regulating using a
7809

For your info also this:
It IS possible using a 78xx series regulator and boost it's output current
by using a simple setup as follows
(copy and paste "image" below with a fixed font, say courier, in notepad)

power transistor, pnp
-----------o-\ /---------------
| | v_/ |
| .1u | | |
| === / 33R |
| === \ +--------+ |
| 2R2 | / | | |
--o---/\/\/\-o---o---- 7809 ---o----o--- 9V
| | | | |
=== +----|---+ === ===
=== | === ===
| | | |
--o------------------------o-------o----o---
3500u .1u 2200u

Explanation:
For low currents the transistor is closed, reg takes it all
If reg current increases to say 0.32A, voltage drop over 2R2 will rise
enough to open the transistor (Vbe rises above say 0.7V), thus having the
transistor take everything above 0.32A output current.
Transistor type has to be capable of handling the output current, in your
case I'd say BD138 or something similar.
This setup uses a SINGLE reg and ONE transistor with few discretes instead
of 4 regs and buffering

greetz


"under buffered"? can you be more specific?

 

A couple of things are screwy here. First, a daisy chain is when you use one
effect to power the rest. A few of the Boss pedals (like the NS2 noise
suppressor) have 2 DC power jacks - one for the incoming DC, and the other
to connect to the next pedal. You have 4 separate regulators, so pulling one
or any or all of them isn't going to hurt anything.

Another thing you should double check is polarity. Some pedals use a centre
positive connection and others centre negative. If you have things connected
the wrong way, look out! Do you have the positive DC connected to the
positive pin on the plug?

 

Well, if the input is above 12 volts DC and the load is only 5mA,
the only thing I can think of is it may be oscillating.
Use the DMM to measure AC voltage at the output. It should
be very low at 5 to 10 mV. You could try a small capacitor on the
output, maybe 1-10 uF. Maybe that will help.

-Bill

 

Each regulator requires 400mA just to function *at all*? I didn't know this.
My wall wart only puts out 400mA. Where would I look up this information?
(more specifically - where did you get it from?)

 

NOPE
I didn't tell you regs need 400mA to function *at all*
(wonder where you read this, jeses)
Regs of this type work with output currents from 0 upto 1A, see datasheets

Each regulator requires 400mA just to function *at all*? I didn't know this.
My wall wart only puts out 400mA. Where would I look up this information?
(more specifically - where did you get it from?)

 

My bad. Thats how I read your previous posting.