Low Noise 9v Volt Regulator Using IC 7809, Any Advice ?

[Roger Martin]

Hi guys,
I have modified my 9 v wall wart with this scheme using IC 7809 :


I have tested it using decimator g-string clone and i got these results :
- Noisy background sound like when you use a flanger pedal effect is present.
- Volt reading is a stable 8.9 v.
- Wall wart is working cooly, not as hot as before it was modified.
- Seems like the whole trimmer setting of decimator g-string clone needs another trimming, its different from the setting when i used 9v battery.
  When i used 9v battery, the decimator g-string clone is fine and doing its job as supposed to. No irritating noise.
  And i believe this 9v wall wart is using wrong scheme design, and it cant control the noise ripple perfectly.

What is wrong with the wall wart scheme, maybe anyone would be so kind to explain or correct it ?

Thank you in advance.

[Arch]

"needs another trimming, its different from the setting when i used 9v battery."

Yes! Use stable voltage and trim trimpot! Forget about battery! On long comes short!
You measure voltage on battery? Think, it differs from 8.9 volt

You install heat sink on the IC 7809?
http://www.floodland.nl/aim/plaatjes/info_78xx_1_1.jpg

[Roger Martin]

"needs another trimming, its different from the setting when i used 9v battery."

Yes! Use stable voltage and trim trimpot! Forget about battery! On long comes short!
You measure voltage on battery? Think, it differs from 8.9 volt

You install heat sink on the IC 7809?
http://www.floodland.nl/aim/plaatjes/info_78xx_1_1.jpg

Oh.....ok, Arch. I havent trimmed it yet using the modified wall wart.
It was already 2 in the morning, i had to sleep and i had to wake up at 5.30.
I thought i didnt want to mess up the old setting coz i had no more 9v battery then.
My 9v battery i used was shorcircuited by accident for 5 seconds once and it leaked and exhausted quickly and i measured it was 7.6 volt !!!
And i trimmed it again using that 7.6 volt battery ! It was ok. It worked out well as it was used to be.
Now the battery is about 0.6 volt !
I guess, i can't leave the 9v battery attached to the decimator for so long time after use, it will drain the volt ?? Although it's not turned on ?
I'll trim it again using wall wart then.

Yes, of course...i installed a heat sink. And the wall wart didnt get as hot as it was used to be.
I got digital multimeter readings, and i always got a stable 8.9v, unplugged from any pedal. I think it's stable now.
It used to be 8.78 v - 8.86 v. It was using some fake unknown volt regulator IC, it was a bad quality i think.

I found out a good schematic, i think.
Maybe you can asses it, Arch. Let me know what your opinion about it.
http://www.tnt-audio.com/clinica/regulators_noise3_e.html

Let me quote :

Zener + Emitter Follower



This is about the simplest discrete-device series regulator one could imagine. Ignore R13 (it lowers power dissipation and can be part of an RC input filter for enhanced ripple rejection). The core is zener diode D1, serving as voltage reference. As ordinary zeners are noisy some filtering is required, but zeners have a rather low impedance, hence D1 is put in series with R14: this reduces DC stability, which we can live with, and massively reduces noise. Q4 is the output of the circuit: a bipolar transistor working as emitter follower. At 30mA output current its output impedance is a useable 1 Ohm, and at higher currents it is even lower. Base-stopper R17 is there to ensure stability. That's it.



And noise-wise we have a winner here, being 20dB lower than the best of the above circuits! The small peaks in the lower frequencies are at 100Hz and harmonics, clearly related to the mains feed, though it is not quite clear to me how these could have gotten in. Probably a slight routing mistake with a ground wire.

Oh, and listening through the headphones?

Ghostly quiet ...

This is awesome, eh ?
It can lower the noise down to 20dB lower than the best of the above circuits !
How do i mod this schematic to produce a 9 volt wall wart ? Any idea, Arch ?

[Arch]

Schematic is BAD for pedal!
Don't have shortcut protection!
(Otherwise it is necessary mount in pedal)

Can you record noise and give the link on it?

(And connect you stable wall wart to decimator and measure voltage. I think, voltage be low then 8.9v. For better stabilizations, 7809 need 12-14v DC on input)

[R.G.]

- Noisy background sound like when you use a flanger pedal effect is present.

Very odd. That description says "oscillation" but there's nothing in the schematic to suggest that oscillation would be a problem. Possibly an interaction of the power and effect together. An oscilloscope would tell you.

- Volt reading is a stable 8.9 v.

That's low for a 7809. They're often just about 9.0-something. Still, within spec.

- Seems like the whole trimmer setting of decimator g-string clone needs another trimming, its different from the setting when i used 9v battery.
  When i used 9v battery, the decimator g-string clone is fine and doing its job as supposed to. No irritating noise.

"9V" batteries are almost never 9.0000V. A fresh alkaline 9V is about 9.45V. They sag as they are discharged, and are considered exhausted when they get down to 7V or so. So if the decimator needs trimmed when the power supply voltage changes, it will need a change when the power supply is substituted in for a battery.  It would be odd for the decimator to be so picky, but it can happen.

And i believe this 9v wall wart is using wrong scheme design, and it cant control the noise ripple perfectly.
What is wrong with the wall wart scheme, maybe anyone would be so kind to explain or correct it ?

I don't see anything particularly wrong with the schematic. I could have missed something, but nothing jumps out. I did check the pinout, and that's the pin numbering Fairchild shows for the 7800 series.

There may be nothing wrong with the regulator/power supply, or the pedal, but a problem in how they interact. They may not play well with each other. Can't tell any more from the info provided. It is very, very odd that a 7800 series regulator would be a problem, as there are millions of these used in similar circuits all over the world. I have personally used several hundred over the years, and designed them into products selling in the 10s of thousands. You may just have had bad luck with this setup.

I found out a good schematic, i think.
Maybe you can asses it, Arch. Let me know what your opinion about it.
http://www.tnt-audio.com/clinica/regulators_noise3_e.html

Let me quote :

Zener + Emitter Follower

This is about the simplest discrete-device series regulator one could imagine. Ignore R13 (it lowers power dissipation and can be part of an RC input filter for enhanced ripple rejection). The core is zener diode D1, serving as voltage reference. As ordinary zeners are noisy some filtering is required, but zeners have a rather low impedance, hence D1 is put in series with R14: this reduces DC stability, which we can live with, and massively reduces noise. Q4 is the output of the circuit: a bipolar transistor working as emitter follower. At 30mA output current its output impedance is a useable 1 Ohm, and at higher currents it is even lower. Base-stopper R17 is there to ensure stability. That's it.

And noise-wise we have a winner here, being 20dB lower than the best of the above circuits! The small peaks in the lower frequencies are at 100Hz and harmonics, clearly related to the mains feed, though it is not quite clear to me how these could have gotten in. Probably a slight routing mistake with a grou
Ghostly quiet ...

This is awesome, eh ?
It can lower the noise down to 20dB lower than the best of the above circuits !

It's important to not believe everything you read on the internet. While simple emitter follower regulators can be used - EH used them a lot inside pedals - they are not without problems, among those problems being limited output current, limited noise reduction from the input, high output impedance, and a tendency to oscillate with certain capacitive loading. Like everything else, they're not a solution to everything, but useful in some situations, not others. The base resistor helps, but does not necessarily force stability. A small resistor on the output helps more, but also makes the output voltage "regulation" worse. And it will be stable with some capacitance, not others.

How do i mod this schematic to produce a 9 volt wall wart ?

You connect your 1000uF capacitor from the wall wart rectifier/filter where it says "v4", with the correct polarity, and you take your "9V" out in parallel with C7.

[PRR]

> Schematic is BAD for pedal!
> Don't have shortcut protection!

The emitter follower does have short-circuit "protection". The absolute maximum short-time current is 21V/320= 65mA. A half-Watt R13 will burn-up at 40mA for several seconds.

Like R.G., I don't see a problem with the first-post drawing plus a LM7809. Such supplies power millions of sensitive audio circuits with little problem.

In cases where audio noise is more important than precise load-voltage stability, you can try 47 ohms and 470uFd on the output of a '7809. This adds another 20dB of buzz reduction, over 20dB over most of the audio band, over 40dB in the worst hiss frequency ranges.

________

I didn't know what a "decimator g-string" was. It is a noise-gate. Very odd for a noise-gate to make noise. Should have ample internal filtering on its low-level stage. I don't know what may be different in a "clone". Noise-gates can be very critical about layout (they need a ton of gain) but wart/battery should not be real different.

[Roger Martin]

Thank you for all the advices poured into this topic, guys.

I have just tested it again and trimmed the 1M trimmer using the above mentioned wall wart.
I did'nt find the oscillation background sound like a flanger pedal in use again now.
But one annoying noise is so dominant in all setting is white noise like shower water or heavy rain is pouring down.
No matter what i tried this and that, the white noise is so dominant. It sounds SSSSHHHHHHHHHHHH..........not a hiss but........HOOOOOSSSSSSHHHHHHHHH
If i turned threshold to out of range of the right position, another noise is  present like a big MMMEEEEEEEEMMMMMMMP......

I measured the volt when plugged into decimator along with a zoom 505II pedal in use is 8.47 volt.
When only using zoom it was only 8.73 volt.
At first, i tried a new 9 volt battery reading 9.93 volt.
I can find a sweet spot for trimmer with nice n clean sound. No noise.
But after using it, the 9 v battery was quickly drainned and reading was 8.46 volt.
That is all i can report to you now.

@ PRR.....
You mean, I can try adding 47 ohms and 470uFd on the output in series ?
And i use 47 ohm with 5 wattage is enough for not burning it ?
Thank you for commenting this topic, guys.

[Roger Martin]

Alright guys, this afternoon i decided to modify this wall wart another shot.
I tried to interpret what PRR said above and imagined some fantasy about it.
I was doing this in hopeless mood and i was so down and pesimisticly bored with this annoying hum noise.
I was just fed up of buying 9 volt batteries.
I tested it, and i realized...using a 9 volt regulator is impossible...
then i was thinking harder and gave it a deep contemplation.....oh shoot !
I pushed the Over Drive effect button on the Marshall amp all along this insane déjà vu test !
It means that i used a double heavy dose of Over Drive effect !
No wonder this wall wart cant cope with such over capacity.
After mod, it gave reading of 8.80 - 8.89 - 8.90 volt.
When plugged longer it once gave 9.10 volt.
And it just worked like a charm after i turned off the Marshall's over drive button.
Then i began to work on setting 1 M ohm trimmer again.
Rolled it clock wise until i met up with a huge buzz sound, stopped and i tuned back counter clock wise 1 roll with a clean sound.
That's it, i used all kinds of clean, over drive, distortion, metal, death metal...wow it worked, man !
I always cranked up volume, gain, treble, bass, contour, effect at max 10.
It was clean like a crystal clear sound.
Of course, if you push the Threshold roll it too much it will suck up the tone. Just find the best spot and it's done.
This decimator is so awesome, it behaves nicely when processing a note, it doesnt suck up and it sustains as long as it's supposed to stop.

And i think this 9 volt battery is so powerful that even the cleanest wall wart cant beat it when used with decimator g-string.
This noise reduction pedal + 9 volt battery can clean this huge noise of a double over drive of zoom and Marshall amp (you know how dirty a Marshall is).
Oh by the way, that flanger background noise was caused by this Marshall's built-in effect button turned on.
I felt all the heavy burdens on my chest were relieved.
To all DIYers who love those noisy heavy fuzz or distortions looking for a great contender of a noise reduction pedal champion...
don't look further far away...decimator g-string is definitely the best noise reduction there is.

[Arch]

"g-string is definitely the best noise reduction there is"

Truth

[Roger Martin]

Arch,
If you are interested in building decimator g-string version II ? 
Let's join us at :
http://www.diystompboxes.com/smfforum/index.php?topic=103574.0

[Frenky_18]

Arch,
If you are interested in building decimator g-string version II ? 
Let's join us at :
http://www.diystompboxes.com/smfforum/index.php?topic=103574.0

Looks like the link or topic does not work... I am realy sad about it because I´ve seen few clones on the youtube and they looks realy reasonable... and can´t find schematics and/or layout of g-string around whole damn internet... not even unverified...