Maxon OD820 hiss

[ninjadave7]

Hey all,
I recently acquired an old Maxon OD820 (which I love...) but when it's turned on it's got a bit of a hiss, enough that I can't keep it on all the time, I have to turn it off during silent or quiet parts. This happens at any volume as the gain is turned up. It's manageable when the gain is all the way down, but as soon as you give it any gas its loud. It's powered off it's own PP2+ output, I don't think it's noise inherited from somewhere else.

So... any suggestions as to where to start poking around to get rid of the hiss?

[52 Bill]

It appears that the gain control changes the amount of signal fed into the first op amp stage and it changes the gain of that op amp stage. If turning up the gain increases the hiss I'd guess that the op amp is noisy. Try subbing a new 4558 and see what happens.

[Mark Hammer]

U2 has nothing whatsoever to do with the clipping quality. Which means that you have the latitude to sub another chip without any mojo-degradation. I would suggest an NE552 or LM833. Both of these are likely to provide lower noise in the circuit than the 4558, given the surrounding components (both of these prefer low feedback and input resistances). That may not "cure" the pedal, but is a step in the right direction.

[ninjadave7]

Ok, good to know, both of you. I replaced both opamps with new TI 4558's, the noise level stayed pretty much the same. I'll try and get a hold of one of the other ops and see what happens.

Maybe this is related, so I thought I'd ask, but I also noticed that when the pedal is on (and I don't run the gain very hot, about 9-10 o'clock), it's amplifying all the "non-effects" noise from anything before it in the chain. By this I mean you tap on the buffer enclosure, or a patch cable, and especially the od820 enclosure, and it sounds like you're tapping on the outside of an empty metal gas can or something. Is this a bad ground, microphonic cable, cold solder, or something like that? Would that contribute greatly to the actual pedal noise problem?

[J M Fahey]

You have 2 suspects here.
1) The FET Q1 . replace it and see what happens.
2) the high frequency switcher, close to the sensitive inputs.
You might not hear the swtching waveform iself, but its artifacts as digital noise, very similar to hiss.
Block the oscillator and use a "real" extra 9V battery for -9V.

[ninjadave7]

Pardon my ignorance then, I'm assuming you mean the high freq switcher on the 1044, I think I have another, or a 7660 laying around, I could try replacing that. Otherwise, how would I block the oscillator and use a "real" extra 9v for -9v? Not sure what you mean!

[Mark Hammer]

What JM sugests is quite do-able, though from the sounds of it, it might invite some difficulties on your part.

That said, the 1044 has an oscillator inside it, and essentially works by accumulating the pulses generated. Those pulses can be artifacts on the power line itself. or it can be stray noise that is picked up by the circuitry the same way it would pick up and amplify other varieties of extraneous noise (radio stations, etc.).

The OD820 uses a bipolar supply (+/-9V) and Maxon has elected to implement that via a 1044, permitting the user to plug a regular 9v wallwart/PSU into the adaptor jack. One could, however, cut the -9v supply line (where D7, C18, and C19 meet), and wire up a 2nd battery such that the normal battery provides +9v, and the second battery provides -9v. The problem this introduces is how one shuts off both batteries when not in use. Normally, the ground end of the on-board battery goes to the ring connection on the input jack so that it goes ungrounded when you unplug. If you run the red end of the battary connector from the 2nd battery to that same ring connection on the input jack, then you have two 9v batteries connected in series and still wired up to the circuit board. Come back in the morning, and your batteries may be dead.

RG Keen has an elegant little circuit to get around this on his New Page 1 site, but it still involves much more than what you have right now, and there may not be convenient space in the pedal for it. The other solution is a little DPDT toggle that turns both batteries on and off. It is simpler and takes up much less space.

But we're getting ahead of ourselves here. None of this prior discussion pertains if the 1044 is not the source of the noise.

So let's turn our attention to more conventional sources of hiss: amplified input noise in high-gain circuits. My experience is that many high-gain circuits will have "protection" against oscillation and other circuit instability, but often have far more bandwidth than they need. And that unnecessary bandwidth includes high-frequency hiss.

In the case of your OD820, after R6/C8, there's not much to trim back on unused top end. There is some treble trim in the last stage, via C2 (20pf), but that only starts trimming back just under 80khz. So, what I will suggest is that you find the solder pads for C2 and R2 on the copper side, and tack on another feedback cap in parallel to trim back that treble even further. Feel free to go as high as 270pf (though 220pf may be more than sufficient). That should remove a lot of audible hiss but still leave plenty of bite intact. The nice thing is that by tacking a cap on the copper side, it's easy to experiment with optimal values and easily reversible.

For your consideration:
with C2 = 20pf, rolloff starts at 79.5khz
with C2 = 100pf, rolloff starts at 15.9khz
with C2 = 150pf, rolloff starts at 10.6khz
with C2 = 220pf, rolloff starts at 7.2khz
with C2 = 270pf, rolloff starts at 5.9khz

Remember that these are standard values. Also remember that they would be added to the 20pf already there (150pf + 20pf = 170pf), such that the real rolloff would begin a little lower than what is shown if you added one of those standard values in parallel. Finally, note that the rolloff is fairly shallow, so plenty of top end will still come through.

I've done this to plenty of pedals, and it has been an effective hiss-reduction strategy.

[ninjadave7]

Beautiful.

Any thoughts on the extra noise from previous pedals/cables? Like I said, when you turn on the pedal, tapping the patch cable connector into it, or the case, or even the case of pedals before it sounds like you're tapping the power tubes in the amp!

[J M Fahey]

Hi.
What I suggested was not a permanent solution but a 5 minute mod to check whether my theory is correct or not.
If nothing changes, back to square 1; if hiss disappears or *substantially* reduces, let´s worry about switching the extra -9V battery.
I chase ghosts every day, and best is to grab them to check whether they are immaterial or just a guy wrapped in some bedsheet
And I *still* mistrust Q1 (paranoid guy ) , so lift one end of C9 , any change?

[ninjadave7]

That makes sense.

[ninjadave7]

Can I sub a 201 or 5088 for that 2sk246 Q1? I know pinout is different, but I can work with that.

[Mark Hammer]

You DO realize that a 5088 is bipolar not a FET, right?

I don't know how quiet a 246 is, relative to a 201, but I do know that a 5089 is quieter than a 5088.

I also know that, despite having only 3 pins, by some miraculous fashion there are some 823 different pinouts for JFETs, so make sure you have a datasheet handy!!

[ninjadave7]

haha i guess I'm showing my ignorance...

I'll try the 201 then, really just trying things to get rid of all the noise.

[mozwell]

I would try other op amps, NE5532, LM833, LF353, and even TL072 instead of the 4558.
With a source resistance of (say) 50k, you can have up to 0.5mV noise on the the op amps output. This is getting "up there" as a % of the guitar signal, & hte higher gain you have, the more noise.

Did you try runing it from a battery ? did the noise decrease? IF so, part of hte noise is from the negative supply ic.
Maybe some supply decoupling will help here. Increase C13 to 1000uF or 2200uF.
In between op amp pin 8 & C13 insert a 22 ohm resistor. On op amp pin 8 place a 22uF & 0.1uF in parallel to pcb 0V. Do this also from op amp pin 4 to C19. caps on op amp pin 8 will have positive to pin 8, caps on op amp pin 4 will have positive to pcb 0V.
increasing C4, & C2 can help roll off some of your "highs". Try different values till you get one you like.

[ninjadave7]

I tried subbing some opamps. The quietest I've gotten it is with a TL072 in the gain circuit, but with the 4558 in the EQ. It's the same amount of noise with/without a battery.