Idea for the ultimate in T-series noise & hum elimination

[Papus]

This is just a thought experiment at the moment, I don't have the resources to test it at present but maybe someone on the forum can try it out.
I come from the world of guitars, and it occurs to me that many of the tinkerers here are more or less purely keyboard oriented, so some of you might not be aware of the array of awesome goodness available in the world of guitar pedals.
Enter the Boss NS-2 Noise Suppressor:

https://www.boss.info/us/products/ns-2/

Here's my idea: take a pre-swell pedal feed from the recovery amp output (pin 6 if I remember correctly) and use that as the trigger signal for the NS-2.
Take a post-swell pedal line out from the headphone jack, power amp board or wherever you get yours from. I get mine from the headphone jack into a D.I. box with 40dB pad engaged to bring it down to instrument level. Unfortunately the headphone jack method means the line-out signal contains all of the power amp hum and noise - this is where the NS-2 shines :)
Run the post-swell pedal signal through the send-return loop of the NS-2.

The result is this:
The NS-2 gate will remain closed unless a signal is present on the pre-swell line out. The gate then opens, allowing the post-swell line out to pass signal via the NS-2's send-return loop.
By using the pre-swell line out as the trigger for the gate, the position of the swell pedal has no bearing on whether the gate opens or closes. This means that if your post-swell signal is near or below the noise floor, the gate will still function correctly even when the swell pedal is so low that an ordinary simple gate would not be able to differentiate between noise and actual low-volume organ signal.

My own critical listening to the noise levels in my T202 indicates that any latent noise/hum is not detectable by the ear when an organ signal is present, even with the swell pedal at minimum. Therefore a noise gate will effectively totally kill all noise and hum when no keys are depressed, and during any key press the noise is totally masked by the organ signal.

As an added bonus, because the noise gate is triggered by pure recovery amp signal, any additional downstream noise created by distortion pedals or other effects is also effectively gated much better than any other gating method.

Somebody please try this method, I'd love to know if it's worthwhile pursuing.

[Brendon Wright]

Yep, I agree that when a note is being played the hum doesn't matter.
With my own T500 I had a DIY noise gate between the preamp output pin (pin 3) and the main input of the amp (pin 8) which did help for a time, but the power amp itself was producing an almost equal amount of hum itself.
(So I'd halved it, I guess, until it stopped working)

It's worth noting that the swell pedal drops most of the hum as well as the notes when in heel down position, so your pre-swell triggering ought to work well as in there won't be much hum when you're playing at low volume: this always used to be a problem on guitar.

The reverb can be prey to hum as well, but reverb's not the sort of thing that works well with a noise gate.

I guess with the preamp on your 122 you could take the signal from BEFORE the T's final amp stage....
maybe where the reverb meets back with the main signal after C604?

The emitter of Q602?

I'm not sure if you'd have to add a load resistor or not, but the signal from those points may have less hum?

[Papus]

Yep, I agree that when a note is being played the hum doesn't matter.
With my own T500 I had a DIY noise gate between the preamp output pin (pin 3) and the main input of the amp (pin 8) which did help for a time, but the power amp itself was producing an almost equal amount of hum itself.
(So I'd halved it, I guess, until it stopped working)

It's worth noting that the swell pedal drops most of the hum as well as the notes when in heel down position, so your pre-swell triggering ought to work well as in there won't be much hum when you're playing at low volume: this always used to be a problem on guitar.

The reverb can be prey to hum as well, but reverb's not the sort of thing that works well with a noise gate.

I guess with the preamp on your 122 you could take the signal from BEFORE the T's final amp stage....
maybe where the reverb meets back with the main signal after C604?

The emitter of Q602?

I'm not sure if you'd have to add a load resistor or not, but the signal from those points may have less hum?

As always your knowledge of the T-series is exemplary!
I lack the requisite soldering skills to tap into an emitter point :(
I'm afraid my skills are limited to doing tag strips and snipping off capacitors.
On my T202 the headphone amp output (which is really just the main amp speaker output padded down) hum is masked by even the quietest notes, so I think I'll be ok with a normal "dumb" noise gate in my effects pedal chain which goes straight into my 122's 1/4" input.

[Brendon Wright]

What about plain ol' pin 12 of the power amp, then?
That's the return from the reverb amp where it rejoins the main amp, it's still got a good strong backwash from the dry signal PLUS the reverb, even though I'd prefer to grab it from the other side of the fat capacitor right in front of it (C604: 5uF. Shouldn't make too much difference but it might?? maybe?? filter out some bass end from the dry signal)

My power amp has actually had a LOT of grief over the years, which is how I've learned to (superficially) read those jolly schematics.
I'm responsible for most of that grief, I have to add.

I'm working towards replacing it with the guts of a dead M3, it will be a sad farewell to something so familiar, but I'm hoping the "new" amp will suffer less from my antics.

- - - Updated - - -

Ahem, anyhow!!
Pin 12 is neither here nor there if your hum's not too bad and you can slip in a good noise gate.

[wayshot]

Fighting hum/noise/crackling on a T is a task that never seems to finish :-) Tried replacing three e-caps on my T's power amp, no change. I THINK I've managed to get rid of some crackling by replacing two transistors and the electrolytic cap on the recovery/nonvibrato board. The parts were sourced from circuit boards in an old noisy UPS with defective batteries. There is still hum, and I plan to replace two PSU filter caps - for the +25V and +15V rails going to the power amp etc.

There's a plan B if filter cap replacement won't reduce hum: Yank the power amp board and replace it with some low-power guitar tube amp with built-in reverb, like the Bugera V5. Another more ambitious plan is replacing the power amp and various other boards with an AO-29, like Brendon seems to be working towards. Doable, but with a few pitfalls like missing reverb and possible level/impedance mismatches between the AO-29 and the remaining solid-state circuitry.

[Papus]

Pin 12 you say, interesting. Is that post-swell?
I never use the internal amp, speakers or cheese wheel Leslie.
All I need is a nice line-level signal post-swell :)

[Brendon Wright]

Pin 12 you say, interesting. Is that post-swell?
I never use the internal amp, speakers or cheese wheel Leslie.
All I need is a nice line-level signal post-swell :)

Yup, it is post swell.
Note that I'm a T enthusiast NOT a trained electronics guy.
By experimentation and looking at the schematics I've identified this spot.
These circuits are VERY forgiving, but if this was a valve/tube amp I couldn't guarantee I'm not stressing some part of the amp.

However, that aside, I'm fairly sure that should be a nice spot to take a signal.
I think that also means you can pull pins 22 to 26, the ins and out of the output transistors (those ones on the big aluminium cooling fins at the back of the chassis).

I've also written down notes on bypassing the autoaccompaniment board, but I need to check which of my other mods it depends on and which it can do without, including the removal of the drum machine.

[Brendon Wright]

Fighting hum/noise/crackling on a T is a task that never seems to finish

Yep, I haven't managed to locate recordings of my T BEFORE the Carsten Meyer mods, but the earlier vids I've got on youtube all showed significant hum.
It's possible the T's original "bass filtering" was solely designed to deal with that very hum.

I never realised how significant that hum was until I got my M101 and leslie 122: the only hum they have is the physical vibration from the low budget step-down transformers I used, otherwise, silent.

At one stage I replaced all of the T's transistors I could find ('til I ran out of replacements) though a lot of the original carbon resistors are still in there. I'm NOT going to replace all those! Repopulating a board badly damages the circuit tracks.

After the obvious ones like the power supply, every step in my "Hum Witch-hunt" I'd done in the past seemed to improve it to a micro-degree.
However, there are so many wires and the potential of so many ground loops that it becomes a dizzying prospect.

I THINK the lower manual may be especially open to hum, I'm not sure why.


I DID always thought that hum was kinda "Rock n' Roll!" in a kind of musical anarchy kind of way.

[wayshot]

I haven't experienced any more hum from the lower manual, but the signal goes through the auto-accomp board, so there's more circuitry involved and more sources of hum. I've noticed one thing, though: There's slightly less bottom on the lower manual than the upper; noticeable when comparing the lowest F keys with just 8' drawbar on lower and 16' on upper. The manual recovery stages and bus amps are identical. Wonder if the LM signal path in the auto-accomp board is the culprit as there are three series caps possibly acting as high-pass filters.

The crackling seemed to increase after removing the click filter/feedback loop caps from the recovery/nonvib board. Maybe the transistors "acted up" because of the changed configuration. Haven't heard any noticeable crackling after replacing Q854 and Q855.

[Brendon Wright]

>Wonder if the LM signal path in the auto-accomp board is the culprit

You can try the LM signal without the auto accomp by shorting pin 4 of the accomp board with pin 6 (counting from left on the multi pin plug/socket)
unfortunately this also bypasses the volume trim knob included on the board.
these two pins are actually the A and B cables behind the LM tabs. That's probably a better place to short them.

I possibly have more hum on the LM because I've added an LM volume knob bypass to add more bass at low volume. The Vol knob is in place of the Accomp board.


>Haven't heard any noticeable crackling after replacing Q854 and Q855

Phew!

[wayshot]

I'll try bypassing it, should be easy. However, I'm somehow opposed to removing or bypassing organ functions. There's a "soul" in this organ that I don't wanna ruin :-) I'm planning to sell it if I get a console organ (A-100 or E-100 preferably) as my apartment is too small for two organs. It must be in 100% working order and not too noisy before selling.

Speaking of hum/noise compared to other organs: Running the Studiologic Numa Organ (when it worked) through the Leslie was almost dead silent. The T-500, definitely not so.

[Papus]

This thread has become a goldmine for me, thanks guys!

There's another option in the pedal market: not a gate, but a true hum eliminator using sophisticated FFT narrowband intelligent filters in a digital DSP:

http://www.ehx.com/products/hum-debugger

The DSP listens to the noise sample during silent passages and continually tweaks the noise print to intelligently filter out only the precise offending frequencies.
If you've ever used the FFT noise reduction algorithm in Adobe Audition (formerly Cool Edit Pro) then you'll know how effective this technique is at stripping high levels of noise whilst preserving the signal perfectly.

[Brendon Wright]

I'll try bypassing it, should be easy. However, I'm somehow opposed to removing or bypassing organ functions. There's a "soul" in this organ that I don't wanna ruin :-) I'm planning to sell it if I get a console organ (A-100 or E-100 preferably) as my apartment is too small for two organs. It must be in 100% working order and not too noisy before selling.

Speaking of hum/noise compared to other organs: Running the Studiologic Numa Organ (when it worked) through the Leslie was almost dead silent. The T-500, definitely not so.

Sounds a mighty fair reason to keep it original!

- - - Updated - - -

This thread has become a goldmine for me, thanks guys!

There's another option in the pedal market: not a gate, but a true hum eliminator using sophisticated FFT narrowband intelligent filters in a digital DSP:

http://www.ehx.com/products/hum-debugger

The DSP listens to the noise sample during silent passages and continually tweaks the noise print to intelligently filter out only the precise offending frequencies.
If you've ever used the FFT noise reduction algorithm in Adobe Audition (formerly Cool Edit Pro) then you'll know how effective this technique is at stripping high levels of noise whilst preserving the signal perfectly.

Now THAT sounds a mighty useful pedal. Might be worth googling for a price. I see there's an article on the page which says it works well with vintage keys!

- - - Updated - - -

US$117.50 with lots of very happy reviews.

- - - Updated - - -

HOWEVER a thread on the Seymour-Duncan forum said the early product got recalled and made the tone of the output sound very digital, especially notable at volume.
On that thread the fixed up version was said to be better and certainly better to listen than the hum you get on a guitar at times. The fellow hadn't been able to really test drive it while at home though.

- - - Updated - - -

ALSO worth noting, is that if it's designed to negate 60Hz Hum it may not be any good in a 50Hz country.

[Papus]

Sounds a mighty fair reason to keep it original!

- - - Updated - - -



Now THAT sounds a mighty useful pedal. Might be worth googling for a price. I see there's an article on the page which says it works well with vintage keys!

- - - Updated - - -

US$117.50 with lots of very happy reviews.

- - - Updated - - -

HOWEVER a thread on the Seymour-Duncan forum said the early product got recalled and made the tone of the output sound very digital, especially notable at volume.
On that thread the fixed up version was said to be better and certainly better to listen than the hum you get on a guitar at times. The fellow hadn't been able to really test drive it while at home though.

- - - Updated - - -

ALSO worth noting, is that if it's designed to negate 60Hz Hum it may not be any good in a 50Hz country.

Yeah the 60Hz vs 50Hz issue is what is holding me back from trying this pedal, especially as I find that a regular dumb single-ended noise gate works nicely on my T.
EHX sells it in Australia via Australian distributors, so maybe it is safe to assume that the pedal is either; shipped pre-configured for 50Hz, or is sufficiently intelligent to discriminate between all manner and frequencies of noise on the fly.

[chbelt]

I don't know why, but as of yesterday the hum in the T500 I've been working on has practically disappeared. I haven't really been targeting the hum, but I did recently cut out the wires going to the cassette. I also did "Skill 1" of the keyboardpartner mods, removing 5 capacitors. The ground wire going to the Rhythm III unit has been disconnected during this work while I repeatedly removed/replaced the top, but yesterday I re-connected the ground and the hum went away. There's a tiny bit when the swell pedal is floored, but it's barely noticeable, unlike before I started working on it. I don't know what to conclude from this, but maybe somebody who's smarter than me will find a clue.