Tonepad Ross comp build high noise floor

Forum dedicated to helping people debug and troubleshoot non-functional pedals or builds. Please use an clear and informative title, indicating circuit and basic problem. Don't forget to mark the issue as fixed if this is the case.
Post Reply
User avatar
danb435
Information
Posts: 5
Joined: 15 Nov 2015, 02:28

Post by danb435 »

Hi all, new to the forum. I have previous electronics experience, although only enough to be scary (years experience but very little troubleshooting to component level). I've completed a Tonepad Ross compressor (started as a Dyna build but decided to switch to the Ross). The comp seems fully functional but the noise floor with serious gain is pretty bad. This is using a LP with a Seymour JB (fairly hot) humbucker. The soldering all looks good, I tested all the component values before installing, used metal film resistors, and the 3080 OTA was from Small Bear (supposed to be a decent one). Because I'm a cheapwad I used Tayda resistors, caps and diodes. Again, I really like this comp, and I understand that a comp by nature raises the noise floor but this seems excessive. Any suggestions? Seems I've read that there is a quieter substitution for the 2N3904 transistors? Much thanks for any assistance. Dan B.

User avatar
stolen
Breadboard Brother
Information
Posts: 129
Joined: 04 Mar 2021, 09:54
Has thanked: 44 times
Been thanked: 55 times

Post by stolen »

Hello Dan!

Don't worry about the Tayda parts, those are fine. Capacitors generally do not contribute noise, and metal film resistors are basically all the same except when there is significant DC across them. 2n3904 do not have excellent noise in this circuit, but they are certainly not the limiting factor - the CA3080 is notoriously noisy, and, assuming there are no build errors, all other noise sources pale in comparison in this circuit. That doesn't mean all is lost though, but it will require some manual optimization.

One reason why OTAs are generally noisy is their extremely low input headroom, typically around 100-200mV p-p, depending on how much distortion is acceptable. This means the signal is typically heavily attenuated before the OTA to provide clean performance. We couldn't find any original schematic with labelled components, so we'll be referring to this drawing, even though it has a different OTA:

https://www.musikding.de/docs/musikding ... schalt.pdf

If you look at the differential input of the OTA, you'll find that both inputs are driven by the buffered input signal: The negative port directly, the positive port via a voltage divider composed of R7 and R10. There's a bunch of biasing and frequency response stuff going on with the high value resistors and caps, we can just ignore that for now. A differential input means simply that the difference between the inputs is considered the OTAs "real" input signal, so that the input gain is 15/(15+2)-1 =~0.118.

This is the attenuation we mentioned; if R10 were shorted, the input gain would increase by factor 8.5, decreasing the noise floor contributed by the OTA by the same amount. Hooray! Naturally, this increases the overall gain of the circuit, but we could simply reduce the OTA's current output scaling factor by decreasing R13 proportionally to 150k/8.5 =~18k.

Of course, this reduces the headroom of the circuit, so that loud input signals are more prone to clipping. Here's where the manual part comes in: You just have to test how much distortion with your guitar is just so acceptable to you. You could hack a ~25k trimmer or so across R10 and mess around with it a little to find your favorite position, then adjust R13 to taste.

There is one last drawback this comes with: We've ignored C3 so far. This cap produces a shelving treble boost by shorting R10 for high frequencies; we just did the math and the corner frequency is at about 9kHz, so it shouldn't be a big deal. Anyways, it influences the frequency response a little bit, and there's no elegant way of preserving that for arbitrary input gains. Then again, we think it'll be fine.

Of course, all of this won't change noise drastically, it's just a moderate optimization. If that doesn't suffice, go for a 13700 OTA, or maybe a proper blackmer cell.

Hope this helps, and all the best!

User avatar
danb435
Information
Posts: 5
Joined: 15 Nov 2015, 02:28

Post by danb435 »

Thanks for the reply. The Tonepad schematic, aside from the actual OTA, looks identical to the one linked in your post. I want to try to keep things as simple as possible and am considering looking at a quieter OTA. I can perfboard an adapter to use a LM13700 but my first choice would be a drop in replacement chip. Are you familiar with the Alfa AS3080E? The few reviews I've read seem positive. Is the LM13700 a quieter chip? Thanks

User avatar
stolen
Breadboard Brother
Information
Posts: 129
Joined: 04 Mar 2021, 09:54
Has thanked: 44 times
Been thanked: 55 times

Post by stolen »

Hey!

Our assumption would be that the AS3080E is a very close clone of the CA3080 for the synth folks who use it a lot in VCOs, where noise is of little concern. The datasheet doesn't mention any meaningful noise spec, and neither does the datasheet of the CA3080, at least the ones we could find. Because of this we'd guess that the Alfa chip does not improve noise performance, but feel free to send them an email if they have any data on this (and if they do, maybe share it here). The chances may be slim though.

For the same reason we don't really know how much quieter the LM13700 will be; we don't have any numbers to do the math with on the CA3080 side. Builders who've done more OTA stuff all seem to share the opinion that the 13700 is a meaningful upgrade, so we'll take their (and our memory's) word for it. We understand that perfboard adapters reduce the "clean look" of a PCB build, but we'd recommend to go for it nevertheless.

All the best,
stolen

User avatar
danb435
Information
Posts: 5
Joined: 15 Nov 2015, 02:28

Post by danb435 »

Looks like I will pursue an alternate OTA. And now....I'm going to admit my ignorance here....what is the 2K trimmer pot in the schematic for and how do I adjust it?
Thanks stolen, you've been a great help.

User avatar
gght
Solder Soldier
Information
Posts: 187
Joined: 18 Nov 2007, 05:32
Location: Houston, Texas USA
Has thanked: 47 times
Been thanked: 10 times

Post by gght »

Googling quickly showed you aren’t the first to run into this. Here are a couple of links.

https://www.madbeanpedals.com/forum/ind ... ic=22959.0

https://www.diystompboxes.com/smfforum/ ... c=100153.0

User avatar
danb435
Information
Posts: 5
Joined: 15 Nov 2015, 02:28

Post by danb435 »

Good info, again I may try swapping to a LM13700 with my existing pedal (not much effort or cost to try). It will be "less than" the cleanest mod but it's just for me and I'm getting too darn old to be as anal about this stuff as in past days. GuitarPCB.com has a modified Ross pc board that looks cool to me, I might save up some pennies to give it a try. In any event, I express my gratitude for the input and advice. Now I'm thinking that my build is probably all correct. Thanks all!

User avatar
phatt
Transistor Tuner
Information
Posts: 1224
Joined: 04 Aug 2010, 05:17
Location: Morayfield SE Qld AU
Has thanked: 8 times
Been thanked: 343 times

Post by phatt »

Hi DanB,
If you want low noise then you might be better served building something that does not use a TC chip in the signal path as the LM3080 is noise prone.

Go here for some clues on the pitfalls of comp circuits and some really good advice.
viewtopic.php?f=1&t=2086&hilit=lm3080+compressor
Down the bottom of page 2 Member *Tonetweaker* kindly uploaded the Telray comp circuit and Morley Pro they will give you ideas to build a far better circuit.
From memory these circuits use the LM3080 in the Feedback path hence the signal does not pass through the chip and noise is greatly reduced. I've never seen these so I can't give valid info but once you understand the finer details of comp circuits the Telray & Morley are likely a better mouse trap design.

*Mictester* was very helpful in setting me straight on the pitfalls of compressors. Well worth reading his observations on comp circuits. :secret:

I've used Marshall ED1, Boss CS2 and Boss CS3 and an optical Moen unit.
The CS3 is at least far less noisy (uses a fancy chip) But I was never happy with any of them. All the ones I've tried are to aggressive for want of a better explanation. And if you want the sustain then they tend to pump and sound wrong.

I now use Mictester's optical comp, a simple circuit that works for R/Roll guitar. Add to that it's low parts count (and only 2 knobs) and easy to build DIY.
Really Simple Compressor design by Mictester, Go here; viewtopic.php?f=13&t=8581

Be aware that everyone's idea and expectation of how they interact with the particular style you play will define the best circuit you need.
i.e. The boss CS3 is very good but it's very sterile for me.
While mictesters circuit actually adds a little grit which just adds some magic and feel to the interaction of the way I play.
Phil.

Post Reply