Gate/Swell Harley Benton NG-100 analysis/trace

[lars-musik]

Some time ago, I traced this cheap pedal because it is the only Slow-Gear/Gator swell circuit that really worked for me. Back then I posted it over at the other forum but the feedback was very sparse.

I would really like to rebuild this - if possible at all in a 1590a. However, my knowledge of electrical things is very limited (I don't even know if the schematic I deduced from my tracing efforts is readable to the professionals) so I am unable to spot:
a) any mistakes
b) superfluous parts (I'd rather save space and use true bypass).

Is somebody willing to lend me an eye and analyse this circuit with me (or rather for me)?

Maybe (most likely considering the price) this is a knock-off of another pedal and a real and true schematic already exists?

The possible and intended outcome will be a fine swell pedal....

Thanks in advance!














Schematic pdf:

https://www.dropbox.com/s/3myw4sk6kxeemyl/NG-100_v0-1.pdf?dl=0

[Kipper4]

I'm watching this. I'd love me a good swell.
Ill pm you lars

[samhay]

Is the schematic difficult to follow - yes.
Does the schematic have errors - probably.
Can it be simplified - certainly. Most/all of the FET switching can likely be done away with.
Will it fit in a 1590A - I'm sure Russ could make it fit, but most people would likely see that as being unnecessarily challenging.

What type of OTA is it? I'm guessing 3080 if it's 8 pin, but I take it the name has been sanded off?

I don't know much about Harley Benton, but if they are anything like Joyo, Behringer, etc, then they didn't originally develop this circuit. If you can find what it is based/cloned from, then you will likely find more info on the circuit.

[lars-musik]

Is the schematic difficult to follow - yes.

It is my first (in my real life I am a biologist)

Does the schematic have errors - probably.

I was hoping someone would point them out

Can it be simplified - certainly. Most/all of the FET switching can likely be done away with.

My hope was that a knowledgable person would take the red pen and would - ah -do it.

1590a

The 1590a challenge will be my pleasure.

What type of OTA is it? I'm guessing 3080

Yes, it is. Not sanded. Obviously I didn't put the designator to that part.

If you can find what it is based/cloned from....

Well, I did search quite a lot before I started that (ridiculous?) tracing endavour. Maybe I didn't find the right keywords, but "gate+3080" or "noise gate + 3080+LM358" etc didn't offer the answer. I again hoped (I am doing that a lot lately) somebody from the forum might recognize parts from the circuit and tells me what it is.


But be it as it is, Kipper4 kindly offered assistance and I'll happily share my Diptrace files woth anyone who is also obsessed with these swell/backwards pedals and wants to join the effort.

[samhay]

I think the problem you are having is that nobody else wants to put the work into tracing a £18 pedal, myself included.

However, there is interest in DIY swell effects. Do you use this as a noise gate or a swell and which effect are you looking to replicate? If it is the latter, do you mind if it is not a clone, but a work-alike (of sorts)?

[PRR]

30-second check of PDF:

There is surely a ground where C16 and R3 come together. Not on the schematic, and I do not see it on your PCB, though it is clear on the photographs of the original.

> I don't even know if the schematic I deduced ....is readable to the professionals

No longer a professional, but I've read a few schematics.

This is an EXCELLENT back-draw of a way-complicated board. I don't think I would attempt it.

But I found an error, and that means it needs more checking, and I would re-draw it for clarity first. Following signal/power in first-drafts is often like reading the number on a stunt-plane doing loop-de-loops.

After filling a whole poster-paper with scribbles, think how to organize it. One tradition is signal flows left to right, power flows top to bottom. Your signal flows center-bottom to dead-center... I had a hard time finding IN and OUT. This plan has only a few explicit power points, but they are essential for function and should "read well at a glance" for second-checking.

Such conventions will also *usually* put transistors "the right way up", which aids dim minds in understanding what it does (or spotting typos).

If Q1 were laid-out signal left/right (it nearly is) and power top/bottom, the missing power ground at Q1's load R3 might stand-out as a loose end.

Q5 Q6 surely form a *symmetrical* flip-flop. See most BOSS pedals. I suspect you could steal a BOSS plan and change any different values. BOSS usually lays-out this section pretty clear.

Likewise Q2 Q3 form a common SPDT FET switch. Steal this layout from BOSS. As the "direct" path is minor, focus on getting the Effect path laid-out pretty/clear, and let the direct bypass line run long or be two labels (like you did for Va).

U3 U4 U5 have separate power connection details lower-right (good), but U2 has power pins with the rest of the chip, and if U1 gets power I do not see where.

I do not see U2 type-number on the PDF. In this case I can guess; and it may be documented elsewhere.

For checking/confirming, it helps to have a sense of "signal flow" and circuit function. This plan IS complex, I agree. I see some of it which may help you. Audio of course flows U1.2 U2 U1.1 and out. "Control" taps off U1.2 to U5.2 to Q7 to U5.1 to U4.2 to Q8 to U4.1 and spanks U2; tapped at U4.1 to spank an LED. If you trace this on your sketch it is a maze; for study (and checking) a straightforward layout would be less eye-twisting. As this path is longer than the audio path, I would draw it out margin-to-margin. Put the audio and switching above that. Power and MISC like LED below that. Most interconnects between these three rows would be labels; long twisty lines are not eye-friendly.

> take the red pen and would - ah -do it.

Hard work. Not inspired.

I too have "gotten into" replicating 20-buck devices. But this one isn't my thing.

Pink-pencil markup:

http://oi60.tinypic.com/4j52zd.jpg

[lars-musik]

@Sam:

I think the problem you are having is that nobody else wants to put the work into tracing a £18 pedal, myself included.

Yes, I can absolutely see that and to be honest: it really felt stupid to put that effort and time into that project. On the other hand: Next time when I dare to open it, I'll take a picture of my "swell circuit corpses" drawer. A catastrophe - about 4 different versions of the HUGE EHX Attack/Decay (I finally got one working with a pcb layout by armdnrdy), different Paia Gators (Gaussmarkov and own layout) and finally the Slow Gear (here).

I won't go into detail, but the "perfect" swell circuit wasn't amongst those. So I was really pleased when I discovered that this super cheapo gate in fact IS the perfect swell circuit.

What to do? I know that the interest for a DIY swell is evident. So. I was really astonished that I couldn't identify the original design. Plus: I have the obsession to build myself a 1590a board. So tracing it must be.

However, there is interest in DIY swell effects. Do you use this as a noise gate or a swell and which effect are you looking to replicate? If it is the latter, do you mind if it is not a clone, but a work-alike (of sorts)?

I want only the swell function. And NO! I wouldn't mind a work-alike. In fact I would love one that works. If there exists another one besides the above mentioned I would build this one too.

______________________

@Paul: Thanks for helping out!  I really appreciate the "30 seconds" you already invested - although I guess it was a bit longer. The points you make are really helpful and need more studying when it is not late in the evening here.

I discovered that some schematics are pretty easy to follow (most of Madbeans for example) and some are not (I am working with a Decimator schematic for another 1590a madness and that one is an pain in the neck).

I was astonished that even when being aware of the conventions (eg signal left to right) it becomes trickier with every part that pops up to stick to these conventions. The result was a schematic that ultimately didn't follow the rules as you can see.

But you gave a lot of precious hints. I will try to abide by them and hopefully come up with a better readable version (if Rich doesn't do it quicker...)

I dread checking all these boss schematic and deduce which parts belong to the switching circuit. If I could rid of these parts before re-drawing I guess the following of the signal path would be much easier for me.




One fundamental question leaves me baffled: Why has Harley Benton such a cheap pedal (that I strongly suspect to be a clone) that works in this aspect so good? What is the original and is a schematic available?

[PRR]

> which parts belong to the switching circuit.

The Technology of JFET Switching in Boss and Ibanez Pedals

[samhay]

There are plenty of reasons for tracing a circuit. If it's a cheap circuit, but you want to make modifications/miniaturise it then I don't see any reason not to spend the time to trace it.
If you say that it does a good swell effect, then you have piqued my interest.

How old is the circuit you traced. Seems odd that you can by a new $20 circuit now that is all through hole components.

As far as trying to find where this circuit came from originally, it is often better to search for circuits with the same controls rather than circuit elements.
First hit is the Carlsbro noise gate - http://www.carlsbro.com/product-view/noise-gate/27/
Probably the same circuit, but doesn't look like these guys developed it either.

[lars-musik]

You are right, I was astonished to find an all through-hole pedal with a CA3080 OTA for that money. But it is new and you can still buy them.

Funny enough, there are a lot of these pedals with different labels out there. Most probably all from a factory somewhere in China, but certainly they are not the inventors of this circuit. Look here:

http://www.effectsdatabase.com/model/bostonengineering/100/ng100

http://www.effectsdatabase.com/model/legacy/ng100

http://www.amazon.com/Electric-Guitar-Noise-Effects-Pedal/dp/B0013NDUXW

http://www.effectsdatabase.com/model/groovin/ng800

But to none of those a schematic is available.

Very good that I piqued your interest, too. Your support will certainly help.

To get rid of the fet switch, I found this http://www.electrosmash.com/boss-ce-2-analysis#link3 with a circuit analyis of Boss pedals and their switching schem. Hopefully I'll be able to apply this on my schematic.