Traced a Behringer UO-300, which is "supposed" to be a Boss OC-2 clone

[bushidov]

So, I've wanted to try my hands at a Boss OC-2. When trying to find schematics, I get some differences between versions. The original BOSS one has parts that are no longer available, and the one that substitutes in a 4027, I tried and get nothing to work. I didn't want to go down the path of the "chopped" version as I wanted both Octaves.

So, then I remembered Behringer always cloned BOSS stuff, so I found a used one for $20.00. Tested it out and it worked great. Tracked Bass really well, which I remember being an issue with the original OC-2. The Behringer UO-300 also has a DP3T switch that selects "Range" of Hi, Mid, and Low.

So, yes, I now have a pretty butchered up PCB, but I do believe I can put it back together. As of BOM, it used all surface mount parts, which made reading the JFETs and such a pain. For the JFETs, they were silk screened with "113", which I assume is J113, but when looking up the markings on the JFETs, they had "M6GN", which on sites, came up as MMBF4393LT1. They have similar stats to the J113, but they are not identical. The BJTs had "846" silk screened, which I guess is a BC846, which is basically a surface mount BC547. As the BJTs are used in the classic BOSS buffered bypass, I would guess that BC547's would work. For Diodes, except for the 1N5817 Schottky for power and the 3.9V Zener for the BOSS bypass LED, the rest were silk screened "4148", so I'm pretty sure those are 1N4148's. For the ICs, again, surface mount components. It used 3x Quad Op-Amps, all being ST Micro's version of the TL064. It also had a HEF4013BT and a HEF4027BT. Pots are all 100K linears.

Anyways, I am posting this here, primarily to understand the differences in this circuit vs the classic OC-2. It does the same thing, for the most part, but things like the 4013 on this one, none of the wiring matches the OC-2, the "Modded OC-2" I see in some forums, or the "Chopped OC-2". I suspect that's why the ones I tried from these previously named schematics never worked.

Anyways, let me know your thoughts. I'd be particularly curious to hear Mark Hammer's opinion on this, as he's a freaking super genius when it comes to this stuff, as well as PRR.

[Mark Hammer]

First, kudos on your diligence in tracing it out.  I rarely have that sort of patience.

I'm no super-genius whatsoever, but I do have opinions (and a rectum, which I understand is similar to an opinion with respect to its distribution across the overall population!).  And one of those opinions is that Behringer has made some poor choices in terms of op-amps.  In this instance, the choice of TL064 strikes me as unwise.  They're good on current consumption, but at the cost of hiss.  The Sharp IR9022, that the OC-2 uses, is also a low-current chip, but I have no idea what sort of noise specs it has, relative to a TL064.

From what I can tell, the UO-300 uses a similar "engine" as Boss to produce suboctaves, but many aspects of the audio path are different.  In particular, Boss does not employ the 3-position filter Behringer uses, and also opts for transistor buffers, rather than op-amps as Behringer does.

The switched-FET approach that both the OC-2 and UO-300 use is known to provide a more natural-sounding octave than the direct of a flip-flop triggering that something like the MXR Blue Box uses.

[ElectricDruid]

Wow. The first thing that strikes me is that as soon as people start designing for SMD components, the a mount of *stuff* goes up! Board space suddenly isn't such an issue, and so you add that extra X and include that optional Y. In some ways that eats into the benefits, but in other ways, it moves the benefits to audio quality or pedal flexibility over simplicity.

But wow...what a lot of stuff!

[bushidov]

Mark, if I were to rebuild this, what quad op amp would you use instead of the TL064?

[Rob Strand]

Anyways, I am posting this here, primarily to understand the differences in this circuit vs the classic OC-2. It does the same thing, for the most part, but things like the 4013 on this one, none of the wiring matches the OC-2, the "Modded OC-2" I see in some forums, or the "Chopped OC-2". I suspect that's why the ones I tried from these previously named schematics never worked.

For the first octave the main difference is the lack of the germanium diode from C1 to Vbias.  This shifts the waveform DC wise before doing the phase-inversions.  Also the value of R3 is large which makes a difference.   As drawn it is more like an Ampeg SubBlaster or EBS Octaver.  See this article,

Octave Article

There's also the variable filter.     The way you have drawn the filter is hard to read  .   It's a Sallen and Key type filter.

For the second octave the circuit here is quite different to any other octavers.  It's a very crude chopper where as the OC-2 uses a complete circuit very similar to the first stage octaver.   It's quite possible this crude method sounds OK after filtering.

As for the detector, U2 etc..  The detector is more like the EBS circuit than the Ampeg or Boss circuit.  One unique feature of the Boss detector is it had a fast roll-off low-pass filter before the detector.  This has been replaced by much simpler filtering in many of the later Octaver "copies".  The time constants for the peak detector are like the EBS Octave.  It also looks like the range switch is tied into the filtering C16, C17 etc.

Overall it looks more like an EBS octave copy than a Boss Octaver copy, except it has the crude extra octave added on.   I guess the two octaves is  why people think it's a Boss OC-2, clearly it is not.

[Mark Hammer]

Mark, if I were to rebuild this, what quad op amp would you use instead of the TL064?

Given that external power is the default these days, you don't need to use low current chips.  You're probably good with 4558 chips.

[bushidov]

Electric Druid, you called it. I do more work in industrial digital stuff (micro controllers, small embedded SBC stuffs), so my analog audio knowledge is pretty new, but in the SBC world, this is actually pretty light weight as far as part count is concerned. However, to compare it to some other classic guitar pedals, this seems a bit overkill.

Mark, I do intent to make a through-hole variant of this at some point, so I'll take your suggestion. I may use the 2058 (Quad version of the 4558) and see where it gets me. I'll probably use a rotary knob from Tayda to emulate Behringer's slide switch.

Rob Strand, I'll have to add you to the genius list. You are correct on what you said there. I was not aware of the EBS Octabass, but I do see the similarities now that I looked at it. Also, that Octave article is pretty educational. It might have been what got the Behringer guys to use the TL064's. Which part was the variable filter that I drew "hard to read"? Is that the part with the DT3P switch? If so, yeah, I was drawing as I was tracing, so that usually means it doesn't turn out pretty at first. I still need to clean that up.

[Rob Strand]

Also, that Octave article is pretty educational.

It's a good article it puts a lot of info in the one document.

It might have been what got the Behringer guys to use the TL064's

Behringer use that opamp in a lot of products.  Even in circuit positions where it is quite detrimental to the noise.  The main advantage is low power.  When you have a few Quad opamps in a circuit the power can go up.

Which part was the variable filter that I drew "hard to read"? Is that the part with the DT3P switch?

So for the low-pass filter, R10, C18, R16, R18, R19, C21, U1C, C16, C17 and the top part of SW1 are all part of the same filter.    The circuit should end-up looking like your filter R13 through to U1B.  If you have a look at the EBS Octabass circuit they have drawn it in a good way.

The second filter is R8, R38, R39, R40, C14, SW1 which go together with U5D, C5, R35, R37.   You should draw those together.  This circuit is also similar to the EBS Octabass.  Use their circuit as an example.   The difference is Behringer switch some of the part values.

You don't need to keep the two parts of SW1 together on your schematic.

I noticed C24 is 100n on the Behringer and 10n on the EBS.  Also Behringer changed R37 from 100k to 10k.    You might want check those values.

BTW,  thanks for posting the schematic.  It's interesting to know what these things really are.

[Rob Strand]

I don't know what happened this morning I must have got lost flicking around on the large scale schematic.

So for the low-pass filter, R10, C18, R16, R18, R19, C21, U1C, C16, C17 and the top part of SW1 are all part of the same filter.    The circuit should end-up looking like your filter R13 through to U1B.  If you have a look at the EBS Octabass circuit they have drawn it in a good way.

R10, C18, R16, R18, R19, C21, U1C, *C19*, *C20*

The second filter is R8, R38, R39, R40, C14, SW1 which go together with U5D, C5, R35, R37.   You should draw those together.  This circuit is also similar to the EBS Octabass.  Use their circuit as an example.   The difference is Behringer switch some of the part values.

R8, R38, R39, R40, C14, SW1, *C16*, *C17*

[bushidov]

Thanks Rob. I am going to try to redraw that. Might take a bit. In doing the tracing, I lost a capacitor, all the diodes, and one of the JFETs, so I have ordered them off Tayda and will be resoldering the Behringer back together. That should bring me back to a functioning pedal. From there, I'll try to redo it as a through hole, true-bypass setup. I can do the slide switch they have with a DP3T rotary switch I ordered from Tayda as well.

Before I tore it apart, I had a lot of fun with the UO300, so I wanted to see if a "DIY" version could be makeable.

[bool]

At a quick glance, you may get a liiittle better noise spec if you replace the U1 (a,b,c,d) with a lower noise part. A TL074?

[Rob Strand]

I lost a capacitor, all the diodes, and one of the JFETs, so I have ordered them off Tayda and will be resoldering the Behringer back together. That should bring me back to a functioning pedal.

I admire people that go that extra 1 mile (on in this case 10 miles) by stripping back to those SMD boards like that and measuring the parts!  Not many people have that sort of commitment to the cause 

I try to use more of a Ninja method. I measure voltages with a sine test signal and estimate the part values from the measurements.  It takes quite a bit of care to come up with the right numbers instead of just loose estimates.

[bushidov]

Hi Rob,

I am trying your suggestion as to re-organize the schematic. However, I am not understanding what you are saying. I tried moving that low pass filter over, but it is not looking like U1B. Do you have EAGLE? If so, I'll send you the file to show me what it should look like.

[Rob Strand]

I don't have eagle.  I've patched together these sections.  I've tried to draw it so the high setting is with the switch in the up position.  Funny thing at first glance the circuit around U5D looks like the switch is reversed since the filter cut-off is dropped in high mode and lowered in low mode.  Maybe you can re-check it.



Edit: On SW1b the Mid should connect to the switch pole (like SW1a):

[bushidov]

Hi Rob, I updated the schematic to look the way you pointed out and the EAGLE layout remained, so everything appears to be good. Thanks for your help on that, because I would have never have guessed that in a million years, but now that I look at it, it does make more sense.



I did check values for C24 and R37. They are what I listed them as: 100nF and 10K respectively. Below is the PCB. Behringer was nice enough to label these values:



I desoldered the cap, just to be sure, but it is indeed 100nF. R37 I measured on the board. It too matches its silk screened values.

Anything else, I should check?

[Rob Strand]

I desoldered the cap, just to be sure, but it is indeed 100nF. R37 I measured on the board. It too matches its silk screened values.

Looks good.

Well from that I'd believe your values.   In some cases Behringer don't copy the original values.  Who knows there may be more than one EBS circuit out there, or, the EBS schematic doesn't represent reality in the first place.  I haven't verified the EBS values against a PCB.

Anything else, I should check?

On the pic I can only see the R8 connection to the switch.  To me it looks like R8 connects in the up (HIGH) position.    Maybe both SW1B and SW1B need the HIGH and LOW swapped?   That actually makes more sense to me circuit-wise.

There's two wires missing: base of Q6 to C30/R51 and base of Q5 to C31/R52.

[bushidov]

Crap, there goes my accuracy reputation. 

Rob, good eye! I see how I missed those traces on the stomp switch. The vias were underneath the transistors. But they ringed out on the multimeter. I should have been able to see them on the backside of the board, as they were there. I just missed them. I think it was because I saw that it was already forming the traditional BOSS stomp-switch circuit, and I stopped looking at that part of the board as meticulous as I was looking at the rest. My bad on that one, for sure.

As of the switch, again, I was bone headed on that. You are correct, the traces are correct, but the labelling was incorrect. Behringer's slide switch has a weird pinout which makes it counter intuitive. That probably contributed to me goofing the labels. But now that I have the board back together, I was able to ring out with the multimeter what things were going to the common pins wherever the slide switch was at, and indeed, the schematic needed relabeled.

Re-posting with the corrections:


So... what else did I miss? I am really curious, because the schematics for the EBS I was looking at didn't match this very well.

[Rob Strand]

Crap, there goes my accuracy reputation. 

You did a good job.   It's extremely hard to get it 100% right, especially when working alone.  I've probably traced more than 100 circuits.  No matter how hard you try little errors sneak in.  I usually have to leave it for a couple of days then check it all again - blows out the time a bit.  If I leave it for months or years then it's a lot more work

So... what else did I miss? I am really curious, because the schematics for the EBS I was looking at didn't match this very well.

It looks pretty good as it stands.  It won't  be a 100% copy.  That's what I've noticed about the Behringer pedals.  The small differences is where there's likely to be an error.  I checked a few places like that and your schematic seems to match-up with the board.