MXR Blue Box design issues

[corythosaurus]

Hey everyone, first of all: I'm a longtime lurker, first time poster. Thanks to everyone here for making an awesome forum.

Anyway, I'm planning on doing a version of the MXR Blue Box (with some mods on the output, etc.). After spending some time working through the circuit, there is some strangeness to the design, and I'd like to appeal to the collective wisdom of other DIYers for your thoughts on it. I've attached the schematic from Tonepad below, which seems to be the only complete schematic out there. Here are some of the issues I see:

• C5 is a polarized electrolytic cap (10uF). The negative end is connected to 4.5V (Vb), and the positive end is connected to the feedback network on IC1a. Now, the R5/R8 network is not biased (which could be  its own issue) but I would imagine it's somewhere around 4.5V. I'm surprised that this isn't blowing up people's electrolytic caps, but maybe it's causing an interesting change in the sound? Either way, I usually try to avoid doing this to an electrolytic, and might connect the negative end of C5 to GND instead of 4.5V. Does anyone have any advice on this?

• Second, there will be a DC current through R23-R6-R7-R9 from 4.5V to GND. I can imagine all sorts of weirdness that this could cause, though perhaps it does nothing of the sort. For one, because of the resistor values (R23=10k, R6=1M, R7=56k, R9=10k), pin 7 of IC1b is going to be at 0.3V, so any large negative swing will hit the GND rail. Is this part of the distortion sound? I was under the impression that the fuzz/octave came from Q1/2/3 and IC2. This DC current could also cause some oscillation/other weirdness with IC1b. I feel like I should put a coupling cap in between R7 and the junction of IC1b pin 7 and R6.

Has anyone out there seen any issues with this, or on similar pedals or gain stages? I was hoping to skip breadboarding this and going straight to a PCB, but I don't want to "fix" some element of it that's crucial to the sound. Am I just missing something here that makes this whole circuit OK? From searching the forum it looks like people have had success with this schem, but that doesn't mean that there's no room for improvement.

Thanks so much!

[MaxPower]

I'm still teaching myself electronics. Anyway, as far as C5 goes, remember voltage is relative. I mean, the 4.5 volts (vb) is 4.5 volts relative to ground. Relative to +9V it becomes -4.5V. So if the other end of the cap is greater than 4.5V, it makes the 4.5V end negative relative to the higher (positive) side. So it shouldn't be a problem.

Or maybe the potential differences in voltages at both ends of the cap are not big enough to damage the cap regardless of polarity?

As I said, I'm still learning so I may well be wrong but that would be my guess.

You might want to give Tina-ti a try. It's very easy to run dc and ac simulations to get an idea about what voltages and currents to expect. LTSpice seems to be more popular but I found Tina-ti to be easier to figure out.

[PRR]

> C5 is a polarized electrolytic cap (10uF).

Zero-volt operation doesn't hurt (much).

More than 1V *reverse* does hurt.

Old-old electrolytics did decay in zero-volt operation. Modern ones last decades that way.

That string is "biased" by the opamp output which should be +4.5V--- actually it will follow the bias on the other input of that opamp.

> DC current through R23-R6-R7-R9 from 4.5V to GND

So? Opamp can deliver current and still do its job. (Or: delivering current is a job requirement.)

> pin 7 of IC1b is going to be at 0.3V

No. The only way the opamp can be happy is by swinging its output to 4.5V. That's the only way both of its input pins are the same voltage. It will hold 4.5V until output loading exceeds 20mA. 4.5V/0.020A= 200-some Ohms. Since the load is K-Ohms, it will do its job. Output at 4.5V. Junction of 56K+10K UN-loaded is 0.68V. With Q1 there, the node is probably clamped closer to 0.6V and Q1's collector is slammed low. Without part-values, I can only guess that Q2 collector is slammed high. However a few milli-volt change at Q1 base (a dozen mV change at opamp output) will throw both the other way. Opamp has gain of 100 so there WILL be LARGE audio swings. Q1 Q2 further hammer the audio into crisp rectangle-waves. These may be sent to the output; they also go to flip-flops for further manglement (I suspect sub-octaves?) for output. Either way is too nasty for ears, so C9 softens the rasp.

[corythosaurus]

>> pin 7 of IC1b is going to be at 0.3V

>No. The only way the opamp can be happy is by swinging its output to 4.5V. That's the only way both of its input pins are the same voltage. It will hold 4.5V until output loading exceeds 20mA.

Oy, that's one of those "I should've known that!" answers. Thanks so much for the help. I'm so used to slapping a coupling cap after every op-amp stage that I didn't stop to work through the actual behavior.

Regarding C5, that polarized cap, I suppose I might as well connect it to GND since I'm building one from scratch.

Thanks for working through the transistor section, too. And yes, that flip flop is supposed to make sub-octaves. If you're interested, the tonepad .pdf has all of the component values available here: http://www.tonepad.com/getFileInfo.asp?id=73