Messages

OR is there a way to bypass the 2nd Minibooster altogether?

This is where most of the drive comes from (the first stage boosts the signal which clips the 2nd stage) so if you bypassed the 2nd stage it probably would tame things a lot more than you want.

I would say biasing is key. Try raising the 2nd stage resistor to ground from 120R to 220R or even higher.

Also make sure they're genuine/known-good 5457s... the type used to develop the original would have a Vgs(off) in the 1.2 to 1.4 range, but if they're not Fairchild/ONSemi brand (or if they are fakes) then they could be wildly different.

Maybe you can try to add a sort of Pre-Gain control, a simple pot in front of the circuit.

The gain control is a volume control after the first stage (which shouldn't really distort on its own) so this probably wouldn't do much more than allow fine-tuning of the gain range - e.g. setting the pregain control to a place where the maximum gain setting is good. Better to try biasing it differently and only try this if tweaking the bias is a dead end.

Ideally you want both the correct ratios (Q2 50-75% higher than Q1) and the trimmers to fine tune it. The trimmers don't compensate for the difference in hFE, that's a  completely separate thing.

One other possible reason. I read recently about the different ways solder mask color is achieved and there was a claim that the materials used for black could become conductive in the wrong situation. I don't know the truth of that but I personally have had a couple cases where I picked up voltages on boards where none should exist. And this has only happened in the last year or so. I can't attribute it to anything in particular but it's for sure caused me to rethink how I have the boards made.

I've heard this as well, maybe from the same articles even... but assuming there's any basis in fact, I would submit that it's more than likely something that happened very early on when black soldermask was first developed and has since been resolved as part of the developmental leaps in PCB manufacturing technology (skyrocketing quality and plummeting costs). Oftentimes those cautionary stories survive long after the need for caution has disappeared.

I see black used all the time in high-end computer motherboards, GPUs, etc. where the slightest bit of soldermask conductivity would cause huge performance issues—applications that are exponentially more complex than our circuits whose technology stopped being leading-edge in 1977!

Charge pumps... TC1044SCPA has dried up and Mouser's expected restock is in January 2023!

MAX1044 and ICL7660 are a little easier to find, but quantities are low and dropping. Fortunately there seem to be plenty of LT1054s, though they're a bit more expensive.

Does it happen if you have a buffered pedal in front? High-impedance signals (e.g. direct from passive guitar) are much, much more susceptible to picking up LFO noise, and it almost always happens on the input since most of the internals and the output signal coming from the pedal are low impedance and therefore better able to reject that type of noise.

I've found that many/most instances of ticking will go away if the input signal has been buffered in advance - which is going to be the normal real-world use case most of the time, but not usually done when testing a build.

You can change the values with no I'll effects as long as you have at least C1 with a value in the uf range.  You will commonly see a 47u electrolytic and 100pf mlcc for C1 and 2. You see multiple caps of the same value in parallel alot too. For better performance, It is also recommended to use low-esr electrolytics if possible.

I don't think I've ever seen a pf value for C2, it's usually either 100n or 10n... my understanding is that it's mainly just because ceramic provides better filtering at certain frequencies, so using an "average value" ceramic in conjunction with a large electrolytic, you get the best of both types. I'm not sure a 100pf capacitor would contribute much to power filtering. More often you see them on the signal input to prevent radio frequencies.

This is mostly just because of the way I drew it. Look at the top one as the gate-source cap for the upper JFET (present in all mu-amp topologies) and the bottom one as the drain-gate feedback cap of the lower JFET that cuts highs. The lower one in each stage also has a series resistor to reduce its effect. Almost all of the CB mu-amp pedals have the same setup for their tube emulation stages, but usually they're lower-value film caps so it's more obvious what they're each doing.

Intensity! I'll take it

The SML , SM9 is not the same as a Metal Charger it has a transistor output buffer. It is the same as a SM7 and there's a schematic http://www.freeinfosociety.com/media.php?id=4071  I know it's an old topic but some people do search like me.

Btw, this is super-duper not true. The SM7 has nothing to do with the earlier Super Metal pedals. The SM-9, SML, MS-10 are all the same circuit, although IC3 and IC4 are sometimes different (either 4558 or 1458 depending on the model). The MS-10 may use a BJT instead of JFET for the input buffer, but this could be a tracing error. Otherwise—notwithstanding that some of the traces have obvious errors that are easily corrected when comparing other schematics—the three pedals are identical inside, even the part values.

The Maxon SM-01 appears to be the same as well, same PCB layout as the Ibanez SM-9. I can't confirm the values but there's no reason to think it's any different.

The Maxon SM-9 reissue from a few years ago is different, however. It only has 4 knobs, replacing "Attack" and "Punch" with one called "Scoop", and it has an optional 18V charge pump. I haven't seen a trace of this one, but just from appearance we can tell it's not an accurate reissue.

I guess we just need to give Steve a call and ask for source of the parts. In case in the future, we will order the parts. At least we know where to go. But again, if he's going to provide us with some directions.

Someone on FB emailed Steve to ask about sources and he had this response:

"The vendor information and part numbers are part of my intellectual capital. What I suggest to customers who are looking for a new source is that they do the homework I did."

I was already thinking about asking about their source for pot dust covers since I use them for all my kits - they fit better than Tayda's. So this was pretty disappointing to see, since he hasn't given any other indication that he plans on capitalizing on his intellectual property (i.e. selling the business including supplier lists). Based on the information he's shared, the inference is that he's just going to leave customers stranded if they were dependent on parts that only SBE sells.

I hope it's just bad miscommunication on his part and there's more news to come. But SBE's relationship with the DIY community goes both ways. He's been a great part of it, legendary even, but we've also given him his job for 20 years, so this is not a great way to close out an otherwise impeccable career.

Yeah, the charge pump is only for PNP transistors while NPN transistors need the charge pump omitted (and capacitor polarity reversed in a few places - the default orientation on the PCB is for positive ground).

Hi Aion!
What's the latest on the Roland Jet Phaser PCB???
Woo hoo!

Ran into a PCB layout issue and so it got bumped to the bottom of the stack for awhile.

It's worth a try. The LEDs seem the most likely culprit. The LFO has a fixed range of modulation (no biasing) and is attempting to modulate the LED brightness within a given range, impacted by the Span pot. If the LEDs are more efficient or sensitive then it seems they'd be more likely to stay on at a certain point when the fixed-range LFO expects that they'd still be modulating.

I didn't have any of these issues in my prototype either which used 2k pots, but I'm going to change the docs so 2k is no longer the default and only mentioned as a possible mod with caveats since it's happened more than once. I would put a 2k or 2k2 resistor across the lugs of the span pots (can solder from the top of the PCB without disassembling) to drop the value back down to the stock 1k. You could try increasing the parallel resistors to 3k9 or 4k7 to get a little more range without getting into lockup territory.

It's a pretty terrible pedal stock, but an interesting concept and something that should have potential. I've thought about doing one just to put it out there in hopes that the community will be inspired to come up with some mods to make it great.