25-Year Pollinator (One Control Silver Bee HamishR Mod)

[MattG]

@bluedmc777's build report, Sheepy Love Bee Sting (One Control Silver Bee) brought this circuit to my attention. After reading the build report, I immediately went to Sheepy Love's Bee Sting Product Page so I could look at the schematic. I initially missed the numerous references across the main Sheepy Love site and the build doc that says this is a "Mods by HamishR" circuit. Most of the Bjorn Juhl overdrive circuits I've looked at use the wacky CA3130EZ single opamp plus a JFET for gain recovery. So looking at the Bee Sting schematic, I saw it uses a good ol' TL072. Still under the mistaken impression that this was a straight trace of the One Control Silver Bee, I thought perhaps Bjorn had abandoned the old architecture in this newer design.

So I copied the circuit into my own PCB. I'm working on a series of builds now to have explicit hardware references for my microcontroller bypass firmware. So with the PCB already designed and submitted to JLCPCB, I randomly took a break at work and was looking at threads on freestompboxes, when I ran across this thread: Bearfoot - Silver Bee Overdrive [traced]. This has the actual trace of the Silver Bee Overdrive, and @HamishR's commentary on the mods he did to the circuit! It was only then I realized the circuit I used wasn't the "real" Silver Bee, but the HamishR modded Silver Bee. I took that as a positive though, as I've enjoyed a lot of his mods (particularly when he collaborates with @Chuck D. Bones). In fact, his Old Fashioned has been a mainstay on my board since I built it.

So anyway...

Lol why is it unremarkable when a build goes as planned?

This one's for you, Don!

After building it up, on first power-on, bypass worked fine, but I got nothing when I engaged the circuit. I first suspected the bypass scheme, since this was my first build with this particular scheme, and first live test of the MCU firmware. But I triple-checked the bypass wiring, verified switching voltages, and manually ran through the switching logic table for the 100th time. I felt pretty confident the switching was right. More on that later.

So I busted out the audio probe and found everything seemed right until the last coupling cap between the effect's last opamp output and the volume control. This is a 1uF film cap. It's certainly possible that I got a bad film cap, but it seemed unlikely. But, I went ahead and un-soldered the cap. Before soldering in a new one, I checked continuity between the cap's solder pads and the neighboring components: one side in fact did not have continuity. I didn't scratch the PCB; and it's a straight and very short trace (about 1cm). I tested the same trace on one of my unpopulated boards - it was fine there. So I'm not sure if I somehow damaged the trace during initial assembly, or if this one board happens to have a manufacturing defect. Either way - I replaced the cap, and left a little extra lead length to use as a fix for the apparently bad trace. See the picture below.

Once I got that sorted out, it worked. But the next problem: the switching made a pretty awful "pop" whenever I changed between engage or bypass. Which is the exact opposite of what it should do: in my past usage of the CD4053 analog switch, I used a simple scheme, which sometimes did have a bit of pop. But in a DM conversation with @jubal81, he suggested a smarter way to wire the CD4053, which allows for an explicit mute state. Thus the circuit can be muted, then switched, then un-muted, resulting in pop/click free switching.

I spent a lot of time reviewing the schematic. First thing I noticed: I had a pulldown to ground resistor on the same net as where I had a pullup to VREF resistor. This wasn't a major problem, but would affect the bias at that stage, reduces headroom, and wastes current. So I pulled that resistor. It didn't fix the popping.

I finally realized: my mistake was forgetting that, within the CD4053 sub-circuit, for the actual audio signals, "ground" is VREF. I was mistakenly using power ground (i.e. 0v plane) for muting. The dramatic voltage jump between 0v and 4.5v (VREF) was causing the popping. Rather than try to cut PCB traces, I just lifted the mis-routed 4053 pins out of the socket, and used a wire to tie them to VREF. Now it has truly pop-free switching!

After all that rigmarole, it turns out I'm not terribly hot on the circuit. Not that it's bad by any means! It's a great-sounding circuit. But I prefer this Barber Silver LTD clone I've been rocking for a few weeks now. Comparing the two, the Silver Bee feels a bit "softer", while the LTD is "crisper". Again - not bad, this just speaks more to my preferences than the circuit quality. The next guy might have the exact opposite opinion.

The enclosure actually came out pretty nice. I recycled the graphic from my Aion Procyon, this time on a Tayda "Ball Silver" enclosure. (Ball Silver for the Silver Bee circuit, get it?)

So, despite not being thrilled with the effect itself, it was a good learning experience, and ultimately does prove out one aspect of my MCU firmware / bypass design scheme.

 

[HamishR]

Thanks for writing such a detailed summery of your build! I agree that the Bee Sting is a kinda specialised OD - It's quite hifi sounding and no MIAB. Initially it was simply an experiment in seeing what would happen if I used a dual opamp to supply both gain stages rather than a single opamp and a transistor. It can indeed change the sound significantly. But it does have a sound I can use. I like both the Silver Bee and Bee Sting more than a Honey Bee - although I'm sure a lot of players would love the HB.

I also tried building a Rat in the same way - using a TL072 instead of the IC + transistor. As you might expect, the result sounds quite different from a Rat because the particular JFET and IC in the Rat are a big part of the sound. The TL072 version is I guess maybe a more "anonymous" sounding circuit but still very usable MHO. I built it with a bass control and it has a more open sound than a Rat.

 

[MattG]

So I spent some time reviewing the original circuit schematic, and one thing that jumped out at me was that it uses red LEDs for soft-clipping (versus the BAT41s that the modded version uses). I suspected the red LED clippers might give me more of the “crispness” I was looking for. So I modded (or un-modded depending on how you look at it) my build accordingly.

The result was as expected. I definitely like this circuit a lot more with the LED clippers in the opamp feedback loop. Now, comparing this to the Barber Silver LTD, it feels like I like one better for some things, and the other one for other things.

I need more time to compare, but I feel like the Silver Bee might be a little more touch-sensitive. I know I tend to be heavy handed when I play, but I found myself using a more relaxed approach with this circuit, to take advantage of its dynamics.

It might just be confirmation bias, but I always feel like all the Bjorn Juhl circuits have a distinctly nice feel to them, and I’ve always suspected it’s due to the fancy feedback scheme that so many of his designs use.

Now I’m contemplating whether swapping the 1n400x hard clippers for higher Vf diodes (e.g. good ol’ 1n4148) might further move the needle in my preferred direction.