Big Muffarray

[digi2t]

This is another, "I shoulda built this a looooong time ago" pedals, but of course, I'm getting to the party late. As with my Superfuzz, lots of identical transistors screams "transistor array" in my head, so I went the Pickle route. I had to search hi n' lo for "just" the right array, and the MPQ2484 nails things down pretty well. The MPQ2483 sounds pretty good too, but it's a lower gain version, in between a 2N3904, and a 2N5088. The MPQ2484 gives me gains a tad above the 2N5088. In any case, it rips pretty good. Here's the verified vero;



The transistor and cap values are a mix of some Triangle, and Ram's Head values, and then tweaked to suit my ears. Using the collector/emitter resistor value calculation of the first three stages, it gives me a total gain value of 412, which is plenty for me, even without any tubes to push. I used two different values for the clipping caps, 0.05uF in the first section, followed by 0.15uF in the second section. I find that the clip frequencies are just right for me. The feedback sustain, even at bedroom volume level is fantastic. The diodes are unknowns. There's no numbers on them, only 3 color bands (white / violet / red or brown), and the letters "TSC" (Taiwan Semiconductor). I can't find any info on then, and even though they show typical forward voltages of .65 to .7, these just DO NOT perform like 1N4148's or 1N914's. I don't know what it is about them, but they add a more jagged edge to the sound. Maybe I'm hallucinating, maybe I chowed down on the locoweed, I don't know, but these sound different. I picked them up from a local guy, who got a whole whack of them from Northern Telecom. I'm still looking for data on them.

The voltages are as follows (E / B / C), going from input to output. Power supply = 8.95vdc, and I measure 8.80vdc after R24;
Q1; 39.3mV / 0.636v / 4.06v
Q2; 31.8mV / 0.627v / 4.06v
Q3; 29.7mV / 0.622v / 3.99v
Q4; 0.826v / 1.404v / 5.01v


I don't have the customary video done yet, but here are some teaser pics of the board for now.







I'm so enamoured with the resistor and cap values, that I've been also testing some other transistors as well on the breadboard. The usual suspects, like the 2N5088, and BC550C sound really good too, but I've found some lesser known (to me, anyway) transistors also rip and snort pretty well in here as well. The 2N3390, Russian KT342VM are very good, and one that really plays (and looks) the part as well, is the BCY59X. This one people, has grown on me. It's a metal can, like the BC108/9's, but on steroids. It's like a BC109, and a 2N5088, got together, and begat an evil redheaded child. I'm already planning to build a second one, right after this one, with the BCY59X's, but maybe slightly different transistor and cap values in the tone section. I like the tone section the way I have it arranged right now, it's got a healthy range, and the sustain is just off the map with what's on my board right now, but who knows what a little more tweaking will reveal.

If you want to build a BMP, but not use the regular, run-of-the-mill transistors, the aforementioned should make your, ahem... pickle crunchy.

I've been playing around with carbon comps, just to get that retro look, but I was short on 100K's, so some carbon films had to fill in. Just the same... Daddy likes.

[armdnrdy]

Well look at that!

It's Verofied.

That's really something! What...I don't know.  because without a schematic...it looks like a cave painting to me but...
could you have used CA3046 or CA3086?

[digi2t]

Here you go sir;



I forgot to add the 1N4001 protection diode. Sorry.

CA308* didn't really fit the bill. The differential pair is not useful here, which makes it 1 transistor short at the end of the day. MPQ2222, or MPQ3904 will work though. I think the MPQ3904 is what the Swollen Pickle used.

[pinkjimiphoton]

that is a thing o' beauty bro!

well done!!

i just built a bmp, too.... tweaked ram's head, may i humbly suggest you consider adding a midrange pot if you haven't already.

to my ear, and in my opinion it is a worthy mod, if you're still at a point you can try adding it.

can't wait for a clip!

[haveyouseenhim]

By having matched trannies, does it have a huge impact on the sound?

BTW, there's some idiot selling carbon comp resistors in the sale/trade section.

[petey twofinger]

your photos are outstanding there.

[Gus]

Something I posted in the past this might help with understanding what is happening
Note what happen with higher beta transistors with the BMP type biasing also compare grounded emitter Q2  to Q3
The 4th stage part of BMP is sensitive to transistor beta

With a high gain transistors you can design in more predictable transistor operating points

[tommycataus]

That looks spectacular, I guess it's another muff on the build list!!

By the way, those shielded cables you are using look much more user friendly than the ones I have. May I ask where you bought them?

[newperson]

thanks for the share.

[digi2t]

Gus;

As always, thanks for the sim. For a visual beast such as myself, it's great to have a glimpse at what's going from end to end.

Tommy;

The cable is RG-174, from China. They sell it for 7.50/30ft. on EBay. I do have higher quality Belden in my shop as well, but it's Teflon core, with an added foil shield, but it's a mofo to get around a tight enclosure. It's like trying to bend a tree branch.
This stuff's core is something akin to PE, so you really have to be quick on the solder gun, or the wire ends up melting out of it. It didn't take me long to develop a technique where I could solder it, without swearing. It's much more flexible than the Teflon core though, and doesn't carry the extra foil, but I find that the shielding is adequate enough for pedals. For an amp though, where a lot bigger signals are flying around, I would use the Belden, but then, space isn't usually a problem there.

I should have a video of this puppy in action this weekend. I've taken Brother Jimi's words to heart, and I've been testing a Flat/Scoop/Big Scoop switch configuration in the Duncan Tone graph thingy. Looks like a viable addition.

[Gus]

The 100K from base to ground are resistors that will adjust the bias of the first 3 stages.

An good experiment would be a fixed 47k or 68k (something under 100K) with a 100K adjustable resistor (0 to 100K) for 47K to 147K or 68K to 168K etc.
You can't make the 100K a low value because of loading the stage before
Try different hfe, beta transistors but adjust the collector voltages to be the same

The last stage needs to have selected transistors or adjusted biasing resistor values to set the collector current you want .  

[digi2t]

Finally boxed. I took the Photonic One's advice, and went with a Body control for the mids. I initially was leaning towards a switch, but the AMZ scheme works great. I used a 15nf cap, rather than a 12nf, for a little more girth in the middle. Duncan Tone Stack tool is invaluable for figuring how to balance things out.

Here's an updated vero, with the added Body control. Pics of the build as well, but video coming soon. Work has been busy of late.

[newperson]

Where did you/can you buy the MPQ2484?  Also what is this from?  Self designed from the BMP?

[digi2t]

Where did you/can you buy the MPQ2484?  Also what is this from?  Self designed from the BMP?

I got it directly from Central Semi. I put in a request for samples. Mouser has them as well, but for some reason, minimum purchase quantity is 220. I contacted Mouser, asking if I could purchase smaller quantities, but no reply to date. I'm still hunting for a cheap, reliable source for small quantities.

"Self designed"? Hmmm.... nope, I'm not that smart. Yes, it's a BMP. Thought I pointed that out in my original post. 

[stonerbox]

Very nice work. That thing looks so damn tasty! 

[IsaacMoth]

Dude that looks great!

[digi2t]

Thanks guys.

I ran into a problem that affects some BMP's, the tone pot "swish". My research on the problem enlightened me to the lack of coupling cap between the third stage transistor, and tone section. Funny that I didn't hear it on the breadboard, or before I boxed it. Oh well.

My question is this; I'm quite happy with the frequency response of the circuit as it is right now. I know I can add a cap between the collector, and the tone section, to kill the DC hitting the pot, but what value should I use? I've read two posts on this problem, one states to use a 10uF electro cap, and the other, a 1uF electro. All the while, my circuit uses 0.12uF between sections, and testing with a 0.12uF, it works fine. I'm just wondering what value would least affect the present performance. I don't want to thin it out, but I don't want it to get flabby either. I've been told that a bipolar is preferred for an audio application, so do I really need an electro here?

Opinions? Advice?

[digi2t]

OK, the swish is gone. I nice Panasonic 1uF metal film took care of things nicely. My vero however took a bit of a beating, since I had to add some more jumpers for the Body control, and the "de-swishing" cap. I'm going to update the veros to make them a bit neater, and to accommodate the Body control, and extra coupling cap, more elegantly. I'll post them in the next couple of days.

Anywho, here's the video of the build. Enjoy.

[Brymus]

Sounds good,some nice tones in that build

[digi2t]

Here is the final vero. This includes the Body control, and supplies an option to install a cap between the third transistor, and the tone section, to eliminate any noise from the Tone or Body pots..



That's it. Sticking a fork in this one. Glad I built it. It rawks!