A build thread.

[Paul_C]

I trust this will be ok, and maybe entertaining for some.

I'm going to try and record my progress as I build a Parallel Universe, based on a layout which is freely available on Parasit Studio's website.

As far as I'm aware it's not a commercially available product, and I have no intention of starting a business selling pedals, so hopefully I'm not breaking any rules by doing this.

I've accumulated the necessary parts (but used one of the pots in another build so I've got some more on order) and drawn out the layout on paper (for some reason I prefer to do this rather than take a laptop out to my workshop) so I'm ready to make a start.

All being well I shall make some progress tomorrow and take a few photos as I go, and I'll document any problems (or mistakes on my part) if they occur - hopefully this thread will help me stay focused

[BetterOffShred]

It's the PU2 probably Eh?

It's Colin's design but I'm sure he doesn't care.

Try multiple 386s, the oscillation is pretty touchy and will be different with each

[Paul_C]

Yes, PU2.

Trying multiple ICs sounds like a gateway drug to "having to" build more pedals to use them up

[BetterOffShred]

Well the 386 is very unique, though it's made in 3 voltage ranges, and all 3 sounded different in my build, and even among each range, different chips sounded different. 386 is used in a lot of things as a gain stage so it's worth having a bunch.

It's a good pedal but you may want to look at what dino did with a load emulator using a 018 transformer so it works right on single coils. It's been years since I built mine, and I have mostly humbucker guitars so I didn't do it to mine.

[Paul_C]

First steps taken tonight - I trimmed the board to size (I have a handy band saw which makes this easy), marked the position of the cuts and drilled them followed by tidying the rough edges with a scalpel (the drill bit was a bit on the large side, if I'm honest, I'll use a slightly smaller one next time), then soldered in the links and the IC sockets (which took longer than it should because I had to take the sockets off again for a couple of missed links and then put them back on - sigh).

No transistors in this build so next up are the resistors, caps and one diode, which I'll install tomorrow evening.

Here are some slightly blurry pics:

[Paul_C]

Finally got back to this after some more transistor swapping has made my Random Number Generator sound more like it's supposed to.

This evening I finished populating the board, with only one error - I put a 47k resistor where a 4.7k resistor should have been - and that was quickly sorted.

Next I've got all the outboard stuff, pots, jacks etc. to attach so I can plug it in for a (hopefully) quick test before preparing the enclosure.

[eatyourguitar]

when collin did it with PCB mounted jacks that was pretty much the best way to keep it from having a lot of problems. I built a PCB with a proper ground plane but flying wires everywhere and it was shit. I already posted my rant about why I don't like it so I will not post the complete explanation twice. all I will say is that it can be fun but actually owning it and trying to tame the beast is not as easy as you might think. it is definitely not something to take on tour.

[Paul_C]

I'll have a look for the rant later

I've got it completed and it makes some sound, but nothing like enough - I've since read that it can be very sensitive to which LM386 is used, so I've ordered a few in the hope that the issue is that simple, but now I know there's a rant I'm beginning to think it might be a while until it's behaving

I built a Random Number Generator which I'm still trying different transistors in as it seems like it's another one where everything has to be just right for it to work optimally - luckily at this point I'm still considering these things as ideal ways of learning rather than reasons to give up building and finding another hobby

A friend asked if I was thinking of making pedals for sale and I quickly let him know that I'm not going to spoil the fun by doing something as daft as that.

[eatyourguitar]

the LM386 will put noise on to your power rails even if you use the right LM386. the TL072 picks up the noise. the high current draw of the LM386 makes it into some kind of RF pollution thing that then gets picked up at the inputs of both the LM386 AND the inputs of the TL072 through the induction of parallel wires acting as transformers or arials. this is why PCB mounted jacks and a proper ground plain and the right LM386 makes the proper pedal sold by collin. it is pretty much impossible to fix on a hand wired pedal but if you succeed then you are a genius of power distribution and RF shielding.

[Paul_C]

I bought some NJM386BDs from a UK supplier, which are listed as being the equivalents of LM386N-4s (the hottest output of the three types IIRC) and with one of those in it works really well (I also put one in my Tone God Finish Line clone and really like that too).

It does seem sensitive to what's in the loop - so far I've only tried an EHX Pitch Fork and currently an Alexander Colour Theory and the oscillation behaved differently with each of them.

I'm no genius of anything, but I'm not getting any noises I'd describe as "unwanted"

The fuzzy bits in this that start at around 33 seconds are the Colour Theory (in the loop) doing the four note pattern, plus my ropey playing (I've been spending a lot of time building and not enough time playing) through the PU II. The other bits were recorded separately.

[digi2t]

Well the 386 is very unique, though it's made in 3 voltage ranges, and all 3 sounded different in my build, and even among each range, different chips sounded different. 386 is used in a lot of things as a gain stage so it's worth having a bunch.

It's a good pedal but you may want to look at what dino did with a load emulator using a 018 transformer so it works right on single coils. It's been years since I built mine, and I have mostly humbucker guitars so I didn't do it to mine.

The pickup simulator on the front end is to give the circuit an impedance that it likes to see when using any sort of buffer in front of it. Without it, the buffer will kill the oscillation section of the circuit, which IMHO, is the best part. Otherwise, the circuit works just as well with single coils, as it does with buckers.

This is the pickup sim I used;


And my vero which incorporates the pickup sim;


Bigger version vero: https://www.aronnelson.com/gallery/main.php/v/digi2t/veros/Parallel+Universe+II/PU+II_NEW.jpg.html?g2_imageViewsIndex=1

Insofar as the 386 is concerned, I had the best results with the N-1 version, which is the lowest power output of the different N models. Using the higher output N-3, or N-4 (I find) diminishes the oscillation capacity. I think the lower the output power from the 386, the better the oscillation and starve sounds are. Of course, YMMV, so socket away.

My unit is a three footswitch build, allowing me to bypass, oscillation, and starve with the footswitches. This works for me, since I can set it up, and then switch stuff on/off on the fly.

All I can say, is that this is one of the most underrated circuits ever. Lots of people turn their noses up on it, but it was one of the first pedals I ever built, dating back close to 10 years now. This, along with my Uglyface, has occupied a permanent place on my pedalboard ever since then, and it doesn't look like anything is about to come along to knock it off anytime soon (read "ever").

[eatyourguitar]

I have that coil but mounting it on strip board doesn't really work. nothing lines up. I have been staring at it for years.

[Paul_C]

Insofar as the 386 is concerned, I had the best results with the N-1 version, which is the lowest power output of the different N models. Using the higher output N-3, or N-4 (I find) diminishes the oscillation capacity. I think the lower the output power from the 386, the better the oscillation and starve sounds are. Of course, YMMV, so socket away.

I've got some N-1s and I'm always curious, so I'll give it a try at some point and report back if things change significantly - though I'm happy enough with it as it is.

Having said that, I noticed that on the layout I used the LM386 is being used with pins 1 and 8 connected and I'm tempted to try adding a switch to cut it from 46dB gain to 26dB to see what that does to the circuit. I'm sure there are lots of knowledgeable folk out there who already have some idea, but part of the fun for me is learning by trying, and I've got an awful lot to learn

[digi2t]

I have that coil but mounting it on strip board doesn't really work. nothing lines up. I have been staring at it for years.

It fit on mine. As illustrated on my vero. OK... it's not "perfect", but it's doable. The side tabs need to be thinned down with a Dremel to fit into the holes, but it will work. You can even enlarge the holes slightly to leave more meat on the tab. Worse case scenario, bend the tabs under, and it will stay put by the pins alone. Inject some hot glue on either side for stability. The xfo outline on my vero is representative of the scale.

[BetterOffShred]

I just made a break out board from a piece of strip board like 7 rows by 3 columns.. got a spot for the wire for each going to the primary. Clipped the secondary legs off, and the whole thing can go on the front of any circuit. Put it in a 1590A