Your hardest /most difficult perf project?

[jimbob]

This is sort of related to the other perf link, but it makes me wonder as i work on projects --What are the hardest to make/perf projects the other guys are doing?

In my case, i have perfed A LOT. Everything from Tube Screamers to the Rat,  s lot at GGG--except the  delays and the obvious boards like sans amp ect.. Just tons of stuff. Are there anyone out there doing some really ridiculous stuff? Crazy insane layouts 80 + parts ect.. on perf with tons of lines? curves? ect..

[mikeb]

I build a lot of my one-offs on perf - the last one was an echowave (i.e. triwave picogenerator + PT2399-based delay). If you break it down into smaller chunks it's not so intimidating, and I'm happier doing things this way rather than put together etch a special PCB design, drill it etc.

Mike

[RickL]

Fender Blender, BMP and Bee Baa in the past, most recently a Shin Ei Octave. Perf's not really that complicated as long as you do the design first, not as you're building it. My first couple of perf projects (Dr. Q and Distortion +) were about double the size they needed to be with flying jumpers all over the place.

[lightningfingers]

Easyvibe, BMP, smallstone(i regret that one) Univibe...etc

I use perf for everything. I have neither the skill nor the patience to make PCBs.

[Peter Snowberg]

My hardest perf would surely be a DSP board that had a 13x13 DSP, six 32 pin SRAMS, about twenty 20 pin latch chips, a few FIFOs, and some glue, connectors, etc.

It was about 6x9 inches and wired point to point with different colors of #30 wire. Plated throughs on pad-per-hole.

[petemoore]

I'm thinking of trying my PCB making skills using the Lazer/Photopaper technique described here recently, hopefully I can find the link when I need it/get the $tartup stuff, not like it's that much...
 Something like a Neutron Filter or Big' 'ol Phaser or something.
 I have a perfed Phaze 90, and did an EZ Vibe, that still only works for a while before the sweep fades and I trip the 'auto sweep fade restart switch' [lol]..a momentary power cut switch...still needs work...used to work...has an LM317 regulator with adjustable pot for voltage controlled PS, but even so doesn't exactly keep sweeping...
 I did a SS phazer on PCB from Fransisco at Tonepad, and the ease of working with it to get a phazer made it Totally worth using it. also I PCB'd a DynaComp, [after a perf build I scrapped / couldn't figure out].
 I didn't do a precise parts count , but can see from a schem where it starts to get too many variables for Semi EZ perfing.

[cd]

If I'm doing a one-off, I'm too lazy to make PCBs.  I perfed a Small Clone once, as well as an Ibanez Chorus, and those two pedals got "stolen" by some bandmate friends.  Which then resulted in a couple of other "thieves" wanting their own, so I perfed a few more.  I managed to stuff the whole thing inside a BB with some planning, too

[DDD]

The most hardest to make project for me was suboctaver, something like B*ss OC-3 with additional frequency doubler function. It's schematic is completely original, and it works quite good and stable, but I think I will never do it again.

[Paul Marossy]

My hardest perf job was a Shaka Tube. Anything more complicated than that, and I want a PCB.  

[aron]

I never got that Dr. Quack to work.

[Mark Hammer]

I think I have a 6-stage FET-based phaser that remains uncompleted.  What made it hard was that I tried to fit it into a confined space.

Obviously there is a point where hard is hard (like Peter Snowberg's DSP behemoth), but for most of the things folks here will want to do, "hard" is often a function of making poor initial choices in chassis size or semiconductors or of having awkward-sized/shaped components than bugger up a clean layout.  For instance, using a 1590B where a 1590BB or even larger chassis might be better suited, using quad op-amp chips when duals would provide more flexibility in component and supply-bus arrangement, using single op-amps and multiplying the number of supply connectons you have to make when duals would be easier, having big ugly electrolytic caps because that's all you have, etc.

I generally find discrete transistor circuits easier to perf because one can often just run a bus along two opposing edges of the perfboard and run collector and emitter resistors out to them (better yet, make the bus itself out of the component leads).

Connecting wire can also make a big difference in difficulty.  If you can score some narrow gauge solid core hookup wires in different colours, that's huge help.  First because the different colours can help you to mentally keep track of things, and second because the narrow gauge and solid core allows you to fit them in tricky spaces.  For instance, sometimes you can run the supply lines to several op-amps lined up in a column by running the hookup wire between the chip and IC socket.  Or maybe you have a feedback resistor snuggled up in the first row of holes beside a chip; solid core small gauge wire will let you run a wire/supply-line between that resistor and the socket.

There is hard to make, and there is hard to tweak.  Some of us fools (and there MIGHT be a hint of autobiography in there)  use perf when we COULD be using our breadboards.  The circuit gets built in a hurry, and predictably there are component values to change.  An easy circuit/build is one that lets you identify post-hoc which one of those many 10k resistors is the one you need to swap out without having to trace the connection on the bottom :?  and one that lets you physically get in there and remove it without too much disruption.  So, what seems to start out as an easy perf-build turns out to be harder than you thought because you didn't leave yourself the space to do all the after-the-fact stuff easily.