discrete belton brick vero layout to allow for modding and tweaking.

[tabbycat]

from a purely 'time and expense' perspective there is no real point (masochism aside) in building a belton brick from scratch when the complete tried and tested finished item can be had for £12 a pop (or even less if you buy in bulk). after sourcing and buying all the individual parts, laying out, soldering and testing it makes no sense at all.

however... from the modding perspective, i'm really curious to get beyond the fixed-and-final values of the gooped brick and into swapping some key resistors for pots, or putting a variety of tone-controlling caps on a selector switches, just to see what kinds of untapped potential can be squeezed from these things. i think the idea has a hell of a lot more to offer that the stock brick presents, in any of its three variations.
am a big belton fan, eqd's ghost echo is my favourite belton build to date. orchasmic. but apart from the dwell mod (recommended if you like deep) the belton fixed unit seals in a lot of the sideways potential for manoeuvre on the reverb end.

so i’ve been thinking that it would be pretty neat to be able to work up a variation of a discrete belton into a separate stomp, with all the key variables on pots and switches, that could then be substituted (temporarily or permanently) into any belton-dependent effect to extend its possibilities.
is that a good or shit idea? i thought it was good when i had it. but that’s invariably the case.

so am making a request here (enthusiastic and abusive comments welcome) in case anyone who also likes the idea and is smart at vero layouts (i am pretty rubbish tbh, a max twenty components being my absolute limit at present) wants to have a crack at it.
or if there are three people as slow as me at layouts, but all up for this, maybe we could divide up the schematic into a pt2399 each and then pool our layouts when done?

anyway, i thought i would throw this out there to see if anyone is interested. lots of potential i think.

as far as i know this is the best shot at a schematic for the brick out there, but am not aware of anyone having built it to date.

[Alex]

It won't be too difficult to draw a layout but it would be big (there's also a voltage regulator and the 14 pins Schmitt Oscillator, if I'm correct).
The main problem is to understand which components would be variable first.

[tabbycat]

http://ada.graudal.dk/bdtr2/bdtr2_top.JPG



http://ada.graudal.dk/bdtr2/bdtr2_bottom.JPG

Alex80 wrote

It won't be too difficult to draw a layout but it would be big (there's also a voltage regulator and the 14 pins Schmitt Oscillator, if I'm correct).
The main problem is to understand which components would be variable first.

pretty difficult for a me to draw a layout anyway. i was looking at it last night and i reckoned that if i did an hour each night this week i could maybe have one pt2399 done by sunday. that's my workrate. i find layouting bloody confusing. like trying to count ants.

re the modding side, was going to go low fi and socket the vero for components that look likely candidates. just work through suspect resistors then caps to see what effects what most.
the links between the pt2399 being a good place to start. those 11k8 between the pt2399 could be 25k pots.
and between pins 4 and 6, the resistor works down in value on subsequent pt2399 from 5k to 3.2k to 1.8k; what would happen if they were all 5k? infinite repeats? self-oscillation hell? parallel universe? unchartered territory, but interesting. but i would be working ('playing') hit-and-miss monkey style as i'm no techie.

have to admit to having a bit of a fetish for this unit, which is curious because most pt2399 delays i find a bit sterile and dry sounding. but i think the modulation is the thing that most people who don't like it are unconvinced by, and if that could be tamed i'm sure there would be more enthusiasm for the brick concept. but until belton offer varied modulation a tweaked discrete build is the only option.

having said that, are you familiar with any 3x pt2399 delays that share a lot in common with this circuit? may be an option to tweak one of them closer to belton specs for the sake of experiment.

[nocentelli]

I've never played with a belton brick or discrete pt2399 reverb, but I've been working on a 2 x PT2399 delay recently: pin 16/15 is the delay input, pin 14 outputs a single delayed repeat. Pin 6 to ground resistance alters the delay length and resolution (delay time and sample rate are inversely proportional) with 50k giving around 500ms, 4k7 will be something around 60-100ms. The belton appears to send the input signal to put chip 1 and 2 in parallel, then send the delayed repeat from chip 1 through chip 3 within even shorter delay. Delayed signal from chip 1 is also.sent through chip 2, but chip 2 is also doing something I don't understand with the output being fed through the 220k to the delay time pin; The oscillator section is just super fast square wave generator made from one inverter (i don't see why a 16 pin chip is needed for that).

It's very easily to massively overdrive the inputs of the PT, literally a few k's less resistance and it's screaming so I'd limit initial tinkering to the pin 6 resistors to mess with the delay time of each chip. The vero I've done for a 2 x PT2399 plus dual opamp is 21 x 27. I'd guess three PT chips and another 16 pin (or 8) for the oscillator would need more like 35 x 35, and this is for a circuit that we're not even convinced is correct? I would always breadboard this kind of thing first, see whether it works and what might make useful tweaking points before thinking of a vero layout.

If you have a look at merlin's schematic for the equinox reverb, you can see much more clearly how delayed signal from two PTs with different delay times is fed back into the inputs of each, and it also includes the dry signal mixing opamps; I'd breadboard this one first to see what tweaking potential exists, perhaps let your findings inform future experiments with a discrete belton:

http://www.valvewizard.co.uk/equinoxschem.jpg

[tabbycat]

^ big thanks for the info, nocentelli. i remember watching your pt2399 experiments clip a while ago, maybe to do with a comment on fsb about one of those seppuku delay things. didn't you go to three at some point? maybe you posted a link to three. i remember it got pretty hectic. i need to work through your yt clips again soon. lots of interesting things in your investigations and experiments.

run down on the pin uses v useful to know. and the map of the delay path. tangled web.

i had a wild intuition about pin 6 being a time thing. the way the resistor value there progresively diminishes from chip to chip. good to hear it from someone who's actually worked it out rationally.

re the oscillation, do you think tweaking the 3m3 would increase or decrease the effect, or just kill the oscillation flat?

re size, i've got three breadboards so i could build one pt2399 network on each breadboard and link them up. but i find breadboards way more confusing than vero to work with. messy and chaotic once you get more than half a dozen jumpers in.
so maybe i will try to work out three separate individual veros for each chip and then join them up after. as a lot of the supporting stuff is common to all three chips (and the equinox). once i've actually got one down, the other two will only be variations of that (less painful than doing each from scratch).

do you think something like a TL062 would do for the oscillation? you said you couldn't see why it need be bigger.

will look at the equinox 2 while i summon the courage to attempt the belton veros. thanks for the tip.