Braids THT in KOSMO format

They’re input and feedback resistors for inverting amplifiers:

image

What’s significant is not their values but their ratios. In the amp on the left above the gain is -R53/R54 = -0.25 (actually -0.249). On the right the gain is -R34/R41 = -1.0. In this section it’s handling a pitch CV so presumably you want very accurate gains for good 1 V/oct tracking. In fact I’m surprised there’s no trimmer in there to fine tune it. Maybe there’s some fine tuning provision in the firmware.

Elsewhere there are for instance -R25/R30 = -49.9/24.9 = -2 (actually -2.004).

Should be able to use other resistor values but you’d want to keep the ratios about the same. With 24k you’d need 48k and 96k but these aren’t standard values. Alternatives would be to use two or four 24k resistors in series, or use trimmers to get the correct resistances. But it seems to me none of these gains should be critical except for the V/oct CV, so 24k and 51k (or, better, 49k) should be okay there, I think. For the V/oct CV maybe use a 24k for R53 and somehow kludge a 2k trimmer in series with it, and then just use 24k for R34 and R41 — but hand select them to choose two whose values match as closely as possible.

5 Likes

Thank You very much @analogoutput, i will try this ! quick and detailed answer : you are a beast…

Thank you for this great analysis, Rich! I only want to add that the calibration is in fact done in software:

To calibrate the unit, disconnect any signal from the FM input, and connect the note CV output of a well-calibrated keyboard interface or MIDI-CV converter to the V/OCT input. Move the COARSE and FINE knobs to 12 o’clock position. Go to CAL. in the options list, and keep the encoder pressed for 1s (this is not an option you want to select by mistake during a performance !). The screen displays >C2 .

Send a voltage of 1V to the CV input. Click on the encoder.

The screen displays >C4 . Send a voltage of 3V to the CV input. Click on the encoder to finish calibration.

Because Braids uses this software calibration procedure, it is compatible with the 1.2V/Oct standard too! Here is a tip: you can very well perform the calibration procedure with another pair of notes 2 octaves apart, and with the COARSE knob in another position. This can be used to shift up or down the range of the COARSE knob.

So, I think it should work with a slightly different ratio for V/oct as well :slight_smile:

4 Likes

I think so too. In fact if you can use it to do 1.2 V/Oct too, that suggests the ratios can be more than slightly different. In fact it now seems mystifying they used such oddball resistor values — just going with 24k and 49k or 51k, and no trimmer or hand-matching, seems like it should be fine.

Though if it were me I’d breadboard it first.

2 Likes

Okayy, thank you dudes for your analysis ! I will try this and report to you!

2 Likes

@sebastian, quick question , i only have a polarised cap for C24, how did you end up putting your polarised cap? Did u put the nagetive pole towards R50? (as indicated in the schematic) ?

Dang! (If that’s a word) I still have a bunch of original Mutable boards including Braids that I’ve just never had the guts or patience to start. But it’s such a cool module.

3 Likes

I think so… negative towards the output jack, if that makes sense… I can check tomorrow

1 Like

That would be the correct way. The audio out signal before the cap is positive.

1 Like

Ty, maybe it will pop when i plug it, then ill know for sure xD

It is late… i finished soldering everything appart from the jacks. But i wanted to give a try at downloading the braids firmware and see if the display, displays something. But after dwloading STM32CubeProgrammer and downloading both the bootloader and the firmware on the chip, nothing happens. But i have so many questions, i dont know anything about these things i am kind of lost to be honnest! i dont even know how to place the jumpers on the back of the pill, i dont know if i have to solder the jacks before lighting the display… But at the moment the display stays dark. I will try some things tomorrow with some sleep :slight_smile:

5 Likes

Maybe tonight but definitely soon I’ll put in an order for the same module release 1.1 as in Claquoss’s build. I just found the link on sebastian’s GitHub.

-Fumu / Esopus

3 Likes

Hm… it should light up even without the jacks connected. The jumpers are irrelevant if you use the ST link V2, at least, I did not even solder the headers for them :wink: The display and encoder should work already only with the programmer connected, no need for the 12V eurorack at this point. When progamming, make sure you don’t delete everything else when you flash the bootloader (there is a checkbox)… also, you can flash the firmware alone and check a box which is something like “run program after upload”, so you can make sure that there is no issue with the bootloader (it will only work until you restart the module). Hmm… Just a few thoughts, of what I had problems with, maybe it can help you :wink:
Apart from that, check the usual, no shorts on the power lines, the correct ICs inserted in the right orientation… are the two boards properly connected? The EL caps sometimes make it hard. I put them lying down in my second build when there was space.

3 Likes

Hi Sebastian,

Thank you for your time and your advices, after a night of sleep i feel ready to get this board working. I think you are right about the boards getting a bit separated appart from each other because of the EL capacitors, maybe i’ll desoder and resolder those.

Second thing i can think of: i mega failed, by thinking it would be easier to use a nano socket and a half to fit the pill, instead of simple female headers… But the pads of the pills are not going all the way down, so maybe it’s not even connected to the rest of the module…

Third thing : i did not bother solder this 6 pin header labeled “SERIAL_PROG” because i can’t find a use for it, but maybe i am missing something?

That is all i can think of for now but maybe other nice surprises are waiting for me!!

5 Likes

Is this pin properly soldered?

4 Likes

looks like it is , but it doesn’t hurt to heat them up if in doubt .

3 Likes

thx for your help guys and HAPPY NEW YEAR!

I reflowed everything suspicious, but still not working… But these pcb’s are rly tight so a lot of mistakes can be made. If u have the motivation, you can check these pics to see if you spot mistakes:

I am also wondering if it could be a defective chip? I’ll try and replace them, but maybe the display should display without any chip involved…

oooh dang just saw these with the switch too haha amazing!!!

I put my order in to JLCPCB for boards and panel. I was about to make a contribution to the dumbassery by putting my order for a version previous to Sam’s suggested switch. I caught it in time. If anyone is in the US and wants a set I think stock will be low but I may have an extra.

-Fumu / Esopus

2 Likes

You should give them a go. They really are not difficult. That being said…I would suggest maybe skipping Stages and Tides because they use mostly 0402. I’ve built most of their modules and really the only issue that I haven’t been able to figure out has been how to program the Grids ATmega328 chip. I still can’t figure out how to do that. :frowning:

2 Likes