Was just scrolling around and I saw this post about the Midimuso - I I created a 3D printed Kosmo Format panel for it that seems to be working well for me so far. Here is a link if anyone is interested:
4 Likes
yes max. after I have done the 3320 hi/low filter Iāll make a voice, I also have after that a plan for a poly module. to save me having to lug around a prophet that breaks all the time hahaha
6 Likes
nice Wesley! one question I have regarding an un modified off the shelf midimuso with eurorack power. can you get enough power to the voltage reference? from my experience eurorack power usually doesnt whack out quite 12v. especially with 3 or more modules. and for the lm317 to do its thing the voltage input needs to be a bit over 1v above the desired reference. which meant for the module im doing around it I had to do a bit of modification to the circuit around the vref to make it work in less optimum power conditions. whatās your experience with it??? its more than fine when powered with a wall wart as your usually in nice and happy 12v zone, but most eurorack power doesnāt quite get there (I have tested all I have had to hand, fc powerdiy, doepfer monster case, befaco skiff, excalibus, tiptop etc
2 Likes
Maybe I misunderstand what youāre saying but I have a TipTop ĀµZeus powering 84HP of Eurorack and it reads 11.99V on my multimeter.
3 Likes
Hi Sam,
Here is the Eurorack power Iām using:
Each one I build has minute differences on the power rails, but mine are a smidge over 12v. That being said, my Midimuso actually wouldnāt track 1v per octave at first - I was having issues until I clipped out the LM317 and replaced it with a new one, and then things seemed to work fine. I worry though, since the midimuso apparently take 500 milliamps or something crazy - Iām worried Iām overtaxing my supply and Iām considering switching to a wall wart.
One thing I noticed when trimming the voltage ref - I was able to expand the range of the voltage trimmer (and get close to 1v per octave) by decreasing the value of the 220r resistor connecting the Vout pin on the LM317 to pin 1 on the 5k trim pot. Be warned though, donāt go too low on that resistor or Iām pretty sure you will destroy the LM317. You could maybe take the 220r down to 100r or 75 in a pinch to squeeze yourself a bit closer to the desired voltage if you are getting like 10.5 instead of 10.6 on the test pad or something. Let me know how you fare!
Another full disclosure, I am having some power issues which I described in this thread - hopefully they are not connected to the midimuso:
best of luck,
-Wes
2 Likes
1 Like
hi fredrik! after quite a bit of looking around ill find a link it seems quite fair game that a headroom of 1v is enough, Iāll try and find what I found regarding it, and it led me to do quite a bit of messing around because getting the problem I mentioned above from the lm317 I went on the hunt for an alternative, which lead me to sticking with an lm317 and figuring out if it can hack it. and so far its been more than fine, even the main one I have been using since November, and another prototype sitting in another module case thats used a lot. I also read in a forum of other peoples experiences saying 1v is the minimum headroom they have achieved, and I guess thats what rob from midimuso has been working off too. however in the midimuso setup anything even a tad lower than 12v is just impossible to get the reference voltage
1 Like
im still trying to understand why 3v was specified, as i saw numerous projects with less headroom. maybe it is when its working at its maximum ratings of consumption. which the midimuso is well under
yes wes thats what I have gone for in the latest iteration. 100r and also (after reading on a robot forum that I mentioned to fredrik but I canāt remember what the heck it was) I isolated the trimpot from ground. its been faring good for about 2 months.
1 Like
but maybe you can unearth a better reason!
1 Like
thoughts??? this is it in its current form (haha current) the thing is 10.6 is nearly at the top of its headroom to the 10k is suffice to get to it.
1 Like
also if there is an alternative to getting a stable 10.6v comparable in cost and simplicity im all ears! it was suggested to try the LM117 but they are clunky bo bunky
1 Like
Yeah, it does say āup toā so low power use might be whatās saved you (and the midimuso folks) but if youāre designing for a dropout around 1-1.5 V you really want a low dropout regulator, which LM317 isnāt (itās a mid-seventies design iirc). Iām still a bit surprised that youāre seeing only a volt, but guess I have to experiment a bit tomorrow 
Thereās a ton of modern LDO options, but seems a reasonable replacement would be a LM1117-ADJ which is widely available, has max dropout of 1.25 V guaranteed at quite a bit of current, and should be available in the same package & pinout.
EDIT: Increasingly convinced that the low dropout is because the LM317 doesnāt regulate, so itās just the voltage drop over a saturated output transistor. To the breadboards!
4 Likes
While we are on the subject - Sam are you at all concerned about trying to run multiple of your tuner VCOs and a Midimuso off the same power supply? I seem to remember arduinos using a decent amount of current (and so does the midimuso), so Iām a little worried Iām gonna use all the current on my supply just off those modules alone. Would it be crazy/impractical to power the Arduinos separately? I have a different 5v source already in my case which I am considering using for the arduino part of the vco circuit.
2 Likes
The resistor and the trimmer form a voltage divider (*) that provides a voltage to the ADJ pin, and the regulator will tweak the output voltage to keep ADJ at 1.25 V. The pin itself only consumes 50-100 uA according to the datasheet, so the main load will be through the resistor & trimmer to ground. If you go too low, you may damage the trimmer, though ā e.g. with 100 ohm the current is 1.25/100=12.5 mA which is a bit too much for a 10k trimmer (see here).
*) Well, in Samās sketch itās not a divider. What was the idea behind not connecting the pot to ground? Seems to me that would turn it into a current regulator, not a voltage regulator, with the regulator trying to keep the current through resistor + trimmer high enough to drop 1.25 V, which doesnāt quite work since 1.25 V/50 uA = 25k. Are we sure itās actually regulating when youāre seeing a 1 V drop, or could that be the Vce drop of the regulatorās power transistor?
3 Likes
You want the LM1117, not LM117 
Seems like LM117 is an LM317 in a metal can
(TO-3 or TO-39), iiuc. Thatās pretty metal, but not very practical (unless you need an extreme temperature range).
EDIT: Older TI chips often use 1 for military grade, 2 for industrial, and 3 for commercial, with the main difference being temperature ranges. Seems TI no longer makes an LM217, but itās available from others in case someone here prefers industrial over metal.
3 Likes
The LM1084-ADJ is a pin compatible, low dropout, replacement for the LM317, its dropout is less than 1V at 1A at room temperature and above.
And Iāll join my voice to Fredrik again and wonder why the trimpot in your schemaitc above is not connected to ground in the normal voltage regulator configuration.
2 Likes
(The LM1084 would work, and is much beefier (5.5 A!) than the LM1117 mentioned earlier, but also more expensive ($0.88/1ku vs $0.33/1ku). The lower amp LM1086 is a slightly cheaper, but still not as cheap as the 1117.)
1 Like
In current capability, but not in size as Iām assuming weāre looking at TO-220 packages for all of these regulators or voltage references.
The LM1084 is indeed more expensive than the LM1117, (e.g. $2.60 vs $1.54 in single quantity at Digikey) but depending on the total current draw, its lower dropout voltage might give us just enough margin to be able to use Schottky diodes for power supply reversal protection which I prefer over the 10 ohm resistors/fuses (or the mostly harmless (to be polite) ferrite beads - although a MIDI to CV module with its microprocessor might be one of the rare modules where ferrite bead actually have a non-zero (but probably still negligible) effect).
2 Likes
I havenāt had a problem personally. powering a midimuso, 2 braids, 4 of the tuner VCOs and 5 or 6 more modules including a safety valve off 1 power supply. never had a problem, probably most definitely over its rating. but its been on for days at a time over a number of years. no melting or going into a wormhole yet!
1 Like