Modular Power Supplies


#102

to confirm the problem, I would try to reduce the modules to the point there wont be any issues. DSP modules would be the first to be removed as these tend to draw lot of power during the power-up. Maybe this way you can identify one module that is causing all the trouble.


#103

Yeah have been doing that all afternoon. After removing Rings, which supposedly draws 120ma, no more issues. Funny though, because rings was never connected to anything when I had those issues, doing to check now if Rings is ok.


#104

I had a similar issue with MI Clouds, so am not surprised if switching off Rings gives you a similar experience. (Is it an ARM M4 on both?)


#105

What is an ARM M4 ??


#106

the chip inside? stm32 family info


#107

I don’t know, how can I check this? This is a micro 8hp version of Rings.

Now I’ve rings in another case to check if it is ok, and seems to work as before thankfully.


#108

hopefully this will solve the stability problems.

Also, don’t be too devastated, I think all of us have had a fair share of technical problems with our modules and I think it’s normal. Eurorack as a format is basically a field of experimental instrument building using the supposedly universal modules that should work together but not always will. It’s far from the stability of ready-made instruments that will go trough the extensive quality control (and still can fail). I would not hesitate to get rid of the modules not fitting into the mix, without growing too much gray hair.


#109

So after that, I decided to put Rings back into the problematic case, but this time to leave some empty HP space around Rings and Ts-L, both of which seem to get really hot, and I suspect that might also be contributing to the overload.

I subsequently pushed the system with all triggers triggering, all modulation sources modulating and made the clock fluctuate super fast to super slow etc. And managed to get through more than an hour without the issue coming up again!!!

Maybe something that wasn’t discussed much when addressing the problem, overheating in the case. What are your experiences with that?


#110

Modules really shouldn’t be getting warm. Are you screwing them into wood or do you have metal rails?


#111

Metal rails and 20 char


#112

I am starting to try to figure out some hiss issues with integrating my pedals + modular. I am curious if there are solutions that will help me achieve a higher SNR than my current system.

My euro (intellijel 7u) has a low noise floor. I can do loopback from my computer (RME out -> MI Veils -> Mannequins RIP -> RME in) and achieve consistently over 80dB SNR. Noise floor is about -100dB signal going out around 0dB peak and coming back -20dB peak.

Pedals are okay in loopback by themselves in a similar loopback configuration (RME out -> Meris Polymoon -> Meris Mercury7 -> Elektron Heat -> RME in). All pedals use their own PSUs, hooked to a power strip under the pedal board. I get ~65dB SNR. Noise floor is -85dB, signal again going out around 0db peak and coming back around 20dB peak. I will note the Heat can have a higher noise floor when pushed, but I feel like that is too be expected somewhat as they are analog saturation circuits.

When used together is when things get more noticeable. In loopback (RME out -> Veils -> Intellijel Pedal I/O Send -> Meris Polymoon -> Meris Mercury7 -> Elektron Heat -> Intellijel Pedal I/O Return -> Veils -> RIP -> RME In) I get a maximum of about ~50dB SNR (Noise floor -80dB, signal going out 0dB and coming back -30dB).

I am curious if there are things I can try to improve this. Can other things like impedance mismatches going in and out of the pedal i/o be part of the problem? Could power be causing issue (I am using a Furman SS-6B filtering strip but it’s not like a fancy conditioner or anything). One thing I have thought of is running the pedal loops independently in and out of the interface and routing them serially in the RME software. That will probably introduce a bit more latency, but I think that is okay for my current use case. The only real issue there is that it will max out my analog I/O on my interface and so I will not be able to monitor with my current setup.

UPDATE: I’ve been thinking about this today, and was thinking about how my method was doing several steps up and down from modular to line to interface with the pedals. I tried to minimize this, and sure enough I can get pretty close to a -80dB noise floor. The chain I’m currently using is RME out -> Meris Polymoon -> Meris Mercury7 -> Elektron Heat -> RIP -> Veils -> RME in.


#113

It’s not unusual for modules to get warm, especially for digital ones that have on-board voltage regulators.


#114

Hmm. The only time it’s happened for me was when I had a bamboo DIY skiff with wooden “rails” and a uZeus that should have had the jumper removed. Otherwise everything seems to stay at room temperature for me.


#115

You won’t necessarily be able to tell from the front panel, because the regulators are on the back and will only heat up the surrounding PCB / air.


#116

The OP did say “Rings and Ts-L both seem to get really hot” so I assumed some panel touching was involved in that determination. But, whatever :shrug:


#117

this is only tangentially related to power supplies, but I’m curious if anyone has ever used 1x5 power cables for Eurorack rather than the typical 2x5. the main reason I could see would be to allow built-in daisy chaining, particularly of low-current modules, since the 2 adjacent pins in each Euro power row are connected anyway, and something only drawing a few mA doesn’t really need two 24AWG cables per rail (as in the normal 2x5 power ribbon cable) to support its draw.

I’m thinking some of these (or similar) would do the trick, just curious if anyone else has tried or if I am missing something as to why this would be unwise, of course bearing in mind that any chained modules would now be sharing grounds and could introduce any noise issues that would also be present in a more typical flying bus board or daisy chain cable situation. I just made a little stripboard power splitter for some of my low-draw modules since I have more modules than power sockets on my Intellijel Mini power supply, so I may not bother trying this in my current case, just curious about the idea!


#118

I’m not even sure if what I’m mentioning here is actually true and does in any way respond to your question but I believe the http://www.konstantlab.audio PWR CHECKER 1hp module uses such kind of connector on the module side. Not sure where I’m going with this though…
I would want to imagine that your above described scenario would work just like a bus cable with only two sockets attached. If that’s something clever to do, I wouldn’t be able to say.


#119

come to think of it, I think @bpcmusic’s 2hp designs use a single row 1x5 power connector, so maybe he has some insight?


#120

This seems to be pretty common practice for smaller modules. Here is a shot of the line from 2HP Modular:

For my own use, I’ve put multiple female power plugs on a single power cable in order to keep the cable nightmare to a minimum when I have a few modules tightly placed together. (Like a micro flying bus cable.)


#121

How exactly would the 1x5 daisy-chaining work? Would you physically arrange for two separate 1x5 pin headers to mate with a single 2x5 socket?

I’m asking because daisy-chaining is already possible with 2x5 cables - just put multiple sockets on the same ribbon, spaced far enough apart to adapt to the varying locations of power pins on various modules.

I don’t see how 1x5 makes daisy-chaining any easier (or harder).

I agree that 1x5 is a good choice, especially for the 2 HP modules mentioned above. There’s barely enough room for 2x5, and the signals are mirrored on both halves.