Muff's Modules & More

komyta

Hello everyone,

Here's a long post about an unexpected behaviour I encountered while using my very little MU system.

First, let me introduce the system :

My power supply is a QPS2 from Synthesizers.com.
The modules I've got so far are :

- Corsynth C101 VCF
- MegaOhm Audio MA20 VCF
- Synthesizers.com Q106 VCO
- Synthesizers.com Q108 VCA
- Synthesizers.com Q109 EG

For the casing, I've got a Synthesizers.com 8-MU wide Rack Frame inside an AllMetalParts 5U, 20 cm deep rack cabinet. love

In the months to come I plan to add one more Rack Frame / AllMetalParts cabinet solution, with another QPS2 power supply. The idea is to have a very portable system, whose two 8-MU cabinets would work by themselves for the times when I need only one of them, hence the two little QPS2 power supplies. The total module quantity should not exceed 12.

I've decided to begin my MU system with filters that would be close to my all-time favourite monosynths, the Roland SH101 and Korg MS20. This is fun!

The filters' ability to self-oscillate is very important for me because I want to be able to use the filters as audio sources or audio-rate modulators when they're not used to filter an incoming signal.

And here comes the problem :
When I first plugged my system in and tried my C101 with the "Resonance" knob set fully clockwise, the filter would not self-oscillate, whatever the frequency at which I would set the "Frequency" knob.
The only way to make it self-oscillate (apart from feeding its audio inputs with incoming sound) was to feed the "Res Mod" input with a DC voltage provided by the Q106 "Pulse" output with the Pulse Width set at 100%. But as the C101 is said to be a self-oscillating filter, I was expecting it to self-oscillate without any help from a fellow module.

After a few hours making sure that I wasn't doing something wrong, I sent an e-mail to Pablo (the man behind the Corsynth brand, AKA Thalassa here on Muffwiggler's), explaining the problem.
He responded very kindly the same day, asking me to do a few tests so he could have an idea of where the problem could come from.
After a few tests and e-mails, it seemed that the module could be faulty. Pablo asked me to ship the module to him, so he could test it himself and repair it if it was needed.

When Pablo tested it, the module was working without problem. He told me that after warm-up the resonance decreased a little bit more than it should, so he changed a few resistors to make it stronger, tested it again to be sure that the level was always the same over time, and sent it back to me.

Here let me write a big THANK YOU to Pablo : he was always very nice, answered quickly and kindly to all my e-mails and paid the shipping for the module from my home to his office and back. Also he was very careful with the module : I'm a little picky about the cosmetic condition of my gear, and the C101 was still as new when it came back to me. thumbs up

So when the module came back home, I plugged it in my rack and… The C101 would still not self-oscillate! very frustrating

I wrote back to Pablo, and had the idea to tell him that I powered my system with a QPS2 switched power supply. Also I asked him what power solution he had. He answered that he used a QPS1 and thought that maybe my QPS2 would not provide enough current for the filter to self-oscillate when my other modules were also plugged in. To be sure, he asked me to unplug all other modules, leaving only the C101 connected to the QPS2. First I was I little sceptical, but I did the test and – surprise – there it was : a gentle, pure but strong sine wave coming out of the speakers! w00t

I then plugged other modules in, one at a time, and discovered that the more modules were plugged in, the more quiet the sine wave would become, before disappearing completely when the fourth module was plugged in.

I reported this behaviour to Pablo, who asked me to measure the voltage at the power supply output when all modules are plugged in. That was (and still is) 14.8V. As Pablo told me, the C101 requires no more than +/-14.5V to be able to self-oscillate. With 14.8V, the module should give a nice 10V peak-to-peak sine wave. So he supposed that the problem comes from the switched power supply, whose ripples could be too high to make it a reliable enough power solution for the C101 in self-oscillation.

That's bad news for me, as I do not plan to upgrade my power solution since it works all right with my other modules, including my MA20 filter in self-oscillation.
The QPS1 would be overkill to power only 12 modules, and quite expensive since I would need to buy modules (Q137, Q103) and cables (2x QDH20, QIC-12) to be able to use it.
The QPS2 price tag is perfect for me ! cool

Also each cabinet would not work by itself anymore, and this is important for me.
Finally, as I have no time for it, I know I'm not going the DIY route.

So I decided to let you fellow wigglers know about that.
Am I the only C101 owner to power it with a QPS2 ?
What do you think ? seriously, i just don't get it

Maybe I'll have to separate from a module that I like very much in terms of sound, possibilities and ergonomics because of this little incompatibility… waah

That would be a shame, because Pablo is such a nice fellow. He gave me great customer support.

diophantine · Posted: Mon May 14, 2012 2:48 pm Post subject:

Quick question while I'm heading back from lunch...

What other modules are you trying to power from the QPS2? You are possibly trying to pull too much current.

komyta · Posted: Mon May 14, 2012 3:01 pm Post subject:

Hi diophantine,

the other modules connected to the QPS2 are only the ones listed above :

- MegaOhm Audio MA20
- DotCom Q106
- DotCom Q108
- DotCom Q109

Only one of each, no more.

Before buying the QPS2 I checked the power consumption of several module configurations using the DotCom SynthInvent software.
Whatever power-hungry config I tried (with only DotCom modules, of course) the QPS2 always seemed to offer high enough headroom...

kindredlost · Posted: Mon May 14, 2012 4:41 pm Post subject:

Sad to read about the QPS2. Sounds like it is less than capable of powering the small system you have. Your best bet is the QPS1. If the rack cab has a back opening, get another rack frame for the Q137. Then you can get by without the Q103, just one QDH20 harness and you will not need the QIC-12 cable. The Q137 will mount in back and you have the modules in front.

Another way to do this is to buy a separate QPS2 and plug both into a power strip inside the cabinet. I'm cheap so I think like this all the time. hihi

Thalassa · Posted: Mon May 14, 2012 5:29 pm Post subject:

Hello Komyta , thanks for your kind words about the customer support smile

The current consumption of the C101 is approx 24 mA so in theory the QPS2 should be enough even with other modules connected. As I told you by mail , seems that when you increase the load on the QPS2 the voltage ripple is too high making the C101 unable to self-oscillate.

whitewulfe · Posted: Mon May 14, 2012 6:33 pm Post subject:

I don't really have any advice to give about the module's behaviour, but given that you're looking at picking up another QPS2, why not chat with Vince about his power supplies? 1amp, and around the $100 USD mark -> http://www.muffwiggler.com/forum/viewtopic.php?t=56165

$25 USD cheaper than a QPS2 as well!

Additionally, as an alternate offering to the AllMetalParts 5U metal cases, Christopher Winkels makes some absolutely wonderfully made 14MU wide fir-core plywood cabinets with a maple veneer. I currently have two, and they go together quite well (and quickly too, even with just hand tools, although I'd recommend actual power tools), and are nice and sturdy. Plus one can stain them whatever colour one wishes! -> http://amherstdesign.wordpress.com/

Thought I'd plug them just in case they weren't know smile

(Errr... I didn't see the shipping costs to the UK/europe, eep! Bad Canada Post, Bad!)

Putte · Posted: Tue May 15, 2012 3:12 am Post subject:

It seems the dotcom power supplies handles a lot less than the information says. Synthinvent tells me I´ve got plenty of power. But about half an hour after patching up most of the system, some modules won´t work normally any more (The Moon modules).
My system is some 80 % dotcom, counting 63 U:s at the moment and a QPS1. Therefore, Synthinvent isn´t completely accurate. But, it tells me I´m only using half of it´s capacity.
In summa, I say play safe and plan for a system that only uses about half of the power the QPS:s are supposed to provide. I didn´t, but I´m going for a larger system and was eventually going to add a second QPS1 anyway.

fyvewytches · Posted: Tue May 15, 2012 3:42 am Post subject:

whitewulfe · Posted: Tue May 15, 2012 6:55 am Post subject:

fyvewytches · Posted: Tue May 15, 2012 7:44 am Post subject:

komyta · Posted: Thu May 17, 2012 3:58 am Post subject:

komyta · Posted: Thu May 17, 2012 4:08 am Post subject:

komyta · Posted: Thu May 17, 2012 4:33 am Post subject:

komyta · Posted: Thu May 17, 2012 5:05 am Post subject:

whitewulfe · Posted: Thu May 17, 2012 6:18 am Post subject:

wsy

Occasionally, you can just get a cranky module or PSU that don't play well together.

I had a quantizer that would *not* power up correctly 9 times out of 10.

I sent it back, they were unable to reproduce the problem. I figured "solder ball".

Nope. Problem back.

Moved it to two other (independently powered) cabs - worked fine.

Moved it back to the Evil Cab. Problem returns. But the oscilloscope and DVM
show that the power supply is in spec.

I started pulling modules; it got better but never got "right". I finally soldered
a 100 uF capacitor across the +5V to the module and now it powers up
correctly 90+ % of the time, even with the "evil" PSU.

So- you might want to try that. Add some bypass capacitors in the 100 uF
range between ground and +15, -15, and +5 *right at the module*, on the back side
from where the MTA156 (power connector) is. Yes, this is
soldering to the module, so make sure it's OK with the seller first.

But, as they say, "It Works For Me."

- Bill

Henfield · Posted: Thu May 17, 2012 9:59 am Post subject:

From what I can surmise here, the problem is with the newer " line lump" type of QPS2, which looks like a computer adapter and can accept International voltage? Has anyone had a similar issue with the older style "wall wart" type of QPS2 that only worked with North American voltage/mains?

komyta