Johey Posted January 8, 2013 Report Share Posted January 8, 2013 I'll continue my monologue from the in a new one, to keep the topic unharmed. Maybe someone can make use of my experiences. To sum it up, I wanted to use the power connector and switch on the MB6582 with the Mean Well RPT-60B Switching PSU we ordered together some months ago. I have finally got it working and it looks nice on the outside and somewhat nice on the inside. :) As I did not find an 8 pin DIN male connector in town, I cut the cable from a C64 PSU. The molex connector was not of the right type for the PSU so I had to dremel it a little. When building the MB6582 I initially skipped all power related components and temporarily used PSU option D, soldering wires from an ATX PSU and provided 9V using an LM7809. Yesterday I installed heat sinks on the SID chips and trying to find out how to reuse the PCB traces in the best possible way for custom installation of the PSU. I finally made it work, I thought.... Today was the day when I hoped to start making music! But no. When putting on my headphones, I heard the noice level had drastically increased. I did so many changes yesterday so I took a break thinking of where to start trouble shooting. First up was to just switch the PSU for the ATX one, and voila! The noice was gone. Ironic, no? This little Mean Well adds noice which the ATX doesn't. I remembered there was a couple of big capacitors in the parts kit for PSU option A and B, but that you can optionally install it anyway. Worth a try I thought and installed the caps labeled C3 and C4. The noice was all gone (well...)! Problem solved. Conclusion: These caps are mandatory when using with a noicy PSU. There might eventually come a photo or two. Quote Link to comment Share on other sites More sharing options...
Johey Posted January 8, 2013 Author Report Share Posted January 8, 2013 Hm. No. It's still noicy. Seems like the noice came back. Must keep troubleshooting... :wacko: Quote Link to comment Share on other sites More sharing options...
Shuriken Posted January 8, 2013 Report Share Posted January 8, 2013 I will probably try the same approach this weekend. Will let you know what the outcome is. Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted January 8, 2013 Report Share Posted January 8, 2013 Have you grounded the SID input connectors? I've run the MB6582 directly from the mean well and didn't notice any increased noise. Nevertheless I am putting a C - R - C filter between the PSU and the other boards anyway. It's just some big capacitors and 1/2 W 10 Ohm resistors, very simple. Quote Link to comment Share on other sites More sharing options...
Johey Posted January 8, 2013 Author Report Share Posted January 8, 2013 The input connectors are grounded, yes. Not doing so creates far much more noise than what I am talking about. It might be so that I have a misleading reference picture of a noise free SID environment. :) I'll record it and put it up for listening tomorrow or so. Maybe I also need more filters capacitors. Quote Link to comment Share on other sites More sharing options...
orange_hand Posted January 10, 2013 Report Share Posted January 10, 2013 (edited) Hi, I am using the following Meanwell PSU for the 5V supply and there is no noise in the audio path :-): http://www.reichelt.de/Schaltnetzteile-Case-geschlossen/SNT-MW25-05M/3/index.html?;ACTION=3;LA=446;ARTICLE=66858;GROUPID=4959;artnr=SNT+MW25-05M;SID=13UO6i3H8AAAIAABMVujEf5a943e29a75bef0f11cdfbba54a9d83 On the 9V supply I am using this transformer: http://www.reichelt.de/Printtrafos-25-50VA/EI-66-34-7-209/3/index.html?;ACTION=3;LA=446;ARTICLE=27407;GROUPID=3318;artnr=EI+66%2F34%2C7+209;SID=13UO6i3H8AAAIAABMVujEf5a943e29a75bef0f11cdfbba54a9d83 And by the way: the noise is normally not filtered by the big capacitors. The small ceramic capacitors, close to the voltage regulators are filtering the noise. Cheers orange Edited January 10, 2013 by orange_hand Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted January 10, 2013 Report Share Posted January 10, 2013 ...the noise is normally not filtered by the big capacitors. The small ceramic capacitors, close to the voltage regulators are filtering the noise. You are correct that the smaller caps filter high frequency noise, but the bigger caps filter AC ripple. AC ripple can cause hum, some people refer to hum as "noise" too. Since the mean well PSU has a switching frequency well above the audio range I don't suspect Johey has a problem with HF noise. Also, he stated that adding big caps reduced his "noise". Maybe he has a grounding issue. Anyway, if the big caps solve the issue, leave them in; problem solved :smile: Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted January 10, 2013 Report Share Posted January 10, 2013 Also not to forget the inherent SID chip noise, that also depends on the active patch (e.g. filter) settings... Many people use a noisegate to get rid of it when the SIDs are not playing. When they are playing, their awesomeness renders it all but inaudible :-D Quote Link to comment Share on other sites More sharing options...
jjonas Posted January 18, 2013 Report Share Posted January 18, 2013 Also not to forget the inherent SID chip noise, that also depends on the active patch (e.g. filter) settings... Many people use a noisegate to get rid of it when the SIDs are not playing. When they are playing, their awesomeness renders it all but inaudible :-D Talking of noise gates, anyone here ever build this: http://tonepad.com/project.asp?id=18 I'm still waiting for the FETs to arrive, but it would be nice to know someone here managed to do it (= for advice), because based on the build reports the schematics / parts list seem not be completely in order. Quote Link to comment Share on other sites More sharing options...
daemonik Posted March 19, 2013 Report Share Posted March 19, 2013 Finished my external PSU based on the Mean Well RPT-60B last week and can't recognize any additional PSU introduced noize. 12V > [2200uF and 330nF] > 7809 > [10uF and 100nF] 9V > 6x6582 5V > [2200uF and 330nF] > 6x6582 and 4xPIC I'll go on and add an active Mixer (OpAmp) stage soon - using 12V/-12V supplies reduced to 9V/-9V by 7809/7909 Quote Link to comment Share on other sites More sharing options...
Altitude Posted March 20, 2013 Report Share Posted March 20, 2013 (edited) You know, that is a REALLY noisy PSU. If you compare it to a Power One that the modular guys are fond of, its not even close.. Power Ones have a 2 mV output noise and ripple, these are 100 and 80 mVs. Even the new school switching 5V regulators are no more than 50 mV PtoP noise wise. Into the digital rail that is that is fine but for the analog side not so much Edited March 21, 2013 by Altitude Quote Link to comment Share on other sites More sharing options...
daemonik Posted March 21, 2013 Report Share Posted March 21, 2013 I'm also not much convinced about using a switching psu! particularly for the analog path! Still I gave it a try... with the option to replace it with a Transformer - Vreg(+/-) and switching 5V regulator. But at the moment I'm fine and can't recognise any noise! What about using a Vreg in series with the swichting psu? does the switching noise propagate thru it? since I only use 6582 the whole analog path is essentially fed by a 7809. further an opamp mixer shall be included which will be fed by +/-9V from Vregs. Then what about the AOUT-NG Module? is it wise to feed it with +/-12V from the switching psu? what do I have to expect? my Midibox CV V1 is also fed with a cheap 15V switching psu at the moment (yes unipolar) - and up to now I'm quite fine with it Thanks for the Power One recomendation :smile: hopefully soon I will start with a modular.... but probably i will go with an adjustable transformer-Vreg psu Quote Link to comment Share on other sites More sharing options...
Altitude Posted March 21, 2013 Report Share Posted March 21, 2013 (edited) using a vreg in series from a switcher might improve things depending on what you use. 78xx ones probably not though since they have notoriously bad ripple rejection. More modern switching regulators do have have very good ripple rejection but will unlikely be in a format that you can simply drop in but there are some that can be used directly (the Recom ones are drop in replacements). The problem with that is if you are going to go to a modern switching regulator, what's the point of using a big switching supply? You can simply feed a switching vReg straight from a standard wall wart, they can take a high input voltage without a heat sink (I'm talking up to 40V!). The way i've powered a NG module was to use an AC supply (12VAC) and split the rails with a half wave rectifier to ~+/- 12VDC and then just power the core with 12VAC with a switching 5V vReg. Everyone is always looking for big complicated power supplies for all this stuff and really, the solution for all of it is pretty straight forward. I power all my devices with a single wall wart or AC/AC transformer (MB6582 runs on 12VDC, SEQV4 with AOut runs on 115V/12VAC) All that being said, there is no reason with enough filtering he shouldn't be able to get the supply noise down to acceptable levels and that will involve a fair share of elbow grease to figure out since the ripple will get worse at load so everything will need beefing up. I know ppl that have tamed computer PSUs for audio stuff. But again, why make a mountain out of a mole hill? Back in the day, Elpac power systems use to make tripple output linear regulated wall warts (+5@.6A, +/- 12@0.2A) called WM063s. I have a couple that I prize but those are long gone and impossible to find Edited March 21, 2013 by Altitude Quote Link to comment Share on other sites More sharing options...
latigid on Posted March 21, 2013 Report Share Posted March 21, 2013 (edited) You can buy triple output PSU modules from Mouser but then you will need to enclose them. I bought the MeanWell P25A14E for +5/+15/-15, I haven't hooked up the 15V rails yet but I hope to power the AOUT_NG inside my SEQ at some point. Regulators have a certain voltage drop to contend with, which makes using a LM805 with a +5 V regulator edit: switching power supply impossible. Okay if your voltage is slightly higher. Edited March 21, 2013 by latigid on Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.