electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links
Forum with support of Syndicator RSS
 FAQFAQ   CalendarCalendar   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   LinksLinks
 RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in  Chat RoomChat Room 
 Forum index » DIY Hardware and Software
power isolation question
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [6 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
e1999



Joined: Apr 25, 2019
Posts: 29
Location: naples

PostPosted: Fri Sep 06, 2019 2:37 am    Post subject: power isolation question Reply with quote  Mark this post and the followings unread

hi guys,
i have a question about power isolation in a project that I'm working on. I'm new to this and any information/ hints/ tips are greatly appreciated.

I've been working to combine several simple circuits into 1 integrated instrument. they include a discrete twin -t voice, white noise, vca, diode env, 4017 sequencer, and preamplifier. each module is on its own protoboard with power coming from a central 12v single supply adapter connected to two 10 pin headers (1 header for positive, 1 for ground).

with 1 audio output connected i don't have any troubles but if I connect more than 1 audio output to the mixer (for example twin-t oscillator and the vca output) there would be a whining oscillation noise in the background. it was impossible to get rid of noise even if I were to power only the two blocks that I connect to the mixer and leave everything else unpowered. I have bypass capacitors on each section, also tried inductors, and changing the order that the grounds for each module were connected in with relation to the power supply.

then i found that using diodes (bat85 and 1n4001 were what i had available) between the main power source and the + and ground for each individual block of the instrument, I have no noise at all.

I've looked at schematics for isolated power supplies for guitar pedals and haven't found anything quite like this so it makes me wonder...

my question is, are there any problems that anyone can see from using diodes in this way?

thanks!


diode power.jpg
 Description:
 Filesize:  95.09 KB
 Viewed:  145 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

diode power.jpg


Back to top
View user's profile Send private message
gabbagabi



Joined: Nov 29, 2008
Posts: 651
Location: Berlin by n8
Audio files: 23

PostPosted: Fri Sep 06, 2019 4:23 am    Post subject: Reply with quote  Mark this post and the followings unread

it is very interessting what you found out (at least for everyone who encountered this noise or even "ghostmodulations" (like myself))

The diodes from +12V going to the module shouldn't be a problem,
as si-diodes have a typical voltage drop of 0.6V,
your modules would work with 11.4V instead of 12V as positive supply voltage.

iam not so shure about the doides going to ground,
cos your modules would work with 0.6V instead of 0V as GND.
Which could be fine in your system, dunno if there would be magic smoke but it would be immediately a problem if you connect your system to something else Shocked


have you tried if your trick is working when you put the diodes only on the +12V?

do you use the bypass caps in combination with a resistor to form a low-pass filter, usually 22µF/10R ?
do you use decoupling caps (100nF) on each IC?

sometimes it makes sense to have two seperate GND, a clean and a dirty one (cGND & dGND Very Happy ).
LEDs (and i guess some logicICs too) are famous for causing such problems, so those would be the first candidates to be connected to the dGND.

((here is one point to think or talk about: it came to my mind that we are always thinking and talking that the current is comming from the +12V, but in fact our current is made out of electrons, since they are of negative nature they coming from GND and going to +12V.
So if we would invert our view for a moment, we could say we have two seperate "power lines", one dirty and one clean.))

edit: some month back there was a thread where "all" problems where related to virtual ground? JovianPyx (and others) has writen a lot here about virtual-ground
eg:
http://electro-music.com/forum/topic-66839.html
http://electro-music.com/forum/topic-70709.html

so far,
my 0.02$
gabbagabi
Back to top
View user's profile Send private message
e1999



Joined: Apr 25, 2019
Posts: 29
Location: naples

PostPosted: Sat Sep 07, 2019 2:34 am    Post subject: Reply with quote  Mark this post and the followings unread

Hi , thanks for the reply:)
Quote:
The diodes from +12V going to the module shouldn't be a problem,
as si-diodes have a typical voltage drop of 0.6V,
your modules would work with 11.4V instead of 12V as positive supply voltage.

iam not so shure about the doides going to ground,
cos your modules would work with 0.6V instead of 0V as GND.
Which could be fine in your system, dunno if there would be magic smoke but it would be immediately a problem if you connect your system to something else


So strangely, according to my multimeter the voltage is actually being boosted.
From post diode ground to post diode +V is now 16v. Also, the new ground is now 6v above the old ground. I think somehow this can't be right, but that is what I'm measuring.

As for connecting to other equipment, so far I've connected to my audio interface, 2 different mixers, 1 that pulls ac power straight from the wall and another that uses a wall wart, krk monitors and via cv to moog 32 synth. no magic smoke yet.

I looked around on the internet and found on some non music just general electrical engineering forums, some bits of information that if I understand correctly is saying that diode in series with ground to isolate ground was something that has been used in the past like 1970's but no longer allowed by code, all this info is in reference to people working on projects with mains power supply much bigger loads, etc... but in any case among other things you have a potentially unstable ground.


Quote:
have you tried if your trick is working when you put the diodes only on the +12V?

yes the trick seems to work if connected only to +12v with the addition of inductor coils on the input to the power of certain modules. I have not tested it yet in this configuration though, running for example the output of the twin-t into the preamp, which was another scenario that was creating noise.

Quote:
do you use the bypass caps in combination with a resistor to form a low-pass filter, usually 22µF/10R ?
do you use decoupling caps (100nF) on each IC?

re: bypass caps, probably a dumb question, but where would you put this bypass cap with resistor? for example in the attached schematic of an electret microphone, there is a bypass cap at the audio ouput, would you put the 10r resistor there too?

re: 100n coupling cap, yes to this and i also put a larger cap and in some cases like the mic preamp an array of capacitors.


Quote:
sometimes it makes sense to have two seperate GND, a clean and a dirty one


yes this is good advice, but in my case even if I have just two very simple discrete circuits and nothing else sharing power - no ic's, no leds, for example twin -t oscillator and filtered white noise burst. I still have this whining noise problem if both of them are connected to the mixer on their own outputs. And if I have only 1 output from my instrument connected to the mixer, I can have literally all modules connected to power and the system is quiet. It seems to be maybe related to some kind of ground loop being formed when more than 1 audio output is connected to the mixer?

maybe op amp buffers on the audio outputs could resolve this as well?


Electret_condenser_microphone_schematic.png
 Description:
 Filesize:  4.17 KB
 Viewed:  2756 Time(s)

Electret_condenser_microphone_schematic.png


Back to top
View user's profile Send private message
e1999



Joined: Apr 25, 2019
Posts: 29
Location: naples

PostPosted: Sat Sep 07, 2019 3:38 am    Post subject: Reply with quote  Mark this post and the followings unread

also, if anyone wants to comment -
i'm curious how with regard to running the ground through a diode, and the problems related such as a shifted ground -
- how is this different than a bridge rectifier?
isn't the ground going through diodes there as well?

if forward voltage is dependent on temperature and current draw, wouldn't a diode bridge be susceptible to these issues as well?
Back to top
View user's profile Send private message
gabbagabi



Joined: Nov 29, 2008
Posts: 651
Location: Berlin by n8
Audio files: 23

PostPosted: Sat Sep 07, 2019 5:34 am    Post subject: Reply with quote  Mark this post and the followings unread

dont have time for more now


kuppel.png
 Description:
 Filesize:  48.08 KB
 Viewed:  135 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

kuppel.png


Back to top
View user's profile Send private message
JovianPyx



Joined: Nov 20, 2007
Posts: 1988
Location: West Red Spot, Jupiter
Audio files: 224

PostPosted: Sat Sep 07, 2019 6:39 am    Post subject: Reply with quote  Mark this post and the followings unread

e1999 wrote:
also, if anyone wants to comment -
i'm curious how with regard to running the ground through a diode, and the problems related such as a shifted ground -
- how is this different than a bridge rectifier?

It's not. All diodes have a "diode voltage" that they "steal". However, in a PSU, the bridge or other rectifier starts out with several more volts than is needed for the regulator and none of the voltages on the bridge rectifier are used as references (such as is the case for ground or a PSU rail). Once the bridge supplies current to the regulators, the diode job is done and the actual ground is formed after the bridge, not before it.

Quote:
isn't the ground going through diodes there as well?

Not in the way you're thinking. Yes, current that ends up at ground had to go through the bridge, and while voltage is lost there, it doesn't matter because the ground is formed after the diodes, so those diodes don't subtract anything from the actual regulated PSU output voltages.

Quote:
if forward voltage is dependent on temperature and current draw, wouldn't a diode bridge be susceptible to these issues as well?

Yes, that's part of the law of physics, however, at the currents and voltages of common PSU circuits, the effect is negligible. Electronics is generally divided into to categories - small signal and large signal. Large signals include the currents and voltages in a PSU. Small signals are audio, CV and other very low current signals. Large signals don't care about teensy weensy signal variations because they are large. Where the temperature matters in a diode (or transistor) are things such as expo current converters. We just happened to get lucky with expo in that the characteristic that causes the conversion arithmetic just happens to be a characteristic rather grossly affected by temperature (while most characteristics are either not affected or are affected in a negligible manner). For the most part we don't need to worry about diode and transistor temperature. In synth electronics, the place where that matters is the expo converters. That is why you see/hear about "tempco" resistors - they correct for the temperature characteristic of a transistor used as an expo current converter.

While I've seen needs for diodes placed in the path of a PSU rail they are quite rare and I've not seen it on the ground. Ground is a sacro-sanct connection of much mystery, but it shouldn't be. Ground is supposed to be a zero volt reference and that is all it should be. This means that you want the ground rail and connections to be as low resistance as you can make it. A circuit that has several different gound potentials is asking for all kinds of weird problems. Grounds as well as PSU power rails need to be distributed as a "star" configuration. That is, there is a central high current capacity connection at the PSU. From this point, ground connections should radiate out like the arms of a star. Daisy chained ground and power connections is another way to ask for noise and other weird problems.

A good PSU with proper ground and rail connections is paramount to a low noise high quality project. Done correctly, there is no need for any diodes to do "isolation" of power, especially in an audio circuit. Other considerations come into play when your circuit is a mix of large signal high power needs as well as small signal. Unless you're mixing stage lighting with your audio circuit, this shouldn't be a problem.

As a side note, there are people who install 10R, 20R and/or diodes on their modules' power connections. This is for what is known as "dummy proofing" the circuit. Those devices can be plugged in power-backwards and the board won't blow (hence dummy proofed). However, there can be odd costs to these components that are there only to protect against something that should, in fact, never happen in the first place.

_________________
FPGA, dsPIC and Fatman Synth Stuff

Time flies like a banana.
Fruit flies when you're having fun.
BTW, Do these genes make my ass look fat?
corruptio optimi pessima
Back to top
View user's profile Send private message Visit poster's website
Display posts from previous:   
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [6 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum


Forum with support of Syndicator RSS
Powered by phpBB © 2001, 2005 phpBB Group
Copyright © 2003 through 2009 by electro-music.com - Conditions Of Use