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 connectors
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [16 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
cyclic



Joined: Mar 15, 2015
Posts: 95
Location: hobart

PostPosted: Sat Feb 13, 2016 11:09 pm    Post subject: power connectors
Subject description: what am I missing?
Reply with quote  Mark this post and the followings unread

ok,

so I'd admit it, I've made all my own stuff, and generally not used standardised connections already, but I'm about to start on a new case.

So far I have used 3 pin 0.1" polarised headers for my power connections. Cheap. Seems to work fine.

I look at MOTM and see massive oversized conenctors for modules that pull, what, 50mA?

I look at eurorack and see TEN pins, for what I can see is at most 4 power rails (except I just use a 7805 on board whenever I need it)

What am I missing?
Apart from being non-standard, is there actually anything wrong with what I use so far?

cheers

lance
Back to top
View user's profile Send private message
roglok



Joined: Aug 28, 2010
Posts: 202
Location: uptown

PostPosted: Sun Feb 14, 2016 4:21 am    Post subject: Reply with quote  Mark this post and the followings unread

there's nothing wrong with using your own connectors. in fact, many people advise against using IDC headers for power (eurorack).

if you want to minimize bleed/crosstalk problems, you should use as much copper as you can fit for power distribution, especially on the 0V lines. in other words: there's no harm in going 'oversized'. to double up the rather small diameter of the euro ribbon cable, two pins of the 10 pin connector are joined for each power rail.
Back to top
View user's profile Send private message
cyclic



Joined: Mar 15, 2015
Posts: 95
Location: hobart

PostPosted: Sun Feb 14, 2016 5:41 pm    Post subject:
Subject description: copper thickness for ground lines.
Reply with quote  Mark this post and the followings unread

ok,

I did so some reading about this once before but it was all a bit too much for me at the time, and I haven't found the reference again. If there is a well known good reference which someone can point me to I'd be happy to try it again.


Yes, crosstalk is a problem for me with some modules.


From what I gather so far though:
4 pins is better than 3, since you double your ground copper.
I guess this also means that you can do a +12/gnd and -12/gnd twisted pair which might also help a bit.

So it seems I should have got 4pin polarised headers instead - and I will do so for the future build.

In general, if I do double ground lines, is it also worthwhile using thicker copper or is this over kill? I've been using either 0.4mm2 or 0.6mm2 single core (old network cable I think) because I've got a big roll of it and its nice to work with for stripboard traces. Power lines are generally about 20-30cm long.

If I do double up and/or use thicker copper on the cables, will the 0.1" header pin connectors then be a problematic bottleneck? Like, they're 'rated' for 3A but I'm getting the feeling here that this is more complicated than the mere power rating. I'd really really rather stick to 0.1" spacing rather than move up to the molex big-uns.

cheers

Lance
Back to top
View user's profile Send private message
AlanP



Joined: Mar 11, 2014
Posts: 746
Location: New Zealand
Audio files: 41

PostPosted: Sun Feb 14, 2016 8:53 pm    Post subject: Reply with quote  Mark this post and the followings unread

I seem to recall low impedance or something being a big thing with power supply lines.

I'm far from an expert, though!
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 1294
Location: Netherlands
Audio files: 30

PostPosted: Mon Feb 15, 2016 1:04 am    Post subject:
Subject description: copper thickness for ground lines.
Reply with quote  Mark this post and the followings unread

EL BO wrote:
I've been using either 0.4mm2 or 0.6mm2 single core . Power lines are generally about 20-30cm long.

10 cm of 0.5mm2 wire has a resistance of about 0.007 ohm.
one IDC connector 0.1" pitch has a max impedance of 15mohm (0.015ohm).
Do the math Wink
Back to top
View user's profile Send private message Visit poster's website
prgdeltablues



Joined: Sep 25, 2006
Posts: 222
Location: UK
Audio files: 12

PostPosted: Mon Feb 15, 2016 3:34 am    Post subject: Reply with quote  Mark this post and the followings unread

I moved from 3pin 0.1" molex headers for power to 4 pin, allowed me to have separate clean = analogue and dirty = digital grounds (they join back at the power supply, obviously). This supposedly helps keep digital switching and LED on/off noise away from your audio signal grounds. Can't say I heard any difference, but it wasn't as complicated to retro-fit to my stripboard layouts as I feared.

The impression I get is that many crosstalk etc problems are caused by insufficient current capacity from power supplies. You might want to disconnect all the modules you're not using and see if your crosstalk problem disappears. If it does, reconnect them one-by one - sometimes a particular module can cause a problem. Or else it could simply be the combined load.

I use 13/02 speaker cable wire for my power distribution.

Peter
Back to top
View user's profile Send private message
cyclic



Joined: Mar 15, 2015
Posts: 95
Location: hobart

PostPosted: Mon Feb 15, 2016 4:18 am    Post subject: Reply with quote  Mark this post and the followings unread

prgdeltablues wrote:
I moved from 3pin 0.1" molex headers for power to 4 pin, allowed me to have separate clean = analogue and dirty = digital grounds (they join back at the power supply, obviously). This supposedly helps keep digital switching and LED on/off noise away from your audio signal grounds. Can't say I heard any difference, but it wasn't as complicated to retro-fit to my stripboard layouts as I feared.

The impression I get is that many crosstalk etc problems are caused by insufficient current capacity from power supplies. You might want to disconnect all the modules you're not using and see if your crosstalk problem disappears. If it does, reconnect them one-by one - sometimes a particular module can cause a problem. Or else it could simply be the combined load.

I use 13/02 speaker cable wire for my power distribution.

Peter


ok, 'd thought about the analog/digital noise with dual power lines but wasn't sure if it was really relevant. I don't have much digital in there yet, but probably will.

I'm glad to hear someone else does use what I'm planning i.e. 4 pin 0.1" headers.

Neither I nor google can tell me what 13/02 speaker wire is. Do you mean AWG? If thats the case you use waaay thicker than I am right now, which must be about 22 AWG.

My current power supply is pretty shaky, so on the new box I am planning on using the MFOS wall-wart supply, with a 1A 12V AC supply I have already.


Quote:


EL BO wrote:
I've been using either 0.4mm2 or 0.6mm2 single core . Power lines are generally about 20-30cm long.

10 cm of 0.5mm2 wire has a resistance of about 0.007 ohm.
one IDC connector 0.1" pitch has a max impedance of 15mohm (0.015ohm).
Do the math Wink



um. ok. facts are good, but I need a reference point!
so if I use 2x20 cm ground wires to individual headers I get total impedance of 14mohm. This seems utterly insignificant to me, especially since the first thing it usually hits is a 10ohm resistor. Is this actually a 'very low impedance' supply line or is it something I should worry about?


thanks everybody!

LAnce
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 1294
Location: Netherlands
Audio files: 30

PostPosted: Mon Feb 15, 2016 5:20 am    Post subject: Reply with quote  Mark this post and the followings unread

I can't give any reference, but some more info:
For the DC currents you use the thickness is ok, problems arise when there is high frequency noise to be led to ground with wires like these.
For dc currents the entire surface will be used to conduct the current, this is different for high frequencies: The higher the frequency the more the current flows on the circumference of the conductor. This is called skin effect.
At a frequency of 50 Hz, the (effective) depth in Copper is about 1cm, at 10 kHz it is 0.66mm and at 10MHz just about 20um!
You may think that 10MHz is of no concern for you, but when you use digital parts in your synth there will be frequencies a magnitude higher as 10 MHz!!
Exactly why it is so very important to use decoupling capacitors to prevent these high frequency signals from flowing thru power lines.
Back to top
View user's profile Send private message Visit poster's website
L´Andratté



Joined: Sep 23, 2012
Posts: 150
Location: Hamburg, Germany
Audio files: 5

PostPosted: Mon Feb 15, 2016 6:31 am    Post subject: Reply with quote  Mark this post and the followings unread

Grumble wrote:
This is called skin effect.
At a frequency of 50 Hz, the (effective) depth in Copper is about 1cm, at 10 kHz it is 0.66mm and at 10MHz just about 20um!


Nice, didn´t know that!
Back to top
View user's profile Send private message
wackelpeter



Joined: May 05, 2013
Posts: 461
Location: germany
Audio files: 10

PostPosted: Mon Feb 15, 2016 9:03 am    Post subject: Reply with quote  Mark this post and the followings unread

Cross-Talk is a nice Topic... Smile

Had some Trouble with my MFOS Noise Coripucia too... which picked up Signals... at least it turned out as a Problem on the power rails...

there are many things to optimize in your System...
depends on what Kind of Problem you have...

Meanwhile i only use low current LED's when i have too many LED's like in my last rack that included a lot of dividers and sequencers i ran These from a separate PSU and all the "Musical" stuff from another...
the LED resistors are then also set so that they have the smallest current and are still lightning bright enough... keeps the current consumption small and the voltage doesn't break in... which is audible at some Point on VCO's...

the usual use of shielded cable for noise sources and one end tied to ground should be Standard... same can help with some crosstalk on oscillators where obviously the Signal with a "hard" edge tend to produce cross talk like rectangle and saw when too Close to other wires or a specific Level of the signal is reached...


also use ferrite beads instead of small resistors, this also reduces the voltage Drops across them...
well you now have to measure the resistance before you plug in your module as there are no resistors left that can be burned when you have a short somewhere...

as the others mentioned decoupling caps at the IC'S are essential and thicker cable is also helpful to some extend...
same goes for separating grounds of analag and digital (or heavy with LED overloaded, like gate sequencers) modules...

At least it's important that your PSU isn't underrated current wise... Pushing it to it's Limits isn't much recommended...

That Skin-effect thing was new to me too and the advantage of using thicker wires for grounds to my modules... Interesting new discovery.

_________________
https://soundcloud.com/bastian-j
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 1294
Location: Netherlands
Audio files: 30

PostPosted: Mon Feb 15, 2016 11:28 am    Post subject: Reply with quote  Mark this post and the followings unread

wackelpeter wrote:
Cross-Talk is a nice Topic... Smile
At least it's important that your PSU isn't underrated current wise... Pushing it to it's Limits isn't much recommended...

In the quest to lower and lower power there is a big chance of introducing problems:
1 - higher impedance's introduce higher noise levels, this is a law of nature and there is nothing you can do to prevent this kind of noise but to lower the impedance's and lower the temperature. This type of noise is especially annoying at the input of amplifiers, because the amplifier will.. well amplify the noise.
2 - higher impedance's are more prone to pick up noise from other sources like the power grid 50Hz or 60Hz hum, depending where you are and also from other switching devices (cross talk!)
Back to top
View user's profile Send private message Visit poster's website
prgdeltablues



Joined: Sep 25, 2006
Posts: 222
Location: UK
Audio files: 12

PostPosted: Tue Feb 16, 2016 2:33 am    Post subject: Reply with quote  Mark this post and the followings unread

Should have read 13/0.2 13 strands each 0.2mm diameter. Equates to 21 AWG.

Peter
Back to top
View user's profile Send private message
cyclic



Joined: Mar 15, 2015
Posts: 95
Location: hobart

PostPosted: Wed Feb 17, 2016 4:52 am    Post subject: Reply with quote  Mark this post and the followings unread

ok,

so my understanding of this discussion so far is:

4 pins are good.

22 awg is generally adequate for my power lines.

separating analog and digital ground is also necessary - hence the 4th pin - AND it would be useful to make the digital ground a multistrand cable rather than the solid core which i currently use to reduce the effects of skinning. maybe even a shielded coax?

reading between a few lines (i.e. I'm not totally sure about this!)

I should put LEDS on the digital ground if the circuit allows me to do so

the ground lnes are actually connected together at the power supply headers, and its only the power lines themselves which are separate.

I should sprinkle decoupling capacitors liberally.


is that a fair summary?



Now, not to change the direction of my own thread at all, but the MFOS wall-wart power supply uses 6x3300uF capacitors to smooth the AC ripple. Since I don;t have any of those hanging about, but I've got a few kicking about in the 6000-10000uF range (pulled from hi-fi power supplies) does it actually matter if I use 3 caps or 1? Apart from the physical massiveness of them of course. Is there some kind of 'speed' issue with 3 vs 1? In which case, can I just whack in a few of whatever pairs I've got handy?

cheers

Lance
Back to top
View user's profile Send private message
prgdeltablues



Joined: Sep 25, 2006
Posts: 222
Location: UK
Audio files: 12

PostPosted: Wed Feb 17, 2016 5:13 am    Post subject: Reply with quote  Mark this post and the followings unread

Not quite sure what you mean by 'ground lines connected at the headers' but the two grounds should connect as far back in the power circuitry as possible - mine come off from the same point at the regulators, and are separate from then on - ie separate busses, separate wires to each board, separate on each board.

Caps - I've used both 3x3300u and 2x4700u with Ray's wall-wart supply without any issues. You need enough capacitance, but don't want to overdo it, or you can end up with problems of in-rush current (current surge if large capacitors are in a discharged state could exceed the capacity of your wall-wart).

Peter
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 1294
Location: Netherlands
Audio files: 30

PostPosted: Wed Feb 17, 2016 5:34 am    Post subject: Reply with quote  Mark this post and the followings unread

EL BO wrote:
power supply uses 6x3300uF capacitors to smooth the AC ripple. Since I don;t have any of those hanging about, but I've got a few kicking about in the 6000-10000uF range (pulled from hi-fi power supplies) does it actually matter if I use 3 caps or 1? Apart from the physical massiveness of them of course. Is there some kind of 'speed' issue with 3 vs 1? In which case, can I just whack in a few of whatever pairs I've got handy?

What prgdeltablues says: too much capacitance may give rise to rush-in current problems BUT there is something more to be careful about when using large capacitors: Simple said, electrolyte capacitors are just two rolled up large pieces of very thin aluminum foil with a insulator between the plates.
Since they are rolled up they can and will also act as an inductance.
You know from inductance's that their impedance will increase if you increase the frequency from the current going thru them.
So, for low frequencies large caps are great, but for higher frequencies you need capacitors with a low impedance at high frequencies.
A pro is that these capacitors can be of a lower capacitance, I use 100nF a lot, sometimes even in parallel with 100pF and 10pF in very critical situations where high frequency switching and audio amplification are in each others neighborhood...
Back to top
View user's profile Send private message Visit poster's website
cyclic



Joined: Mar 15, 2015
Posts: 95
Location: hobart

PostPosted: Wed Feb 17, 2016 9:31 pm    Post subject: Reply with quote  Mark this post and the followings unread

thanks guys,

i'm sure this has all been covered a thousand times before, but reading online hasn't been the same amount of help as you guys answering my specific questions. I reckon I've got a much better handle on a few issues now.

cheers

lance
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [16 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