Author |
Message |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Sun Apr 10, 2016 2:32 pm Post subject:
single and split power supply best practices? Subject description: How to power multiple IC's that require different voltages |
 |
|
I hope this is the right forum to post this question concerning powering my synth. If not, please advise.
I am currently building a Lunetta type synth. Most of the IC's are CMOS and run on a single +9v supply as would be expected with a lunetta. I am powering it with a simple 9v wall wart or battery. I want to add some envelope generators, VCA's and VCF's that are OPamp based to the circuit which require a split supply and virtual ground reference. Therefore, I am splitting the 9 volts into +4.5v/ 0v/-4.5v for these OPamp IC's. I know EG's VCA's and VCF's are not typical with Lunettas but I want to experiment anyway.
My question is this:
What is the best way to power multiple IC's that require single supply and split supply power?
Is it best to just buss the +9v and 0v to the circuit and then create a separate voltage divider for each IC that requires the +4.5/0/-4.5 split supply?
Can i use just one voltage divider and split the 9v supply first, then buss this +4.5/0/-4.5 to the circuit powering all the IC's that require the split supply AND THEN pull from the +4.5 and the -4.5 (which is 9v) for the IC's that require a single 9v supply?
Thanks for any suggestions! _________________ Jason |
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1310 Location: Netherlands
Audio files: 30
|
Posted: Sun Apr 10, 2016 2:49 pm Post subject:
|
 |
|
There is a short answer and a long one...
The short answer is: it depends.
The long answer is: it all comes down to the amount of power you use on each voltage.
If you use the virtual ground just for reference, and draw very little current from it, I would go for a 4.5v reference voltage ic like a RF194 from analog devices or the like.
But if you plan to sink or source from this virtual ground you might be better of with two (or maybe three) power supplies. |
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1310 Location: Netherlands
Audio files: 30
|
Posted: Sun Apr 10, 2016 3:03 pm Post subject:
|
 |
|
There might might be another solution, one I have used in the past.
Once I needed a negative power supply of a few volts, and used a voltage inverter., but you can only use this if you draw very little current from the negative voltage. |
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5797 Location: Moon Base
Audio files: 709
|
Posted: Sun Apr 10, 2016 3:04 pm Post subject:
Re: single and split power supply best practices? Subject description: How to power multiple IC's that require different voltages |
 |
|
You've got several options depending a bit on what your plans for the future are.
The BEST way would be to create a dual supply with an actual ground. You could either use a center tapped transformer
for this but you can also get great results with a single AC wallwart. You can have a look at the MFOS wallwart supply
on how to do this. If you also want to use batteries you would have to use 2 batteries wired in series where the connection
between them will be your ground/0V. Most if not all CMOS chips will run on 18V but check the datasheet to be sure. You would
probably have to add a switch to choose between AC or battery power. You could also run the CMOS of 9V and add AC coupling
caps at inputs/outputs of circuits that need the dual supply.
Another option would be to create a fake ground but the succes rate of this depends a bit on the circuits you use.
To do this you can sometimes get away with just a simple voltage divider made with 2 resistors but you will get a more
stable 'ground' if you connect an opamp to it configured as a voltage follower. If you use multiple circuits that need a dual
supply I would advice you to create a fake ground for each circuit which you don't connect together but you will have to
add AC coupling caps to in/outputs of the circuit to remove any DC offset voltage.
Keep in mind that if you use a single 9V supply and create a fake ground your (dual supply) circuits will only have -4.5V/+4.5V which
doesn't leave a lot of headroom. So you might run into some problems if you connect the outputs of CMOS chips (9V) directly to it
without attenuation. On the other hand they are squarewaves so it might not matter very much. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Sun Apr 10, 2016 4:22 pm Post subject:
|
 |
|
Grumble wrote: |
But if you plan to sink or source from this virtual ground you might be better of with two (or maybe three) power supplies. |
Please elaborate what you mean by two or three power supplies? _________________ Jason |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Sun Apr 10, 2016 4:27 pm Post subject:
Re: single and split power supply best practices? Subject description: How to power multiple IC's that require different voltages |
 |
|
PHOBoS wrote: | If you use multiple circuits that need a dual
supply I would advice you to create a fake ground for each circuit which you don't connect together but you will have to
add AC coupling caps to in/outputs of the circuit to remove any DC offset voltage. |
not sure i understand you here. I will need to look up AC coupling caps and DC offset voltage. heard the terms before but dont know what they mean... _________________ Jason |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Sun Apr 10, 2016 4:34 pm Post subject:
|
 |
|
Great feedback already! Love it! Let me give some more info on what I am trying to power.
Everything is on breadboards at the moment. I set a goal to try and get the most out of a 9 volt powered synth as I possibly can. I realize this goal has its limitations, but wanted to see how far I can push things within this restriction and master this level of building before advancing on to dual batteries or the MFOS wall wart supply (which I have built and played with a bit). So for the moment I am focusing on powering everything from 9 volts or less.
As far as how much power I will be drawing and if I will be sinking/sourcing from the virtual ground, or just using it as a reference, I do not know. Mostly due to my limited knowledge on such things and also because I'm still adding and removing "modules" as I breadboard/discover them. Hopefully if I list out what Im building it may give a better picture on how to solve my supply power question.
So far this is what I have. 9v powered IC's which make up a kind of sample/hold and melody/sequencer thingy:
4046's
4066
40106
4017
4051
4040
simple VCA's from 2N5457 tranny's
IC's running on 5 volts which is basically atari junk (would love to get this to just run off 9 volts so i can ditch the 5 volt regulator):
78L05 regulator
7555
74HC164
74HC86
Now for the OP amps:
LM324 envelope generator and a VCF (based on MFOS noise toaster)
I plan on using a couple of these chips as you can get an envelope and VCF out of the same chip (thanks Ray!). I am trying other OP amp VCF circuits, but with less luck.
So I have the 4000 series IC's all on one breadboard and the 74HC series IC's on another breadboard and the LM324 OPamp stuff on yet another breadboard. I want to power all the boards simultaneously so I'm using a TLE2426 "rail splitter" to send -4.5/0/+4.5 down the busses on the breadboards, mainly for the LM324. From those busses I just pull off what I need for each IC (most IC's are being powered from the +4.5v buss and -4.5v buss, which should appear as 9v to the IC, right?). However, problems seemed to arise when I tried tying each of these separate breadboard "modules" together. I can only imagine that it has something to with splitting the power and expecting some IC's to "see" it as split and other IC's to "see" it as single. Not sure.
Hope this gives a clearer picture of my set up/questions. I'll research the info in the responses I've received already and see where that gets me. _________________ Jason |
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1310 Location: Netherlands
Audio files: 30
|
Posted: Mon Apr 11, 2016 2:41 am Post subject:
|
 |
|
Just Say Noise! wrote: | Grumble wrote: |
But if you plan to sink or source from this virtual ground you might be better of with two (or maybe three) power supplies. |
Please elaborate what you mean by two or three power supplies? |
Depending of how much power you need from each voltage, you might be better of with three (separate) power supplies (+4.5volt, -4.5volt, +9volt), because if you draw a considerable amount of current, the virtual ground will move away from the desired voltage. |
|
Back to top
|
|
 |
synaesthesia

Joined: May 27, 2014 Posts: 291 Location: Germany
Audio files: 85
|
Posted: Mon Apr 18, 2016 9:43 am Post subject:
|
 |
|
Just Say Noise! wrote: | (would love to get this to just run off 9 volts so i can ditch the 5 volt regulator) |
You also could run everything at 7V and create a virtual ground for the opamps somewhere in the middle. However, as PHOBoS said, this leaves you little headroom.
Rather than that, I would try to replace the HC types by other CMOS ICs, so you can use 9V. The 4030 can replace your 74HC86. The 4015 has parallel outputs as well, but different inputs than the 74HC164. Would that work in your circuit? |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Tue Apr 19, 2016 8:37 pm Post subject:
|
 |
|
synaesthesia wrote: | replace the HC types by other CMOS ICs, so you can use 9V. The 4030 can replace your 74HC86. The 4015 has parallel outputs as well, but different inputs than the 74HC164. |
I'll check out the 4030 and the 4015. Looking at the datasheets it appears that I may be able to get a 9 volt version of the atari junk console with the substitutions you suggest. It will take some messing around though as the 74HC164 and the 4015 do not have identical pinouts, and I'm just learning about shift registers. Thanks for the tip! _________________ Jason |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Tue Apr 19, 2016 8:48 pm Post subject:
Re: single and split power supply best practices? Subject description: How to power multiple IC's that require different voltages |
 |
|
PHOBoS wrote: | you will have to
add AC coupling caps to in/outputs of the circuit to remove any DC offset voltage. |
I understand coupling caps now! At least sort of. I spent the last couple days researching AC coupling caps thanks to PHOBoS tip. I've seen those caps in schematics everywhere but never really knew what they were for. Realizing that AC and DC are flowing together through the circuit has really opened up a new understanding for me. Have a lot more to learn. _________________ Jason |
|
Back to top
|
|
 |
Just Say Noise!

Joined: Nov 14, 2011 Posts: 17 Location: Portland, Oregon
|
Posted: Tue Apr 19, 2016 9:06 pm Post subject:
|
 |
|
Grumble wrote: |
Depending of how much power you need from each voltage, you might be better of with three (separate) power supplies (+4.5volt, -4.5volt, +9volt), because if you draw a considerable amount of current, the virtual ground will move away from the desired voltage. |
Ok, so this is getting to the heart of my question, I think. But its where I am a little fuzzy understanding it. Do you literally mean "separate" power supplies? As in, a 9v DC wall wart for the IC's that require 9v and ground and another separate 9v DC wall wart that is then split into +4.5v, 0v and -4.5v? Seems redundant to have multiple wall warts, so perhaps I am misunderstanding you? _________________ Jason |
|
Back to top
|
|
 |
Grumble

Joined: Nov 23, 2015 Posts: 1310 Location: Netherlands
Audio files: 30
|
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5797 Location: Moon Base
Audio files: 709
|
Posted: Wed Apr 20, 2016 10:39 am Post subject:
Re: single and split power supply best practices? Subject description: How to power multiple IC's that require different voltages |
 |
|
Just Say Noise! wrote: | I understand coupling caps now! At least sort of. I spent the last couple days researching AC coupling caps thanks to PHOBoS tip. I've seen those caps in schematics everywhere but never really knew what they were for. Realizing that AC and DC are flowing together through the circuit has really opened up a new understanding for me. Have a lot more to learn. |
ok , so now let me explain my reasoning behind using seperate grounds and coupling caps. As you now probably know using
a fake ground will add a DC (offet) voltage to your output signals. If you would use the same ground for all your circuits this
wouldn't really be a problem but as you know a fake ground can't soure/sink as much current as a real ground. So if you would
use one ground for all the circuits it will become unstable causing crossmodulation between modules, so an LFO might affect
a VCO without it being directly connected to it.
Using seperate grounds will prevent this but because of tolerances in components they will not all be at the exact same level,
so you will have slightly different offset voltages. This might not always be a problem but it can become very noticable if you start
amplifying signals. So adding coupling caps will remove any DC voltage and isolates the circuits from each other. This will also solve any
problems you can have when connecting ciruits that run from a single supply to ones using a dual supply. However there is a catch,
a coupling cap will act as a high pass filter so if you make them too small you will loose low frequencies, but if you make them large
it will become slower in removing the DC voltage. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
 |
mike page

Joined: Sep 26, 2016 Posts: 134 Location: norwich, uk
|
Posted: Thu Feb 09, 2017 1:32 pm Post subject:
|
 |
|
Hi all,
Ive got the same situation. Ive got my lunetta running on +9v, 0 and some simple filter and such circuits running on +5, 0,-5.
Can I just run the lunetta on the -5, +5 of the dual supply?
Will everything be happy?
(sorry if the answer is in the above text and I missed it)
MP |
|
Back to top
|
|
 |
PHOBoS

Joined: Jan 14, 2010 Posts: 5797 Location: Moon Base
Audio files: 709
|
Posted: Thu Feb 09, 2017 1:53 pm Post subject:
|
 |
|
Since you are using a symmetrical supply this should IMO work fine since it will cause the CMOS signals to be biased around 0V,
even though they aren't directly connected to it. If you would for example use a -6V, +9V supply than the CMOS signals would be
biased around 1.5V creating a DC offset voltage.
However it does depend a bit on the circuits you are using and how you connect them to eachother. Synth circuits that create a
gate signal often have a 0V..+5V output and connecting this to a CMOS chip that is powered by -5V/+5V will most likely not work
because the 0V is not low enough to be recognized as a low level. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
 |
JovianPyx

Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Thu Feb 09, 2017 7:32 pm Post subject:
|
 |
|
mike page wrote: | Hi all,
Ive got the same situation. Ive got my lunetta running on +9v, 0 and some simple filter and such circuits running on +5, 0,-5.
Can I just run the lunetta on the -5, +5 of the dual supply?
Will everything be happy?
(sorry if the answer is in the above text and I missed it)
MP |
The inclusion of 0 in +5, 0, -5 is just being specific that it has a real ground rail. This is probably more commonly denoted as +5,-5 or +/-5. It is also orders of magnitude better than using a virtual ground and a 9v supply. I think that +5, -5 is enough to make a very workable capable synth. Just know that for some applications headroom is a real consideration - which is why the modular folk use +/-12 or +/-15. _________________ 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
|
|
 |
Cfish

Joined: Feb 24, 2016 Posts: 477 Location: Indiana
|
Posted: Fri Feb 10, 2017 8:34 pm Post subject:
|
 |
|
Can't believe I had not noticed this thread.
I have a couple of questions.
When using the partial wave rectification used in the MFOS schematic link posted by PhoBos, is the noise level elevated from using a center tapped full wave rectified supply???
It's not lunetta, but center tapped +-15v transformers are getting hard to find.
I had been thinking of trying to calculate the capicitors to do exactly that. Especially for low consumption modules.
My initial quick calculations showed the capicitors would not be cheep to fill the dips in the rectification. And size would get pretty large for 1/2 amp supply
worth. However my curiosity is still aroused.
Edit. Also, what happens if you use to large of capicitors in a supply like this? I have a few 8800uf 370v capicitors lying around. I suspect they might cause a volt drop if the transformer is unable to charge the cap within a wavelength and might cook the transformer. |
|
Back to top
|
|
 |
JovianPyx

Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Sat Feb 11, 2017 7:27 pm Post subject:
|
 |
|
Cfish wrote: | Can't believe I had not noticed this thread.
I have a couple of questions.
When using the partial wave rectification used in the MFOS schematic link posted by PhoBos, is the noise level elevated from using a center tapped full wave rectified supply??? |
Noise can come from a power supply if it's not well designed for the application. The fact that the PSU is fullwave doesn't cause it to add noise nor does having a center tapped transformer. You can safely apply that PSU assuming it's specifications meet or exceed the needs of the application circuit.
Quote: | It's not lunetta, but center tapped +-15v transformers are getting hard to find. |
If the load current isn't huge, you might consider a half-wave design. Can still do +V and -V with a solid stable ground and you can do it without a center tapped transformer. I've made numerous projects that use a half-wave PSU with regulated outputs using an AC wallwart transformer. My current mini-supply makes +8v -8v and +3.3v from a 9V AC 1000 mA wallwart. You do need bigger filter caps on a halfwave supply. My rule of thumb is double (or more) the size needed for fullwave. YMMV and the cap size really depends on the current draw - the more - the bigger.
Quote: | I had been thinking of trying to calculate the capicitors to do exactly that. Especially for low consumption modules.
My initial quick calculations showed the capicitors would not be cheep to fill the dips in the rectification. And size would get pretty large for 1/2 amp supply
worth. However my curiosity is still aroused. |
I use >= 3000uF per rail, and most the stuff I've done draws between 150mA and 350mA. I would think that your 8800uF mentioned should be plenty, but you really don't need them to be 370v. 35V caps will be a lot smaller.
Quote: | Edit. Also, what happens if you use to large of capicitors in a supply like this? I have a few 8800uf 370v capicitors lying around. I suspect they might cause a volt drop if the transformer is unable to charge the cap within a wavelength and might cook the transformer. |
Too large filter caps can cause problems with "in-rush" current. When the caps are fully discharged, the initial application of power will cause a current pulse when the PSU is powered on which can be problematic if the rectifier diodes cannot withstand the current. It is possible to make them too large and from a practical perspective, but there's no point in making them any larger than is necessary for the load circuit. Larger uF and V caps are also more expensive. Sure, if you have caps laying around you want to use up instead of buying new ones, try it. If you're unsure of in-rush, then make the rectifier diodes bigger (i.e. carry more amperes). The inrush current problem will occur on parts that the current goes through, namely the rectifier diodes.
HTH _________________ 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
|
|
 |
Cfish

Joined: Feb 24, 2016 Posts: 477 Location: Indiana
|
Posted: Sun Feb 12, 2017 4:23 pm Post subject:
|
 |
|
Thanks JovianPyx very helpful. |
|
Back to top
|
|
 |
mike page

Joined: Sep 26, 2016 Posts: 134 Location: norwich, uk
|
|
Back to top
|
|
 |
|