Author |
Message |
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Thu Jul 04, 2013 1:18 pm Post subject:
What I'm making and ?about power supplies? Subject description: Advice Request |
|
|
I have started building this thing... a suitcase of wonders in space time -
it currently includes 2 of my PT2399 delay designs (pictured below) that each have a modulation circuit inspired by the Little Angel, and 3 pre-amps (not pictured): 1 for an electret microphone and 2 for contact microphones, other instruments can be plugged straight into the delays
I took these pictures earlier today before I finished wiring the delays and added the preamps
As you can see there is a lot of space to use on the box still. I have a bunch of ideas for things I want to add, probably some oscillators, an active splitting matrix, distortion and some filters. I'm going to try Ray Wilsons state variable filter based on the LM13700
So, onto my question - as the project gets bigger, I'm wondering what I should do about power supply requirements. Some of the things I have built are fine with a standard single supply, I'm currently running 12v at 500ma and I am regulating it down to 9v and 5v for the different boards (I realised that I hadn't thought about using a single regulator for all the 9v and 5v requirements, each board has either a 7809 or 7805 depending on it's needs, not sure how wasteful this is). I'm not sure how much load there currently is on the supply, how do I measure it? I'm sure it is nothing at the moment, but I guess I would like to understand this
The other question I have is that the state variable filter will require a dual supply, and I will probably upgrade other parts of the circuit to do this later as well (the op-amps I use work with a single supply, but I they seem to be louder and sometimes clearer with dual) but other parts will only run on a single supply - so is it ok to just use half of a dual supply for some parts of the system?
I am eyeing up the MFOS power supply at the moment as well...
(this next bit is a little story that you might find pointless so skip it if you have better things to do)
on another subject, the box i'm housing it in was a lush find, I'm taking part in an event in Brighton and hadn't been before, found some cool shops like one called Snoopers Paradise which had loads of ancient forgotten relics from the past - kinda expensive but lots of lovely weird stuff, I mean LOTS! behind it however there was a flea market in a gorgeous old building which is where I found that wooden box. It is a weird one, inside it had a wooden reel which had some electrical wire around it - I was interested because the wood on top was the perfect thickness for mounting parts as well, and it looked tasty so I bought it for £10. when I got back to the event I inspected the box a little closer and inside there was an empty tape spool and an old electret microphone, wtf? I tested the microphone later and it works really well (used Nic Collins electret circuit) so that was cool - but I have been wondering what the box was originally for - I'm thinking it must have been used in a courthouse or someplace, would have originally had a mini reel to reel in it for recording voice dialogue or something... anyway I was pretty smitten _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Fri Jul 05, 2013 3:44 pm Post subject:
|
|
|
I finished my first plans and everything is working fine
did my power supply question make sense? _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5581 Location: Moon Base
Audio files: 705
|
Posted: Fri Jul 05, 2013 4:00 pm Post subject:
|
|
|
yes, you can combine both single and dual supply circuits, just keep in mind that your 'ground' levels are different. So you'll have to use decoupling
caps between signal lines (in-/outputs)of your circuits (which you'll probably allready have on the single supply circuits).
Use the dual supply as much as possible so you can avoid using virtual/fake grounds. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Sun Jul 07, 2013 1:35 am Post subject:
|
|
|
ok cool
I have 5 boards in there at the moment, each has an individual regulator - how big a deal is this as they all currently run on either 9 or 5 volts
Can anyone help explain how to measure load? _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
analog_backlash
Joined: Sep 04, 2012 Posts: 393 Location: Aldershot, UK
Audio files: 21
|
Posted: Sun Jul 07, 2013 4:32 am Post subject:
Re: What I'm making and ?about power supplies? Subject description: Advice Request |
|
|
SineHacker wrote: | I inspected the box a little closer and inside there was an empty tape spool and an old electret microphone, wtf? |
Maybe it's connected to the NSA PRISM program... |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Sun Jul 14, 2013 7:39 am Post subject:
What I'm making and ?about power supplies? Subject description: now also about filters/buffers?? |
|
|
From Prisms good-old analog days?
I added a pair of low pass filters today, a rendition of the LPF from Ray Wilson's Noise Toaster... they kinda work:
tested them on breadboard and they worked really, really well! built on the veroboard and they worked fine - though I tested with guitar and a CMOS drone thing.
When I built it into the suitcase I quickly found that it didn't like the outputs from my contact mic buffers, my electret mic or the delays... so yeah everything already built into the box.
The delays also have a buffer circuit so I guess it could be my dodgey buffers, but then the electret mic doesn't and that won't work either - inputing any of these things is just giving me distortion and thin tinny sounds.
bypassing the buffer completely and just plugging a contact mic into the filter works fine, and I did test it out with the things I tried at breadboard stage with good results
Sooooo... I'm wondering if there is a simple circuit I can add before the filter that will allow it to process generic inputs? filter schematic attached:
_________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Sun Jul 14, 2013 7:41 am Post subject:
|
|
|
just thought of something, I connected the ground of the input and output jacks to the virtual ground, could that be causing the problem being as the other sections of the suitcase are all powered from the same supply? _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Sun Jul 14, 2013 8:17 am Post subject:
|
|
|
Yep that fixed it
anyone wanting to try building the filter, I also removed the 10k resistor near the output pulling to ground _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 16, 2013 1:34 am Post subject:
|
|
|
I had a little spacetime jam with it
https://soundcloud.com/ginkoband/suitcase-mini-jam-ting
(not sure if I can add a soundcloud player to the forum)
I haven't really got on top of power supply issues yet, the box is suffering from power supply ring noise at around 9900 - 10000hz (however the frequency of the ring shifts as more/less modules are introduced)
I'm not sure if I should tackle this at the start of the supply or if it would need dealing with after each regulator... _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Tue Jul 16, 2013 6:30 am Post subject:
|
|
|
How are your bypass caps? |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 16, 2013 7:13 am Post subject:
|
|
|
Right, well a bit messy haha I have 100nf caps across the input and output of each regulator, then some have a 100mf cap and a diode across the regulator input, I think the wall wart I am using has been noisy before when I have used it with guitar pedals - I should add that all the modules are always on and just have a bypass switch, so I was slightly confused with the ringing changing in pitch _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 16, 2013 9:49 am Post subject:
|
|
|
***there are 6 boards which each have different power requirements. As I said earlier I wasn't sure at first how to approach this so at the moment I still have a regulator on each board. Also, as I didn't start with a dual supply, I have some parts working with virtual ground (I know, don't hit me!!) not sure if this is the culprit of the ring noise.
I think for now I am going to try a filtering capacitor and resistor at the dc socket before the supply gets distributed to the rest of the boards _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Tue Jul 16, 2013 9:51 am Post subject:
|
|
|
I'd expect the change in pitch is due to change in current draw.
Normally there are larger caps at the start; don't know if your individual boards have that, but commonly I see 10uF on each rail to ground at each separate PCB. And on power supplies, usually even higher, 100uF or more before regulators.
Also, for most ICs you ought to have a small (I've seen .1uF or .01uF used) capacitors at the power pins between power and ground, as close as possible to the power pin in question (e.g. pin 14).
This is from observation, I can't say the bigger ones are required or not, or causing the ringing because I'm not the power supply guru around here. But I'd give some of that a shot to see if more bypass caps help or solve the problem.
The other possibility that occurs to me is.... I'm assuming a DC output supply, which means it's a switching supply. The switching can cause noise, although I wouldn't expect it to change pitch if that were the problem.
You might take a look at Ray Wilson's Wall Wart power supply which uses an AC wall transformer to get 12 or 9V. |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 16, 2013 11:02 am Post subject:
|
|
|
I was considering using Ray Wilsons supply, (by the way elmegil I think we actually had pretty much this exact same conversation in an older thread on a drone instrument!!)
I added a 220mf cap in parallel with a 47nf across the supply and that seems to have improved it somewhat, as in I can't hear it now. Something occurred to me though, I tested the suitcase in uni last and the mains in uni are really dodgey anyway so I'm wondering how much of an effect that had... It can't hurt that they are there though! _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Tue Jul 16, 2013 11:40 am Post subject:
|
|
|
SineHacker wrote: | I was considering using Ray Wilsons supply, (by the way elmegil I think we actually had pretty much this exact same conversation in an older thread on a drone instrument!!) |
I wouldn't be surprised at all. I have a lot of other stuff going on, so my medium term memory is pretty bad... |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 30, 2013 4:09 pm Post subject:
|
|
|
Sooooo,
using something similar to the MFOS dual supply, I want to provide ~9v to some parts of the circuit and ~5v to others;
specifically I want to build a VCA using the LM13700, the chip needs the ~9v supply but the CV input should be ~5v
I want to use a LFO for the CV input, so if I make a triangle osc with a NE5532 and power it with ~5v, that means the amplitude of the triangle will range from -5v to +5v right? that is the first question... (**I had the VCA kinda working but I powered both the VCA and the LFO with ~9v and noticed distortion near the peaks of the LFO waveform which got me thinking about this**)
second question, can I simply divide down the ~9v supply to power the ~5v parts of the circuit using some resistors? something like a pair of 1meg in series between +9v and ground, with +4.5v coming from between the two resistors (and likewise for -9v and ground) or should I just use 7809/7909 and 7805/7905 regs at the supply just to provide rails for all points
I can post up some schematics to clarify these bits if it helps _________________ aidanrtaylor.co.uk Last edited by SineHacker on Wed Jul 31, 2013 2:12 am; edited 1 time in total |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Tue Jul 30, 2013 4:54 pm Post subject:
|
|
|
waaaaaiiiitaaaaminute...
thinking about it, can I just run both chips on the same ~9v supply and divide down the amplitude of the triangle instead? would that make more sense? _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Tue Jul 30, 2013 5:16 pm Post subject:
|
|
|
Yes that would make more sense |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Wed Jul 31, 2013 2:12 am Post subject:
|
|
|
any tips on how to do that? _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Wed Jul 31, 2013 2:20 am Post subject:
|
|
|
this is what I got so far:
_________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Wed Jul 31, 2013 5:02 am Post subject:
|
|
|
My brain does not have enough bandwidth right now (trying to get a technical certification, taking the test this afternoon) to go back and figure out what you're trying to accomplish with your triangle, but it is normal even when you are working with signals that aren't rail to rail (e.g. 5V) to still use the higher voltages for your op amps.
I'll try to look at this again in the next couple of days when I have some brainage back. |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Wed Jul 31, 2013 5:12 am Post subject:
|
|
|
Good luck!
I'm using an integrator to comparator for the lfo if that helps, i am sort of controlling the amplitude with a pot already that works as an lfo depth, that just takes the output from the lfo on one ear, couples to ground on the other and the wiper goes to the cv input on the LM13700
I have a fairly complete schematic of this I think _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Wed Jul 31, 2013 7:58 am Post subject:
|
|
|
Here they are
_________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
elmegil
Joined: Mar 20, 2012 Posts: 2177 Location: Chicago
Audio files: 16
|
Posted: Wed Jul 31, 2013 10:02 pm Post subject:
|
|
|
So basically you just need to scale your LFO output so that it's 10Vpp (+5 / -5, 10V between them).
When you run the LFO as the schematic, what are the peak voltages?
The simple thing to do would be to put in a voltage divider that would take it down to 10Vpp and then buffer that with another op amp configured as a follower (input into + terminal and just a wire from output to - terminal.
I'm not experienced enough to say that doesn't come with some other downside, but it's what I'd try out on the breadboard to start with.
Edit: of course it's probably better to do the scaling WITH the op amp, so it has the right gain. But then you'll need two op amps because the only way to scale down requires you to use the - input... unless you don't care about the phase of the LFO, which maybe you don't.
http://www.radio-electronics.com/info/circuits/opamp_basics/operational-amplifier-gain.php
Gain is -R2 / R1 -- let's say your Triangle is 15Vpp (chosen to make life simple), you'd want 1/3 to get 5V, so you'd do something like R2 = 10K and R1 = 30K.
If you do care about phase, you'll need another op amp inverter with unity gain (R1 = R2, commonly in the 10K range) before the output. |
|
Back to top
|
|
|
SineHacker
Joined: Mar 09, 2010 Posts: 99 Location: United Kingdom
|
Posted: Thu Aug 01, 2013 2:06 am Post subject:
|
|
|
Ok, I think I can make this fairly simple then and just change the resistor values of R11 and R12 in the VCA schematic. I have a digital oscilloscope which has a few too many features sometimes and I haven't figured out how to measure the peaks of waveforms yet (if I can) but I'll figure that out next, cheers! _________________ aidanrtaylor.co.uk |
|
Back to top
|
|
|
|