Powering a Distortion Device With Phantom Power

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Is there any difference between 1 and 2?
Besides being cheaper, the SparkFun modules are much better performance.

I would need to to use a ±15V supply.

You could still just build your own the way the Silicon Chip I/O module does. That is basically the same as that design in #6 but it doesn't use a stupid expensive buffer for the derived gnd. Even the Silicon Chip circuit has extra stuff that isn't really needed if you start out with a regulated wall wart type supply, like the resistor/zener combination to limit to 30V. That was probably because they wanted to try to make things as fool proof as possible and allow some leeway in what power adapter people connected, but you have control of everything in your project, so just find a power adapter with the voltage you want. Start with a regulated supply and you don't need the additional linear regulators like that Haraldswerk board has, just make a voltage divider and add an op-amp buffer and call it done.
Given that you are using existing modules, rather than starting with a handful of parts and trying to make your own all-in-one circuit, it seems like just buying a dual supply would be more straight forward, unless you already had some power supply you were really set on using.
 
At this stage I think I'm going with ccaudle's suggestions (SparkFun I/O, AdaFruit PS).

@ccaudle You mention that if I decided to use the AdaFruit supply that I might need to add "some extra capacitors, or maybe inductors and capacitors if it has noise that feeds through the power connections."

What would that look like?
 
There are a number of dual output small wattage DCDC SMPS. That might be the path-of-least-resistance for a small hobby project. Power it with an old laptop brick and make a dual capcitance multiplier (google it - very simple circuit) to filter the output. If you pick the right SMPS and with a little luck you might end up with a simple and very high quality supply.

For an example check out MeanWell MDD06. I just found this using parametric search on Mouser. I've never used it (it looks new - there are a lot of new small DCDC converters) but it should work. The smaller ones can have issues with the voltage drooping a little and being a little noisier. But something like the +-15V version after capacitance multipliers, you could get clean +-12V at least. You would have to just try it and see.

Note that the biggest problem with your use case is that the load is very small. One of the issues with SMPS is that you have to load it enough to keep it from modulating the output from underload (almost every post about SMPS here warns about this). So for your case you want a small SMPS. Check the AU sites for small 5W-ish dual SMPS. Particularly the MeanWell ones are a good bet. Look at au.mouser.com.
 
What would that look like?
Capacitor to gnd, series inductor, and capacitor to gnd on each side. CLC low pass filter. If it comes to that you will probably need access to a 'scope to verify the noise behavior so you have an idea of the frequency range you need to reject.
 
What about using one of these (as suggested by Bo Deadly):
Too small. The really small ones will not put out the +-33mA at +-15V. After CMs you'll only have maybe +-20mA at +-10V or something like that. Look for 5W-ish one but you could try a 3W part. And make sure it's "isolated".

But again, for the small ones you will definitely need CMs.

If you made a PCB through OSHPark with a barrel connector, PCB mount isolated SMPS, CMs and pin headers, that could be a very nice little small reusable supply.
 
Yes, that one might work.

CM means Capacitance Multiplier.

Here's a version for bipolar supply:

1653489193912.png
But you would need to adjust a bunch of stuff. For one, C1 and C2 cannot be larger than what is on the datasheet. In practice they would be like maybe 47uF but 10uF would probably work equally well. Then adjust R1+R3 and R4+R2 to load the SMPS enough to keep it from modulating the output due to underload (as mentioned previously). Then adjust C3 and C4 to make a sub-1Hz RC filter with whatever said resistors end up being. The 1000u on the outputs is fine but going lower would probably work ok too if you have say 470uF laying around.

The CM is going to be important for really quiet performance. Especially for a little single-ended distortion circuit that has no power supply rejection.
 
Would the power supply as it now stands (wall wart -> DC/DC converter -> CM) be able to power a stereo version of the unit?

It would need to power 4 SparkFun modules and 2 Colour modules.
 
Can the SparkFun I/O modules be run from a single supply

They can, in the sense that any circuit can with some external modification. The ThatCorp chips need a low impedance at the reference pins, so you would need a buffer after the resistor divider that generates the mid-supply voltage. You would also have to consider what that does to the input and output voltages, and check where you have coupling caps. That turns into a system consideration, to be able to use anywhere with any other device you would probably need caps on the input and output.
If you will only ever use it in one system, you could for instance leave off output caps if you know the next device has input caps, or leave off input caps if you know the device which is driving it already has output caps. Slightly dangerous, because in a couple of years when you have forgotten all that, you connect to different devices then have to figure out why everything is suddenly distorted when it was working fine the day before.

What about using one of these (as suggested by Bo Deadly):

Sure, but that only helps if you have some power supply you were already set on using. Otherwise you are paying for a wall adapter, then another $12 for the DC-DC converter, plus a couple of dollars for some capacitors, and you still have to wire it all up, compared to just spending $15 on a bipolar wall adapter.

If you are going to that much trouble you might as well just skip all the hassle with switching regulators, get one of these 12V AC adapters, and build a half-wave rectified supply with a couple of 12V regulators:
12V AC adapter for Australia

Couple of diodes for half-wave rectification plus some electrolytic caps will get you +/- 17V, 78M12 and 79M12 regulators, couple more caps and done.

be able to power a stereo version of the unit?

You really need to decide what you want to do. You started out asking for a phantom powered circuit, which is a few mA at most, and are now up to a stereo circuit with additional input and output buffers, which has nearly tripled the current requirement per channel, and now double that for stereo.

The VERY basics of any project is to first sit down and decide what you want, then go through all of the datasheets and add up the current required for each part of the circuit, then add in the current required to drive the next downstream stage, and add all that up to get your total power requirements.
THEN you start looking for a power supply solution after you know the voltage and current requirements for your project.

So take a few minutes and think about what you really need. Do you need input and or output level controls? Is that going to require any additional buffering? You started out with a simple unbalanced circuit, and now have balanced input and output buffers. Most unbalanced gear uses different operating levels that balanced gear. Have you considered what you want your default input and output nominal levels and headroom to be? Where in that nominal input/output range do you want noticeable colouration/distortion to kick in? Do you want to control that strictly with the send levels of the upstream device and the input attenuation/make-up gain on the downstream device? Or do you want a gain control before and after the colour circuit so you can control the distortion level independent of the operating audio levels (kind of like the gain and master volume controls on a guitar amp)?
 
They can, in the sense that any circuit can with some external modification. The ThatCorp chips need a low impedance at the reference pins, so you would need a buffer after the resistor divider that generates the mid-supply voltage. You would also have to consider what that does to the input and output voltages, and check where you have coupling caps. That turns into a system consideration, to be able to use anywhere with any other device you would probably need caps on the input and output.
If you will only ever use it in one system, you could for instance leave off output caps if you know the next device has input caps, or leave off input caps if you know the device which is driving it already has output caps. Slightly dangerous, because in a couple of years when you have forgotten all that, you connect to different devices then have to figure out why everything is suddenly distorted when it was working fine the day before.



Sure, but that only helps if you have some power supply you were already set on using. Otherwise you are paying for a wall adapter, then another $12 for the DC-DC converter, plus a couple of dollars for some capacitors, and you still have to wire it all up, compared to just spending $15 on a bipolar wall adapter.

If you are going to that much trouble you might as well just skip all the hassle with switching regulators, get one of these 12V AC adapters, and build a half-wave rectified supply with a couple of 12V regulators:
12V AC adapter for Australia

Couple of diodes for half-wave rectification plus some electrolytic caps will get you +/- 17V, 78M12 and 79M12 regulators, couple more caps and done.



You really need to decide what you want to do. You started out asking for a phantom powered circuit, which is a few mA at most, and are now up to a stereo circuit with additional input and output buffers, which has nearly tripled the current requirement per channel, and now double that for stereo.

The VERY basics of any project is to first sit down and decide what you want, then go through all of the datasheets and add up the current required for each part of the circuit, then add in the current required to drive the next downstream stage, and add all that up to get your total power requirements.
THEN you start looking for a power supply solution after you know the voltage and current requirements for your project.

So take a few minutes and think about what you really need. Do you need input and or output level controls? Is that going to require any additional buffering? You started out with a simple unbalanced circuit, and now have balanced input and output buffers. Most unbalanced gear uses different operating levels that balanced gear. Have you considered what you want your default input and output nominal levels and headroom to be? Where in that nominal input/output range do you want noticeable colouration/distortion to kick in? Do you want to control that strictly with the send levels of the upstream device and the input attenuation/make-up gain on the downstream device? Or do you want a gain control before and after the colour circuit so you can control the distortion level independent of the operating audio levels (kind of like the gain and master volume controls on a guitar amp)?

Thanks, yes, all good advice. The problem I have is that my general electronics knowledge is a bit limited which means that I don't know what's possible and/or the "path of least resistance" for what I'm trying to do. But I'm learning a lot which is a good thing!

What you've said in the last paragraph is good advice which I'm going to think about. The audio I/O levels can be controlled by the pre/post devices with the distortion being controlled by the input level. There are two additional things that I think would be handy. As this is a fairly subtle device some sort of bypass and possibly some sort of input level indicator would both be useful (probably not essential though). I have a box of old meters that I pulled out of some old gear - cassette players mostly but also some very old electronics stuff that might be interesting to add as a rough signal indicator.
 
Here's an idea. I've used this one for tube-like distortion. If one puts feedback around a CMOS inverter, and feeds the input with audio through an input resistor, it acts as a semi-linear inverter...semi-linear being the important bit. It will have gain, just as does an operational amp. With sufficient input voltage and sufficient gain, you get a really warm, bluesy distortion added to the output audio. I used two inverter sections to make the distortion symmetrical, and coupled the input and output with caps. The originals were enclosed in an oversized 1/4" plug and powered by a watch battery.
In regards to your second schematic for unipolar power, I would use a jelly bean amp voltage follower for 1/2 V+; your bipolar transistor circuit already has a reasonably low output source impedance, similar to a jelly bean op amp.
Just food for thought.
 
Here's an idea. I've used this one for tube-like distortion. If one puts feedback around a CMOS inverter, and feeds the input with audio through an input resistor, it acts as a semi-linear inverter...semi-linear being the important bit. It will have gain, just as does an operational amp. With sufficient input voltage and sufficient gain, you get a really warm, bluesy distortion added to the output audio. I used two inverter sections to make the distortion symmetrical, and coupled the input and output with caps. The originals were enclosed in an oversized 1/4" plug and powered by a watch battery.
In regards to your second schematic for unipolar power, I would use a jelly bean amp voltage follower for 1/2 V+; your bipolar transistor circuit already has a reasonably low output source impedance, similar to a jelly bean op amp.
Just food for thought.
old CMOS tricks... but keep the rail voltage low since both the P and N MOS devices will be conducting at the same time for linear operation which can result in dissipation issues.

JR
 
I have no idea why you guys are making this so complicated. Just replace the discrete op amp output buffer with the output buffer from a transformerless microphone:

There are a couple substitutions to reduce current draw so it will actually run on phantom power, but LTSpice says it performs pretty much the same.
 

Attachments

  • Untitled.png
    Untitled.png
    16.3 KB
I have no idea why you guys are making this so complicated. Just replace the discrete op amp output buffer with the output buffer from a transformerless microphone:
Exactly. The output buffer is not what gives the sonic character, it's the JFET stage. Actually, a buffer may not be strictly necessary, just the JFET stage and impedance balanced output arrangement.
 
Just having a look at DC-DC Converters here:

Products | Power Supplies Australia

According to Bo Deadly (above) I'm better off getting a smaller wattage one (around 5W) so which would be the best choice?
I'm assuming you consider it for phantom powering, right?
P48 has a power limit of 0.16W, so the 1W type MAY work, and none of the others (they probably would not even start).
OTOH, the JFT thingy draws 0.32W, so there's no way it can work.
So I would seriously consider Solomute's suggestion (post #36) for decreasing current draw.
Even more seriously I would forget about phantom powering.
Note that I haven't even considered losses in the converter (about 30%) and in the regulation circuitry.
 
I'm assuming you consider it for phantom powering, right?
P48 has a power limit of 0.16W, so the 1W type MAY work, and none of the others (they probably would not even start).
OTOH, the JFT thingy draws 0.32W, so there's no way it can work.
So I would seriously consider Solomute's suggestion (post #36) for decreasing current draw.
Even more seriously I would forget about phantom powering.
Note that I haven't even considered losses in the converter (about 30%) and in the regulation circuitry.

No, I'm thinking of running off a DC wall wart.
 
Back
Top