DC-DC converter for the Supergreen pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

lagoausente

Well-known member
Joined
Oct 2, 2006
Messages
288
Location
Spain
I'd consider one of the numerous 12 V to +/-15 V DC-DC converters for the audio supply. A pack of six 1.5 V cells will provide pretty long power life. Adding a pair of 10 ohm + 10 uH/220 uF filter for each channel should get rid of the switching noise.

Phantom power is more of a problem--you might consider my 48 V DC-DC converter design (see some thread in the Drawing Board, still untested though), two stacked 24 V DC-DC converters or a +/-24 V DC-DC converter with floating outputs.

I´m considering buying just a +/- 24, and regulate then to +/- 20 for the Supergreen. It also need +5V and +48. I have a few doubts:

- what out current should be ok?
- Better two 24 converters, or one dual +/-?
- In any case would get 48 V, could take it for the phantom, and at the same time fot the +/- 20 regulation?

I have found this candidates:
http://www.switching-powersupplies.com/shop/dc-dc-converter/module-type/page1.html
this ones have the price online, and seems easy to order, but I´m not sure if the current specs are enough, for example the DKE15A-24 model gives +/- 313 ma.
The other candidate could be: http://www.motien.com.tw/products/v7l.html
There are diferent models, the higher wattage gives +/- 625 ma.
Would like you opinions, or suggestions about other converters.
 
What out current should be OK?
You need to look at the circuit (supergreen_schematic.pdf). What eats how much current? We have:
* 1x NE5532 (datasheet says 16 mA max.)
* 2x NE5534 (datasheet says 8 mA max.)
* 1x bias string R19/R23 (simple calculation yields 1.7 mA)
* 2x collector current for T1/T2 (less simple calculation yields 1.3 mA)

So all in all we have 36.3 mA maximum quiescent current. Add a few mAs to accomodate increased current demande when driving long cables etc. and you get 50 mA per channel.

Phantom power is 10 mA per channel maximum (15 mA when the input is shorted to ground). Not sure what you need your 5 V rail for so you'd need to do the math yourself.

Better two 24 V converters, or one dual +/-?
A dual is more convenient and there is little reason to believe that a pair would do the job better.

In any case would get 48 V, could take it for the phantom, and at the same time fot the +/-20 V regulation?
No, if the output is +/- 24 V it can't be 0/48 V at the same time...

I have found this candidates:
http://www.switching-powersupplies.com/shop/dc-dc-converter/module-type/page1.html
Haven't looked at all the specs closely but they have a wide input range which is convenient for battery power.

Samuel
 
:thumb: Thanks!

You need to look at the circuit (supergreen_schematic.pdf).

Well, I must understad it, to know what have in account, for example, what you comment, about the bios, T1 and T2.
I´m just starting to read The Art of Electronics, I think it will take me long time, but.., fine if I can be going on with the Supergreen project at the same time.


So all in all we have 36.3 mA maximum quiescent current. Add a few mAs to accomodate increased current demande when driving long cables etc. and you get 50 mA per channel.

I thought would more, so that +/- 24V with +/-313 ma would do the job,DKE15A-24 http://www.switching-powersupplies.com/shop/dc-dc-converter/module-type/dke15.html

Quote:
In any case would get 48 V, could take it for the phantom, and at the same time fot the +/-20 V regulation?

No, if the output is +/- 24 V it can't be 0/48 V at the same time...

I think the Baby Animal uses a 48V smps, not? Maybe I would need an implementation for the Supergreen schem to do the same?
What about this: one +/- 24V converter, plus one +24V single?
 
[quote author="lagoausente"]I´m just starting to read The Art of Electronics[/quote]
That's a good basis. Read it, then read it again, then try to understand the workings of discrete/hybrid circuitry such as the 990 or the Green/Supergreen (or any of the discrete mic preamps that can be found in the Meta). There's no substitute for (a lot of) practice; I find it a helpful exercise to calculate the DC points for most bipolar circuitry that's posted on this forum. In that way it's quite like learning to play an instrument, really.

Next step is to get yourself an oscilloscope (a used 10-20MHz one should be fine, and should not cost more than $50-$100). Build some circuits, test 'em and see if your measurements match your expectations (and if not, why not).

[quote author="lagoausente"]I thought would more, so that +/- 24V with +/-313 ma would do the job,DKE15A-24 http://www.switching-powersupplies.com/shop/dc-dc-converter/module-type/dke15.html[/quote]
Looks OK, although the noise voltage is a bit on the high side. Depending on the switching frequency, a simple LM317/337 shoud be able to filter much of that out. Just get one; it looks like these modules are cheaper than even a single Supergreen channel.

Be warned that simply adding capacitance to the output of the converter to reduce noise does not work; many of these DC-DC modules have a maximum output capacitance.

[quote author="lagoausente"]What about this: one +/- 24V converter, plus one +24V single?[/quote]
As long as you use isolated converters, that should work (and the Mean Well converters that you link to appear to be isolated models).

Good luck,

JDB.
 
Thanks jbdakker, really hepful.

Next step is to get yourself an oscilloscope (a used 10-20MHz one should be fine, and should not cost more than $50-$100). Build some circuits, test 'em and see if your measurements match your expectations (and if not, why not).

I have been thinking just this. On ebay maybe?



Looks OK, although the noise voltage is a bit on the high side.

I have been considering also to get a pair of Li-ion 19-20 volt batteries, to feed the +/- 20 of the supergreen, but maybe the protection circuit of this packages maybe add more noise than the converter?
Other posibility are NiMH 9V batteries, but maybe a little low capacity, 150mah are typical.


Depending on the switching frequency, a simple LM317/337 shoud be able to filter much of that out. Just get one; it looks like these modules are cheaper than even a single Supergreen channel.
What modules do you refer to? the Dc converter or the LM317?

Just another question, have you tried the LT1533 in your portable project?
Probably would be much quieter than the Mean Well, but I really have no idea how to implement a circuit for a +/- 20V.
 
And what about a pure sine wave inverter? Do them use switching? Would this solution give less R&N?
Doesn't look like a clever solution to me--the DC-AC converter will be switching (and likely a pretty bad one) so you have the noise radiation. The rails might be more quiet due to the filtering action of the transformer and smoothing capacitors but that is easily duplicated for a DC-DC converter by the addition of a little RLC filtering before the linear post-regulators you plan on. Check AN101 from Linear for some info on reducing switching noise with linear post regulators.

Samuel
 
[quote author="Samuel Groner"]
And what about a pure sine wave inverter? Do them use switching? Would this solution give less R&N?
Doesn't look like a clever solution to me--the DC-AC converter will be switching (and likely a pretty bad one) so you have the noise radiation. The rails might be more quiet due to the filtering action of the transformer and smoothing capacitors but that is easily duplicated for a DC-DC converter by the addition of a little RLC filtering before the linear post-regulators you plan on. Check AN101 from Linear for some info on reducing switching noise with linear post regulators.

Samuel[/quote]

The AN101 tells a out 20x lower spikes with ferrite beads. Measures about 900uV.
Mean Well converter has 80 mV R&N. Should I expect can get only 900 uV with the linear post regulator+filters? Or the final result will depend on the converter noise amplitude? I mean, if you have a DC converter that has 120 mV R&N, so after regulation and filtering, would get more noise that if the converter has 80MV?
I have no reference about how noisy really is 80mV, or 900uV on a preamp. But is still higher than the 100uV showed on AN70.
The ultra low noise chip series, seems, at first, fine for audio for the way it reduces the swithing noise even when lose some efficiency. Are no vendor taken advandage of them for comercial DC converters?
 
[quote author="lagoausente"]I have no reference about how noisy really is 80mV, or 900uV on a preamp.[/quote]
There's no way to tell, since the Mean Well folks don't specify the noise spectrum (all they say is that the switching frequency is ~150kHz, which is well outside the audio band).

Just buy one, and measure the difference in mic pre output noise with the switcher connected versus a known quiet power supply. An AC millivolt meter specced for the audio band would be best, but a decent sound card plus RMAA or other such software should do as well. Don't overdo the filter until you know that you'll need it.

You might also want to consider if it isn't easier to just buy a load of NiMH AA batteries. 30 2200mAH NiMH cells can supply a Green/Supergreen for several hours, at a price comparable to the Mean Well modules. Add five or six 9V batteries for the phantom supply and you should be good to go. It's not hi-tech, but it works. I use a similar stack of batteries for my current portable mic pre; the only reason that I am working on a switcher for the HDD recorder is that it needs so many different supply lines.

JD 'KISS' B.
 
You might also want to consider if it isn't easier to just buy a load of NiMH AA batteries. 30 2200mAH NiMH cells can supply a Green/Supergreen for several hours, at a price comparable to the Mean Well modules. Add five or six 9V batteries for the phantom supply and you should be good to go. It's not hi-tech, but it works. I use a similar stack of batteries for my current portable mic pre; the only reason that I am working on a switcher for the HDD recorder is that it needs so many different supply lines.

That was another solution I was considering. There is a charger for 36V NIMH packs.
There are some little cons. NiMH suffer high sefl-discharge. I have read that batteries have also inherent noise, due to the chemical process inside. Quote from other forum:
Jerry @ AERCO has mentioned that batteries are not perfectly clean. He mentioned low voltage noise that results from inconsistencies in the chemical battery reaction rates and that low frequency noise is very difficult to filter. I'm really not sure how significant that is and I didn't press for details. I suspect different battery chemistries have different noise characteristics depending on discharge rates, etc. His preamps have a fairly trick looking 48v circuit that features some hand wound coils.

I thought also about Li-ion packs, but since they come with a protecion circuit, can´t know if this ciruit can be adding eve more noise than the Mean converters.
Other issue is the bandwith of the batties noise. Maybe even if the DC converter would me more noisy maybe the band freq would be totally out of the audio band, and maybe the batteries noise, even lower, could be on the audio espectrum?
I got quite confused with a post of a guy, that claims that high voltages like +/- 18 is only neede for use with decks that need hihg line level, and claims that with two AA batteries is enough to get a 1.5V line level signal. Do you agree?
Here is some perspective from recent personal experiences.

The need for 'high' voltages (higher than batteries can easily supply without resorting to DC boost circuitry) or not in a pre really depends on the LINE input signal voltage (dBu) requirements of a particular deck model.

Most decks require more or less ~ 1.5 volt rms (~ +3 dBu) maximum line input level for unbalanced input, or +9 dBu for balanced line in.

For these decks, I make several models of preamp running WITHOUT DC-to-DC boost. This insures having lowest levels of circuit noise as batteries are inherently VERY low noise suppliers of power, and are way quieter than most any AC power supply source. These miniature preamplifiers produced have amplifier circuitry with full rail-to-rail output ability, so using only two AA 1.5 cells gives a true bipolar +plus/-minus 1.5 volts DC supply, and this is the maximum output swing voltage of the preamplifier. These have great advantage to run a VERY long time with two AA L91 type batteries running one model for 200+ hours, and simpler preamp running 500+ hours.

An example of a deck needing very high voltage line inputs is the SD 744 model that requires up to +24 dBu line signal.

A preamplifier suitable for this deck requires having +plus/-minus 15-18 volt supply rails (30-36 volts total). So little practical chance for running directly off batteries. Instead, usually a 6-12 volt battery is used to power DC switching boost supply, and then this supply also requires very careful noise filtering engineering so not to contaminate the audio signals.

I also make a twice size 4 channel preamp model for 744 type of deck that's unique in quietly running the DC boost supplies 15+ hours off 4 D alkaline cell power sled.
 
[quote author="lagoausente"]I thought also about Li-ion packs, but since they come with a protecion circuit, can´t know if this ciruit can be adding eve more noise than the Mean converters.[/quote]
What makes you think that ? Personally I would never use Li packs for portable apps due to their finicky nature (and tendency to catch fire when severely mistreated), but the protection circuit adding noise is not really a concern.

[quote author="lagoausente"]I got quite confused with a post of a guy, that claims that high voltages like +/- 18 is only neede for use with decks that need hihg line level, and claims that with two AA batteries is enough to get a 1.5V line level signal. Do you agree?[/quote]
Last time I checked this thread was about mobile power for a (Super)green mic pre.

Sure, it's possible to build a mic pre that runs directly off batteries. There's even one or two to be found in the Mic Pre meta (and at least one that's not there -- ISTR CJ toying with a SM-57, a mine shaft and some rocks). Will they have the same sound/performance as a more power-hungry +/- 18V pre ? Probably not. Will it matter for your applications ? Nobody knows except you.

Most decks require more or less ~ 1.5 volt rms (~ +3 dBu) maximum line input level for unbalanced input, or +9 dBu for balanced line in.
[...snip...]
These miniature preamplifiers produced have amplifier circuitry with full rail-to-rail output ability, so using only two AA 1.5 cells gives a true bipolar +plus/-minus 1.5 volts DC supply, and this is the maximum output swing voltage of the preamplifier.
That's nice, but 1.5Vrms equals almost 4Vpp, which is not going to happen with only two AA cells, unless you're using push-pull (which he isn't) or transformers.

FYI: when posting random snippets of text, it helps if you post links to the source material so that those who wish to do so can get some context.

JD 'context is good' B.
 
I posted the entire text, but here is the link: http://taperssection.com/index.php/topic,82504.0.html



Last time I checked this thread was about mobile power for a (Super)green mic pre.

Sure, it's possible to build a mic pre that runs directly off batteries. There's even one or two to be found in the Mic Pre meta (and at least one that's not there -- ISTR CJ toying with a SM-57, a mine shaft and some rocks). Will they have the same sound/performance as a more power-hungry +/- 18V pre ? Probably not.

Thats what i wanted to know, if the performance would be the same on those.
But still want to feed the Supergreen pre. Now, I´m curious how could affect puttin less voltage to the Supergreen.

I thought about the Li-ion because the low-self discharge. I thought that since it comes with the protection circuit, I would´n need to worry about it.
I thought in those external batteries for laptops, some give 19V.
perhaps expensive, but maybe it´s life would be more than NIMH?
Look this ones, come with charge status leds:
http://www.batteryspace.com/index.asp?PageAction=VIEWPROD&ProdID=1529
I have read something like "digital battery" on similar items on ebay, so hope this package doesn´t have any swithing inside..
 

Latest posts

Back
Top