Pink Noise using tl072...can't get it to work..

[benb]

If you had a perfect 24V-12V converter, this could be 12V 14mA.

There are places in audio where DC/DC converters, aka switching regulators, aka switch-mode power supplies (SMPS) can reasonably be used to good effect, but as part of a Phantom powered circuit doesn't seem like such a good place. The switching noise might be as loud as the "wanted" noise, and the switching noise surely won't have the desired spectrum.

Maybe I missed something. I had a "noise" source on a chip. Ate 9V at a couple mA. A 9V battery lasted for years. Pink-filter was passive; actually it was the E-V Speaker Test curve with less power below 100Hz or above 5KHz. The particular "feature" was that a short digital "noise" source repeats; this repeated every 1.7 seconds. The repeat caused a mild "thump". So even if there was another hiss in the system, this source was easy to identify.

This was surely the infamous and no-longer-made MM5837 noise generator chip. I've never thought of the repeat as an advantage before, but clearly it is in this case. The "good" news is you can program or buy a microcontroller that generates the same or similar sound output.

 

[Gene Pink]

> loaded down to near 28 volts

"Near"?

This is one time when Matching Theorem applies. P48 is 48V from 3.4K. Max load power is 24V, and at that point you can have 24V/3.4K= 7mA. (Yeah, 28V won't be a lot different.)

Z matching works out if you short pins 2 and 3 together, but I find that arrangement attenuates the differential signal to an unacceptable level.

Most efficient way to derive phantom, center tap of a transformer which is inconvenient. Using voltage dropping, energy burning resistors to sum, these are losses that need to be figured in, and this thread is about using resistors.

Admittedly, I don't remember the details, this is just from memory so bear with me here, but a long time ago I did the heavy math on something powered by phantom.Tweaking all the variables between the sum resistors, the load they make on the op amps, the voltage and current available to push that current into the load with a reasonable amplitude left over without distortion, many iterations, a lot of tweaking, it just wasn't going to work, not enough power available. I do remember that the best output level I could get, was the phantom at the XLR was loaded down to about 28 volts. Never built it, but that's where I ended up.

Thinking about it from the live production perspective, using a 40 ch. transformer spllitter, if those transformers all had a center tap, buss all the centertaps together, and power available from any one channel might be capable of arc welding , as it would be drawing from all channels with +48 in parallel.

Gene