+/- 15v supply to AOUT NG

[widdly]

Can anyone see any potential problems running the AOUT NG from a +/-15v supply?

The TL074's can cope with a supply up to +-18volts so they should be fine.  I imagine the calibration would be the same for either supply voltage.

[/tilted/]

I can't see any problem with the hardware side.

The TLs are of course able to go to +/-18v, the filter caps are rated to well above this,

The DAC is powered by the Core.

The only problem I can see is an operational one. You may find that as a result of your increased supply rails, you have wider "steps".

In fact, They would be around 25% wider.

Of course, I'll happily hand over to Seppoman on this...

[widdly]

I'm not sure the steps would be bigger. 

AFAIC, the opamps are amplifying the voltage from the DAC.  The gain of the opamp is defined by the resistor ratio.  For the first CV it will be 1 + R1/(R9 + P1)..which is between 1 + 27k/(15k + 5k) = 2.35 and 1 + 27k/15k = 2.8 times the output voltage from the DAC.

It's the bipolar section that I don't quite follow.  I imagine the resistors here might need changing if the supply voltage is higher but I don't quite follow how it works.

Anyway I can always use a couple of 12v regulators but it would nice if I could swap resistors instead.

[/tilted/]

Is this because you plan to use it in a modular with +/-15v rails? Or just because you don't want to get regulators?

If you are going for 15v rails, wouldn't you then calibrate the NG so as to get the widest sweep range?

If not, why go for the higher rails?

My impression of the NG was that your max/-max valuse would be as close as possible to your rails (in bipolar mode) and your min would be 0v.

Wouldn't it then follow, that if you increased your headroom by increasing the rail voltage, and of course the number of steps between max, 0, -max would stay the same, then each step would get bigger?

[seppoman]

first, it's perfectly safe to use +-15V. In unipolar mode, the output scale doesn't change at all because the DAC generates its output from an internal voltage reference and the gain of the TL074s isn't influenced by supply voltage.

The bipolar mode: The TLV output range is 0...4.096 V. The "bipolar option" just adds (-2.048V) to the output voltage to transfer the range  into -2.048...+2.048V. The resistors around P9..P16 are chosen for getting a good but narrow adjustment range for zero offset calibration.

By changing to +-15V, the adjustment range moves slightly downwards. In case all used parts have no tolerance (theoretically), the range would be -3.46V...-1.92V. So ideally this would be sufficient to be able to select -2.048 V. But as all parts have some tolerance in real life, it could happen that you can't reach -2.048V anymore on some channel. To be on the safe side, I would advise exchanging R33..R40 for 2.0k ones which would give a theoretical adjustment range of -2.83V..-1.21V.

About what's the ideal output range of an AOUT: At least in unipolar mode, usually you want the output to be conform to the V/Oct standard, i.e. "more is better" isn't true here :) For bipolar mode, I suppose keeping that standard doesn't hurt either - honestly I don't have enough personal experience with modulars that I can tell you much on that topic.

When you've found a suitable solution, please post some feedback with your experiences to this thread :)

S

[widdly]

Thanks for the replies. 

For bipolar , around +/- 5v sounds reasonable.  I usually find that in a modular there is too much CV rather than too little. 

I'll post here when I get this built.