Opamp headroom, replacing Toshiba TA7322, a few questions combined.

Hi guys! (and girls, and aliens, and ghosts)  8)

I have a couple of questions, but first a few preliminary statements.
1. Yes I have done a google search, and searched this site for the answer to my questions. I know I am probablly missing something, so if this has already been covered in depth, please provide a link, as I must be using poor search terms
2. All of this questions have arisen as I am planning on doing some mods to a Yamaha M508. I would like to play around with different opamps in a few channels.

So anyways, here we go.

The M508 has 3 Toshiba Ta7322 chips (9pin SIL package) and 1 AN6552 chip. I was planning on making some SIL to DIP/SOIC adaptors. The Ta7322 chips run at +-24v while the AN6552 runs at +-18v

Of course, finding a 24v op-amp is not the easiest challenge, so I began to think...

Since the supply voltage is limited by the AN6552's supply, that determins the total headroom correct? In most opamp circuits, I rarly see more then +-18v supplies. I know the Ta7322 is a slightly different beast, but would there be any consequences that Im missing if I build a few adjustable voltage regulators to run the circuit at +-18v and throw in some op-amps? Seems fairly reasonable and rational to me, but I might be missing something obvious.

Another question that has been bugging me for awhile...How is headroom determined in db versus the supply voltage? Over at ESP, I have seen Rod meantion it, but couldnt find the formula. I am very curious to see how it works.... Whats for instance a good supply voltage to hit if you want +22db of output headroom? (This is unreleated to the yamaha mixer, just thinking in general and design terms).

Thank you all for your time and dedication!!

well, I was thinking of doing this on one.. The +/-25V (not 24 I don't think) comes in and goes to two 10 Ohm fused resistor.  Maybe those drop it to 24V, I don't remember. In any case, OPA604 will work on +/-24V, you just need to make an adapter and maybe remove some parts.  Or you could up those fused resistors (just stick higher value resistors there to drop the voltage, use Ohm's law and your opamp consumption to figure out how much V will drop).  If you get it down to 22V you have a lot more options with opamps. 

btw, the 512 I had here had a 4558 in the EQ section, and ran it at +/-16VDC.  Dropped it from the 24-25 with zeners.

Anyway, didn't really answer your questions I think, but it should be able to run at lower voltages.

If you want to replace the OpAmp, you need pay attention what are the real voltages at the V+/V- pins in the circuitry, but not to the datasheet’s data (btw, the maximum supply voltages for the TA7322 are +30V/-30V). Vintage OpAmps just could not work at low supply voltages.
You can use any modern low noise OpAmp and change supply voltages in the circuitry in accordance with its supply voltages (for example by means of suitable low power zener diodes). You will not deteriorate headroom due to the fact that input OpAmp works with millivolts and it does not matter whether supply voltages for the OpAmp are +24V/-24V or +12V/-12V.
The best replacement options (ultra-low noise) are LME49990, OPA1611 and LT1115.
Do not forget that TA7322 has passive components for phase compensation that do not need for the modern OpAmp. You need remove them.
The AN6552 is the 4558. One of the best general replacements is 4580.

It is fairly easy to work out max output from rail voltage. A perfect op amp will be able to swing its output from one rail to the other so with a +-18V supply it could output a 36V peak to peak signal. The RMS value is just this divided by 2.828 or about 12.7V rms. To convert this to dBu first divide by 0.775V to get 16.4, then take the log and multiply by 20 which gives 24.3dBu.

Using +-24V supplies the answer becomes  26.8dBu.

In practice, real world op amps cannot swing from rail to rail and maintain output drive and low distortion so you can knock a couple of dB as a safety margin.

Cheers

Ian