DIY: Calibrating AD/DA, Part Three

In conclusion to parts One and Two, I have finally settled with a calibration level that I am satisfied with. After a few times of calibrating, using, and re-calibrating, I settled with the manufacturers factory reference of [0vu=+4dBu=-16dBfs]. I ended up trying several different reference levels in the process.

The topmost photo shows the analog voltage level that I chose. I found a chart in Bob Katz' book detailing the dBu to voltage correlations, and given the author, I took his word for it. Needless to say, my research into other websites revealed an identical result, leaving me to conclude that 1.228 volts is a standard.

So, how did I do it?

The first thing I did was to RTFM...and consult the friendly legacy-products support wing of Apogee's website.

Although the manual and the support files differ in their emphasis on which should come first...i.e. should the AD or DA be calibrated first...both list sequences that are relatively the same. I was able to determine that the DA section should be calibrated first from Frank Wells.

In order to calibrate the DA and subsequently, the AD, I needed a standard tone signal that equals 0vu/+1.228v. It just-so-happens that Pro Tools LE, one of my DAWs, has a "plug-in" that generates noises and tones.

Here is (above) PT's standard signal generator.

I created a new "test" session to perform the calibration duties, created a few sets of mono aux tracks to set up identical signal generators proceeding out through digital and analog outputs.

One interesting thing I found was that the internal levels of PT, in particular, the meters...do NOT correlate to accurate measurements at the digital and analog outputs. Such has been the cause for harsh criticism of DAWs in general, and heated debates on popular gear-junkie forums.

In other words, the zero (0.0) level in PT has an elusive meaning...One might think that the 0 level on the PT meters would equal either 0 peak or 0dBfs at the outputs...but they do not.

But before I get into that, I must mention how I made connections between the PT hardware and my multimeter, as well as my new PSX-100 and my multimeter.

First I ran a TRS to XLRm converter-cable from the main analog left output of PT to another XLR cable (for length extension). At the male end of the XLR extension cable, I used my knowledge of the standard pin structure of modern audio gear to determine which pins to hook my Fluke™ up to.

In modern XLR world, the standard is:
Pin 1=Signal Ground, Pin 2 = (+) or Hot, and Pin 3 = (-) or Cold.

So, given that an audio signal is an alternating current signal, I can measure the potential difference between Pin 2 and 3 to record the output voltage.

Shown above and below is the practical application of this knowledge, hooking red to (+) and black to (-).

When I had established the connections, I set the multimeter to detect AC voltage and then went back to PT and experimented shifting level and sine wave frequency to derive a proper and sustainable 1.228v across Pins 2 & 3 of the analog outputs.

Another interesting thing I found was that finding the right relationship of sine wave frequency-to-level was rather difficult.

I did manage to notice in the process that with the PT level constant, say, 0.1+, the output voltage varied relative to the selected sine wave frequency.

I really want to know why this is, so anyone who can tell me gets a free sandwich. As of now, my search for this answer is ongoing...

I finally found a sustainable sine wave frequency to level relationship of 980 Hz @ -13.9 (in the signal generator) with a PT internal fader level of 0.0 (Unity Gain).

This internal setting gave me a consistent 1.228v level at the analog output across Pins 2 & 3.
(Referencing back to the top,) I copied the setting of the same signal generator and applied them to two other mono aux channels. For these channels, I routed their outputs to the ADAT outputs of the PT hardware, and then onward (digitally) to the DA input section of the PSX-100.

Following the instructions, I set the Apogee unit to Calibration Mode, monitoring the DA side of the PSX-100's metering section. I then used a "tweaker" to turn the left/right trim pots so that each side read -10 on the Apogee meters (viewing DA) and 1.228v on the MM connected to the DA L/R outputs.

Then, I took the analog output originally used in setting the +4dBu/0vu reference and fed it to the PSX-100 AD inputs, adjusting the AD trim pots this time to make the Apogee Meter to read -10. [-10 = -16dBfs]


When all the tweaking was finished, I switched from calibration to normal mode and monitored the DA section. Sure enough...


A +4dBu signal equals roughly -16dBfs on the Apogee meters. AD/DA Calibrated.

Thanks for tuning in,

Andrew Wayland