Signal Generators, Voltages, and Connectors

[andYz00m]

Hey there,

Im deep into calibrating a bunch of gear and I'm starting to doubt/wonder about my signal generating methods.

I understand that 0VU (+4dBu) is 1.28V RMS. When measuring a balanced signal (TRS cable from a balanced output) I have been measuring across positive and ground leads.

Is this correct? Most signal generators are unbalaced output (hot/ground) which makes it easy. I never trust the "voltage" output of my generator and always measure with a multimeter to make sure.

Problem: I am attempting to calibrate a stereo bus comp (API2500) as I needed to replace the meters. The calibration steps begin with a 0VU signal at 2k. But at this level, the signal is too hot for the trim pot to adjust for 0VU on the meter. Im using the balanced output of my interface to send the 2k RMS sine test tone (which is 600Ohm) at 1.28V @ 2Khz into the XLR inputs of the unit.

If my multimeter isnt true RMS will I get inaccurate readings? I've been using my mm for a long with this same method.

What am I missing?

I feel like a noob idiot! It's salso possible something is funky with this unit...

 

[john12ax7]

You might have 6dB too much signal. In a balanced scenario the signal exists across hot and cold, so measure that to verify, don't measure hot to ground.

Also you should double check what actual voltage the api is expecting. 0VU can sometimes mean different things.

 

[andYz00m]

Thanks, John. I have a note out to the fellows there to see if they have some more info.

But I'll double check my output and see whats up.

 

[Tubetec]

Heres a few calculator tools I often use for calibration purposes , you might find them usefull.

http://www.sengpielaudio.com/calculator-db-volt.htm
http://www.sengpielaudio.com/calculator-db.htm
http://www.sengpielaudio.com/calculator-amplification.htm
I use a Fluke RMS multimeter with db measurement as the reference for voltage and frequency , I use that to set up an arb generator (FY6800) with precise levels from -50db to 0db at specific frequencies , I then store these as presets , so its really quick and easy to put out a preset known level to any device I like , without the need to recalibrate each time .

I always use a BNC to bannana plug adapter with a screened cable to a croc clip on the ground and a mini hook test probe on the hot wire . Regular multimeter test cables arent much good for AC measurements , especially at lower levels . I also made a bannana plug to XLR , hot/cold wires go to the inputs of the meter and the screens are connected only at the XLR end to pin 1 .

 

[Gold]

If my multimeter isnt true RMS will I get inaccurate readings? I've been using my mm for a long with this same method.

That could be part of the problem. Inexpensive DMM’s often don’t measure frequencies above a few hundred hertz accurately. Try measuring 60hz and then 2K. If the meter reads significantly different for each frequency it’s the meter that’s off.

 

[ruffrecords]

A balanced signal exists between hot and cold - nowhere else. The 'ground' lead is not a ground it is a screen. It does not carry signal current.

I designed a simple battery operated 1KHz generator with an XLR balanced output that can be switched from 0dBu and -40dBu. Distortion is not brilliant but for calibrating levels it is great. I can send you a PCB for free. All the details of the design and its construction are on my web site under the DIY tab:

Custom Tube Consoles - DIY

Scroll down and select the PocketOsc folder.

Cheers

Ian

 

[ccaudle]

If my multimeter isnt true RMS will I get inaccurate readings?

No, true RMS helps for non-sine signals, but your meter will be accurate for sine waves. What might be an issue is how accurate your meter is at 2kHz. You did not say what make and model of meter you are using, so check the manual specs page for the upper frequency limit.
The meter may still be useable above the specified limit, but accuracy not guaranteed. For example, my meter will continue responding above 10kHz, but the specs say +/- 1% 45Hz to 500Hz, +/- 2% 500Hz to 1kHz. So my meter would probably still be fine at 2kHz, but the vendor is not willing to guarantee it will still meet the 2% accuracy spec at that frequency.

 

[andYz00m]

This is all great info. Thank you so much!

Ian, I have two different signal generators which are unbalanced that I have been using in the last. But with this last batch of gear, Im calibrating in the studio so I used the computer as my signal gen, hence the balanced output and my confusion on how to measure.

I would love to build one of your pocket oscillators though! PM sent.

Multimeter im using is Kleintools MM400

I've been thinking about upgrading but hadnt had a reason yet.

 

[Tubetec]

Heres the accuracy table of the fluke I use ,

 

[Gold]

Unbalanced test equipment is kind of a PIA to use with balanced In/Out if it’s not battery operated. You should investigate what the ground situation is with the oscillator. If the shield is connected to ground at the electrical outlet it’s hard to integrate with balanced gear. You have to transformer isolate the audio out or run it on batteries.

 

[warpie]

Wouldn't a 1:1 (600:600?) transformer at the output of the generator solve the issue? Probably the level would be slightly different but this can be measured.

EDIT: I just noticed that Paul suggests the same.

 

[andYz00m]

I wonder if I can modify Ians design to have the switch change from 0dBu to +4dBu @1K. This is what I use the most...

Also, if I am just using the signal gen in protools and outputting from my interface balanced (which I know the output is 600Ohm), shouldnt this work?

Here is the info on my meter.

 

[Gold]

(which I know the output is 600Ohm), shouldnt this work?

Are you sure about that? It would be very unusual for a modern interface to have a 600Ω output impedance. It's more likely around 50Ω.

 

[andYz00m]

Using an Apollo rack unit. Turns out I have no idea where I got my 600Ohm number from. It's 100Ohm output.

Jesus, I really need to check my brain storage corruption.

 

[andYz00m]

After dusting out the brain and properly measuring the voltage, unit calibrated properly and easily. Thanks everyone.
Sometimes I realize that being self taught means missing small basic fundamentals!!

 

[ccaudle]

Here is the info on my meter

You left out the relevant part just above that, the table with the voltage measurement accuracy. It is +/- 1.2% with a note at the bottom: "Frequency Range: 50 to 60Hz"
So your meter probably responds at 2kHz, but with indeterminate accuracy.

Unbalanced test equipment is kind of a PIA to use with balanced In/Out if it’s not battery operated.

Have you found problems with just connecting the signal out to hot in, and sig generator ground/shield to cold in? I wouldn't run really long cables that way, but since this is benchtop test equipment I would think 1m or shorter cables would be in use. Should be fine if you are sitting right next to a radio transmitter.

Also, if I am just using the signal gen in protools and outputting from my interface balanced...shouldnt this work?

Yes, should work fine. Just make sure you are aware of any places your interface has for level adjustment so that you know the true analog output level at the connectors. I'm not familiar with the Apollo interface you are using, is it integrated with Pro Tools well enough that Pro Tools will actually show you the analog level in dBu, or is Pro Tools showing dB FS relative to full scale digital, and you have to know the conversion between digital full scale and analog output level?

 

[amplexus]

I see you’ve worked it out, but for future or anyone else wondering- this is the device I use in the shop for feeding a DUT a balanced signal from my unbalanced generators. It’s got cal. Trims to keep unity gain from in to out, and i feed the same signal to both inputs so i can calibrate stereo gear faster and more accurately.
They’re discontinued but easily found on ebay and reverb etc…

https://www.broadcaststore.com/store/model_detail.cfm?id=12229

 

[Gold]

Have you found problems with just connecting the signal out to hot in, and sig generator ground/shield to cold in? I wouldn't run really long cables that way, but since this is benchtop test equipment I would think 1m or shorter cables would be in use. Should be fine if you are sitting right next to a radio transmitter.

You are grounding the cold. It probably won’t work depending on the input topologyand if it does work it won’t be an accurate measurement.

 

[andYz00m]

Yes, should work fine. Just make sure you are aware of any places your interface has for level adjustment so that you know the true analog output level at the connectors. I'm not familiar with the Apollo interface you are using, is it integrated with Pro Tools well enough that Pro Tools will actually show you the analog level in dBu, or is Pro Tools showing dB FS relative to full scale digital, and you have to know the conversion between digital full scale and analog output level?

Im measuring the voltage across hot/cold of the TRS output at my patchbay. This way I can use the signal gen and have as many outputs with the same level as I want. Nice for stereo gear.

I have a protools template with the different outputs on the signal gen so I can mute/unmute to get the right output level instantly. Easy peazy.

That match maker box is super cool!

 

[andYz00m]

MM info