Balanced DI Issue

*excuse all the ignorance*
So I’ve been trying to build DIs with vintage transformers...UTC O1s, Triad JO-11’s, etc. Recently I got a signal generator and a scope. Hook the gen up with a 1K sine at 1V to the high imp (50k-ish) primary, then hook channels 1 and 2 of my scope up to the secondary, hook their grounds up to the ground of the transformer. Issues I’m having:
1. Scope is showing very low voltage, around 100mv...why? I have put my decade box at 50k across signal and ground of my gen, trying to force the primary to “see” a 50K load, and that doesn’t help at all.
2. Comparing channels 1 & 2 on the scope, they are NOT flipped phase UNLESS I ground the center tap...why is that? I’m under the impression the DI should be balancing the signal, flipping the polarity of what goes to the hot and cold of the XLR (channel 1 and 2 of the scope).
What am I missing here?
I’m sure this is operator error. Help! Thanks!

What does 'ground of the transformer' mean?  Is the transformer single primary and single secondary?

Maybe post a diagram of how you've hooked everything up?

Sorry, here you go...grounds to scope are tied together.

pmercer said:

hook channels 1 and 2 of my scope up to the secondary, hook their grounds up to the ground of the transformerand 2 of the scope).

Click to expand...

Your setup is wrong. Maybe your scope can be used in differential mode, if yes, you would see a single trace that represents the differential voltage across the secondary, which is what you want and use.
If not, you need to ground one side of the secondary and measure the voltage at the other side.
With your setup, the two channels receive a variable part of the output signal that is meaningless.

So in my diagram below, I should be taking two separate measurements? One with pin 1 grounded and pin 3 signaling to the scope, and then vice-versa? Should I be grounding pin 6 (transformer ground) as well? And, if the transformer is balancing the secondary, the phase on those should be flipped, correct?

pmercer said:

So in my diagram below, I should be taking two separate measurements? One with pin 1 grounded and pin 3 signaling to the scope, and then vice-versa? Should I be grounding pin 6 (transformer ground) as well? And, if the transformer is balancing the secondary, the phase on those should be flipped, correct?

Click to expand...

you should ground pin 1 and measure pin 3. No need for an other measurement.

How will I know if the output is balanced? Eventually Pin 1 of the transformer will go to Pin 2 of an XLR, and Pin 3 of the transformer will go to Pin 3 of an XLR. For the audio to be balanced, Pin 2 and Pin 3 of the XLR should have the same signal, but reverse-phase, correct?

pmercer said:

For the audio to be balanced, Pin 2 and Pin 3 of the XLR should have the same signal, but reverse-phase, correct?

Click to expand...

No. Not really. What we commonly call "balanced" is in fact floating. That means that the signal at one end is referenced to the signal at the other end.
You could make it real balanced, non-floating, by connecting pin 2 of the xfmr to "ground" and collecting signals at pins 1 & 3, but there is absolutely no reason to do that. Very old-fasioned engineering practice inherited from telephone lines.

abbey road d enfer said:

Very old-fasioned engineering practice inherited from telephone lines.

Click to expand...

Yes, and everything you said is, of course, spot on.  But just to clarify for the OP, the term 'balanced' comes from those old times when each side of the transformer was 'balanced' to ground.  Almost no audio devices that we still see and all love from back then actually had a transformer's centre tap grounded, most was left floating and was technically a 'differential' signal.
These days, with electronic inputs and outputs, the term balanced applies since each leg (pin 2 and 3) is balanced to ground.

Anyway, ground one side of the transformer to make measurements if you can't measure differentially with your scope.  But don't leave that side grounded for actually using it as a DI.  Keep it floating.  The mic pre you plug into will be grounding the centre, the phantom 6k81 resistors will be doing that even if the mic pre is a transformer input device. 
Confused yet?  .  Hope not!

> 1V to the high imp...primary, ....secondary,... showing very low voltage, around 100mv...why?

A 60,000:600 transformer is 100:1 impedance, 10:1 of voltage. So 1V in *should* come out at 1/10th volt. How could it be otherwise??

We do this (because it is easy, and) because Microphone levels are historically much lower than Guitar (or fuzz-box!!) levels, and we need a major step-down to avoid input overload. (Though modern studio techniques have forced huge "mic" input swallowing ability....)

And because you can't run a 50k-500k guitar signal more than a few dozen feet without treble loss. Working at lower impedance allows much longer lines, so the board-op can get away from the racket.

"Balanced" windings were often CT-to-ground when working directly under AM radio transmitters. And the CT was used to connect additional telephone circuits when there were more paying customers than pairs in the cable. You can connect a third conversation across the CTs of two telephone pairs with hardly any interference. (In fact the 'phantom pair' has lower resistance.) Four pair support four primary conversations, two phantom conversations, and a seventh pair from there (which may not be great quality). This was of great economic importance to telephone companies, before "carrier" systems became cheaper than pair-tricks.