Does the ap ever lie?

[pucho812]

This is perplexing for sure. Testing a stereo compressor with the ap, in this case a portable one. Using 1k sine unit in bypass shows same level in and out on both channels, when compressing I get same level both sides and so forth. By all accounts the unit tests to spec.  But when the same compressor gets hooked up directly to an apollo twin(the 2channel interface by ua), I get one side/channel working perfectly, the other channel is down in level by 6dB.  This is with testing using 1k sine wave. We swapped wiring, which has no fault, we wired it as a loop bypassing the compressor and level  I/o on the ua was good. But compressor gets connected and boom down in level on one side.    Very odd.

 

[JohnRoberts]

probably a quirk between single ended vs differential output modes.

Does bypass sw (in compressor I ASSume) make a difference?

If you add a transformer between comp output and interface input does it make a difference?

Something may be weird with either the compressor output or interface input... (like one leg shorted or something like that).

Scope or meter both + and - output lines wrt 0V  in good and bad modes. 

JR

 

[pucho812]

We found and fixed the issue. Off pcb wiring to an xlr jack flipped cold with ground. Once corrected was all good. Interesting that the ap did not show it.

 

[Gold]

The AP has transformer coupled inputs and outputs. It wouldn't show that fault.

 

[pucho812]

But wouldn't the unit which is electronically balanced I/o drop in level and this the ap pick it up?

 

[JohnRoberts]

But wouldn't the unit which is electronically balanced I/o drop in level and this the ap pick it up?

There are several variants on electronic outputs...

If ground is swapped with cold... wtf is cold? Do you mean audio - ?

If signal on both + and - out, but only looking at + to ground, you could be down 6 dB.

Lots more ways to screw up using 3 wires than 2 wires.  ;D ;D

JR

 

[Gold]

If signal on both + and - out, but only looking at + to ground, you could be down 6 dB.

If  audio - and ground were swapped you were probably grounding one side of the balanced out as well.

 

[JohnRoberts]

If signal on both + and - out, but only looking at + to ground, you could be down 6 dB.

If  audio - and ground were swapped you were probably grounding one side of the balanced out as well.

Transformer and some electronic balanced outputs can adjust for one leg being grounded (or not), but ignoring one active leg will generally result in a loss of signal.

JR

 

[pucho812]

John wtf is cold? The weather, we might get below freezing tonight. Very unusual for l.a.  In all seriousness, yes the audio -, in this case pin 3 on the unit had been flipped with ground on the input of channel one, of the stereo unit. The rest of the inputs and outputs are correct.
The ap portable did not show this when I ran tests on the compressor , but in use gave the volume discrepancies.  All has been fixed and back in service it goes.

 

[PRR]

> wtf is cold? The weather, we might get below freezing tonight. Very unusual for l.a.

Come out to balmy Maine. 42 F and rising.

 

[Monte McGuire]

The AP has transformer coupled inputs and outputs. It wouldn't show that fault.

Not the new APx555. My old 2322 (System Two) has transformer coupled outputs, but the APx555 does not have transformer coupled outputs. This bit me recently, as I thought I had tested the common mode performance of a circuit, using the 2322, but I had actually not, because of the 2322's XFMR outputs, and the APx-555 pointed this out quite clearly. I'm actually quite happy about that, but still, it brings about the larger issue of the original poster - does test equipment lie, and if so, how?

The best answer is that any testing procedure, using whatever sort of gear and procedures, is going to return results according to the natural laws of physics, math etc. If we fail to understand what is being "tested" and then fail to understand the meaning of the "results", then things will go wrong. It seems likely that even with extreme vigilance, things will fail occasionally, not so much the fault of the test equipment, as much as the misunderstanding of the meaning of the results obtained.

My only suggestion is to never blindly trust any measurement, unless you understand what the number represents, how it was obtained, whether or not it is plausible, and whether it agrees with other results obtained using different equipment, procedures or approaches. Even still, one might get things wrong, but at least the "small failures" can be avoided.

 

[Gold]

Pucho said he was using an AP Portable One. That model has transformer coupled inputs and outputs.

And you can't beat the vintage sound 8)

 

[pucho812]

Monte I understand what is being tested just fine. The  gear on the other hand is new to me. I am well versed in the  ap system one and system two but nrver used a portable one up until I started my new gig two weeks ago.

 

[JohnRoberts]

Monte I understand what is being tested just fine. The  gear on the other hand is new to me. I am well versed in the  ap system one and system two but nrver used a portable one up until I started my new gig two weeks ago.

As I posted earlier there are multiple variant balanced/differential outputs. These can all behave differently when one of the two legs is grounded or floating, (or swapped with ground).

Even transformer outputs can misbehave too if one leg is floated.  A center-tapped transformer output will be more tolerant of the open circuit, but less tolerant of shorting one side. 

The variants and gotcha's involved with using these variant interfaces are enough for a whole chapter in the book of what is...

IIRC there is a Rane reference showing the sundry wiring connections for interfaces.

JR

 

[Gold]

Even transformer outputs can misbehave too if one leg is floated.  A center-tapped transformer output will be more tolerant of the open circuit, but less tolerant of shorting one side.

In the Portable one manual nearly the first page says signal always goes between pins 2&3 whether balanced or unbalanced. In the setup page you can change the output impedance and drive one or both legs for balanced or unbalanced.

 

[Audio1Man]

Hi Guys
The AP's are different Outputs types depending on the basic models. The System One, System Two, Cascade, and
2700 have an Output Transformer. It can be placed in floated (Balanced) or grounded (Unbalanced) output modes.
The ATS2 and APX models are IC Driven in Balanced mode. They are CT grounded. The unbalanced mode is
1/2 the output voltage, about 8 volts. The BNC connectors are above chassis ground and are actively driven to
remove small ground differences, Any ground hum or other signals will reduce the total unbalanced output from
its 8 volts level to 8v - (Hum or ? levels).

The INPUTS are fully balanced in XLR / Dual Bananas in all of the analyzers. The BNC in some models are
grounded and others models are floating.

As allways making any measurements you must know your test system. 

Duke