Neutrik NA2F-D0B-TX XLR to RCA Adapter Review

[solderdude]

True, but what about the built-in high-pass filter?

The 'high-pass filter' is just the natural roll-off from the transformer. It is not an actual filter.

When driven with 200ohm source resistance it should be loaded with 2k or higher.
When driven from 600ohm it should be loaded with 10k or higher.
In both cases there will be a roll-off in the lows but slightly different response in the upper frequency range.

 

[PeteL]

The 'high-pass filter' is just the natural roll-off from the transformer. It is not an actual filter.

When driven with 200ohm source resistance it should be loaded with 2k or higher.
When driven from 600ohm it should be loaded with 10k or higher.
In both cases there will be a roll-off in the lows but slightly different response in the upper frequency range.

Are you fully sure of that tough? I understand that there is a cutoff where the transformer will saturate but it looks to you like just the natural roll off? I know all kinda stuff can be found online, but this article make sense to me and explain how it could be desirable to implement a high pass on small signal Audio Transformers.
The very last line of his conclusion is:

"I consider the inclusion of a high-pass filter before the transformer drive circuit to be mandatory, whether a NIC is used or not."

Transformers for Audio

Driving audio signal transformers seems to be a straightforward process, but there are some interesting and quite fascinating issues that are poised to cause the user much grief.

sound-au.com

 

[pma]

I am no expert.
At low levels, the distortion is due to hysteresis. At some stage (the knee) it starts to distort due to overloads, two causes, producing distortions.
my 2 pence.

I mean the ”S” 50Hz distortion plot curve shape near 2V. You can see change in even/ order distortion ratio there. Saturation, probably.

 

[solderdude]

Are you fully sure of that tough? I understand that there is a cutoff where the transformer will saturate but it looks to you like just the natural roll off? I know all kinda stuff can be found online, but this article make sense to me and explain how it could be desirable to implement a high pass on small signal Audio Transformers.
The very last line of his conclusion is:

"I consider the inclusion of a high-pass filter before the transformer drive circuit to be mandatory, whether a NIC is used or not."

Transformers for Audio

Driving audio signal transformers seems to be a straightforward process, but there are some interesting and quite fascinating issues that are poised to cause the user much grief.

sound-au.com

There is no 'high-pass' filter (actually would be bandpass) as in: There is no deliberate low-pass filter using capacitors etc. It is just roll-off caused by 'the transformer'.
A combination of inductance/winding resistantance, saturation and source/load impedance.

 

[pma]

It does make more sense, but then it's just a regular DI box, which there are plenty, with better performance, more features, more flexibility. The other way around is a bit more unique with less options to choose from. I do wonder if these are bidirectional since it's 1:1 but the shown filter may suggest otherwise. It does make sense to have a High pass at the input to avoid hitting the trnsfo with subsonics, but this is a bit high cutoff as measured. Can you also measure the effect of source impedance on that filter?

After making measurements, I will probably use the nice box to put inside a balanced xlr-xlr divider. The Neutrik RCA CINCH has same dimensions as their panel XLR (I have used dozens of both in my preamps). Good shielded metal boxes are expensive if you buy in pieces. I like the good quality of the box. Budget badly produced cheap products are uninteresting to me.

 

[Vladimir Filevski]

There is no 'high-pass' filter (actually would be bandpass) as in: There is no deliberate low-pass filter using capacitors etc. It is just roll-off caused by 'the transformer'.
A combination of inductance/winding resistantance, saturation and source/load impedance.

I concur.
Previously I didn't pay attention at the measured LF roll-off, thought there must be a capacitor, because it seems to be quite steep. Wrong! Amir measured LF roll-off at about 6 dB/octave - which is according to the theoretical circuit model, so no additional Low-pass filter, indeed.
But it will be interesting to measure the low-pass behavior with 200/600 ohms at input and 2k/10k ohms at output load. Distortion too.

 

[pma]

I concur.
Previously I didn't pay attention at the measured LF roll-off, thought there must be a capacitor, because it seems to be quite steep. Wrong! Amir measured LF roll-off at about 6 dB/octave - which is according to the theoretical circuit model, so no additional Low-pass filter, indeed.
But it will be interesting to measure the low-pass behavior with 200/600 ohms at input and 2k/10k ohms at output load. Distortion too.

Too low inductance - my guess.

 

[paschulke2]

Good luck finding an audio device that is happy to drive just 46 ohm DC resistance. We are talking speaker level impedances here folks, not line level!

Are you trying to test your readers? The DC resistance of a transformer (which - by definition - can't transform DC) is totally unrelated to the AC input impedance of a (loaded) transformer. Have you ever measured the DC resistance of a (toroidal) power transformer and wondered why it doesn't trigger the circuit breaker?

 

[KSTR]

Klaus, would you have an explanation for this (50Hz plot):

View attachment 293451
The transformer is driven from Zout close to 0ohm with 250mA current capability. At the knee, ratio of even/odd harmonics is changing. Thank you.

The kink at ~2Vrms... mhh, without detail spectra -- or better yet, a distortion time-domain residual**) -- stepped over level and frequency it's hard to tell what's going on there.

**) 50Hz filter is PITA for a passive analog notch filter, notably as our 50Hz EU mains frequency tends to disturb, so I rather use 60Hz or 40Hz. I often resort to record the output and then filter it digitally, applying a notch filter two times, second pass on time-inverted output of first pass, then time-invert again. This makes the notch linear-phase to not disturb the waveform of the distortion for the lower-order harmonics.

 

[pma]

The kink at ~2Vrms... mhh, without detail spectra -- or better yet, a distortion time-domain residual**) -- stepped over level and frequency it's hard to tell what's going on there.

**) 50Hz filter is PITA for a passive analog notch filter, notably as our 50Hz EU mains frequency tends to disturb, so I rather use 60Hz or 40Hz. I often resort to record the output and then filter it digitally, applying a notch filter two times, second pass on time-inverted output of first pass, then time-invert again. This makes the notch linear-phase to not disturb the waveform of the distortion for the lower-order harmonics.

This is the simultaneous record of transformer input/output. I think that 50Hz test frequency is a non-issue, because the input is distortion free and all the distortion is in the transformer output (1:1 transformer). I would guess the point of core saturation.

 

[Cbdb2]

I have not read this entire thread. Has any one noticed this is a mic level 1 to 1 transformer?

Theres a table on page 13. They only show one line level transformer, it will take +19dbu.
Edit: Sorry this download can be found here: https://www.neutrik.com/en/product/na2f-d2b-tx

NTE1 1 : 1 200 : 200 200 / 2k, (600 / 10k) -3dbu General purpose, splitting, XLR inline
NTE4 1 : 4 200 : 3.2k 200 / 10 K -7dbu Mic input step-up
NTE10/3 1 : 3 200 : 1.8k 200 / 10 K -7dbu General purpose mic input step-up
1 : 10 200 : 20k 200 / 50 K -6dbu
NTL1 1 : 1 10k : 10k 600 / 10k +19dbu Line input
NTM1 1 : 1 200 : 200 200 / 2k +7dbu Mic input, splitting
NTM4 1 : 4 200 : 3.2k 200 / 10k +9dbu Mic input step-

 

[Cbdb2]

Are you trying to test your readers? The DC resistance of a transformer (which - by definition - can't transform DC) is totally unrelated to the AC input impedance of a (loaded) transformer. Have you ever measured the DC resistance of a (toroidal) power transformer and wondered why it doesn't trigger the circuit breaker?

Yes, and the input impedance of a transformer is the transformer ratio times the load impedance. So for these 1:1 transformers the load on the driving device doesn't change with or without the transformer.

 

[PeteL]

I have not read this entire thread. Has any one noticed this is a mic level 1 to 1 transformer?
file:///C:/Users/joflo/Downloads/07%20Section%20Accessories%20and%20Patch%20Panels%20PG%20EN%20202202-V22.pdf

Theres a table on page 13. They only show one line level transformer, it will take +19dbu.

I don't think anybody here can access your C drive... That would be scary.
Mics on RCA don't really make sense. It is operating at what is historically the consumer unbalanced line level standard which unity is set at - 10 dBV with a very limited 5 dB headroom, which, granted, is insufficient. 19 dBU is professional standard Balanced line at +4 and 15 dB Headroom. Mic levels are between -40 and -60 dBUs. and well unbalanced mic pres do exist in the consumer world and I guess this could work, but mic level it is not.

 

[KSTR]

Yes, and the input impedance of a transformer is the transformer ratio times the load impedance. So for these 1:1 transformers the load on the driving device doesn't change with or without the transformer.

Not quite because for low frequencies the primary inductance and the effective drive impedance (amp impedance + coil DC resistance) become dominant over the reflected load impedance.
At the -3dB corner frequency of that R-L primary intrinsic 1st-order filter the load impedance for the amp is R + jR which is pretty low and reactive, close to 100R magnitude many times.

 

[KSTR]

I understand that there is a cutoff where the transformer will saturate but it looks to you like just the natural roll off? I know all kinda stuff can be found online, but this article make sense to me and explain how it could be desirable to implement a high pass on small signal Audio Transformers.

Rod's article on transformers is an excellent resource.

A high-pass is a good idea if only to keep DC offset and infrasonic garbage away.
The simple solution with a series capacitor needs proper value trimming to avoid the resonance peak while keeping the value as small as possible.

 

[PeteL]

I have not read this entire thread. Has any one noticed this is a mic level 1 to 1 transformer?

Theres a table on page 13. They only show one line level transformer, it will take +19dbu.
Edit: Sorry this download can be found here: https://www.neutrik.com/en/product/na2f-d2b-tx

NTE1 1 : 1 200 : 200 200 / 2k, (600 / 10k) -3dbu General purpose, splitting, XLR inline
NTE4 1 : 4 200 : 3.2k 200 / 10 K -7dbu Mic input step-up
NTE10/3 1 : 3 200 : 1.8k 200 / 10 K -7dbu General purpose mic input step-up
1 : 10 200 : 20k 200 / 50 K -6dbu
NTL1 1 : 1 10k : 10k 600 / 10k +19dbu Line input
NTM1 1 : 1 200 : 200 200 / 2k +7dbu Mic input, splitting
NTM4 1 : 4 200 : 3.2k 200 / 10k +9dbu Mic input step-

So that would be the NTE1 transfo inside the box. It doesn't provide any more info on it's behaviour tough, but I think we are getting a clearer picture.

 

[OShag]

Very useful review. I have been looking at adapters as a possibility for some components. This indicates a penalty in performance. Also, it is my understanding is it is not good practice to use a simple adapter, as it shorts grounding pin on the xlr balanced side.

 

[Ken Tajalli]

This is the simultaneous record of transformer input/output. I think that 50Hz test frequency is a non-issue, because the input is distortion free and all the distortion is in the transformer output (1:1 transformer). I would guess the point of core saturation.

View attachment 293522

I am thinking out loud now! could the windings be vibrating, and at 2V creating a cancellation?
I wonder!
try it at 40Hz and 60Hz, see if the pattern holds out.

 

[Gruesome]

I have not read this entire thread. Has any one noticed this is a mic level 1 to 1 transformer?

Theres a table on page 13. They only show one line level transformer, it will take +19dbu.
Edit: Sorry this download can be found here: https://www.neutrik.com/en/product/na2f-d2b-tx

NTE1 1 : 1 200 : 200 200 / 2k, (600 / 10k) -3dbu General purpose, splitting, XLR inline
NTE4 1 : 4 200 : 3.2k 200 / 10 K -7dbu Mic input step-up
NTE10/3 1 : 3 200 : 1.8k 200 / 10 K -7dbu General purpose mic input step-up
1 : 10 200 : 20k 200 / 50 K -6dbu
NTL1 1 : 1 10k : 10k 600 / 10k +19dbu Line input
NTM1 1 : 1 200 : 200 200 / 2k +7dbu Mic input, splitting
NTM4 1 : 4 200 : 3.2k 200 / 10k +9dbu Mic input step-

Thanks for linking to the selection guide. The fact that this is a microphone level adapter was pointed out several times, but every time it was posted it was overruled by the majority opinion that that can not possibly be. I'm pretty certain the same will happen to your post.

 

[PeteL]

Thanks for linking to the selection guide. The fact that this is a microphone level adapter was pointed out several times, but every time it was posted it was overruled by the majority opinion that that can not possibly be. I'm pretty certain the same will happen to your post.

It's rightly appropriate (Not saying it's great) for these line level at least, from my trustee great sounding Sony MX-P21 mixer that I use for quite everything, along with a multitude of consumer grade audio equipment.

If all the faders are set at zero, this device will cover the full range of this meter without hard saturation.

And then will start clipping roughly at the same time as this led turn red. (yes distortion creeps in before that, but Amir's measurments don't reach a hard clipping.)