Mic Input Transformer Sweeps - Vintage and Modern

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

BluegrassDan

Well-known member
GDIY Supporter
Joined
Mar 17, 2009
Messages
507
Location
Elizabethton, TN
A buddy of mine lent me a couple input transformers to try out. Tacked them in, did some quick listening, and ran some sweeps.

ADC A5508
Peerless 4813
Cinemag 1:10
Cinemag 1:5 (the one I currently use in my preamps)

I gain matched these only roughly. The ADC and Peerless were approximately 4dB hotter than the 1:10.

Sweep were run through a line-to-mic attenuator at 150 ohms. No loading on the secondaries - straight to the first 6072 gain stage.

 

Attachments

  • Input TX Sweeps.jpg
    Input TX Sweeps.jpg
    227.3 KB · Views: 108
Nice!  I have one of those ADC - posted measurements here earlier - interesting to see that top end peak.  So you are going straight to the grid, no resistor to ground?
 
I've done a few more wiring configurations and have come up with even better results.

This is the Peerless wired to a lower ratio (seems to be 1:7). The frequency response is much better now. Only down 0.8 dB at 20k.

The ADC has the best response with a 60k resistor across the secondary to even out the ringing. But it still drops like a fly after 10k.
 
Loving this thread.  So by reversing secondaries, you are talking about swapping because one leg is going to be tied to ground in the input circuit.  In the case of the ADC pin 7 or 8, I would tend to think 7 is tied to grid and 8 to ground.  Which way measured better?
 
BluegrassDan said:
The ADC has the best response with a 60k resistor across the secondary to even out the ringing. But it still drops like a fly after 10k.

In general all of your vintage sweeps are rolling off much earlier and steeper than I ever measure here.  Not sure why, but why I made the suggestion.  You’re getting results up top that look more like early 1930’s inputs.

The ADC in particular may be designed for a resistive secondary load, many winders used to offer loaded and unloaded varieties. 

I just tested an RCA bridging input that has a 15dB resonant peak if unloaded.  It wants a matching condition 50K; even 70K is +1. 

Anything you test with a split secondary winding that shows a dip at one spot in the treble is interesting to look at in PP, pretty much always response is better than SE.  All UTC A / HA / LS, lots of ADC, etc etc. 

There’s an ADC with split secondary used SE in one Gates preamp, uses one sec and leaves the other floating.  I tried connecting it both series and parallel and both look much worse.  Looks best of all in PP, best used that way. 

Have also seen an LS-10 with one bad secondary half.  Response using just the good half was much worse.  Response IMPROVED with a resistive load across the floating BAD secondary, which I take as more evidence there’s always stray capacitance leakage to consider. 
 
The only way I have to measure, which seems more true to reality, is to go out of my DA convertor, through a pad (150ohm), and into the preamp. The output of the pre goes back into the AD convertor and to my computer.

Whenever I bypass the pad and simply turn down the output by about -40dB, the frequency sweeps are generally more linear up top.

But, it seems like that wouldn’t reflect how the input transformer interacts with a microphone.

I’m wide open for suggestions.
 
BluegrassDan said:
The only way I have to measure, which seems more true to reality, is to go out of my DA convertor, through a pad (150ohm), and into the preamp. The output of the pre goes back into the AD convertor and to my computer.

Whenever I bypass the pad and simply turn down the output by about -40dB, the frequency sweeps are generally more linear up top.

But, it seems like that wouldn’t reflect how the input transformer interacts with a microphone.

I’m wide open for suggestions.

I don't know what the answer is, other than to measure multiple ways and read between the lines.  Neither way is like a direct mic connection.   

If I take DA out direct to transformer it's a 200 ohm source.  One measurement.
If I take that into a 20dB U pad it's a different treble response.
If I take that into a 40dB U pad it's yet another treble response.  This one is the most treble loss, BTW.  Something about the increased series resistance?  Capacitive losses in the connection increased because of it? 

But flipping the secondary should always be considered it it looks like too much early rolloff, or there's a crazy resonance. 

Another one, still don't understand the severity of it:

https://groupdiy.com/index.php?topic=66350.msg840421#msg840421:

 
EmRR said:
Winding capacitance to ground.  ALWAYS compare.
So they can and often are asymmetrical? Maybe this explains the poor HF response I have been having with some OEP 1:10 input transformers?

Cheers

Ian
 
ruffrecords said:
So they can and often are asymmetrical? Maybe this explains the poor HF response I have been having with some OEP 1:10 input transformers?

Cheers

Ian

I haven't looked at many modern ones, but there are pretty clear warnings about it in old books about the transformers of the time.  Some manufacturers specify grid and ground, some don't. 
 
BluegrassDan said:
Here are two sweeps of the Peerless with different situations. The red line is with a -30dB PAD line-level attenuator (150 ohm) between the interface output and the transformer input. The blue line is straight from the interface output (5 ohm) with the levels attenuated -30dB digitally.

Wow, so I've never seen that bottom end response but I've also never driven from so low an impedance. 

Since you have 5 ohm output (lower than), I'd simply use a pair of series 68R or 82R resistors. 

The old books say you should expect top and bottom resonances (the smiley face EQ) with 'lower than' source Z, and you see it done on purpose to extend response in systems with many series amps.    I have seen at least one modern mic/line to grid transformer make smiley face EQ if wired 600 and driven 150. 
 
Ok, more data for the fun of it.  I took my ADC A5508 and hooked it up to my Amber 5500.  Here are the settings and results:

Source Z: 150 ohms
Level: -40dBm
Frequency: 1kHz

Balanced cable to clip leads (ugh, I know) terminals 1+4 tied with clip lead, terminals 3+6 tied with clip lead.  Positive signal input to 1+4, Negative to 3+6, Shield to pin S.  I believe this is correct for 150 ohm input.

Out of the transformer secondary, unbalanced but shielded cable with pin 8 + and pin 7 Gnd. 

Input Z of the Amber: 100k ohms
Signal level: -18.6dBm

I don't have a sweep function, so I took a few data points.  Referenced to the level at 1k I found:

20Hz is +0.06dBm
100Hz is +0.15dBm
2kHz is -0.3dBm
5kHz is -1.9dBm
10kHz is -6.1dBm
15kHz is -11.1dBm
20kHz is -16.3dBm
 
EmRR said:
Wow, so I've never seen that bottom end response but I've also never driven from so low an impedance. 

Since you have 5 ohm output (lower than), I'd simply use a pair of series 68R or 82R resistors. 

Me neither. I regularly see a peak around 25KHz in 1:10 mic input transformers when driven from a low impedance source and this usually flattens out when driven from 150 ohms. But I have never seen that kind of bass response.

Cheers

Ian
 
mjrippe said:
Ok, more data for the fun of it.  I took my ADC A5508 and hooked it up to my Amber 5500.  Here are the settings and results:

Source Z: 150 ohms
Level: -40dBm
Frequency: 1kHz

Balanced cable to clip leads (ugh, I know) terminals 1+4 tied with clip lead, terminals 3+6 tied with clip lead.  Positive signal input to 1+4, Negative to 3+6, Shield to pin S.  I believe this is correct for 150 ohm input.

Out of the transformer secondary, unbalanced but shielded cable with pin 8 + and pin 7 Gnd. 

Input Z of the Amber: 100k ohms
Signal level: -18.6dBm

I don't have a sweep function, so I took a few data points.  Referenced to the level at 1k I found:

20Hz is +0.06dBm
100Hz is +0.15dBm
2kHz is -0.3dBm
5kHz is -1.9dBm
10kHz is -6.1dBm
15kHz is -11.1dBm
20kHz is -16.3dBm



So, your findings are similar. It drops off steeply in the 10k area.

The best frequency response I obtained was by:
bridging 3&4
input + pin 1
input - pin 6
output + pin 8
output - pins 7 and S
 
BluegrassDan said:
So, your findings are similar. It drops off steeply in the 10k area.

The best frequency response I obtained was by:
bridging 3&4
input + pin 1
input - pin 6
output + pin 8
output - pins 7 and S

Yup.  For the folks playing along at home we are talking about this transformer: https://groupdiy.com/index.php?topic=74774.msg947296#msg947296

When I go to my shop tomorrow I will wire it up with the primary in series and shield tied to 7 as you suggest here and see what I get.
 
mjrippe said:
Yup.  For the folks playing along at home we are talking about this transformer: https://groupdiy.com/index.php?topic=74774.msg947296#msg947296

When I go to my shop tomorrow I will wire it up with the primary in series and shield tied to 7 as you suggest here and see what I get.

I'll look forward to seeing your results.

From an audio standpoint, it has an interesting "bigness," but lacks that last little bit above 12k to make it useful in a modern preamp. Could work as an effect, or with an overly bright microphone I suppose.
 
Back
Top