Transformer experts! CJ, PRR, etc. relates to grounded center tap

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
After quickly reading the thread, my initial thought is that somehow the signal source you're using is "floating" (perhaps a generator that's not grounded via its AC power) and has considerable voltage with respect to this gear - which would appear as common-mode on the transformer primary. Of course, with the primary CT grounded, the CM voltage would see very low impedance (same current in opposite directions in primary) but, when the CT is ungrounded, would couple to the secondary in spite of the pri to sec Faraday shield. I'd start by looking at your signal source with a scope grounded to the unit under test.

Incidentally, the difference in DCR of the secondary "halves" is likely due to this being a 3-layer design. The first layer is half the secondary (the lower DCR part), then a Faraday shield, then the primaries (possibly bifilar wound), then a Faraday shield, then the other half of the secondary. Since the halves need to have the same number of turns, but the winding diameter is bigger for the outer layer, the DCRs will be quite different.
 
Oh well Bill got in just before me ,
My comment was going to be :
Could the primary be bifilar wound , without a centre tap capacitance might change appreciably.
 
After quickly reading the thread, my initial thought is that somehow the signal source you're using is "floating" (perhaps a generator that's not grounded via its AC power) and has considerable voltage with respect to this gear - which would appear as common-mode on the transformer primary. Of course, with the primary CT grounded, the CM voltage would see very low impedance (same current in opposite directions in primary) but, when the CT is ungrounded, would couple to the secondary in spite of the pri to sec Faraday shield. I'd start by looking at your signal source with a scope grounded to the unit under test.

Signal source is my recording system DA converter, sent through a 20dB mic pad. Everything on the same power trunk and ground. Spectrafoo Complete is the test generator/reader software. I've tested hundreds of raw transformers and complete transformer coupled equipments with this method, have never seen this other than in a couple 1930's RCA mic input transformers. Absolutely every other grounded center tap I've ever tried lifting has not shown anything similar, 99% of the time they show no response change at all. It's very curious.
 
Must be some peculiarity with the arrangement of the windings causing this effect . Only way to be sure would be a teardown CJ style , which wouldnt be an option in this case as they are in good working order . Any clues visually or is it in a sealed screened container ?
 
@emrr
IMO, you made a mistake somewhere in the measurement or the measurement setup is faulty, because it seems to me that it is impossible to measure +40dB relative to nominal output signal on the microphone transformer secondary regardless of its turns ratio. You also concluded well in the first post that the spikes at 60Hz, 20 and 40kHz are some kind of interfering signals, not transformer resonances. They cannot be so sharp. Also, noise in the midrange is seen in the measurement, which together with the present interference indicates that the problem is maybe somewhere in the common ground of the measuring setup and the microphone preamplifier.
Edit:
There is another possibility which is that the preamplifier is unstable and oscillates when the central terminal of the primary winding is not grounded. This would explain the large measured signal components at 60Hz (poor PSRR and overloaded power supply), 20 and 40kHz (first and second harmonic of oscillation signals).
 
Last edited:
It seems obvious to me that there is a large common-mode voltage being applied to the inputs. Bear in mind that a conventional balanced "U" pad doesn't attenuate common-mode at all - only differential signal. Also note that, as long as the CT is grounded, common-mode will be rejected right in the primary windings. But, if the CT is lifted, now this large CM voltage will appear on all primary windings and capacitively couple to the transformer secondary (i.e., a big output signal). And this would be especially true if the internal Faraday (pri to sec) shield were to also somehow be lifted or tied to the also floating primary CT. Disconnect your signal source from the transformer inputs and look at each of them with respect to the preamp's ground.
 
@emrr
There is another possibility which is that the preamplifier is unstable and oscillates when the central terminal of the primary winding is not grounded. This would explain the large measured signal components at 60Hz (poor PSRR and overloaded power supply), 20 and 40kHz (first and second harmonic of oscillation signals).
Seems the most likely with the sagged and dirty response baseline. We are talking about 1930's technology.

It seems obvious to me that there is a large common-mode voltage being applied to the inputs. A conventional balanced "U" pad doesn't attenuate common-mode at all - only differential signal. Also note that, as long as the CT is grounded, common-mode will be rejected right in the primary windings. But, if the CT is lifted, now this large CM voltage will appear on all primary windings and capacitively couple to the transformer secondary (i.e., a big output signal). And this would be especially true if the internal Faraday (pri to sec) shield were to also somehow be lifted or tied to the also floating primary CT. Disconnect your signal source from the transformer inputs and look at each of them with respect to the preamp's ground.

Look at "each of them"....the input transformers? The other two had CT's grounded, the offending did not, all behaved the same once the CT was grounded, all show continuity to the B-/Chassis connection.


The question raised, is why have I never seen this sort of spike with the hundreds of other floating transformer inputs tested in this exact manner? Why have I never seen any change at all when disconnecting 99.9% of other grounded input center taps? What's special here? It would only seem to be something in the particular transformer based on all available comparative data. It'd be great if someone could find a fault in my system, but I know I can hook up one of a dozen or so similar amps in the same setup with no fault shown.

In the attached pics you can see the normal connection in one pic, and the floated connection in the other, as well as the potted can shielding of the overall construction. This is brute force broadcast heavy shielding intended for use near an AM transmitter. The drawing is back on the first page. The circuit is a triode connected cathode biased 6SJ7 feeding a second of the same.

The second report I made later references a set of earlier 1934ish units that were restored and tested years later, with more than a dozen units to compare. Those were particularly goofy in regards to hum field shielding, with many things that had to be exactly right to keep hum or buzz out, from the era with shield plates between each tube, external tube shield box, and transformers in 1/2" thick solid machined shield cases.
 

Attachments

  • DSC_0001 1.JPG
    DSC_0001 1.JPG
    223.2 KB · Views: 52
  • IMG_5692.JPG
    IMG_5692.JPG
    197.8 KB · Views: 35
  • IMG_5690 floated T2 CT.JPG
    IMG_5690 floated T2 CT.JPG
    211.7 KB · Views: 52
Last edited:
Does the issue still exist when there are the usual pair of centre-tapped, ground referenced 6K81 phantom resistors on the input?
 
By "each of them," I meant each of the balanced line pair that would drive the transformer. What would explain the symptom is the center-tap of the primary staying tied to the Faraday shield when it's lifted. Just to clarify, the Faraday shield is a one-turn sheet of copper foil placed between primary and secondary layers (or several if it's a multilayer construction) to prevent capacitive coupling from the primary winding to the secondary. Some people confuse a Faraday shield with the external "can" shield, which usually functions as a magnetic shield.

The scenario I imagine is that 1. there is common-mode hum and 20 kHz (from where I could only guess) on both "generator" input leads (i.e., the common-mode); and 2. the center-tap of the winding remains connected to the transformer Faraday shield; 3. when the center-tap is lifted from ground, the common-mode voltage at the center-tap now rises because it's not grounded any more; and 4. because this voltage is also on the Faraday shield, it capacitively couples (typically over 100 pF) to the high-side of the secondary and is amplified just as if it were signal.

Perhaps the internal connection between the Faraday shield and the transformer terminal has come loose. When the CT is grounded, there serves little purpose because the CM voltage is extremely low - but when the primary floats, it's needed to prevent coupling, but it's not functioning.
 
Thanks, OK, that matches the thought back on page 1. Here’s a pic of another RCA from the era that seems to corroborate. The official drawing on this one is NOT drawn this way, it’s only on the can.

It appears then (I assume) the usual input transformer does not connect faraday to primary center tap and thus does not exhibit this behavior. A change in design theory based on expected needs/usage.
 

Attachments

  • 80715135-4468-4D52-946B-0C8002A3D0F2.jpeg
    80715135-4468-4D52-946B-0C8002A3D0F2.jpeg
    81.8 KB · Views: 38
Last edited:
The schematic printed on that transformer is unusual - it could even mean that terminal 7 is connected to the core stack, which is definitely not the same as a Faraday shield between windings. In any case, a Faraday shield is almost always brought out as a separate connection in my experience. It would then behave as I described earlier, like a transformer without a Faraday shield - which would make it exquisitely sensitive to primary common-mode voltage (and whether the CT is grounded).
 
Back
Top