What are the Pros and Cons of Using Output Transformers in a Tube Preamplifier Design?

[caesar]

What are the upsides and downsides of using Output Transformers in a tube preamp design? What are the tradeoffs with other design approaches?

Thanks in advance for the replies

 

[ArnoFenn]

Low output impedance. I got them for this reason in my LCR based phono pre....and reduction of series caps.

 

[analogsa]

Provided the transformers are really good, which often means a custom design, there are no downsides. A low output impedance is achieved without resorting to poor sounding topologies such as cathode followers. Compared to a good transformer even the best cap sounds coloured and less coherent. Due to the inductive load, a transformer coupled stage is capable of double the voltage swing compared to a resistive or active load. This may seem unimportant but somehow appears to improve perceived dynamics.

The transformer can also be a part of a balanced or differential topology and allow for bal > se and se > bal conversions for free, i.e. withought the penalty of adding extra stages.

 

[gadawg58]

I'm enjoying my BAT Rex 2 Preamp which uses output transformers. I've had a Ref10 in my system which uses caps but not at the same time as the BAT so I can't really offer a great comparison ... too many other changes. Would love to try it again one day! Theoretically the output transformers should allow me to drive amplifiers with lower input impedances than the ARC10 could comfortably but the amps I really want to try either have input impedances high enough that it doesn't matter what you drive them with or likely too low for even the BAT. I enjoyed my time with both preamps so I wouldn't say either is better ... Just my 2c.

George

 

[DonH50]

Off-the-cuff, and an objective viewpoint (yes, I know, sorry)... I am sure to forget a few pros and cons but here's a few quick thoughts.

Pros:

1. Impedance transformation; can take a high-Z (impedance), high-V (voltage), low-I (current) tube output and generate a low-impedance output with lower voltage but greater output current to drive cables and the amplifier load with less change in performance.

2. Ability to generate a differential (and balanced) output for greater noise rejection.

3. Isolation to reduce noise coupling and ground loops between (among) components. This also helps in the event a tube arcs and tries to apply B+ (high DC voltage) on the output; the transformer will prevent sustained DC on the output (but you'll still get a nasty pop).

4. Intrinsically AC-coupled (except for rare balun designs, and most tube preamps I've seen are AC-coupled anyway). This could be a pro or a con, I suppose.

5. Potentially simpler and less costly than adding an active buffer stage especially for differential outputs.

Cons:

1. Potential noise pickup by the transformer itself ( better line-level signal transformers are shielded to reduce this issue). Both LF hum and RFI.

2. Roll off of the highest and lowest frequencies (most good ones will have response well above the audio band; lowest frequencies may be an issue if you want sub-20 Hz output). Also introduces phase shift over frequency though likely inaudible.

3. High-level signals can saturate the transformer and lead to very high (and ugly-sounding) distortion. This can be a big issue for power amps, less so for preamps since signal power is generally low.

4. Stability; have to be careful in the circuit design to handle any feedback loops and watch for any HF ringing induced by the transformer's inductance interacting with internal, cable, and load capacitance. There is some load sensitivity but rarely an issue for a line-level stage (again more so an issue with tube amplifiers driving speaker loads).

5. Cost, though see (5) under "Pros" -- could go either way.

FWIWFM - Don

 

[Al M.]

Off-the-cuff, and an objective viewpoint (yes, I know, sorry)... I am sure to forget a few pros and cons but here's a few quick thoughts.

Pros:

1. Impedance transformation; can take a high-Z (impedance), high-V (voltage), low-I (current) tube output and generate a low-impedance output with lower voltage but greater output current to drive cables and the amplifier load with less change in performance.

2. Ability to generate a differential (and balanced) output for greater noise rejection.

3. Isolation to reduce noise coupling and ground loops between (among) components. This also helps in the event a tube arcs and tries to apply B+ (high DC voltage) on the output; the transformer will prevent sustained DC on the output (but you'll still get a nasty pop).

4. Intrinsically AC-coupled (except for rare balun designs, and most tube preamps I've seen are AC-coupled anyway). This could be a pro or a con, I suppose.

5. Potentially simpler and less costly than adding an active buffer stage especially for differential outputs.

Cons:

1. Potential noise pickup by the transformer itself ( better line-level signal transformers are shielded to reduce this issue). Both LF hum and RFI.

2. Roll off of the highest and lowest frequencies (most good ones will have response well above the audio band; lowest frequencies may be an issue if you want sub-20 Hz output). Also introduces phase shift over frequency though likely inaudible.

3. High-level signals can saturate the transformer and lead to very high (and ugly-sounding) distortion. This can be a big issue for power amps, less so for preamps since signal power is generally low.

4. Stability; have to be careful in the circuit design to handle any feedback loops and watch for any HF ringing induced by the transformer's inductance interacting with internal, cable, and load capacitance. There is some load sensitivity but rarely an issue for a line-level stage (again more so an issue with tube amplifiers driving speaker loads).

5. Cost, though see (5) under "Pros" -- could go either way.

FWIWFM - Don

Thanks for this informative synopsis, Don. As to point 2 under Cons, roll-off in the high frequencies apparently is often a problem with transformers, and the transformer quality is decisive. But to some listeners HF roll off seems to have appeal; I suspect that the "warm" sound of many tube amps is to a significant extent related to transformer HF roll off. So for some, a flaw becomes a virtue.

My Octave amp has a transformer with wide bandwidth (linear to 80 kHz), and no audible roll off. But then, the father of the designer of Octave amps had a transformer company, and that specialization in quality is what the son started from. Octave still make their own transformers of course.

Not sure about the low frequencies, but my low bass anyway comes from subwoofers that are driven by the preamp, which has no output transformer. So I wouldn't know if my power amp has, through its output transformer, a limitation there or not.

 

[DonH50]

NP!

The only name-brand audio transformer I have any experience with (and none really recent) is Jensen. I used one on my preamp for a while to provide a balanced output. IIRC its response was from around 10 Hz to about 40 kHz or so. A few years ago I was looking for a subwoofer isolation transformer (stupid ground loops) and they had a special LF transformer that would work to 1 Hz!!! Impressive... I think the top end on that model was a few kHz, plenty for a subwoofer, and I don't know if they had (have) a model that works from 1 Hz to 20+ kHz. At that time (again, couple of years ago) they had branched out from my long-ago experience in the pro audio world and had a wide range of pro and home audio transformers for a multiplicity of applications from MC step-up transformers to line-level to the aforementioned subwoofer-rated model. Expensive (well, for me, not for WBF) but really high quality.

Some of the mic transformers I looked at only went down to 50 Hz or so so you have to check specs.

I had a Denon step-up transformer for a while for my 103 but am not sure if Denon or someone else made the transformer inside.