autoformer attenuator and step up transformers


can a step up transformer (like a jensen 1:4 for example) be used in front of an autoformer attenuator (such as the one made by dave slagle (intact audio) and used in the Emia attenuator box) to increase the gain of the signal going into the passive attenuator? would you expect this type of set up to sound good or do you foresee any issues with this line of thinking? thank you. 
salbert
If the transformer steps up the voltage provided by the source by a factor of 4 (which is a 12 db increase in gain), it will result in the source component seeing a load impedance of 1/16th of the impedance that exists "looking forward" from the output of the transformer.

If the autoformer is set so as to step down the voltage that is applied to it to a significant degree, it would result in the impedance "looking forward" from the output of the transformer being significantly greater than the input impedance of the amplifier (or whatever component is being driven). But since what you are apparently trying to accomplish is a net overall increase in voltage, presumably the attenuator would often be set to not much lower than 1:1. For a 1:1 setting of the attenuator and a 4:1 transformer ratio the source component would see a load impedance of 1/16th of the input impedance of the amp. Which in many cases will result in a poor impedance match and consequently in degraded sonics, depending on the impedances of the specific components and on how those impedances vary as a function of frequency.

Also, while I’m not certain, it seems conceivable to me that sonics could also be adversely affected by things like ringing, phase shifts, etc.

Regards,
-- Al

If I were to use a step up and autoformer combo I would place the 1:4 after the autoformer and arrange it so it is only used when it is absolutely needed.  

dave
hi dave. would you please kindly explain why you would place the step up transformer after the autoformer and not before? also,why would you only use it if absolutely necessary?

my second question to dave is whether your autformers sound better or work more optimally when used on their lower half to the volume control (from 7 o'clock to 12 o'clock i.e. greater attenuation) vs the upper half of the volume control (from say 1 o'clock to 5 o'clock i.e. less attenuation)? i read that being on the lower end of the volume control drops the output impedance but if that is a non issue, does it matter with regards to sonics which setting on the autoformer i use?
hi al. thank you for replying to my post. since i am not very good at figuring out impedance issues i will go ahead and tell you about my equipment and you can tell me if you forsee a problem. also, do you know why dave from intact audio suggested putting the step up after the autoformers?
i am using a revox b77 reel to reel player(http://www.reeltoreel.de/worldwide/B77.htm) , the intact audio autoformers, and monoblocks with 137k input impedance. thank you again :)
It looks like the output impedance of the B77 is nominally 390 ohms, although if I’m interpreting the datasheet correctly it may rise to as much as 1.5K if the output level is adjusted downward (which you presumably would not be doing). While those numbers most likely correspond to mid-range frequencies, given that the B77 is a solid state design chances are they do not rise greatly at other frequencies. (Tube-based output stages commonly have coupling capacitors at their outputs that in many cases cause the output impedance to rise substantially at deep bass frequencies, although there are some solid state designs for which that is also the case).

137K divided by the factor of 16 that I mentioned would result in the B77 seeing a load impedance of 8.6K, which seems reasonably comfortable in relation to those numbers.

If I recall correctly, though, some autoformer-based passive attenuators can provide an output voltage that is up to about 6 db greater than (i.e., twice as much as) their input voltage. If that applies to your unit, and if you were to set the volume control at the top of its range, the 8.6K number would be reduced to one-fourth of that amount (2.15K), which may be too low to be optimal in relation to a 390 ohm output impedance.

Regarding the reason Dave suggested putting the transformer after the autoformer, I don’t really know. But a guess would be that he might be anticipating that the sonics provided by the autoformer would be compromised if it is asked to handle voltages as high as 6 volts or so (corresponding to 4 times the 1.55V max output of the B77, 1.55V being roughly in the neighborhood of the max voltage of many unbalanced line-level outputs).

Regards,
-- Al

Placing the 1:4 in front of the autoformer would increase the source impedance to 6.2K which would then be reduced by whatever attenuation is needed.  This added 12dB of gain would be offset by anything over 12dB of attenuation adding lots of compromise into the system when it need not be there.  I would guess the 1:4 is needed for the occasion when one needs to "turn it up to 11" and if that is the case it should only be used when needed and not be in system all of the time.  
Having a 390 ohms source drive the autoformers is fine and the output impedance from the autoformers will always be 390 ohms or lower so the source will have no issues driving the step up when needed however going the other way around and feeding the autoformers from a 6K+ source impedance is not placing them in the best situation.  If you find you constantly need the extra 12dB of gain then a better solution would be a simple active stage.  

dave 
hi dave. thank you for the detailed explanation. so far, i have been able to reach a volume that is quite loud in my room without using a step up transformer but rather turning up the autoformer volume control within 3 clicks of maximum (i.e. at 3 o'clock). this leads me to my second question which i really hope you would take the time to answer as well:

do your autformers sound better or work more optimally when used on their lower half to the volume control (from 7 o'clock to 12 o'clock i.e. greater attenuation) vs the upper half of the volume control (from say 1 o'clock to 5 o'clock i.e. less attenuation)? i read that being on the lower end of the volume control drops the output impedance but if that is a non issue, does it matter with regards to sonics which setting on the autoformer i use?

the perfect position for the autoformers is full volume then they are simply a shunt element.  The worst performing taps are in the 0 to -6dB range but they still have 100Khz+ bandwidth and what ringing there is is also above 100khz which is far enough outside the audio band that it doesn't bother me from a technical perspective.  In any event the behavior of the autoformers at -3dB will far exceed that of any step up that you put in front of it and if just a slight amount of gain is needed (up to +6db) again the autoformers can be configured to seamlessly shift into step-up mode so you not only get 11 but possibly 12 on the dial.  The important part here is that when you leave the "gain" mode you also leave the compromises incurred from that mode.

dave
hi dave. thank you for the explanation. i will go ahead and use the autoformers without any step up as you recommend :)

i saw on your website that the autoformers skip the -3.75 dB tab.

what attenuation are you actually getting when you are on the setting that should correspond to -3.75 dB?
-5dB