Rerurn to Romy the Cat's Site

Romy the Cat wrote:

In my upcoming DH Milq MF DSET I will have an Innerstage transformer and output transformer. The loading of output tube has known effect to harmonics vs. transients, it is know, it is predicable and it is manageable. The Innerstage transformer might or might not have a resistor in secondary that impacts the loading of the driver stage. The loading of the driver stage also has the harmonics vs. transients consequences. So, what I wonder: how different sonically the impact of Innerstage loading from the output transformer loading. My Innerstage transformer is reconfigurable 1:1, 2:1, 3:1, and 4:1. So, would the change in the Innerstage loading twice have the same sonic effect as the change of the output transformer twice (let discard the gain change or power change)? Does anybody have this experience to share?

A site visitor commented to my question above and I think the information and the view are valuable to share it:

"I guess I'm suggesting that applying a load to the secondary of a transformer never helps the sound of a properly designed transformer. There are four cases that I can think of where people purposely load a transformer and I'll brush on them briefly.

The output transformer:  The nature of this beast requires loading in order to deliver current, so this is the one case where loading is a necessary evil.  The whole purpose of this transformer is to deliver power to the load.  Every other audio coupling transformer is essentially there to couple voltage.

Secondary Loading to damp ringing:  I personally believe the lower the load (higher ohmic value) the better.   A 1:1 interstage feeding the grid of a tube essentially works 100% in the voltage realm since no current flows through the transformer.  It is only when you apply a load that current starts to flow and then losses become an increasing problem.  Often times people will look at the frequency response of a mis-applied transformer and think they are getting clever when they add a resistor or R/C network across the secondary to "fix" issues with frequency response (typically ringing).  The problem is these networks absolutely suck the life out of the music.  Here is the thing most people do not understand.  For a given source and load impedance a transformer can be designed to match those impedances without ringing.  Since a transformer manufacturer cannot offer devices for every possible combination they typically design a one size fits all approach and rely on secondary networks to explain  their choice of compromise.  One other thing secondary loading does to a transformer is impact the phase response.

Secondary loading to provide a proper load to the source:  This can often be a "kill two birds with one stone" approach where the added load both provides the proper load for the source and damps any possible ringning.  In the voltage realm the best example of this is the MC SUT.  I'll link you to a post I made about this that should cover the basics, sadly it seems to be falling on deaf ears.

Loading a transformer by a filter:  Many people find that lower impedance filters sound better than higher impedance ones and if transformers are involved, the low impedances provide a larger load. This like the output transformer becomes a necessary evil, yet it is still important to remember that a 4:1 to drive a 600R LCR should be designed quite differently than a 4:1 used as a preamp output which will be terminated by the cables and a 100K input resistor of the following amp.

My entire approach to magnetic design comes from collecting the needed info to make the proper choice of compromise for each specific situation.  I just did a pair of 8K nickel outputs that someone plans on using for a pentode.  This design had to be drastically different than the same 8K that would be driven by a triode.  Why you may ask?  

Well the 8K driven from the pentode has 80K source impedance and the 8K driven by a triode a fraction of that.  Sure both will "work" and to this day I have yet to see anybody make this distinction in transformer design.  I'm not sure whether the transformer makers are scared that if this info gets out they will need to double their product line of if they fundamentally do not get the difference in operation of the two since that is a circuit thing and they are magnetics guys."

*** Are you implying that in inner-stage transformer the sound  differences between the various ratios (discarding gain) will NOT be  ONLY because the change of loading?

Rerurn to Romy the Cat's Site