THE STORY

Like many other times, It all started with a thread in one of those easy-to-spend-a night forum. 
This time was Bert Doppemberg's  Lowther's Forum in Egroups.

The period was jenuary or thereabout, the  thread was started  By brian Clark   describing his sound chain. During the description the spoke of a "trasformer volume control" based on  Sowter components.

The idea was very simple, susbtituting the usual volume control with a x-former having  multiple taps in the secondary winding. Each taps was connected to a suitable rotary switch to allow volume adjusting.

With a quick exchange of e-mail with Brian Clark we got more details: he was using a nice & cheap 14level solution ( 94xxx serie: steps of about 3db from 0 to 42 db). There was also the chance to have some taps in the primary section to have also a volume boost COOL!. You can also use it to do a passive preamp!.

OUR CHOICE

After a little consultation with Andrea  the dice was thrown and we decided to experiment this solution. we decided that the first need was to  increase the number of taps in the secondary section. This was necessary because we were afraid that  in our high efficiency - low power sistems 3db steps were  too high to manage power in our listening rooms.

The last  doubts were concerned about overall performance of this x formers: it'is not easy to build pieces like that. Expecially band response, distortion withing the extreme of the band are keypoints.

These were  the main needs we passed via e-mail to Sowter.
Mr Brian Sowter himself spoke to us. He  was very kind and helpful giving us a lot of details and ideas. We were discussing about a couple of items made on our own specifications when, seeing the high request for a "off-the shelf" model with the highest secondary number of taps available according to manufacturing needs and highest performaces, he decided to put in the list a new model number : Type 9300 . This attenuator  was what we exactly need.

These are its main features:

•  impendance 10k:10K
• 50db attenuation in 2 db steps
• 10k, 2.5K, 600ohms, 150 ohms primary imputs
• mu-metal core

Mumetal is a special mixture of metals componets to have the highest permeability an lowest distortion. This manifactuiring technology is used by sowter to build recording  studio line trasformers (!) . The case is simple &  cylindrical,  so no frivoluos details only performances!.
Here is a photo taken from Sovter's site

As you can see a lot of wires are coming out from this little boy, and these will be you main problem. ;).

Here is the connection diagram (from sovter datasheets)
 
 

Here is a photo of the prototipe of this x-former attenuator near the DAC "minimo"

Here are some ideas to adapt this solution  to you listening pleasure:

• Use a very good switch (we've used Elna gold pin  24 pos..).
•  AVOID current flowing. Remember that this is an attenuator not a trasformer.  Mu metal can easyly saturate and be damaged by current flowing in windings. So check the working condition.
• In DACs with current output, can be used at the same time  to increase voltages on the (low) l/V resistor  and to adjust volume  you can spare one stage and have beeter volume control (cool).
• You can use this as a passive preamp using 2500, or 600 ohm taps and allowing that the previus stage as adequate output impendance. For example if you use the 600 ohm tap you can have a 16 time voltage amplification.
• It decouples grouds! this is a nice feature to take away those nasty  noisy grounds like DACs, CDs etc.

Listening Impressions
 

Frankly we were a bit unsatisfied with the stepped resistor volume control (see the project in this site) based on holco resitor. The main reason was that changing volume level caused also a slight modification in the frequency response.

Andrea has integrated this volume control in Euridice's Preamp to the  chain we used to test this solution was this:

• CD Motor: CD Pro philips unit
• DA Unit:    Built in Cd pro + DAC with 1541 (will be published soon).
• Preamp: Euridice with 6c45P
• Power Amp: Mono-Triodino III (2a3+6sn7) see the details in my homepage
• Speakers:  Side Vivace (side horn with PM2A)
• Signal cables: DIY silver-teflon Cables
• Power cables: Megahertz "silver-Tef" cable

These are (in our  poor english) the impression we had:

• "Trasparency & ambience": There is a great sense of clearness.  This is probably due to ground decoupling and ambience recovery.
• "The Band is the same"  Another of the most notable things is that is you increase level all the part of the spectrum move in the same way . One thing that we've realized is that with the stepped resistor attenuator we used previously (project in these pages) in certain condition  the sound started to "shout" (very high level).
• "Whispering with body" at very low level - those nice at midnight ;)  -  with previous voulume control sound seemed  very "thin", this will be another  surprise: you can have a great "body" listening  also at low level .
• "Pump up the Volume": We don't know why but from the time we've put this pots in the chain we're listening at higher level than before. The tonal balance seems more natural & open also power flows more gently. At the end we find ourselves listening ( without fatigue) a level is so high we coulnd't speak .

Rating pro and cons we could say that the main  "minus" of this pots are:

• too much taps  at the lowest attenuation levels: We would say that having 2 db taps from -50db till -40 db is not so useful: the sound volume in these position  is *very* low: you won't use them. better would be to have 3 db steps till 40 db.
• the cost (but  don't think to spare a lot with a  good switched resitor pot
• the low level impedance (10k) that doesnt' allow "universal match".

Some Last Minute's Measure's
 

Thanks to the availability of time (and wive's holidays) Yesterday me & Andrea  took some measures on these little boys using a Philips 50Mhz oscilloscope & and an Exact signal generator. Here is the raw scan of one of the measures's set we've taken:

These  trasformers are almost flat starting from 15hz (My scope has problem  to go below) well  beyond 48khz. There is a little "hole" arond 50khz , a little tip then  a very sligt decrease till  60khz ending with the  peak  at 90kz due to internal resonance . COOL

Very good Mr Sowter: these are really very good responses for this little jewels!.
We've taken data at 10db steps and some more random sample to discover that the Frequency response  has no change with attenuation!. Very good once more!!!
 

To gain a more deep wiev in what was happened to our listening chain we've repeated the same measure to the steppend ressitors attenuator we were using. All resistors were holco 1%.
The most notable thing we've discovered was that the response was flat till about 30khz then a steadly decrease begun.  This response changed with attenuation, being slighlty worse at higher level of attenuation.
 

Andrea stated: "Okay:  that's was  what our ears said us for weeks ..."
"do you need to switch on the scope to be satisfied? " ...  just let's put all in place, put on some good Cd and enjoy..... we've lost the afternoon with these measures...  now it's time to play and have a good Beer!!!"

Me : "Okay you' right but what's the use of the scope then???". BTW do you know someone who needs a good resistor switched attenuator? (laughing!)
 

That's all folks, Listen - HAVE a GOOD BEER and ENJOY!
 

Back to my  homepage