Continued sidetrack discussion from "Tiniest Portable amp I can build" - nikongod microtransformer-based impedance step-down box

Jun 7, 2011 at 11:40 PM Post #46 of 102
Okay, one last set of measures (even I think I am starting to beat a dead horse!):
 
This time I changed the measurement methodology to probably the most realistic setup yet. Not sure why I did not do this from the start, but, the SoundBlaster has a headphone out jack (with analog attenuation). So, I routed the headphone out (through the device under test where appropriate) to the mic input for all tests. There is no digital attenuation on the signal (all outputs full) and the mic set at 50% (0db). When performing the loopback, the analog attenuation is heavily applied. When using the transformers, there is no analog attenuation (wide open). Input levels for all tests are about -3.6db at the mic.
 
So, lets see what we get:
 

 
Freq:
 

 
THD: 

 
IMD:
 

 
Dynamic Range:
 

 
Noise:
 

 
 
This particular set shows more similarities than differences, so, you can draw your own conclusions. Mine lean towards stating the transformers are allowing more signal through (less attenuation) and adding little to no distortion or frequency artifacts. At least that is what this set of measures show, which, I have certainly learned can be all over the map across differing samples and totally open to interpretation. Pretty much like any other statistical characterization. 
tongue_smile.gif

 
So, at the end of the day, I am not sure if I have helped or hindered the argument, but, there it is.
 
Consider this horse dead.
 
 
Jun 11, 2011 at 10:35 AM Post #47 of 102


Quote:
Nice work Eee Pee!
 
What did you do with the center taps (black wire)?
 
Most importantly, what do you think of the sound?


Thanks.  
 
Using your picture as a reference, I cut the black wires as short as possible.  Is that not right?  
blink.gif

 
Everything sounds like it's gotten a smoothing over.  I'm using 50 ohm Sennheiser HD595s and tried the iPod Touch and it simply wasn't even close to loud enough to be able to get into it.  Then I plugged it into the iMac which goes a little bit louder and enjoyed it for hours at a time for a couple days.  I also tried it with the DNA Sonett amp with both impedance settings (120 ohm or 28 ohm) and while it went louder, it got uglier and uglier as it did.  Then I got some new tubes for that amp, so the little box sat off to the side until about an hour ago where it's back attached to the computer again.  Figure I'll do a couple few hours this way, then take the box out of the chain and check it out again and see what differences I hear.
 
biggrin.gif

 
 
Jun 11, 2011 at 10:29 PM Post #48 of 102
The DNA sonnet already has transformer coupled outputs. I'd lean away from trying to put the 2 together. 
 
The key is considering how you "stress" the amp the least. 
The Sonnet with its transformers is already running the tubes under a fairly light load, so the transformer means it has to swing a bunch more voltage to make the headphones run. 
Opamps (or OTL tubes) load the output devices HEAVILY and trading voltage swing for current & impedance can make a huge difference.
 
Jun 12, 2011 at 6:13 PM Post #50 of 102
 
Quote:
It'd be interesting to try that with a Bottlehead Crack amp.


Get transformers rated for more power (like 3-5W, or an honest 1W) and it WILL run efficient speakers - a 600:8ohm transformer should get you just short of 0.5W into 8ohms with that amp (yes, I tried!) which is more than many people think. Certainly enough for many single drivers. 
 
My auto-former connected transformer schematic is still up in the Crack thread, post #383
 
 
http://www.head-fi.org/forum/thread/476650/crack-bottlehead-otl/375#post_6987382
 
You could be more conventional and hook the transformers up as transformers too, but I wanted to be "more bottlehead" and do it like they do in the SEX amp. 
 
Jun 12, 2011 at 11:16 PM Post #51 of 102
Using your picture as a reference, I cut the black wires as short as possible.  Is that not right?


No, that's the correct way for the full step down (1k to 8ohm).

I am thinking of building another one using the center tap (blue-black or black-green, take your pick) and halving the step down (500 to 8ohm). This should accomplish more or less the same thing. The amp will see a slightly larger load than the 1k version, but, still a good deal easier than a direct low impedance load. The usable volume should go up by a good bit as well. Certainly worth trying at $12USD plus altoids tin. The truly enterprising would add a switch and make both versions selectable. Might add a dollar or two to the cost of the build, but, what the heck....

I just need for my local RatShacks to restock as I cleared them all out of transformers! :D
 
Jun 13, 2011 at 12:20 PM Post #52 of 102
The center tap to one of the ends is 250:8ohms. 
 
Voltage ratio equals the turns ratio, which is halved when you use the center tap. 
 
Impedance is proportional to the SQUARE of voltage ratio or turns ratio.
 
Jun 26, 2011 at 1:05 PM Post #55 of 102
Nikongod Box MkII
 
Rainy day here in Florida. Woodworking is hard enough in 90% humidity, impossible in 100%.  
wink_face.gif
  So, the main project on the bench will have to wait.
 
Anyway, I had some time on my hands and went ahead and added a new part to the mix:
 

 
And came up with this:
 

 
Yup, this allows either the end tap or the center tap to be utilized with a quick flick of the switch (switchable primary of 250 or 1k ohms):
 

 

 
Sound is about the same in over all quality at either setting. Perhaps a bit more down in the bass at the 250ohm setting, but, it is subtle. In any case, there is certainly less attenuation at the 250 ohm setting than at the 1k ohm setting (as expected). I think this is the definitive build only from the overall cost stand point and general flexibility. No matter the source, you should be able to get a usable setting and corresponding useable range of volume. Fun stuff!
 
BTW, I am NOT a big fan of the chocolate Altoids. Can't recommend them. 
frown.gif

 
Jun 28, 2011 at 2:03 PM Post #56 of 102
might have to give this a try
 
Jun 28, 2011 at 2:29 PM Post #57 of 102
Jun 28, 2011 at 4:37 PM Post #58 of 102
just finished it. since i couldnt find the schematic i went off of jdkjakes picture in the first post of this thread. ive got it hooked up to my macbook pro and i have my headphones plugged into the jack with the red and white wires and my source plugged into the green and blue jack. the mids are warmer and the highs are somewhat rolled off but its hard to compare because its quiet even at full volume. 
 
Jun 28, 2011 at 5:37 PM Post #59 of 102


Quote:
 the mids are warmer and the highs are somewhat rolled off but its hard to compare because its quiet even at full volume. 



Yep, we agree.  Did you cut the black wires, or do you still have them on the transformer?  You can build the 250 ohm setting if you still have them.
 
I have yet to get around to that... I'm so lazy...
 
Jun 28, 2011 at 6:08 PM Post #60 of 102


Quote:
Yep, we agree.  Did you cut the black wires, or do you still have them on the transformer?  You can build the 250 ohm setting if you still have them.
 
I have yet to get around to that... I'm so lazy...


i didnt cut them yet thank god. i just read the post above that says theres less attenuation with the 250ohm setting. how do i wire up the switch for 250ohm? i can kind of see in the picture but its hard to tell
 
 

Users who are viewing this thread

Back
Top