A-B TESTS    
 

 

Headphone Amps

A/B Comparison #8

Beta 22 Bijou

.

DAC: Cambridge Audio  DAC Magic - via SPDIF
 
Source: Various FLAC Music Files
11/21/2010
 
Test conducted using a switch that connected either amp "A" or amp "B" to the headphones; amps were fed in parallel off the DAC.
I carefully matched the volume levels of the two amps using pink noise and an osciloscope


LCD-2

A little tricky setting matched levels here again. Matching levels between the two amps at 1000 Hz did not result in the same apparent loudness from each amp. Nor did trying to set matching levels using a 100 Hz tone.  I ended up using pink noise and a 'scope set to a 2 second sweep to act as an integrator.  This got me levels that sounded identical between the two amps.  Right away, this told me that one of these two amps is not ruler flat when driving the LCD-2.  (Guess which one had slight frequency response variations-  I'll give you a hint:  it wasn't the Beta-22)   SO: listening-  The Bijou emphasized the bass and lower mids just a little, but also rolled off a bit at the lowest extreme of the bass range. This could be related to the 50 ohm impedance of the LCD-2, maybe the Bijou couldn't quite provide the current down low while at the same time having a higher source impedance resulting in lower damping factor.  So the bass, while not "wooly" was a little more "full" but not as deep or tight. Lower mids were also just a little emphasized, making the sound a tiny bit "warm" - this lower midrange emphasis was not as strong as the slightly raised bass level, but it was audible.  Highs on the Bijou sounded slightly "sweetened," the typical harmonic character of a tube amp.  However, the very upper extreme treble was a little lower in level from the Bijou, and the Bijou also seemed to blunt transients just a little on the LCD-2's.  There was something more going on the midrange that I can't grasp- certain sounds seemed to be brought out a little from the mix by the Bijou, while other sounds seemed to recede a little. However this did not seem like a simple  frequency response artifact- rather, it seemed like the harmonic content of the sound was being changed a little at certain frequencies, which was audible at times, and at other times this change in timbre was masked when the music  was "busy" in terms of lots of sounds happening at once. This isn't really a satisfactory description of the differences in midrange I heard between the two amps, but it's as close as I can get.  In simpler terms: the midrange sounded a little different between the two amps.  Overall, on the LCD-2's, I liked the Beta-22 more. 

It should be noted that the Bijou drove the LCD-2s to a normal level without having to resort to increasing the NFB. I would say that the Bijou had no real problems driving the LCD-2, but the match was less than ideal.

HD-800

Setting matched levels was easier, although for consistency I used the same pink noise method as with the LCD-2.  Again, the bass and lower midrange were a little "warmed up" on the Bijou, with ultimate low bass a little rolled off.  However, this was not as pronounced as on the LCD-2.  Also, the midrange timbral changes I heard using the LCD-2 were not evident.  The midrange on the Bijou sounded just a little warmer and that's it, no other hard-to-describe difference. Highs from the Bijou were "sweetened" a little, again the signature of tubes, but the transients and upper treble did not seem very much different between the Beta-22 and Bijou; on the LCD-2 I could hear differences.  Overall, it was clear that the Bijou "liked" the HD800 more than the 50 ohm LCD-2.   Even so, I preferred the Beta-22 to the Bijou on the HD800's, because I liked the more neutral bass balance of the Beta-22.

MY CONCLUSIONS

There were easily audible differences between these amps, especially on the LCD-2's.

Both amps sounded "clean" and highs were never harsh from either amp on either set of 'phones.

LOUD levels were attainable with both amps, although the Beta-22 could not only play LOUD it could produce RIDICULOUS levels. (Hearing-harmful levels, levels at which I would never actually listen.)

I may revisit this test with some other headphones, like AKG-701's or maybe Beyer DT880.  I want to see if the Bijou might have a better synergy with other phones.

 


The Bijou is a DIY tube amp  designed by Alex Cavali and with some evolution by Regal and others on the boards. It is an OTL amp of unique design, being similar in topology to the Futterman OTL design.  In this case, all the complications the Futterman amps use to maintain zero volts DC on the output is forgone and a capacitor is used on the output to block the DC.The amp is also unique in having a continuously variable global negative feedback pot. The power supply is regulated, supplying 250 volts to the output tube plates. A neat feature of the power supply is the use of a rectifier tube rather than silicon diode bridge- the advantage here is that the heater of the rectifier tube is elegantly used as a delay timer, and full B+ is not applied to the output tubes until the rectifier filament heats up- saving the output tubes from the wear and tear of applying high voltage to them before their filaments are heated.

My Bijou uses "Regal's mod," and I am using Bugle Boy 6DJ8 input stage tubes and ECC99's as the output tubes.  For this test, I kept the NFB pot at minimum. (Which is also how I listen.)

Photos: My compact Bijou build.


 

Copyright © 2011