Possible improvement to Mojo

[kp93300]

Hi yeo,

thanks for the explanation.
I add in the 1n 4148 last night,
 With the 3 transistor and the diode mod, i have to re adjust myself to the listerning experience. There seems to be so much of information being presented.
regards kp93300

[yeo]

hi guys

i installed the 3 transistor mod few nights ago. heard some improvement but something not quite right though the realism and dynamism improved further!

had a proper listening last night. found out i mistakenly swapped left/right channel the previous night, thus the "not quite right" experience. you must be wondering what kind of tin ears i have right? ha!

anyway, the 3 transistor mod is superior over the 2 transistor mod but the margin is not as great compared to 2 transistor mod over "no mod".

the refinement, smoothness, control and dynamism is really something!

i love nat king cole's voice and with this latest mod, i have never ever heard him sound soooooo good in any system at all.

one of our most frequently used test track is cassandra wilson's "strange fruit" from her "new moon daughter" album. very taxing on your system as the bass and her baritone voice could overwhelm your system (and your ears!) easily. every time we play this track, my system handles pretty well but with the 2 and then 3 transistor tweak, the control and "ease of flow"... marvelous! this is a good test for your system too. if you find her voice and the bass meshed up, you know something is giving up and couldn't handle all that energy.

so all in all, i urge everyone to mod their monicas! all credit goes to paul for this excellent tweak!

yeo

[NicMac]

Hi Yeo,
You and Paul have convinced me that I should introduce the three transistor tweak on my Mojo's.
My boards are becoming a bit messy with external V-regs et al. and I was wondering if you are planning/considering to make a Mojo2 board with the transistor tweak introduced and possibly headers for the (optional) use of discrete voltage regulators.
I would certainly get a couple of these to tidy up my bird-nest Monica's.
Thanks,
Nic

[weaver]

Hi Yeo,
You and Paul have convinced me that I should introduce the three transistor tweak on my Mojo's.
My boards are becoming a bit messy with external V-regs et al. and I was wondering if you are planning/considering to make a Mojo2 board with the transistor tweak introduced and possibly headers for the (optional) use of discrete voltage regulators.
I would certainly get a couple of these to tidy up my bird-nest Monica's.
Thanks,
Nic

+1

[Paul_Hynes]

I could do with one too as it might encourage me to pull my finger out and assemble my Monica DAC.

Regards
Paul

[Paul_Hynes]

Hi Yeo

Regards
Paul

[Paul N]

Hi

I've just noticed that I made a slight error on the three transistor (CFT) circuits. The two resistor's values are interchanged. The top one should be 2k2 and the bottom 1k. I've corrected the schematics.

This isn't that critical, I believe (Paul H suggested the given values as a starting point) and as several of you have had good results with the "wrong" circuit, I wouldn't worry about it.

Apologies

Paul

[yeo]

wow! i hear you guys.

ha! i thought of stopping any working on monica and starting off on a new dac but the old girl just refuses to die! and with paul's help, managed to seduce us all over again!

yeo

[Paul N]

Hi All

Now things- for me- are getting weird! I suggested the changes to Mojo mainly to try and help Yeo get his mojo back! Things have been a bit dead around here this year, and Yeo was showing signs of losing his enthusiasm. Unthinkable! 

I hoped that the changes would sound better, and hoped to try (still do) something similar with Twisted Pear’s Buffalo DAC.  But I never expected that Monica would compete with Buffalo whatever I did. The gulf between them- to my ears- was too big.

Well, I’ve done a lot of listening now, to the three transistor mod (CFT), and I’m consistently preferring Monica! Until yesterday, I was pretty certain, but.... For, instance, listening to a Mingus track,  Buffalo brought out inner structure of hf sounds much better- cymbals, maracas etc (and violin harmonics on other tracks). But it's a bit like the effect of a supertweeter- emphasising very high treble (10-20kHz only- I don't believe in all the hype about supersonics- clue is in the name!)- which is known- and used by speaker designers- to falsely augment detail. Actually, in this case, I suspected this detail was genuine- NOS dacs generally have a problem with hf, I've read (including Monica I thought, though much less than most). But, on the track I was listening to, Monica sounded more "organic"- more real, better timbral reproduction. And Mingus' bass accompaniment was much better with Monica- fuller, more distinct, much more real than with Buffalo (I'd noticed the vast improvement the CFP and CFT gave Monica before in the bass)

Last night I removed the thick film SMD (1k and 2k2) resistors, and fitted some metal films (still SMD). Since then, all I hear from Monica is good-  hf detail and all!

It’s possible that I’d cracked one end of one of the original resistors. A SMD must lie flat of the surfaces it’s soldered to.  Otherwise it’s easy to snap off its end contact. I only used them as CFP’s can oscillate, and I wanted to keep the layout compact. But the Zetex seem fine, so I’d recommend other people use good quality small (0.25W?) through hole resistors instead. Keep the leads short.

Yes, we need upgraded pcbs. I’d suggest one for the digital circuitry and the DAC- with a layout to accommodate the various three legged regs we now use instead of the diode chain/CCS, and one for the I/V circuit. It should be possible to connect the two closely together. I suggest the inclusion of a preset at the bottom of the CCS chains on the I/V board. This could be shorted out for use with Monica, or used- with suitable diodes – to set alternative input voltages for other SE current output DACs. I’m sure this circuit will be of interest to many, when the word gets around.

Moving the DAC chip off the I/V board would lessen the board’s backward compatibility I know- but we’ve come a long way again, and need- IMO- to make best use of what we now have

Paul

[ole]

I will receive the Zetex transistors in a few days so I can try this. About the resistors - I know it's an age old topic, but anyway: which type would be preferrable here? Metal film, carbon film, wire wound? I'm towards carbon films, but perhaps one of you know better?

Ole

[leo]

Went straight to the 3 transistor + 1N4148 version,  I used BC337/327  just to try it out, used metal film resistors .
Initial impression, top end more zingy, vocals more forward ,  bit more clout to bass.
Only just added it so more listening needed 

[sph]

It’s possible that I’d cracked one end of one of the original resistors. A SMD must lie flat of the surfaces it’s soldered to.  Otherwise it’s easy to snap off its end contact. I only used them as CFP’s can oscillate, and I wanted to keep the layout compact. But the Zetex seem fine, so I’d recommend other people use good quality small (0.25W?) through hole resistors instead. Keep the leads short.

Will the smaller 0.125W resistor do?

[yeo]

yeah, 0.125w is no problem.

yeo

[Paul N]

Hi Ole

I've no idea. Take an empirical approach, and report back!  

Hi splv

The resistors are across the base-emitter junction of each transistor. This is forward biassed and therfore has about 0.6V across it- just like, e.g. a 1N4148 diode would.

The power developed in the resistor is thus V²/R = 0.36 mW. Very low! So the resistor can be a a very low wattage type

The resistors act to "speed up" the circuit. I've yet to read an explanation of this, to be honest. Can any one help?

Paul

[Paul_Hynes]

Without getting into complex semiconductor physics search “Hole Storage in transistors” on the Net for some background information. The base-emitter resistor is there to reduce the effect of hole-storage in the base region by providing a path for the stored base charge to discharge. This speeds up the switching performance of a transistor.

Regards
Paul