Mezmerize DCB1 Building Thread

diyAudio Chief Moderator
Joined 2002
Paid Member
Thinking of it (+12dB gain, great sounding and neutral in character, using +/-10V PSU, no coupling caps and designed in compliance with B1/Pass concepts) I came to this idea (in attachment) based on Mr. Pass' proposal from BAF2013. Having LSK/LSJ JFETs available makes this possible.
J310 CCS provides about 20mA each, LSK/LSJ JFETs work at Vds of 7.5V and with Id at about 80% of Idss. P1 enables dynamic matching/distortion profile optimizing and P2 sets the DC offset.
Preamp's max. output is about 7V_peak of clean swing.
Zout is determined by value of R9 and it's not critical (R9/R6 ratio determines the gain of the circuit) since we have a Mesmerize buffer stage next to it.
I think they will work great together. AFAIC, it would be very difficult to find a better sounding gain stage...

Hmm BJT folded cascode JFET input symmetrical JC style basically, had such class in the list, yours is nicely refined, will have to breadboard and then build p2p but I have a hunch that the distortion profile with level will be different than DCB1 while being an excellent gain cell in itself. I wanted to test TO-92 Mosfets in differential symmetry with resistor loads also...
 
Hmm BJT folded cascode JFET input symmetrical JC style basically, had such class in the list, yours is nicely refined, will have to breadboard and then build p2p but I have a hunch that the distortion profile with level will be different than DCB1 while being an excellent gain cell in itself. I wanted to test TO-92 Mosfets in differential symmetry with resistor loads also...
I didn't measure the distortion but Mr. Pass did - you can see it in his BAF 2013 presentation ( 2013 Burning Amp Speaker: Nelson Pass - YouTube from 27:36 to 29:10).
THD is very low, clean H2, just like B1.
Anyway I did include the P1 so anyone can tune it according to personal taste.
Besides really great sound, the circuit provides great versatility - many different device types can be used, it can be made to perform a lot of different tasks under different conditions. I play with it for the last three months and I'm still thrilled... :)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I didn't measure the distortion but Mr. Pass did - you can see it in his BAF 2013 presentation ( 2013 Burning Amp Speaker: Nelson Pass - YouTube from 27:36 to 29:10).
THD is very low, clean H2, just like B1.
Anyway I did include the P1 so anyone can tune it according to personal taste.
Besides really great sound, the circuit provides great versatility - many different device types can be used, it can be made to perform a lot of different tasks under different conditions. I play with it for the last three months and I'm still thrilled... :)

Thanks for your data Juma. Will evaluate.
 
Member
Joined 2006
Paid Member
Thinking of it (+12dB gain, great sounding and neutral in character, using +/-10V PSU, no coupling caps and designed in compliance with B1/Pass concepts) I came to this idea (in attachment) based on Mr. Pass' proposal from BAF2013. Having LSK/LSJ JFETs available makes this possible.
J310 CCS provides about 20mA each, LSK/LSJ JFETs work at Vds of 7.5V and with Id at about 80% of Idss. P1 enables dynamic matching/distortion profile optimizing and P2 sets the DC offset.
Preamp's max. output is about 7V_peak of clean swing.
Zout is determined by value of R9 and it's not critical (R9/R6 ratio determines the gain of the circuit) since we have a Mesmerize buffer stage next to it.
I think they will work great together. AFAIC, it would be very difficult to find a better sounding gain stage...

I like it ... a lot... but where can I find the lsj fets ?
 
Having a slight issue with my Mez DCB1. Currently using 800ma slow blow fuses, and it has popped three so far - all on first startup after the unit has been cold for a while. The transformer is an Antek AS-1215, 100va. The unit is hot-rodded with 10ohm resistors. I seem to recall purchasing that particular value of fuse after reading through this thread, but all the parts were bought several years ago so I may have overlooked something. Should I use a slightly higher value fuse? I also have a bag of NTC thermistors for an amp build. If I understand their function correctly they're used to slow down inrush current on startup. I currently have a terminal block in place for the DCB1. Could I just use one of those on it like the FirstWatt general power supply?
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Having a slight issue with my Mez DCB1. Currently using 800ma slow blow fuses, and it has popped three so far - all on first startup after the unit has been cold for a while. The transformer is an Antek AS-1215, 100va. The unit is hot-rodded with 10ohm resistors. I seem to recall purchasing that particular value of fuse after reading through this thread, but all the parts were bought several years ago so I may have overlooked something. Should I use a slightly higher value fuse? I also have a bag of NTC thermistors for an amp build. If I understand their function correctly they're used to slow down inrush current on startup. I currently have a terminal block in place for the DCB1. Could I just use one of those on it like the FirstWatt general power supply?
That Antek model says 0.88A current rated, so use 1A.
 
OK, I finally finished my Mesmerize Blue! For input selection I used a circuit I posted here earlier (thanks AndrewT for assistance), and an Antek xformer (50VA 12V) and Par-Metal chassis. I also got one of those Chinese $10 stepped attenuators, which is remarkably good!

I'm attaching a few pics here. I previously posted photos of the board powered up. When I built my push-button selector I had initially planned for the whole thing to fit behind the front panel, but I couldn't get the layout to fit, so I split it into 2 boards:

View attachment 221467

Here is the chassis pretty much wired up. I ran out of my favourite shielded wire, so have only wired 3 of 6 inputs so far.

View attachment 221468

And here is why I built it: last year I built a set of Linkwitz Pluto speakers, but it seemed to me that they were not sounding quite right, and that my "preamp" (pre-outs of an old Arcam integrated amp) was the reason (I don't think the TL072 liked driving long interconnects):

View attachment 221469

Finally: The sound!
Yes, it makes a difference. Lows frequencies are definitely deeper and tighter, which is interesting because the Pluto eq/xover is AC coupled and has a built-in LF roll-off (15 Hz I think). More importantly, I had felt that the Plutos were missing some high-end "sparkle", which I attributed to some HF roll-off due to cable capacitance and a high-ish output impedance. The high end is back! I probably need to adjust the tweeters down a bit now.

Salas, thanks so much for this design and all your help, and others here (notably St. Andrew) also for the freely offered assistance. This is definitely the best thing I have ever built!


Hi nezbleu, do you have a drawing of your switchlogic that you finally used?
I would like to copy that :).

Regards
 
How hot can you rod it?

Hi!

First proper post here at DIYAudio. Hoo-rah.

A while ago BigE brought his Hypnotize DCB1 over and I was flabbergasted by the performance that I set out to build my own. Luckly, I acquired a pre-built Mez board from a well-season DIY'er. It's the black board, and was fitted with the stock value parts.

I have yet to procure the incidentals: box, heatsinks, transformer, RCAs, IEC, fuse, molex connectors, etc.

So to my main question: How high can this board be biased?

I have read that 200mA is common with some MUR820 and a resistor change, but if a higher can be achieved I'd rather go as far as I can with the build rather than biasing lower and forever wondering 'what if'.

Thanks for reading, and I look forward to your answers!

kwad
 
Thank you, Salas. I've read many of your posts from this Mez thread.

Is it counterproductive because of the extra work involved in heatsinking, fitting bigger caps and resistors? Is the sonic benefit not worth the trouble? Does it sound worse if biased beyond 200ma?

With 50VAC, 2x15V toroidal transformer, all I must do is follow your advice in post #1118.
 
Salas, can the 4700uf caps on the PS be increased to allow a .22uf film cap to replace the 100uf cap as in the Hypno? If so, double?
I modified one of my boards to replace the pair of single caps with a pair of double caps.
Added in an intervening R to make an rCRC filter to feed the supply rails.

I LOWERED the cap values to get the 4 caps into the area available.