Waking the dead. I just commissioned my Bruce Rozenblit (Glass Audio 1991) which is nothing really fancy.
It sounds better than my X-PRE. Unfortunately it's as ugly as sin.
So, as I've not been able to find a schematic of the X-PRE I took this opportunity to whip out the guts out of mine (round can version) and have a quick look.
To take the X-pre apart:
The boards themselves are 67mm x 180mm.
Height between top and bottom of can is 70mm.
There are two sets of PCB "rails" in the can about 57mm apart - they run the length of the can (180mm)
The X-pre only uses the lower rails - the upper PCB is attached by 35mm screws/standoffs (including what I can only presume is an EMI barrier - earthed to upper PCB) to the lower near the front and located by the RCA sockets (standard cheap lo-fi parts).
Both my boards are marked "Issue 03".
The valves in mine are Electroharmonix 6992EH dated 2010.
The quad op amp is a Motorola MC33079.
Forgot to check the part #s of the two transistors and two TO220 devices on the top board.
Suspects are BC327/337 and 7815/7915 based on the X-CAN PSU
All rectifiers I looked at were 1N4001 types. 😛
Interestingly the links between top and bottom do the following
a) 12VAC to top
b) lower inputs to top board!
c) inputs and power to selector switch once past the valves
IE it's just a power supply and audio cables (Clever MF designers trying to get around the very small component budget.)
I'm still trying to work out why you would bother with the OP-AMP given the 6922 can do Rp of 2K at 10mA. Perhaps not with Vp at 30V?
Upgrade thoughts to follow (when I scan some notes)
On One thread to bring them all the existing threads I've been able to find are:
It sounds better than my X-PRE. Unfortunately it's as ugly as sin.
So, as I've not been able to find a schematic of the X-PRE I took this opportunity to whip out the guts out of mine (round can version) and have a quick look.
To take the X-pre apart:
- Remove the selector and volume knobs
- Undo cap screws on front
- Undo screws retaining rear RCA sockets
- Undo rear cap screws - noting that top screw retains chassis earth eyelet
- Slide guts out the front. (LED is firmly fixed in the front panel)
The boards themselves are 67mm x 180mm.
Height between top and bottom of can is 70mm.
There are two sets of PCB "rails" in the can about 57mm apart - they run the length of the can (180mm)
The X-pre only uses the lower rails - the upper PCB is attached by 35mm screws/standoffs (including what I can only presume is an EMI barrier - earthed to upper PCB) to the lower near the front and located by the RCA sockets (standard cheap lo-fi parts).
Both my boards are marked "Issue 03".
The valves in mine are Electroharmonix 6992EH dated 2010.
The quad op amp is a Motorola MC33079.
Forgot to check the part #s of the two transistors and two TO220 devices on the top board.
Suspects are BC327/337 and 7815/7915 based on the X-CAN PSU
All rectifiers I looked at were 1N4001 types. 😛
Interestingly the links between top and bottom do the following
a) 12VAC to top
b) lower inputs to top board!
c) inputs and power to selector switch once past the valves
IE it's just a power supply and audio cables (Clever MF designers trying to get around the very small component budget.)
I'm still trying to work out why you would bother with the OP-AMP given the 6922 can do Rp of 2K at 10mA. Perhaps not with Vp at 30V?
Upgrade thoughts to follow (when I scan some notes)
On One thread to bring them all the existing threads I've been able to find are:
- X-PRE musical fidelity preamp. need help modifying Last post 26th August 2010
- Upgrade the Musical Fidelity X-PRE Last post 3rd May 2007
- MF XPre troubleshooting (no real posts)
- Various photos and schematics for other X-products 2008
Last edited:
Why do we have a quad op amp ? Anyone's guess - I've yet to find a schematic. (Took top/bottoms shots so should be able to confirm/deny X-10D / X-CAN copying when I can be bothered)
Interestingly the owner's manual claims power supply current to be 500mA but the heaters alone should be pulling 395mA.
Now it's still a very pretty box with high S.A.F. but clearly sonics are limited.
Plan A was to use the 12V supply to generate HV via a tiny mains transformer, rectify with 6x4 and use a 5687 at 20mA with some CCS. No space for output transformers, I think so caps only. Use the 12V to drive the 6.3 heater on the 6X4
But the power budget is bust - and by a long way. 20mA at 240V times two was 800mA. Add another 800mA of heater and there's 18W in the box.
I can't seem to find a smaller rectifier.
Plan B.
Keep the ECC88/6DJ8. But only one. 2Kohm plate at 10mA c.120V. Still use CCS so c. 20mA at 200V (if choke loaded to shut up the diodes) for 300mA @12V.
If clever I'd series the heater and transformer (as my power transformers will produce 230VAC). Would need a little shunt past the heater as it only wants 375mA and the B+ trannie will want 600mA at 6.3V)
I might find some 10K/10mA line drive transformers (Lundhal LL1660/10mA?) I could stuff into that chassis - then B+ trannie could live on 375mA at 6.3V
Interestingly the owner's manual claims power supply current to be 500mA but the heaters alone should be pulling 395mA.
Now it's still a very pretty box with high S.A.F. but clearly sonics are limited.
Plan A was to use the 12V supply to generate HV via a tiny mains transformer, rectify with 6x4 and use a 5687 at 20mA with some CCS. No space for output transformers, I think so caps only. Use the 12V to drive the 6.3 heater on the 6X4
But the power budget is bust - and by a long way. 20mA at 240V times two was 800mA. Add another 800mA of heater and there's 18W in the box.
I can't seem to find a smaller rectifier.
Plan B.
Keep the ECC88/6DJ8. But only one. 2Kohm plate at 10mA c.120V. Still use CCS so c. 20mA at 200V (if choke loaded to shut up the diodes) for 300mA @12V.
If clever I'd series the heater and transformer (as my power transformers will produce 230VAC). Would need a little shunt past the heater as it only wants 375mA and the B+ trannie will want 600mA at 6.3V)
I might find some 10K/10mA line drive transformers (Lundhal LL1660/10mA?) I could stuff into that chassis - then B+ trannie could live on 375mA at 6.3V
Plan B
Current Plan B starting point.
Current sources are IXCP10M45S .
Roger Modjeski [1], suggests using only 5mA but I'd rather get the lower Rp (2Kohms @ 10mA).
His commentary on the 6DJ8 derating [2] does suggest that 80V-100V is more than enough for Vp. I need to see what the IXCPs want but I'm guessing they should be happy with 20V or so.
I should just be able to squeak B+ of 110V from 120VAC derived from a back-to-front 12VAC/240AC transformer
Current Plan B starting point.
Current sources are IXCP10M45S .
Roger Modjeski [1], suggests using only 5mA but I'd rather get the lower Rp (2Kohms @ 10mA).
His commentary on the 6DJ8 derating [2] does suggest that 80V-100V is more than enough for Vp. I need to see what the IXCPs want but I'm guessing they should be happy with 20V or so.
I should just be able to squeak B+ of 110V from 120VAC derived from a back-to-front 12VAC/240AC transformer
Attachments
Last edited:
Conventional biasing would require about 1.8 or 1.9v at 10mA - about 180ohms. Getting a 10Hz 3db point would require a 90uf capacitor.
And in film and foil you're looking at 60 x 120mm motor run caps. Which I don't think will fit!
Current challenge is then picking a bias diode. I'd like Reff < 10ohms or less but it's a challenge. With gm=~ 10mA/V 1V swing input will produce 0.1V of cathode swing. Which is less than a db of effective feedback (20log10(1/0.9)).
That 10ohms is fairly non linear for typical devices eg ROHM SLR-323 red so another 10ohms of real resistance would assist.
So the hunt goes on either for a really small film-and-foil or for lower Reff diodes
And in film and foil you're looking at 60 x 120mm motor run caps. Which I don't think will fit!
Current challenge is then picking a bias diode. I'd like Reff < 10ohms or less but it's a challenge. With gm=~ 10mA/V 1V swing input will produce 0.1V of cathode swing. Which is less than a db of effective feedback (20log10(1/0.9)).
That 10ohms is fairly non linear for typical devices eg ROHM SLR-323 red so another 10ohms of real resistance would assist.
So the hunt goes on either for a really small film-and-foil or for lower Reff diodes
Last edited:
Mundorf MKP 250V 82uf : 47mm (diameter) x 61mm (length)
Auricap 200V 33uf 46mm x 46mm
Breadboarding
Found some 230/12.6 CT 150mA transformers in the cupboard.
One will autoformer for 6DJ8/ECC88 or 6922/E88CC heater, the other step up to 200VAC from 12VAC
Only having 300mA is just another reason to upgrade to 6922!
Breadboarded up a power supply with bits lying around. I will need a bigger step up as I can only really get about 6mA from B+
Found some 230/12.6 CT 150mA transformers in the cupboard.
One will autoformer for 6DJ8/ECC88 or 6922/E88CC heater, the other step up to 200VAC from 12VAC
Only having 300mA is just another reason to upgrade to 6922!
Breadboarded up a power supply with bits lying around. I will need a bigger step up as I can only really get about 6mA from B+
Attachments
Haven't done much - some other (non audio) projects intervening.
Found another upgrade at RockGrotto - mostly parts-rolling.
Found another upgrade at RockGrotto - mostly parts-rolling.
the thread lives again!
The big project is done.
Still cleaning up that (and the "to do" list that's built up while doing Just One Thing(tm) )
Meanwhile over in DHT world Mogliaa has, after much effort, come to a similar architecture, but with a discrete (and undoubtably better) active load.
The schematic is on the Bartola pages here.
I've got to see how much power it uses but it looks like an interesting upgrade path. At first glance it seems to be using 4mA total.
Unfortunately the 01A & Coleman regulator is right out from space and power considerations.
The big project is done.
Still cleaning up that (and the "to do" list that's built up while doing Just One Thing(tm) )Meanwhile over in DHT world Mogliaa has, after much effort, come to a similar architecture, but with a discrete (and undoubtably better) active load.
The schematic is on the Bartola pages here.
I've got to see how much power it uses but it looks like an interesting upgrade path. At first glance it seems to be using 4mA total.
Unfortunately the 01A & Coleman regulator is right out from space and power considerations.
- Status
- Not open for further replies.