SomeThing is not right with the crosstalk numbers. Either your methodology is flawed, or there is a problem with your Mini3.   I would inspect the soldering on the Mini3 with a loupe. High resistance at the phone jack or ground channel amp could elevate the crosstalk numbers.

It would be easier if you did what Nikongod asks.   Then post your findings.

Measure resistance between RCA signal input jack(the center pin) and circuit ground.

Check the resistance between the input RCA jacks and ground.   68v is a bit high for your power supply. Is it 48v dc or ac?

Allow the tube heaters time to warm before taking measurements.    

Measure between the grid and cathode for voltage.   Measure voltage between the gate and source on the mosfets.

Your power supply is 48v?

Check some voltages and compare the 2 channels.   Power down and check the resistance from input RCA jacks to the grids, and to ground. Again, compare channels.

Yep, wire according to the diagram you posted earlier, while looking at the bottom of the socket.

Nikongod is right. The sockets should be wired looking at the bottom of the tube. They are backwards.

Quote: You are not finished in this pic? It looks like nothing connected to pin 1(plate) and a wire going to pin 9(should not be a wire on pin 9)   Up close pics would help.  

how about pin 4 or 5 on the tubes. One should be about 12v and the other should be zero.

Check some voltages. Start at the power inlet.  Make sure the polarity is not reversed. My first guess would be cold solder joints.

If you are facing the part, the pins would be gate drain source, from left to right.

You double check the pinout on the mostest? The pinout on the devices does not match the pinout on your board.

Some of the power supplies struggle a little to fire up a cold amp, so the power supply may cycle on and off a few times until the heaters get hot.   It takes the heaters more than a few seconds to get hot enough to warm the cathode sufficiently to begin emitting electrons, completing the...

Anything around 24 gauge or bigger should work fine for the source and drain. Smaller stuff would be ok for the gate.

R3 and R9 will be soldered very close to the MOSFET gates?   Looking good.

While the IRF510 output MOSFET here has internal protection devices built in, it could be possible for the drain to remain "hot" even if the device failed, and the entire circuit were inoperable. A problem with the power supply, be it a Cisco or alternate DIY variety, could pose lethal to even...

  Chinese Speaker Maker Edifier Purchases Famous Electrostatic Headphone Maker Stax  

^ Can you post some pics, with close ups?   Its kinda hard to see 15.4v on one side of R9 and 46v on the other.

What kind of amps have you listened to in the past?  

Unless my eyes deceive me, the MiniMax and MosfetMax are identical circuits. The kit from Beezar can be ordered complete with the enclosure already milled and silkscreened. The kit from Glass Jar can be ordered with an optional Hammond enclosure, but no mention of any machining, etc.  ...

^ Never tried the software you linked to, but I use True RTA sometimes, which recommends the UCA202 in the read-me files.... http://www.trueaudio.com/rta_abt1.htm (they have a free evaluation version)... I used the 0404usb briefly while testing a prototype earlier this year. Slightly more...

It's basically a Millet Starving Student.