Biasing KT88 question/issue on Luxman

True!

In this amplifier the cathodes of the preceding 12AU7 are dc-coupled to the control grids of the KT88's, so there are no coupling capacitors between the preceding stage and the KT88's. But if one of the coupling capacitors prior to the preceding 12AU7 would be bad (those are the two 0,22 uF capacitors) it would mess up things.
 
Just tightened the original kt88 pins ( new sockets I got look cheap), checked all caps ( all read good with ESR meter) deox'd the level pot and vr1 & vr2. Will turn it on tomorrow and take measurements and try switching tubes. Thank you all for the comments and links. It has been very helpful.
 
I've only ever had experience with one Luxman tube amp, the MB-3045, but its a lovely sounding piece of gear. If thats any indication these should also be very nice pieces.
 
Ok things looking pretty good. I think the issue was the KT88 pins or the deox on the VR1 & VR2s. No parts replaced and cathode to ground voltage stable.
Plate voltage (pin 3) about 463V, Screen Grid (pin 4) 467V
Cathode (pin 8) to ground voltage I adjusted to .5V.
The cathode current with .5V and 10 ohms would be 50mA. Using Andy's video, looks like the Genalex KT88 (if I am looking at the correct line) has a max DC cathode current of 175mA. So I am below 1/2 and they should not be running too hot.

Does this all sound good?

The two VR pots needed to be adjusted kind of in sync to get each KT88 to match.
 
50ma * 463v = 23.15 watts of plate+screen dissipation, which is well within the capabilities of a tube rated for 40 watts plate+screen. Would be interesting to know what Lux had in mind for this but with no voltages or anything on the schematic there isn't really any way to know.

so long as both cathodes read the same voltage I'd be tempted to run with it lacking better info
 
actually looking at the datasheet, 50ma for very similar voltages seems right

1692890737166.png

this amp is triode mode, not UL, and probably runs in AB2 mode but for voltage and current purposes at idle its pretty similar. The Ia+g2(0) line is idle current plate + screen with zero signal, 50ma per-tube in the 460 plate volts column.
 
gadget73 said:
50ma * 463v = 23.15 watts of plate+screen dissipation, which is well within the capabilities of a tube rated for 40 watts plate+screen. Would be interesting to know what Lux had in mind for this but with no voltages or anything on the schematic there isn't really any way to know.

so long as both cathodes read the same voltage I'd be tempted to run with it lacking better info
Click to expand...
Does this mean with the 2 together I have 46W. Spec for the amp is 60W. Company literature mentions going for less than the 75W in the McIntosh MC275 to extend tube life.
Maybe I should go to 65ma?
 
Mark 1234 said:
Does this mean with the 2 together I have 46W. Spec for the amp is 60W. Company literature mentions going for less than the 75W in the McIntosh MC275 to extend tube life.
Maybe I should go to 65ma?
Click to expand...

The output power rating of the amp is not the same thing. What's being determined here is the plate dissipation of the individual tube. Cathode current times voltage across the tube = the dissipation of the tube itself. I'd stick with what you currently have since it will give you the output power as advertised. Biasing the output tubes hotter may actually decrease the output power.
 
Your amp is fixed bias, it will idle lower than cathode bias. Cathode bias tends to idle hotter to compensate for the bias shift as power output increases.

This also most likely runs in AB2 mode, which will produce considerably more power than an AB1 mode amp. The "2" means the grid draws current and swings positive at parts of the cycle. Being direct coupled off a cathode follower like it is, thats a possibility here.
 
that doesn't tell us what they intended for idle current. Power output and idle plate dissipation are relatively unrelated.
 
You must log in or register to reply here.
Back
Top Bottom