last month (i believe), somebody mentioned pentode characteristics glueing together with ccs.
i first started with ccs+ led bias; first thing i noticed was enorm gain, but shape was heavy deformed. (screen resistor was connected to B+ originally)
screen resistor moving to source of fet helped, it has kind of "selfcentering". (connecting to anode dropped gain a lot...or touching it)
second weird thing - gain roll off from >5khz (i thought mosfets has to be wrong, because you guys use dn2540 which has way low Ciss)
my mosfets i have almost for free 🙂, but has 1nF Ciss.... so i added 50k "gate stopper" , and shape was even worse.
then i changed components, removing led -> 1k; screen resistor 100k; screen cap 1,5u foil; gate resistor ->0R.
this yields-undistorted waveforms (Vpp 200V scope load, or ~130-150Vpp 100k+u47 series// input 1,8Vpp)
-screen resistor helps a lot to correct the shape, 50k too low, 330k too much. (distorted bottom or upper halfwave) you might be best with 150k.
-and most important, no such huge roll off.
B+ ~260V; G2 ~105V; A ~135V; Rk ~1,8V
question is, why i see so many led bias schematic? what is wrong?
i first started with ccs+ led bias; first thing i noticed was enorm gain, but shape was heavy deformed. (screen resistor was connected to B+ originally)
screen resistor moving to source of fet helped, it has kind of "selfcentering". (connecting to anode dropped gain a lot...or touching it)
second weird thing - gain roll off from >5khz (i thought mosfets has to be wrong, because you guys use dn2540 which has way low Ciss)
my mosfets i have almost for free 🙂, but has 1nF Ciss.... so i added 50k "gate stopper" , and shape was even worse.
then i changed components, removing led -> 1k; screen resistor 100k; screen cap 1,5u foil; gate resistor ->0R.
this yields-undistorted waveforms (Vpp 200V scope load, or ~130-150Vpp 100k+u47 series// input 1,8Vpp)
-screen resistor helps a lot to correct the shape, 50k too low, 330k too much. (distorted bottom or upper halfwave) you might be best with 150k.
-and most important, no such huge roll off.
B+ ~260V; G2 ~105V; A ~135V; Rk ~1,8V
question is, why i see so many led bias schematic? what is wrong?
Attachments
Nothing wrong with cathode LED bias. Nothing wrong with anode CCS. Put the two together, especially with a pentode, and you get problems. This is because you cannot independently choose grid and anode bias voltages and anode current for a triode. You choose two of these and get whatever you get for the other one. Or you can use a resistor and set a linear combination of two, plus one other. Basically you only have two degrees of freedom, not three.
For a pentode you effectively have two different CCS both trying to set the current - the result is that the voltage 'bottoms' on one of them so you get distortion.
For a pentode you effectively have two different CCS both trying to set the current - the result is that the voltage 'bottoms' on one of them so you get distortion.
Hi,
Change the LED to a R and it will work just fine. It is a gian monster with high imp output. Please check out JC labs, he did some great designs on this. IMO you should go for a more high GM tube like E180F or E280F.
Best regards
Karsten
Change the LED to a R and it will work just fine. It is a gian monster with high imp output. Please check out JC labs, he did some great designs on this. IMO you should go for a more high GM tube like E180F or E280F.
Best regards
Karsten
can somebody explain why it had so big hf roll off...
(also with u47 g2 cap, it had lf roll off <100hz)
what will be better, ccs+Rk or led +Rp ?
wanted to use it as ltp with hi voltage gain (allinone), but disapointed it still needs hi imp next stage
(also with u47 g2 cap, it had lf roll off <100hz)
what will be better, ccs+Rk or led +Rp ?
wanted to use it as ltp with hi voltage gain (allinone), but disapointed it still needs hi imp next stage
It had a big HF rolloff because the anode is a very high impedance point so it doesn't take much capacitance to affect it. Also high gain means high Miller effect from stray anode-grid capacitance (mainly in the wiring, for a pentode). As you have built a stage with a very high output impedance, the next stage needs an even higher input impedance. Using a plate/anode resistor will reduce it a bit, but if you want low output impedance don't use a pentode.
Like, for example, the capacitance of your scope probe. Yes, this is a stage which MUST be buffered.
- Status
- Not open for further replies.