At 13:42 07/03/2003 +0100, you wrote:
Hi LF-ers...
Just a question for any user of the G3LNP loop (Radcom 99). I replaced
2N3819 (not available in Slovakia) fets with BF245. The current of the
complete preamplifier is only 10mA with 6V on drains of fets and 13V on NPN
TR3/TR4. I'm not sure if it's enough.
73 de Rich OM2TW
The FET bias sounds OK, but there is a flaw in the design of the G3LNP
preamp - with the biasing scheme used, only one of the output NPN
transistors will be biased "on", depending on which FET has the higher
drain voltage. This will lead to increased distortion. The solution is to:
-Remove the 1kohm/1uF between the output transformer centre tap and ground
(C4 and R5 in the LF experimenters handbook), and ground the centre tap
directly.
- Add 2kohm resistors in parallel with 0.1uF capacitors in series with the
emitters of each output transistor and the transformer terminals.
The 2 output transistors are now independently biased, so small differences
in FET characteristics will only slightly unbalance the output stage. The
output stage current will be quite low, about 3mA in each transistor, which
will lead to distortion products if large signals are present at the
output; reducing the 2kohm resistors will increase the bias current.
470ohms will give about 12mA per device, which would increase the available
output roughly 12dB before distortion occurs.
The drain voltage of the FETs will depend on the individual FET Idss, and
the FETS should be selected for drain voltage of about half the supply
voltage (Idss ~ 3mA), so 6V sounds about right. The collector voltage on
the NPN output devices will be equal to the supply voltage. The emitter
voltages should be about 0.6 - 0.7V less than the drain voltages.
Hope this makes sense...
Cheers, Jim Moritz
73 de M0BMU
|
