Re: PA 10 HF linear
I didn't measure it accurately, but the analog meter on my 12.0 vdc power supplytoggle quoted messageShow quoted text
showed around 15ma when I first powered up, this rose to around 60ma when TX was grounded.
The difference of roughly 60-15=45ma was due to the BS170 quiescent current in the driver,
I saw about 2.7v on the BS170 gates.
Looking at the BS170 biasing circuit, I'd expect the total quiescent current through
the BS170's to be 3*0.7vbe/33ohms = 64ma, not the 45ma I wrote down.
My power supply meter could well have been off by that much.
I suggest you leave R211 and R215 fully counter-clockwise (minimum) until you figure out
where that extra 200ma is going.
Does the current fall to around 15ma when you do not have TX grounded?
Is anything getting hot (especially the BS170's in the driver, the 7805 regulator, the IRF510's)?
The gates of the IRF510's should both be zero volts dc, are they?
What voltage do you see on the gates of the BS170's in the driver?
When you get around to driving that amp with some RF, here's a few notes:
Gain of this amp has been said to be around 26dB, equally divided between the BS170 driver
and IRF510 final. We expect an output of 10watts = 10*log10(10000mw/1mw) = 40dBm.
Drive into the amp should be around 40dBm - 26dB = 14dBm = 10**(14/10) = 25mW,
but that must be efficiently coupled into the roughly 200 ohm input impedance
(not measured, and will vary with frequency) of the BS170 driver.
Most of us will just put a 68 ohm resistor across the input in parallel with that 200 ohms,
so our source of RF will see a load of 68*200/(68+200) = 50 ohms.
And then increase the drive up from 25mW to compensate for the loss of power
in the 68 ohm resistor. The 68 ohm resistor will draw about 200/68 = 3 times
as much current as the 200 ohm resistor, and thus three times as much of the
available power. Total power from the exciter is thus 4 times what goes into
the 200 ohm driver stage, so about 4*25mw = 100mW for 10 watts out.
I have not yet made careful measurements of my power gain at various
frequencies, but that should be in the ballpark.
Be careful that the big nut used as a spacer between the board and the heatsink
does not short out the cathode 1n4004 diode. Turn that nut such that a flat edge
faces the diode. Though if it were shorted, I'd expect this to cut BS170 bias current
from my 45ma (or 64ma) down to zero, without damaging anything.
On Sun, Jan 27, 2019 at 10:16 AM, Toni Cossio wrote:
I have completed the assembly of the "PA 10 HF linear".