Yes, that’s the signal I’m concerned about…the desired signal. I appears that
somewhere north of S9 + 20, the product detector might saturate, maybe sooner,
as I haven’t looked at it in detail.
From: minima-bounce@xxxxxxxxxxxxx [mailto:minima-bounce@xxxxxxxxxxxxx] On ;
Behalf Of Ashhar Farhan
Sent: Monday, February 22, 2016 10:27 AM
To: minima@xxxxxxxxxxxxx
Subject: [minima] Re: HF1 AGC
Joe,
The overall gain before the crystal filter is what one needs to bother about.
The gain after the crystal filter is only focused on the tuned signal.
There is about 30 db before the crystal filter. If you see how the gain has
been distributed, you will realize that there are also high losses everywhere.
the first mixer uses ordinary 1N4148 diodes. It has about 8 db loss. The
bandpass filter that follows the first IF amp (centered at 45 MHz) has 4 db
loss. Hence, the overall gain at the output of the 45 Mhz filter is just 4 db.
The second diode mixer has another 8 db loss. followed by the second IF
ampliifer at 10 MHz with another 15 db of gain. The overall gain by now is just
9 db. Then, given that most builders will use cheap microprocessor crystals,
the crystal filter will have another 4-5 db of loss. Finally the detector has
-7db of loss.
Hence, the total loss is -1 db (LPF) - 8db (1st mix), - 4db (vhf bpf) - 8db
(2nd mix) - 4db (crystal filter) - 7db (detector) = a total of 33 of cumulative
loss. balance this against 45 db of gain, we have hardly a 11db gain marigin.
I ran some extensive simulations on the cascade.exe that comes with EMRFD and I
have also measured the IIP3 of this receiver. It comes to a very respectable
+18dbm. That runs circles around my old FT-817 and even an Elecraft K2.
- f
On Mon, Feb 22, 2016 at 8:42 PM, Joe Rocci <joe@xxxxxxxxxx
<mailto:joe@xxxxxxxxxx> > wrote:
Farhan
I’ve been following the new interest in the HF1 and it appears there might be
some traction developing.
Overall, it appears to be a sound architecture, but one of the things I’m
concerned about is that there appears to be around 50 dB of net receiver gain
before the product detector, but no AGC to prevent it from overloading on
strong received signals. Do you see this as a potential problem, and do you
have any suggestions for how to implement AGC if it’s desired/necessary?
Regards
Joe
W3JDR
