[Hackrf-dev] hardware: RF section coupling capacitors

[Michał Byrecki]

Howdy all,
That would be my 1st post here. So, at the beginning - I'd like to pass
my best regards to all the developers that contribute their work to the
HackRF One project.

I've been using HackRF mainly on a HF bands with a dipole antenna W3DZZ
to receive short wave signals (80m band) plus some satellite weather
images reception etc. being transmitted on SW. Today I've realized that
my tiny shiny black box is dead bricked with RF reception. I've quit the
warranty claim idea. There is a possibility to dig a bit into the
hardware of HackRF & have it fixed fast by myself.

Following the Jared video on YT I've opened the box. Too bad, there are
no descriptors on the PCB that would speed up finding the components
location, but nevertheless I have found mosfet Q1 overheated. As it
provides the power for a RF amplifier U13, I quickly thought of a busted
RF amp circuit. In fact, the resistance on the output U13 (pin 6, pin 5)
was approx 25Ohms. Luckily it did not blow the mosfet yet ;) I've
removed the U13.

To make sure the only fault would be 1st RF amp, I've connected a
temporary FM antenna (actually a probe of my multimeter) to the pad no.
6 of the removed U13. All seem to be working fine, I'll order & apply
the new RF amp.

Now, let me get back to the design idea. Every dipole antenna is not
resistant to the electromagnetic interferrence. It will likely absorb
the noise emitted by a thunderstorms, etc. Usually for these cases You
protect the RF section by a two-way transil diode. Here, in HackRF One
that is provided by D1 diode. However, in my case, the diode is yet OK,
while 1st RF is busted. Why? In parallel to the diode, there is a
circuit branch of C64/switch U14/C58 and the input of 1st RF amp. C64
and C58 have infinitive resistance for DC, but not for RF. I believe,
the total impedance of that branch circuit is relatively low to the
impedance diode and thus the diode did not blow (wasn't actitaved), but
the U13. In other words, the amount of charge being transferred thorough
this circuit makes matter. In my opinion, the values of C64 and C58 are
overscaled (especially C58 in the receiver section):

- the higher frequency You're on, the lower coupling capacitance You
will need. So once the capacitance is OK for short wave, it will surely
be OK for higher frequencies. 
- transmitters require "tougher" couplings than the receivers, as there
is much more charge being delivered to the antenna while transmission
than collected from while reception.

I am not gonna go into modelling details now. Just go google any ham
transceiver schematics like Kenwood, etc. They are found reliable and
working, so we shouldn't reinvent the wheel. If You look at the circuit
diagrams, You will find out, that the capacitance used for TX is usually
abt. 500pF (which 200 times less than used in HackRF).

My suggestion is to lower C64 to 500pF and C58 to 500pF or less. If
lowered, the D1 will more likely go short circuit, than the RF amp would
be busted. That are my three cents...

Brgds,
Mike SP5BYK
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist9.pair.net/pipermail/hackrf-dev/attachments/20160522/55b42976/attachment.html>