Re: A simplistic explanation (long)


Hi Roy,

Thanks for the kind words and enthusiasm. Several of us at Silabs have thought that the Si47xx chips would be very fun chips for DIYers and it's gratifying to see experimentation take hold.

I've reviewed your USB experiments and it looks like you are making good progress. I've been working on a similar setup and your write-up with photos is very useful (I can probably avoid having to remove the metal shield). I'm trying to use a USB interface similar to our evaluation boards, which would be compatible with our eval software. I'll post more when I get something working.


--- In ultralightdx@..., "Roy" <roy.dyball@...> wrote:

Hi Scott

Welcome to the group.

Thank you for putting us straight on the operation of the Si4734 and
congratulations on your part in designing one of the best little radio
chips on the planet. It has been fun trying to work out the operation of
the chip with just logic and an old version of the programming manual
plus the data sheet.

Ultralight DXing is a fantastic sport and I hope you get to enjoy it
from the other side of the chip. Gary has put so much energy into
squeezing as much performance as possible from these little radios and
has a natural ability to write up the results of his experiments keeping
the group informed and spurred on in the quest of more distant radio

I will be interested in your findings on the implementations of the
antenna and front-end circuits as this seems to change on each model
Tecsun brings out changing the performance.

I have written a small application via USB interface that enables
control of the Si4734 in the G8 by intercepting the control lines. I
initially used the I2C implementation but soon re-wrote using the 3 Wire
(Legacy) protocol because Tecsun for some reason use the 3 wire
interface to control the Si4734 and then have a separate I2C bus to read
and write the EEprom that saves channel info (could not work that out).
If you get a chance the software is here along with a pictorial
explanation of how to break into the control lines and connect the USB

I suppose the biggest question the group will have is how can we turn
off the AM soft mute. Although this is a great feature for the average
user it just works against you in AM DX work. I can turn the soft mute
off in my software but lose the change in power down.

The PL-380 I have seems to have the soft mute turned off by default. It
has Ver 20 firmware. I notice that Ver 40 firmware has been released and
has a less aggressive default soft mute settings according to the data
sheet but I have not seen one of these chips yet.

I feel like I could ask you a many questions and hope I can over the
course of time.

Cheers for now

--- In ultralightdx@..., "sdwillingham" <sdwillingham@>

Hi Roy, Gary, Jim, and others:

I've been following this group for a week or so with a special
interest in your experiments with radios based on the Si4734 chip. My
interest comes partly as a radio enthusiast and casual occasional DXer,
but more specifically as one of the designers of the Si4734. I am
posting in hopes that I can clarify a few of your questions about the
Si4734's operation.

First, the AM front-end of the Si4734 is a tuned-tank circuit with a
fairly conventional LNA and quadrature mixer as shown in the block
diagram. There is no sampling at the RF signal. The I/Q signals at the
(low) IF are converted to digital form where channel filtering, image
rejection, demod, and audio processing are performed. The analog
front-end circuits are controlled and calibrated by an on-chip

As Roy has pointed out, the loopstick (or air-loop) inductance is
resonated with an on-chip varactor, which tunes in small discrete steps.
This tuning is done each time the frequency is changed. Roy is
absolutely correct that it is vital to repeatedly re-tune the radio as
one evaluates different antenna changes. The chip's tuning method
actively adapts to the inductance, bringing the tank to resonance
regardless of the change. In other words, the varactor value is not set
simply by a lookup table corresponding to frequency. The chip measures
and peaks the resonance upon each tuning event.

Regarding the inductance range listed in the data sheet: this is a
range over which Silicon Labs guarantees _every_ chip will successfully
resonate all broadcast-band frequencies. For a particular chip and/or
band of frequencies, the range will generally be larger. Since the LC
product is inversely proportional to frequency-squared, the chip can
successfully tune much larger inductors at 500 kHz than at 1700 kHz.

I have a Tecsun PL-300wt and have ordered a PL-380. I hope to find
some time in the near future to open them up and poke at the
implementations of the antenna and front-end circuits.


--- In ultralightdx@..., "Roy" roy.dyball@ wrote:

Thank you Gary for taking the time to carry out the tunning
procedure. I
also achieved similar results but with different values of RSSI S/N
my local stations.

It is interesting that result 3 is higher (6dbu in your case) than
result 2 with the coil in exactly the same position. I believe if
tried the same thing on a conventional radio (non Si4734) result 2
result 3 would stay the same.

I think this highlights the need to steep off and back on to
whenever you make any position adjustments to the loopstick coil
as shown it is possible to have two different readings with the coil
exactly the same position.

Using the method of stepping off and back on frequency before noting
signal strength I am finding that on the low end of the MW band my
highest signal readings are with the coil slightly to one side of
centre. While at the high end of the band my best signal readings
with the coil slightly from the end maybe less than a quarter of an
inch. I find my best overall results are with the coil at the end
seems to be consistent for me with different coils and different

I am eagerly awaiting my 7.5" and 4.0" X .5" Amidon 61 rods and
of assorted Litz wire to add more empirical results to the group.

Cheers Roy.

--- In ultralightdx@..., D1028Gary@ wrote:

Hi Roy,

The local station signal strength (RSSI) and signal to noise ratio
readings have been completed on the local station here, as you

On the PL-310 with the 7.5" Amidon ferrite bar and 81-turn Slider
with the coil at the maximum AM sensitivity position (554 uh), our
station KSUH-1450 has a signal strength (RSSI) reading of 80 dBu,
a signal
strength (S/N) reading of 25 db. When the coil was shifted to one
inch from the far end of the ferrite bar, the RSSI reading changed
72 dBu,
and S/N was the same at 25 db.

After moving the frequency to a weaker station and then
changing back to KSUH-1450 (at the same coil position), the RSSI
78 dBu, and
the S/N was still 25 db. The RSSI shifted between 79 dBu and 78
dBu a
couple of times before settling at 78 dBu, in this final test (as
when the Pl-310 is changed to a new station).

I hope this information is helpful to you, Roy.

73, Gary

Join { to automatically receive all group messages.