There is a number of complaints on the Internet about the performance of Si4735 in shortwave receivers, where it is used to demodulate SSB. For example:
The specific problem seems to be the AGC curves, both amplitude and timing.
I will describe my experience with Tecsun PL-880.
1. The initial gain is too high, resulting in the noise being amplified to the maximum amplitude even when there is no signal.
2. The attack time is not fast enough. Combined with the first problem, this results in strong distortion to the point of clipping in the beginning of each word.
3. The decay/hang is too fast.
Tecsun/Kaito support is not responsive regarding this issue. SiLabs AN332 document states that the encrypted firmware patches (such as used to demodulate SSB) are not written by the customer (Tecsun). This is why I am posting the complaint here.
When there's no signal or weak signal, the AGC will be at the max gain.
I think the soft mute (based on SNR or RSSI ) function will lower the audio volume. To Lower the max AVC gain can also help reduce the audio volume in this case.
The AGC attack time can be set to a small value after each tuning in order to let it adjust the gain quickly.
I would like to find out if the Si4735 can receive SSB am . I am not able to find the documentation on this subject . Also how does one find out about the firmware patch ?
Thank You !
Yes, it can receive the SSB at the AM band range.
I'm sorry that the patch for the SSB is dedicated to some customer.
on NY days i investigated the TECSUN PL880 and made all I could to remove the hard audio attack on SSB. Found some wrong design in the LF-part of the PL-880 circuit, now it is much better. You must reduce the 15kOhm on the CLC2011 down to 6k8, feedbacks to 33kOhm, and insert on controller board 3k3 into the L/R lines.
Nevertheless, there is still a hard cut limiting during the first 200msec observed in the 4735 L/R-outputs. Looking on the signals, i think, the signal for SSB must be lowered (on chip) for some 6 dB to avoid any cuttings.
The SSB attack time must be set to some msec, 200msec are to long, think 10msec will fit. And then, the cutting might be eliminated by soft mute for 10msec, leave SSB level unchanged. Would sound smooth and nice.
Question: is it SiLab to modify the soft or TECSUN to modify their soft?
I would very much like to have more information on the modification you made to your Tecsun PL-880 to fix the AGC problem. I see the schematic for the CLC2011 and understand what you are doing, but it would be helpful if you could provide more info. Which resistor is the 6.8k, is that Rg? Are the 3.3k resistors inserted in the 4735 L and R outputs (break the trace and insert the resistors)? Did you use SMD parts?
I hope you are ok and hope that you receive my message. I cannot find anyone else who has opened the PL-880 and done a hardware modification.