SEEK behaviour related to band limits

[G8PTN]

Hi,
This is not necessarily a firmware bug, but a possible feature of how the SI4732-A10 works.

A user in another forum noticed an interesting issue with the hardware seek behaviour on the SI4732-A10.
The seek frequencies appear to be based on the seek spacing and the lower band limit.
goshante/ats20_ats_ex#44 (comment)

I replicated the issue on the ATS-MINI firmware which actually has a fixed AM seek spacing of 5kHz.

Example 1:
Lower band limit = 531 kHz
Manual tune to a frequency on a 5kHz boundary, e.g. 600 kHz.
With a debug print statement added in void showFrequencySeek, start “Seek Up”.
Although not all frequencies get printed, they all follow 531kHz + (5kHz x n), where n is an integer value.
606
611
621
626
636
641
651
656
666
676
681

Repeating the same test where the lower band limit is on a 5kHz boundary.
Example 2:
Lower band limit = 150 kHz
Manual tune to a frequency on a 5kHz boundary, e.g. 200 kHz and start "Seek Up"
Again, although not all frequencies get printed, they all follow 150kHz + (5kHz x n), where n is an integer value.
210
215
225
230
240
245
255
265
270
280

The workaround implemented by the user was to set the lower band limit to a common multiple of the seek spacings available.
Example:
Assuming available seek spacings of 1kHz, 5kHz, 9kHz. 10kHz, a lower band limit of 450 kHz would be valid for all spacings.

I have not seen any direct reference to this in the application notes, although it is possibly implied by the recommended ranges and spacings,

Notes

The SI4732-A10 seek behaviour seems to depend on the configured lower band limit as well as the seek spacing. It may be the case that the upper band limit also influences the behaviour.

If the lower band limit is not on a valid boundary for the spacing, then the tuned seek frequencies, will not be as expected.

All tests with effectively the same patch applied in the latest ats-mini firmware related to the SI4755 library SI4735::seekStationProgress

I noticed that there are additional patches in the Chinese firmware (SI4735 library) relating to adjusting the frequency based on the step. I think this is covered in the ats-mini firmware (stepAdjust) and do believe it solves the hardware seek issue identified above.

The behaviour in FM mode may be the same? Not tested.

73, Dave

[max-arnold]

That is a bummer... I'm not sure that the stepAdjust fixes the hardware seek, because it is for manual tuning only.

[G8PTN]

Hi Max,
The stepAdjust corrects the frequency as soon as you start tuning.
The Chinese patch updates the frequency at the end of the seek detection, but will not fix the issue.
The workaround that the user applied works, but depends on the seek sizes that are going to be used.
There is also the issue of the Auto capacitor as previously covered. That is why I said it may be better to use the firmware SCAN.
The downside is that the firmware scan is probably slower.
#163 (comment)

Another option is just to fudge the band limits during the SEEK based on the selected spacing.
I think further testing is required to understand the exact behaviour and if it applies to FM and well as AM.
There will be some way to fix it, at least from a user point of view.

73, Dave

[G8PTN]

Another option is just to fudge the band limits during the SEEK based on the selected spacing.

Hi Max,
A little more experimentation. The seek behaviour does seem to be dependent on the minimum seek limit for both AM and FM modes.

An example patch for the ATS-MINI firmware "void doSeek()" using the current fixed spacings (FM=100k, AM=5k) is shown below. This simply adjusts the minimum seek limit up to a valid spacing boundary if required.

The hardware appears to automatically limit the maximum seek frequency to a valid value, but the firmware could do the adjustment prior to the seek if necessary.

Any changes to the ats-mini firmware would need to use the configured spacing values and take account of the special case for the 200kHz step that was recently added to the code.

This example patch does not cover the automatic capacitor setting that currently occurs in seek mode.

73, Dave