52840 chip precision and consistency problems

[ardenr]

The first chip silkscreen for 2050ST second chip silkscreen is 2109XN third silkscreen is 2052TJ, transmit power configuration for 8dBm actual test from the chip side, test results for the first chip 7.35dBm, the second chip for 5.94dBm, the third chip for 6.61dBm, the problem: why would there be such inconsistent test results, in addition to the chip is different The rest of the various environmental factors are the same, please help answer, thank you.

first	2402	2480
0	-0.4	-0.4
4	3.39	4.08
8	6.84	7.35
second
0	-1.16	-1.05
4	1.52	2.43
8	5.16	5.94
third
0	-1.03	-1.17
4	2.56	3.08
8	6.19	6.61

[Jared]

Hi,

Not sure what silkscreen you are referring too? Is this a marking on the tape and reel? I guess this is on a custom board, has the board been tuned?

regards

Jared 

[ardenr]

The silkscreen is the mark on the chip as shown below, what does your reference to custom board, has the board been tuned, mean?

[ardenr]

[Jared]

I see. 

ardenr said:

The silkscreen is the mark on the chip as shown below, what does your reference to custom board, has the board been tuned, mean?

If you're using a custom board then you have do an antenna tuning to get optimal performance. Variation in performance of custom boards that hasn't been tuned is expected. See this white paper for more on antenna tuning.

best regards

Jared 

[ardenr]

Jared said:

If you're using a custom board then you have do an antenna tuning to get optimal performance. Variation in performance of custom boards that hasn't been tuned is expected. See this white paper for more on antenna tuning.

best regards

Jared 

Why the same test the same three chips just batch different results will have such a big difference, this is the question I want to know.

[ardenr]

What does RF power accuracy±4 mean in the picture? Is the actual test value between 4 and 12 when I output 8?

[ardenr]

The data is based on direct testing on the chip side, why the gap is so large.

[ardenr]

• Please help to answer the question, thank you.

[Jared]

Hi,

I misunderstood you. If the board has been tuned or not is irrelevant for this case as you are connected directly to the radio. I suspect that most of the variation is due to how you're connected, more specifically how you have soldered the connection for each sample. Have you tried measuring it from the connector? Do you see similar variation?

regards

Jared