SP1018G - 2 MHz SPI SCLK signal faulty recorded

Logic Analyzers
1

Hi. I recognized by not properly working frame decoders a problem with the recording of an SPI 2 MHz SCLK signal. While a scope shows the signal correctly ScanaStudio 5.1.39 with an SP1018G shows something different (see snapshot 1).

image
image2012×1364 237 KB

It occurs sporadically at different signal positions but always several times in each recording. Sometimes the clock blocks of two bytes are even put together (see snapshot 2).

image
image1325×540 13.7 KB

This is the reason why the frame decoder delivered garbage. First I suspected the frame decoder itself but removing all active decoders didn’t change a thing. I also tried different sample rates without any effect. Looks like an SP1018G internal sync problem.

2

Hello @Olaf ,

This is quite odd, and we couldn’t reproduce this.

One thing: i see your signal (on the scope) is 0 - 0.7V (very small amplitude) and i see on ScanaStudio you have 2.5V threshold. Have you tried lowering the threshold in ScanaStudio?

by the way, the new probes are finally ready (allowing to set 1MOhm impedance, i send you a message regarding that soon).

1 Like
3

I was able to reproduce this. i think this was introduced in latest firmware update. i’ll dig into this and get back to you shortly.

4

I can confirm 100% this was introduced with V5.1.39. Let me continue to investigate

1 Like
5

Actually it’s a 5 V signal, you cannot see the pre-scaling in that snapshot.

I was still playing with the PullUp/PullDown tied always to 3,3V, but you said already the new HW will soon be ready. I’m lighting a candle and saying a prayer.:candle:

1 Like
6

@Olaf ,

Just letting you know i finally fixed the bug. was a tough one related to DDR throughput optimizations i’ve made in last update.

Tomorow (in a few hours) a new version will be released, please let me know if it fixes the issue for you.

Thanks for pointing that out!

1 Like
7

Thanks for quick fixing. I will try it after Easter vacation.
Happy Easter