LA104 SPI Problem: MSB and LSB about la104 HOT 18 CLOSED

Please send screen shot

from la104.

from la104.

Where is the picture?

from la104.

Here they are.
The SPI output was created by a ESP32 set to SPI Mode0 and MSB

As Said: The Bit order of the MOSI is 1 0 1 1 1 0 1 1, which is indeed BB for MSB as shown by the decoder, but it is configured as LSB and then it should be DD.

from la104.

When you switch to MSB, will it show 0xDD?

from la104.

Yes it does

from la104.

Bit order in decoder:
https://github.com/gabonator/LA104/blob/master/system/apps/test17_official/source/Analyze.c#L29
MSB/LSB constants:
https://github.com/gabonator/LA104/blob/master/system/apps/test17_official/source/Analyze.h#L12
Menu:
https://github.com/gabonator/LA104/blob/master/system/apps/test17_official/source/Menu.c#L994

gInputSubParam[IN_SPI][3] holds the information what kind of signal we are going to decode:

memcpy(&param_str[LEN * 3], (gInputSubParam[IN_SPI][3] ? " MSB " : " LSB "), SIP_PARAM_STR_LEN);

where 0 means LSB and 1 MSB

#define  LSB           0   // 
#define  MSB           1   // 

SPI_AnalyzeFrame does the decoding part:

                if(CLK_Phase == 1)                         // 
                {
                    if(SCK_Now == CLK_Polarity)
                    {
                        AnalyzeBuf[j * 3 + 0] |= BIT_Order ? ((Status & 0x01) << k) : ((Status & 0x01) << (7 - k));
                        AnalyzeBuf[j * 3 + 1] |= BIT_Order ? (((Status >> 1) & 0x01) << k) : (((Status >> 1) & 0x01) << (7 - k));
                        if(k < 7) k++;
                    }
                }
                else                                       // 
                {
                    if(SCK_Now != CLK_Polarity)
                    {
                        AnalyzeBuf[j * 3 + 0] |= BIT_Order ? ((Status & 0x01) << k) : ((Status & 0x01) << (7 - k));
                        AnalyzeBuf[j * 3 + 1] |= BIT_Order ? (((Status >> 1) & 0x01) << k) : (((Status >> 1) & 0x01) << (7 - k));
                        if(k < 7) k++;
                    }
                }

k is incrementing from 0 to 7, it holds the bit index. This expression BIT_Order ? ((Status & 0x01) << k) : ((Status & 0x01) << (7 - k)) should evaluate the bit weight, probably all of these 8 partial values are later summed together to get transmitted/received value. When decoding LSB frame, the first bit should give value 1 and 128 for MSB respectively. For LSB the Bit_Order is zero (false), so the second term is evaluated ((Status & 0x01) << (7 - k)) which is wrong.
I recommend changing BIT_Order ? ..... into (BIT_Order == LSB) ? ... which will invert the condition to fix this problem.

from la104.

Try this:
17analys.zip

from la104.

I renamed the original analyser.elf to enalyser.elf___ and took the content of your 17analys.zip and put it into the device as analyser.elf
Restart of the device. analyer started.. I am afraid. Exactly the same display.. BB on setting LSB ;-(

I am not familiar with the whole code but what about just inverting the Bit Order
......AnalyzeBuf[j * 3 + 0] |= BIT_Order ?.....

to
.....AnalyzeBuf[j * 3 + 0] |= ! BIT_Order ?.....
?
could that help ? (of course for all 4 ocurences)

from la104.

!BIT_Order, BIT_Order == 0, BIT_Order != 1, BIT_Order == LSB, BIT_Order != MSB, all are equal. But ternary operator is a bit tricky, because you never know, if the negation operator !a ? b : c is applied on the first argument, or on the result, so it is always safer to put the condition in parenthesis (!a) ? b : c

from la104.

I didn't notice that there are 8 places with this code, not four. Try this
17analys.zip

from la104.

yeah you are abolutely right with the parenthesis.
AND ! Tadaaa. I seems to work now. I owe you something

from la104.

Funny enough after YEARS of the LA104 being on the market I am the first to notice??? Hmmm :-)
And congrats to you for your work on this alternative firmware. Bombastic work !

from la104.

I would really like to know how many units did they sell and how many are actually used... but thank you for reporting this issue

from la104.

Glad that you could help me :-)

from la104.

You are welcome :) But why are you on friday night fighting with logic analyser and not somewhere out with friends and drinking beer? :)

from la104.

Poor me.. I am sick at home while they have one after the other :-( (pity is welcome)

from la104.

Fixed here a8904d5

from la104.