Kisssys board issue w/ uSD. M4 long boot time, high MCU

[Mullet]

So I'm starting to receive other parts of my project namely my uSD. It's a SanDisk Ultra 16GB HC I. I've finished my Kisssys board and have 3.3v running to it via the soldered wire. No Pololu 5v step up. I don't appear to have any solder bridges or the like. When I plug in the uSD into the slot and plug in USB to my Mac, boot up takes longer than normal. Also, the uSD doesn't mount on my desktop. When I open QGC and connect I'm getting a 100% MCU load. Normally, I'm getting around 55% MCU without the uSD card plugged in. With my BLDC board the uSD card mounts on the desktop immediately. I hope I don't damage my board.

I'm going to try and use a solder rework station to see if it helps out with re-flowing, etc. Any other ideas?

[Mullet]

After a little testing, I'm thinking something is wrong with the expansion ports of this particular M4.

I have a BLDC expansion board that is working fine with another M4 that I own. When I plug that M4 with BLDC board attached on boot up via USB I get the solid red LED and a quick pulse of the charge LED. Then all other LEDs come up. If I take that same expansion board and plug it into the "bad" M4 the main red LED lights up, but continuously dims and brightens. It makes a high pitched noise. It doesn't boot up fully. Also, if I take the "bad" M4 and plug in the Kisssys board I get a solid red LED, then after a minute or so it boots. This is without the uSD card plugged in. With the uSD card plugged in, it looks like it's stuck on boot with a solid red LED. Not sure what I did wrong, but now the board doesn't seem to want to work with external boards.

I think it might be time to reach out the Jussi to have the M4 repaired.

[aBUGSworstnightmare]

Have you seen this post: viewtopic.php?f=40&t=4218 ?

It's what your describing!
Do a simple test:
1.) Connect the M4 to the expansion board
2.) connect power to the expansion board --> your M4 should boot without any issues
3.) connect it to your PC via USB
4.) remove power from the expansion board --> should stay fully functional

Now that you verified this unplug everything and:
1.) connect bare M4 to your PC via USB
--> if RED LEDs comes on and CHARGE LED flickers it's affected by the boot problem! This is not related to any external board!

[JussiH]

What Jörg said....

[Mullet]

OK... so I just ran thru the battery test and I could "jump" start the Kisssys / M4. However, with the uSD plugged in I'm still getting the 100% MCU in QGC. It's a little bit odd, because my other M4v2 does the red charge LED blink on start up via USB and it's working normally with the BLDC. I'm avoiding plugging in the Kisssys into the "good" M4 until I know it's 100% and not going to possibly ruin another board. I decided to give the Kisssys an Isopropyl alcohol bath. Perhaps some flux is interfering somewhere. I'd like to avoid removing R26, but I'm leaning towards removing it if it doesn't hurt.

[JussiH]

It is perfectly safe to remove R26. In fact this is has been done on all boards shipped since Jan 1st.

Do the Kisssys expander work with the "normal" M4?

[Mullet]

OK I will remove R26.

I'm afraid to use the Kisssys board with the "good" M4. I don't want to take the chance in ruining it.

After giving the Kisssys board a bath: The first time I tried using the board with the "bad" M4 it actually worked with just USB power. MCU was around 65%. I think it actually wrote data to uSD. Since then, I've tried to do it again with no results.

Another test I've run and have not had success with: Use battery to jump start "bad" M4 with BLDC board. Then hook up USB. At this point I'd expect uSD to mount on my desktop like the "good" M4 / BLDC does. It does read at 56% on MCU in QGC. I'm thinking there is something wrong with SDIO pins on M4.

[Mullet]

Here are some logs with the M4s attached to the BLDC with uSD plugged in... the "bad" M4 is generating a lot of data in only a few minutes. The "good" M4 is only generating a tiny bit of data.

BLDC GOOD M4.zip

(129.92 KiB) Downloaded 503 times

BLDC BAD M4.zip

(2.02 MiB) Downloaded 563 times

[aBUGSworstnightmare]

O.K., just had a brief look at the logs!

1.) The setup of both boards is different! One seems to run a QUATOS firmware (Mavlink SYS ID: 103, AQ S/N: 33313037-33334711-00400031, you call it 'GOOD') and doesn't use the onboard Cypress RX, the other one (Mavlink SYS ID: 119, AQ S/N: 33313037-33334718-00250025, you call it 'BAD') uses the Cypress and seems to be on PID firmware.

Some questions that I have:
1.) PICTURES! - I want to see pictures of both setups!
2.) which uSD cards do you use? You need at least class 10 cards --> if one of them is slower it will not work as expected
3.) What is the intended use? When you want do use the M4 with brushless DC motors you need to connect the LiPo to the boards ADC input pads. THIS IS MANDATORY IN CASE OF BLDC USE!!!
4.) Be sure to set SPVR_VIN_SOURCE = 1 (in Parameters) to get correct reading for LiPo voltage. Both logs show only 1S (ADC_VIN = 4.2V)
5.) Looking at the Log files theres is no radio connected in both cases since radio quality is always low --> connect/configure your radio setup and do some more 'logging' of the channels

Suggestions:
Please check the above. In addition, please assure that both boards run the same firmware (at least during testing). From what I see in the logs, both boards are at least logging something, regardless of MCU load. In addition, exchange the uSD between the two setups. If your 'GOOD' board then shows a high MCU load you need to get a better uSD card.
Only post some of the new logs then since the files will get large in size pretty soon.

[Mullet]

Ok I will make the boards the same firmware-wise and post back with logs. For the sake of the test I'll run both with Quatos and EXT BLDC motor firmware. I'm using the same uSD card with both M4s -- SanDisk Ultra 16GB Class 10. The "bad" board has the CYRF jumper soldered. The "good" board doesn't... yet. I will be using a Spektrum radio for both. I'm not at the phase of pairing a radio yet or even close to flying.

My intentions in using the BLDC for testing was to add more data for troubleshooting. The "bad" M4 will be used for a LadyBird config using a KISSSYS expander board and the "good" M4 is to be used with brushless motors at a future date.

For the test, can I just run USB power and have the BLDC board connected? I know this is not normally how the BLDC is configured. If not, I can solder some wires to make the BLDC operate normally.

Note: I have taken R26 off of the "bad" M4. Start up is faster, even with uSD inserted into KISSSYS board. MCU is still at 100%. No logs show up on uSD card.

Does KISSSYS need to operate at 5v for uSD to work correctly? I have jumper wire for 3v3.