Thread Tools
Oct 06, 2015, 12:24 PM
Registered User
Okay, mostly for me I will try to sum up everything.

I have resoldered everything to perfection on the SToRM32 board, and checked for bent pins on the Pixhawk Serial 4 connection.

I have redone the measurements with sniffing with USB adapter and the results are:

Case 1: The setup like this:


Sniffing the SToRM32 Tx:
With Pixhawk and SToRM32 connected: No signal
With Pixhawk disconnected: Signal
With SToRM32 disconnected(Pixhawk connected instead): No signal

Sniffing the SToRM32 Rx:
With Pixhawk and SToRM32 connected: Signal
With Pixhawk disconnected: No signal
With SToRM32 disconnected(Pixhawk connected instead): Signal

Case 2: The setup like this:


With Pixhawk and SToRM32 connected: Signal
With Pixhawk disconnected: No signal
With SToRM32 disconnected(Pixhawk connected instead): Signal

Case 3: The setup like this:


With Pixhawk and SToRM32 connected: No signal
With Pixhawk disconnected: Signal
With SToRM32 disconnected(Pixhawk connected instead): No signal

Case 4: The setup like this:


Sniffing the SToRM32 Tx:
With Pixhawk and SToRM32 connected: Signal
With Pixhawk disconnected: Signal
With SToRM32 disconnected(Pixhawk connected instead): Signal

Sniffing the SToRM32 Rx:
With Pixhawk and SToRM32 connected: No signal
With Pixhawk disconnected: No signal
With SToRM32 disconnected(Pixhawk connected instead): No signal

Case 5: The setup like this:


With Pixhawk and SToRM32 connected: Signal
With Pixhawk disconnected: Signal
With SToRM32 disconnected(Pixhawk connected instead): Signal

Case 6: The setup like this:


With Pixhawk and SToRM32 connected: No signal
With Pixhawk disconnected: No signal
With SToRM32 disconnected(Pixhawk connected instead): No signal
Last edited by Hatmpatn; Oct 06, 2015 at 02:03 PM.
Sign up now
to remove ads between posts
Oct 06, 2015, 01:23 PM
Registered User
Wow good work summarizing all that.
Quote:
Originally Posted by Hatmpatn

Case 4: The setup like this:


Sniffing the SToRM32 Tx:
With Pixhawk and SToRM32 connected: Signal
With Pixhawk disconnected: No signal
With SToRM32 disconnected(Pixhawk connected instead): Signal

Sniffing the SToRM32 Rx:
With Pixhawk and SToRM32 connected: No signal
With Pixhawk disconnected: Signal
With SToRM32 disconnected(Pixhawk connected instead): No signal
Case 4 seems to stand out to me to suggest something is wrong.
"Sniffing the SToRM32 Tx:" should have produced the same results as Case 5
"Sniffing the SToRM32 Rx:" should have produced the same results as Case 6.

EDIT: I was going to suggest that there must be a short between Tx and Rx on your STorM32 but other cases suggest there is one on between Tx and Rx on the pixhawk. Maybe test for a short between Tx and Rx pins on both sides just in case.
Last edited by Fran426ft; Oct 06, 2015 at 01:34 PM.
Oct 06, 2015, 01:42 PM
Registered User
Yes, this is too messy for me to understand. But isn't it weird that in case 4, sniffing the Rx line, the SToRM32 emits a heartbeat?
Oct 06, 2015, 01:50 PM
Registered User
Have you've double checked that? And confirmed the heartbeat is definitely the STorM32 heartbeat i.e (FE 09 ** 47 43 00 06 00.....) the only way you'd get a heart beat on an Rx line is a short to a Tx line somewhere along the way.
Oct 06, 2015, 01:53 PM
Registered User
I'll double check Case 4, 5 and 6 now
Oct 06, 2015, 02:01 PM
OlliW
wau, great work and excellent summary
it's sad however that this got such a big effort
anyhow

when these results are as they are quite some things are pretty weired, and inconsistent. Cases which do not look correct to me:

(a) Case 1: Sniffing the SToRM32 Tx: With Pixhawk and SToRM32 connected: No signal
here one should see the STorM32 heartbeat

(b) Case 3: With Pixhawk and SToRM32 connected: No signal
one should again see the STorM32 heartbeat
but is consistent with (a)

(c) Case 4: Sniffing the SToRM32 Tx: With Pixhawk and SToRM32 connected: Signal
this is a bit strange since the two Tx work against each other, I suspect that because of the 1k on the STorM32 the Pixghawk Tx wins and you see Pixhawk heartbeats

(d) Case 4: Sniffing the SToRM32 Tx: With Pixhawk disconnected: No signal
strange, you should see the STorM32 heartbeat

(e) Case 4: Sniffing the SToRM32 Rx: With Pixhawk disconnected: Signal
IMPOSSIBLE !!!! Since only Rx are sniffed, there can't be any signal !!!!!

(f) Case 5:
this is all OK, when we assume that the Pixhawk overides the Storm as explained in (c)
HOWEVER, this findings are inconsistent with (d), (f) "proves" that in (d) one should have also seen a signal

(g) Case 6:
this is all OK
HOWEVER, this findings are inconsistent with (e), (g) "proves" that in (e) one should not have seen a signal

I make now the assumption that you mixed up (d) and (e), and that the explanation in (c) is correct.
This leaves (a) and (b). For some reason the STorM32's Tx can't drive the Pixhawk's Rx.
This could be because of a damage of this Rx pin. However, you said that you also tried the other serials and observed failure too.
Another possibility would be that the Pixhawk Rx require, because of some internal circuitry for instance, larger currents than the STorM32 Tx with it's 1k output can deliver. However, when this should hold true for ALL Pixhawks, in contrast to that it seems to work for some.

I speculate that it's nevertheless the latter reason. Maybe you could post a good closeup photo of the resistors close to the UART pins, so we can determine its values (you may measure them also with a multimeter). If the Tx resistor is the culprit, one possible solution could be to short cut it. If you're good at soldering you could solder a short piece of wire across, or another 0603 R on top of it.
Oct 06, 2015, 02:04 PM
Registered User
Ok, double checked Case 4,5 and 6. I was wrong, sniffing the Rx line generates no signal at all. Sniffing the Tx line generates signal in everycase.
Oct 06, 2015, 02:24 PM
Registered User


Which resistor should I short? "222" says the red text
Oct 06, 2015, 02:35 PM
OlliW
so, these are 2.2k resistors ... I peeked into the Pixhawk schematics, and it has some protection on all serial pins, not sure what sort of (it's a chip with 120 R's, haven't checked whats inside the chips)

as suggested you could try to short the 2.2K, or solder another R on top of it, to reduce the effective resistance ...
a "brutal" method would be to (carefully) remove that resistor altogether, and to use the two pads as kind of solder bridge and close it
Oct 06, 2015, 02:48 PM
Registered User
I have labeled the image, would you please tell me what number of the resistor it is that I should short?

Oct 06, 2015, 03:00 PM
OlliW
4
Oct 06, 2015, 04:48 PM
Registered User
Great spot OlliW. I'm so hoping this is the solution for Hatmpatn. This has been really frustrating me so I can only imagine he's pulling his hair out. I checked my V1.3.1 Goodluckbuy board and it has 1k resistors and works without issue.
Oct 06, 2015, 05:26 PM
Registered User
Omg, it f***ing works!

I'm in love with Olliw and Fran426
Oct 06, 2015, 05:29 PM
Registered User
So did you short the resistor? I am in same situation.

Thanks
Oct 06, 2015, 05:29 PM
Registered User
Yes, short the resistor!


Quick Reply
Message:

Thread Tools