Re: zBITX Crashes - Suspect Memory Leak
#zbitx
Jody, I see something similar, though the sbitx program seems to keep running. After a while (anywhere between 15 minutes and an hour) the tcp connection dies. I had the web interface running and an open ssh session, and the web interface reverted to the login page and the ssh session terminated. Rebooting brought everything back to normal. This has happend a couple of times. Receiving was still OK though.
73 Frank PA3CNO?
After letting my zBITX sit unattended in Receive Mode for a while, the main RasPi CPU crashes.? I have never timed it because I'm not just sitting there watching it.? But it seems like it only takes about an hour or so to crash.? I am not able to get a SSH connection to the main CPU at that time.? The display CPU is still running at that time.
?
While the device is warm whenever this occurs, I do not think that it is a thermal issue because if I disconnect and reconnect the power it reboots and runs fine.
?
I have never stopped the sbitx application to see if something else on the RasPi could be causing it.? I have SSH and VNC enabled, but I am not using either of those connections.? I do not have anything else running on the RasPi.? So, I really do not think that it is some other RasPi appllication.?
?
I think that it is probably a sbitx application memory leak.? ?Don't see how something like a character array overflow would occur after just sitting there unattended for that period - nothing is being done on the GUI during that time.
?
This crashing occurred whenever I had version 3.021 installed,? And it is still occurring with version 3.052 installed.
?
If you are running on batteries, the batteries may be depleted before this crash occurs.? You will never see it.? I am using a 7.5vdc power supply.??
?
72,
Jody - K3JZD
?
|
Re: New zBitx - New User - CW operation
Hi Gordon, yes I have been following Jody¡¯s efforts pretty closely. As someone who worked in real time embedded systems for most of my career, I have no doubt the CW timing issue is fixable. Jody has a good latest approach, I think. I am not sure when the right time to try that code personally is.?
Perhaps a more serious issue is how insensitive the receiver is. It¡¯s pretty bad even for being a simple receiver. Not sure that software can fix that.?
Also I really want to be able to touch a signal on the waterfall and have the frequency jump there¡ just a matter of software.?
Steve AK0M
|
Re: zBITX Crashes - Suspect Memory Leak
#zbitx
|
I have used htop and top to track memory usage. ?The original 32bit software would have memory usage grow and then reset back to the minimum. ?I did that monitoring, not writing to a file. ?I believe it is possible to pipe the output to a file.?
73
Evan
AC9TU
|
|
Re: zBITX Crashes - Suspect Memory Leak
#zbitx
I wondered why the zBitx crashed after some time... Makes sense, because the PI Zero 2W would NOT work as a 'normal' internet browsing setup, after a few minutes and some websites past, the memory would be filled 100%.
Good luck!
Martin, PD0ZZ
|
is there a list we can get on? do we have a round about cost? want to get in and hopefully get these before the 90 day pause is lifted
|
Re: zBITX Crashes - Suspect Memory Leak
#zbitx
My go-to for debugging this sort of thing is valgrind, but that's likely to be really slow when running on the actual hardware (and may have memory issues given the small headroom?).
I've been meaning to ask - have people been compiling this on the device, or does anyone have a cross-compilation environment set up? It's on my list to figure out, but if anyone already has something put together that might speed up development.
On Wed, May 14, 2025 at 1:57?PM Gordon Gibby KX4Z via <docvacuumtubes= [email protected]> wrote:
I¡¯m not sure how One tests for this?
Do you write a small program that every five minutes asks for various amounts of space and reports which ones it was able to be given?
And then just watch it? Or does that only report the heap? I don¡¯t know a lot about the different ways Memory is handled
But sounds like you¡¯re definitely onto something that can get fixed!
Gordon?
? After letting my zBITX sit unattended in Receive Mode for a while, the main RasPi CPU crashes.? I have never timed it because I'm not just sitting there watching it.? But it seems like it only takes about an hour or so to crash.? I am not able to get a SSH connection to the main CPU at that time.? The display CPU is still running at that time.
?
While the device is warm whenever this occurs, I do not think that it is a thermal issue because if I disconnect and reconnect the power it reboots and runs fine.
?
I have never stopped the sbitx application to see if something else on the RasPi could be causing it.? I have SSH and VNC enabled, but I am not using either of those connections.? I do not have anything else running on the RasPi.? So, I really do not think that it is some other RasPi appllication.?
?
I think that it is probably a sbitx application memory leak.? ?Don't see how something like a character array overflow would occur after just sitting there unattended for that period - nothing is being done on the GUI during that time.
?
This crashing occurred whenever I had version 3.021 installed,? And it is still occurring with version 3.052 installed.
?
If you are running on batteries, the batteries may be depleted before this crash occurs.? You will never see it.? I am using a 7.5vdc power supply.??
?
72,
Jody - K3JZD
?
|
Re: zBITX Crashes - Suspect Memory Leak
#zbitx
I¡¯m not sure how One tests for this?
Do you write a small program that every five minutes asks for various amounts of space and reports which ones it was able to be given?
And then just watch it? Or does that only report the heap? I don¡¯t know a lot about the different ways Memory is handled
But sounds like you¡¯re definitely onto something that can get fixed!
Gordon?
toggle quoted message
Show quoted text
On May 14, 2025, at 14:47, Jody - K3JZD via groups.io <k3jzd.jody@...> wrote:
? After letting my zBITX sit unattended in Receive Mode for a while, the main RasPi CPU crashes.? I have never timed it because I'm not just sitting there watching it.? But it seems like it only takes about an hour or so to crash.? I am not able to get a SSH connection to the main CPU at that time.? The display CPU is still running at that time.
?
While the device is warm whenever this occurs, I do not think that it is a thermal issue because if I disconnect and reconnect the power it reboots and runs fine.
?
I have never stopped the sbitx application to see if something else on the RasPi could be causing it.? I have SSH and VNC enabled, but I am not using either of those connections.? I do not have anything else running on the RasPi.? So, I really do not think that it is some other RasPi appllication.?
?
I think that it is probably a sbitx application memory leak.? ?Don't see how something like a character array overflow would occur after just sitting there unattended for that period - nothing is being done on the GUI during that time.
?
This crashing occurred whenever I had version 3.021 installed,? And it is still occurring with version 3.052 installed.
?
If you are running on batteries, the batteries may be depleted before this crash occurs.? You will never see it.? I am using a 7.5vdc power supply.??
?
72,
Jody - K3JZD
?
|
zBITX Crashes - Suspect Memory Leak
#zbitx
After letting my zBITX sit unattended in Receive Mode for a while, the main RasPi CPU crashes.? I have never timed it because I'm not just sitting there watching it.? But it seems like it only takes about an hour or so to crash.? I am not able to get a SSH connection to the main CPU at that time.? The display CPU is still running at that time.
?
While the device is warm whenever this occurs, I do not think that it is a thermal issue because if I disconnect and reconnect the power it reboots and runs fine.
?
I have never stopped the sbitx application to see if something else on the RasPi could be causing it.? I have SSH and VNC enabled, but I am not using either of those connections.? I do not have anything else running on the RasPi.? So, I really do not think that it is some other RasPi appllication.?
?
I think that it is probably a sbitx application memory leak.? ?Don't see how something like a character array overflow would occur after just sitting there unattended for that period - nothing is being done on the GUI during that time.
?
This crashing occurred whenever I had version 3.021 installed,? And it is still occurring with version 3.052 installed.
?
If you are running on batteries, the batteries may be depleted before this crash occurs.? You will never see it.? I am using a 7.5vdc power supply.??
?
72,
Jody - K3JZD
?
|
mine was order #3783 placed on April first.? I received it May 12th without and import/tariff issues to the USA.??
|
Yes. My bad. Time goes by so quickly when getting older - except waiting for radio delivery.
?
# 3869 (from April 2) is scheduled for DHL delivery on Fri 16 MAY.
|
Re: Z bitx poor sensitivity
I got my zbitx 2 days ago and only had about an hour to play with it.? I thought I understood receiver sensitivity, but not I'm questioning myself.? I tried to contact my dad on 80, 40, and 20 meters yesterday daytime.? We are at about 75 miles so often only one band will work on any given daytime.? 80M is NVIS at night is very reliable.? ?The really strange thing was he could hear me on 40M, but I couldn't hear him on any band.? ?And he was running 500W at the end.??
?
He's got a 160M horz loop, and I 80M horz loop.? Receive of a antenna is reciprocal of transmit, right?? It is a little hard to tune my antenna as I had to put another tuner ahead of my tuners for the lower scale SWR meter.??
?
I guess the basis of my question is, even if the receive isn't sensitive on the zbitx, how can a receive be sensitive enough to have a high noise floor but not receive a signal that should be well over a the noise I hear very well. Or is this an SDR thing?? This is my first ham radio that isn't a old tube rig.??
?
An education is appreciated
|
Assume you meant may 16th...
?
?
"POWER, WE DONT NEED NO STINKIN' POWER" - Every KX2 Owner?
Always respect the air waves,?
73 de KI7PBR, Gary
?
?
|
Update On Shipping Delays from HF Signals
I ordered my zBitx on April 3rd, 6 weeks ago.? I sent sales an email and got this response?
?
"Right now, we expect your order(#3941) to ship by early next week. We ran into some unexpected hiccups with sourcing some components which slowed us down a bit"
?
Fingers crossed I should have mine next week.? That will be 7 weeks from order date.? Clearing a spot on my work bench to bring it up.? Got a road trip in a couple weeks that I hope to have it ready for?
?
Excited and anxious,?
73 de KI7PBR Gary
|
Re: zbitx, compiler optimizations
#zbitx
On May 8, 2025, at 11:51, Peter, DL4PIT <pr@...> wrote:
Is there a reason why the build script is not using optimization? (for development side "-g" is perfectly understandable, but for "production" side?)
Maybe compiler optimization could at least contribute to the puzzle pieces for optimizing CW response.
I wrote a Makefile in??and made further changes in??and yeah I have it building without debug info by default. But if you type ¡°make?SBITX_DEBUG=1¡± then you get ASAN too, which finds all the memory problems at runtime (buffer overflows and things like that). ?But it seems to me that ASAN might be causing a memory leak itself (which becomes critical more quickly on zbitx with so little RAM), and it also deoptimizes it even more, so that¡¯s only for development, not to actually use. And I have ft8lib as a submodule (because it also needs fixes and improvements) and the sbitx Makefile also builds ft8lib if it hasn¡¯t been built yet. ?Also with ASAN, if debug is turned on.
|
Re: zbitx, compiler optimizations
#zbitx
Peter, I've been looking at your note again about time spent in i2c_delay. It is a simple busy-wait loop that waits a constant 400 'delayTicks' (set in i2cbb.c). I put some timing code in for debug and found 400 delayTicks to be just 5.8 microseconds on my sbitx RPI4. Since the delay is meant to make the i2c bus work right I assume that the _time_ value is going to be constant across processors though the delayTicks maybe could be adjusted for slower processors.
Long story, shorter ... What if you change i2c_delay() to get rid of the busy-wait.
static void i2c_delay() {
delayMicroseconds(5.8); // maybe put this value up where delayTicks was set in i2cbb_init()
}
which lets wiringPi use its own timing mechanism to wait the right time, on all platforms?
I haven't done any profilings like you did, but would love to hear if you profile it again and it changes the picture at all. It is called almost 9 million times in your sample! (if wiringPi uses a busy-wait then I should find a different timer).
NOTE: THIS HAS BEEN ONLY CASUALLY TESTED!
Mike - KB2ML
toggle quoted message
Show quoted text
On Fri, May 9, 2025 at 12:48 PM, Peter, DL4PIT wrote:
I found that the program is running in an empty time delay loop (i2c_delay)
for nearly 60% of the time:
% ? cumulative ? self ? ? ? ? ? ? ?self ? ? total
time ? seconds ? seconds ? ?calls ? s/call ? s/call ?name
58.73 ? ? 39.57 ? ?39.57 ?8844853 ? ? 0.00 ? ? 0.00 ?i2c_delay
--
Mike KB2ML
|
Re: zBITX CW Mode - Proof of Concept
That seems like a pretty rapid voltage drawdown.? ?There are lots of crappy 18650 batteries out there.? ?The ones promising a very high capacity seem to be the worst.? I bought 2 from Amazon that were highly rated in the reviews.? But one of them goes down much faster than the other one. I'm going to replace them.
?
That could be what is causing your CW problem.? ?Turning the transmitter on and off rapidly could be pulling down the voltage rapidly and causing strangeness.? The 5.0v and/or 3.3v regulators might not be handling it well.? ?I used a 5A 9vdc supply that I had laying around initially, but I did not like the voltage drawdown that I saw whenever I keyed the radio.? I'm using a 10A supply now.? That's a lot for a QRP radio.
?
In addition to trying it with new batteries, you could try an adjustable buck-boost module to run it from your 12v supply.? Adjust it to provide 7-8 volts.? Some of them are pretty dirty RF wise and can impact the receiver, but they are cheap and it might be worth a try.??
?
72,
Jody - K3JZD?
|
Re: zBITX CW Mode - Proof of Concept
yes I do have a delay of 1000ms.
my batteries are rated 5000ma, and I see a drastic voltage? drop within minutes of using the radio on a fully charge batteries. (from 8.4 to 7.3 to 6.9) IDK if its due to a high current draw or a bad batch of batteries.
I dont have any other means to test my radio, I wish the radio was 12v compatible.
?
|
Re: zBITX CW Mode - Proof of Concept
Hi Miguel,
?
Now I have mine running from a 7.5 vdc power supply., which about what 2 fully charged 18650's was providing.? ?I'm not having any erratic behavior problems at that voltage. I have been sending a little CW on it periodically as I'm testing a change to see if the memory leak or variable overrun goes away.? ?
?
Do you have your Delay set to 1000 or close to that ?? ?I never tried lowering that Semi-QSK Delay time on mine.? A 1 second lag before it drops back into Receive Mode is fine with me.
?
I usually do not worry about CPU usage unless it pegs at 100% for more than an instant.? ?I figure no use having some percent of it unused.?
?
Off hand, I can't think of anything else to suggest.?
?
72,
Jody - K3JZD?
|
The current offering is to purchase a complete kit that includes a case, mic, RPi, and screen. ?The board only is just the board. ?There was a separate offering to upgrade the board only to the equivalent to the complete kit, though the cost advantage goes away. ?I do not see that offering now. ?Better off to buy the complete radio and add a 4 or 8 gb ram RPi4.?
73
Evan
AC9TU
|
At one point, when RPi boards were hard to come by, and HF Signals had to pay premiums to secure the boards we convinced them to sell radios without RPi boards (or displays) included at a reduced price.
Yes, a comparable RPi 5 costs only $5 more than an RPi 4 with the same amount of memory, but Farhan has already indicated that the benefits of the RPi 5 are not needed to smoothly operate the radio, and those discussions (in part) led to the design of the zBitx with its RPi Zero 2 W board (the Zero 2 W is essentially a smaller form-factor RPi 3).
The discussion is in the archive, it shouldn't be hard to find.
Turning the radio into a nice ham radio PC may be something you are interested in, but I suspect HF Signals is content to sell radios to be used as radios.
The 'wasted' cost of the 2 meg RPi 4 in the sBitx is what, $35-40? If you replace an RPi 4 with an RPi 5, you may be introducing thermal issues inside the radio - I can't remember how the discussion on that topic wound up...
In my opinion, you'd have better luck getting HF Signals to omit the RPi 4 for a discount, rather then getting them to offer upgraded versions with either more memory, faster CPUs or both.
Good luck!?
toggle quoted message
Show quoted text
On May 13, 2025, at 16:52, AE9J via groups.io <merrow.jeff@...> wrote:
? I hate wasting money on a Raspberry Pi that I immediately need to replace. ?With a little extra ram it makes a pretty good dedicated ham radio PC. ?
?
Also the raspberry Pi 5 is already on the market and literally just a few dollars more.
?
?
|
Cory Lytle KO4SWI
A Kiddoo
Gary-KI7PBR