Re: zBitx +-2 MHz transmit spurs? --> SHOWSTOPPER for transmitting
Steve, that is really admirable work!!
Without this kind of investigation, I¡¯m not quite understanding how this project proceeds.
I remember spending weeks and months trying to figure out Spurs in my sBitx. ? (Some of them I caused!!!) ? I learned quickly that absolutely anything that touched the analogue plus V voltage was a huge risk. ? ?I stayed away from it after figuring that out , and added a lot more bypass in various places. ?I believe I also dropped the drive to the SI 5351. ?Adjusted idle current to reduce no linearities?
Your idea of reducing gain at HF where it¡¯s not needed is a smart one. ?
It would be a fantastic thing to come up with solutions that made almost all of these units compliant. I don¡¯t know if we can get folks out there who have already bought the units to check there or two preemptively applied the fixes, but it¡¯s pretty important because this is eventually a great radio, and it does not need to have its reputation sullied by Spurs like this?
Gordon kx4z?
toggle quoted message
Show quoted text
On May 23, 2025, at 20:42, Steve Beckman via groups.io <ssteven3sb@...> wrote:
? I got my ZBitx earlier this week, and put it on the test bench to check it out.
?
My setup is a Rigol DSA815 1.5 GHz spectrum analyzer, along with a newly build coupler with a -40 dB tap.
?
Here's what I'm seeing on 20 Meters - 14.000 into a 50 Ohm dummy load, about 5 Watts out:
There are spurs at 12 MHz and 16 MHz; the 12 MHz spur is borderline. there's also a spur at 2 MHz which is only about 30 dB down from the carrier.
?
I've briefly checked the other bands, and am seeing that 17 Meters is worse - results similar to those reported by others.
?
The observations by Joerger DB2OO appear to be right on. Harmonics of the transmitter output frequency are mixing with the 40 MHz IF. In the case of 20 Meters - the third harmonic of 14 MHz is 42 MHz, which when mixed with the 40 MHz IF is creating the 2.000 MHz and spurs at 14 - 2 as well as 14 + 2. When the transmitter is moved to 14.500, the third harmonic is moved to 43.500, which when mixed with the 40 MHz IF creates a spur at 3.5 MHz, as well as spurs at 14 - 3.5 and 14 + 3.5, which I've confirmed with spectrum analyzer measurements.
?
An aside - when I took apart my ZBitx to start experimenting, I tried to repeat my initial measurements, and I got the mess below:
A number of additional spurs were present - and a couple are above the limit. I checked the setup, and finding nothing wrong, I just put one of the M3 x 40mm cap cabinet screws back in place, and the new spurs went away.
?
The heat sink has a DC connection to ground through the tab of the TO-220 5 Volt regulator U2, a 7805. That regulator is on the far side of the board relative to the finals Q1 and Q2. My theory is that the mounting tabs of Q1 and Q2 are forming a capacitor with the heat sink through the insulators, coupling some RF energy into the heat sink. Lacking any other ground points, that energy finds ground at the 5 Volt regulator, which happens to be pretty close to the 40 MHz crystal filters, a mixer stage, and a couple of IF amplifier transistors Q3 and Q4.
?
By the way, the four cabinet screws have a black finish which is non-conductive. They pass through plated holes on the main board, but there isn't a positive mechanical connection at those points. In order to continue bench testing, I needed to have the main PC board positively grounded to the heat sink. I made a little grounding strap shown in the picture below:
It's a short piece of solder wick, soldered to the ground side of C107. It passes closely by the source pin of Q1, but that pin is at ground potential. While testing, I attach the strap to the heat sink with a 6-32 screw and nut as shown below:
<Ground Jumper Installed.jpg> Anyone wishing to try to duplicate my tests needs to be aware of this grounding issue.
?
When the case is installed, I will slip a star washer between the lug and the heat sink, then place the side panels and install the case screws. Tightening the case screws should create a positive contact between the lug and the heat sink. I did get some stainless M3 x 40MM cap screws that stand a better chance of providing a ground path at the other three corners. A no-cost alternative might be to sand the black finish off the screws provided.
?
I've focused my attention on the circuits surrounding the Diode Ring mixer comprised of dual diode packages D3 and D4. Changing C31, the DC blocking capacitor at the output of the SI5351, had no effect on the spurs. Providing a better termination for the mixer RF port at the low-pass filter had no effect. Moving to the transmit IF amplifier transistor Q3, changing the emitter bypass capacitor C22 from 0.1 uF to 1000 pF had a measurable effect, as shown below:
The spurs at +/- 2 MHz from the carrier have dropped about 10 dB, and are no longer a concern. Unfortunately the spur at 2 MHz is still too big, although it has dropped in level as well. The spurs seen on the 17 Meter band are also still too big.
?
The function of Q3 is to amplify the 40 MHz IF signal during transmit. It does not need the large value of emitter bypass capacitor that's currently installed. Reducing the value of C22 reduces the gain of Q3 at lower frequencies (which reduces Q3's gain in the HF bands where it's not desired) while maintaining good gain at 40 MHz. The result is less RF energy is getting into the IF mixer port along with the 40 MHz IF, reducing the amplitude of the spurs.
?
This is as far as I have gotten. The Q3 stage may need to have an LC band-pass filter added to it's input or its output to better filter out the RF being picked up from the transmitter stages. I'm hopeful that a simple solution is going to be found soon, that results in a clean transmitter with plenty of margin for spectral compliance.
?
73; Steve, N3SB
?
?
?
|
Re: zBitx +-2 MHz transmit spurs? --> SHOWSTOPPER for transmitting
I got my ZBitx earlier this week, and put it on the test bench to check it out.
?
My setup is a Rigol DSA815 1.5 GHz spectrum analyzer, along with a newly build coupler with a -40 dB tap.
?
Here's what I'm seeing on 20 Meters - 14.000 into a 50 Ohm dummy load, about 5 Watts out:
There are spurs at 12 MHz and 16 MHz; the 12 MHz spur is borderline. there's also a spur at 2 MHz which is only about 30 dB down from the carrier.
?
I've briefly checked the other bands, and am seeing that 17 Meters is worse - results similar to those reported by others.
?
The observations by Joerger DB2OO appear to be right on. Harmonics of the transmitter output frequency are mixing with the 40 MHz IF. In the case of 20 Meters - the third harmonic of 14 MHz is 42 MHz, which when mixed with the 40 MHz IF is creating the 2.000 MHz and spurs at 14 - 2 as well as 14 + 2. When the transmitter is moved to 14.500, the third harmonic is moved to 43.500, which when mixed with the 40 MHz IF creates a spur at 3.5 MHz, as well as spurs at 14 - 3.5 and 14 + 3.5, which I've confirmed with spectrum analyzer measurements.
?
An aside - when I took apart my ZBitx to start experimenting, I tried to repeat my initial measurements, and I got the mess below:
A number of additional spurs were present - and a couple are above the limit. I checked the setup, and finding nothing wrong, I just put one of the M3 x 40mm cap cabinet screws back in place, and the new spurs went away.
?
The heat sink has a DC connection to ground through the tab of the TO-220 5 Volt regulator U2, a 7805. That regulator is on the far side of the board relative to the finals Q1 and Q2. My theory is that the mounting tabs of Q1 and Q2 are forming a capacitor with the heat sink through the insulators, coupling some RF energy into the heat sink. Lacking any other ground points, that energy finds ground at the 5 Volt regulator, which happens to be pretty close to the 40 MHz crystal filters, a mixer stage, and a couple of IF amplifier transistors Q3 and Q4.
?
By the way, the four cabinet screws have a black finish which is non-conductive. They pass through plated holes on the main board, but there isn't a positive mechanical connection at those points. In order to continue bench testing, I needed to have the main PC board positively grounded to the heat sink. I made a little grounding strap shown in the picture below:
It's a short piece of solder wick, soldered to the ground side of C107. It passes closely by the source pin of Q1, but that pin is at ground potential. While testing, I attach the strap to the heat sink with a 6-32 screw and nut as shown below:
Anyone wishing to try to duplicate my tests needs to be aware of this grounding issue.
?
When the case is installed, I will slip a star washer between the lug and the heat sink, then place the side panels and install the case screws. Tightening the case screws should create a positive contact between the lug and the heat sink. I did get some stainless M3 x 40MM cap screws that stand a better chance of providing a ground path at the other three corners. A no-cost alternative might be to sand the black finish off the screws provided.
?
I've focused my attention on the circuits surrounding the Diode Ring mixer comprised of dual diode packages D3 and D4. Changing C31, the DC blocking capacitor at the output of the SI5351, had no effect on the spurs. Providing a better termination for the mixer RF port at the low-pass filter had no effect. Moving to the transmit IF amplifier transistor Q3, changing the emitter bypass capacitor C22 from 0.1 uF to 1000 pF had a measurable effect, as shown below:
The spurs at +/- 2 MHz from the carrier have dropped about 10 dB, and are no longer a concern. Unfortunately the spur at 2 MHz is still too big, although it has dropped in level as well. The spurs seen on the 17 Meter band are also still too big.
?
The function of Q3 is to amplify the 40 MHz IF signal during transmit. It does not need the large value of emitter bypass capacitor that's currently installed. Reducing the value of C22 reduces the gain of Q3 at lower frequencies (which reduces Q3's gain in the HF bands where it's not desired) while maintaining good gain at 40 MHz. The result is less RF energy is getting into the IF mixer port along with the 40 MHz IF, reducing the amplitude of the spurs.
?
This is as far as I have gotten. The Q3 stage may need to have an LC band-pass filter added to it's input or its output to better filter out the RF being picked up from the transmitter stages. I'm hopeful that a simple solution is going to be found soon, that results in a clean transmitter with plenty of margin for spectral compliance.
?
73; Steve, N3SB
?
?
?
|
Thanks all!? I¡¯m not worried about simple issues. I don¡¯t CW. My use is voice and I want to try digital My big issue is I¡¯m sort of a nomad right now. I rent a small apt and work 2 jobs, (trying to retire at 60). So that limits my use. My goal of a small unit such as this is travel. When I go to help my mom around her place. My gal and I travel a lot. National parks and such. A perk of working in the medical field. I work a week and off a week.? I¡¯d like to get into POTA and SOTA type activations. I don¡¯t have the internet at my place, so that would be a damper on updates and certain uses of the radio. I do have access to the internet though. Just not daily.? I may just place an order and give it a shot. I¡¯ve surely wasted $200.00 on worse things!?
Gents,
Thanks for the info! I thought the review was great, BTW Dave! I look forward to fiddling around with this radio and honestly, weight isn¡¯t a huge issue for me as I am loading my e-bike down with the equipment and taking that to the parks. I also have a G90 setup for portable ops and a uSDX, so this is just another fun radio to experiment with.?
? As for the power issue, I was looking at maybe a Bioenno battery instead of using the 18650¡¯s. They have a 9V version in several different Ah capacities and also have a power pole to DC plug adapter that is the correct size for the zBitx.?
Our club also had a battery box build day recently and I have built several boxes with 15Ah, 20Ah, 3Ah and ?3 others that use Milwaukee, Ryobi or Ridgid batteries.?
On Fri, May 23, 2025 at 09:39 AM, W1WCI wrote:
POTA a perspective, many activations with zBitx.? Usually get five contacts in before taking a break for heat dissipation.? My fingers are my gauge. Running the radio at 75% drive usually change in a second set of batteries after the first drops to 6 volts.? Typically one and a half pairs of batteries per outing, the second set is usually still 7+ volts at the end. Did a five park outing on six batteries.? A lot less gear than using my Sbitx.
Tim - W1WCI
?
Thanks for that input.? I think it helps set expectations quite well.?
?
As a sbitx owner I think I'd find needing a break after five QSOs for the rig to cool down a bit frustrating, yet the sbitx is three times the cost of zbitx, takes up more space and weighs more, etc.
?
--
Regards,
Dave, N1AI
|
Gents,
Thanks for the info! I thought the review was great, BTW Dave! I look forward to fiddling around with this radio and honestly, weight isn¡¯t a huge issue for me as I am loading my e-bike down with the equipment and taking that to the parks. I also have a G90 setup for portable ops and a uSDX, so this is just another fun radio to experiment with.?
? As for the power issue, I was looking at maybe a Bioenno battery instead of using the 18650¡¯s. They have a 9V version in several different Ah capacities and also have a power pole to DC plug adapter that is the correct size for the zBitx.?
Our club also had a battery box build day recently and I have built several boxes with 15Ah, 20Ah, 3Ah and ?3 others that use Milwaukee, Ryobi or Ridgid batteries.?
On Fri, May 23, 2025 at 09:39 AM, W1WCI wrote:
POTA a perspective, many activations with zBitx.? Usually get five contacts in before taking a break for heat dissipation.? My fingers are my gauge. Running the radio at 75% drive usually change in a second set of batteries after the first drops to 6 volts.? Typically one and a half pairs of batteries per outing, the second set is usually still 7+ volts at the end. Did a five park outing on six batteries.? A lot less gear than using my Sbitx.
Tim - W1WCI
?
Thanks for that input.? I think it helps set expectations quite well.?
?
As a sbitx owner I think I'd find needing a break after five QSOs for the rig to cool down a bit frustrating, yet the sbitx is three times the cost of zbitx, takes up more space and weighs more, etc.
?
--
Regards,
Dave, N1AI
|
Yep. Putting the gps module inside is the best solution for portability. If there is no way to cram it in as-is, most of the case is 3D printed and with the stl files, it shouldn't be too difficult to remix a little extra space here or there.
|
It DOES go below 9V. I missed that at first!
|
I have been wondering the same.
?
one SEEMS to tick all the boxes and has some variability for other devices if needed.?
- 9V
- at least 3A (Is 5A too much?)
- Barrel plug is the right size and polarity
I'd like it if it would go below 9V to be on the safe side, but I wonder if someone at HF Signals could chime in here with perhaps a list of acceptable options??
Chris
?
?
|
Re: Unreliable WIFI on zBitx?
Joe and Jody, thanks for the additional details on your set ups and usage.? I was all excited to have gotten four hours out of it this morning before wlan0 crashed.? Rebooted and got five minutes before it happened again.? Ugh.
The 32/64 bit information is interesting though.?? I might have to look into that.?? I haven't actually taken the cover off yet, but this might be my incentive to do so!
73,
Steve KF8KI
toggle quoted message
Show quoted text
On 5/23/2025 2:07 PM, Jody - K3JZD via groups.io wrote: One of the things going on, which may or may not impact overall performance, it that (as I understand it) the zBITX is using a 64 bit RasPi Zero 2.? But as released, there is a 32 bit RasPi OS running on it.? I have to wonder if it might be more efficient if it were running a 64 bit RasPi OS.? ?If anyone tries that, please post your results.
72,
Jody - K3JZD
|
I have wondered if putting one of these in the radio tied directly to the pi. I use this on a raspberry pi pico (no wifi) for a WSPR TX. They are tiny. I would think there is room in the zbitx(now I haven't open it myself. to look)? It's about the size of a nickel?cone.?
I haven't received my zbitx and haven't had my hands on it yet, so maybe my question is moot.?
?
With inexpensive and reliable GPS Modules such as the GT-U7, is there a way to implement it into the zbtix for time sync or maybe even location??
?
The GT-U7 can ouput NEMA data via serial USB or TTL. I have used this module (with Arduino) and it has been sensitive enough to use inside my house.
?
My first thought is to use an Arduino as an NTP server for the zbitz to sync to. However, I am curious if a more direct way could be possible so it can be used without networking.
?
Ed
?
?
|
zBitX power and receive turnover
Still on a steep learning curve with this remarkable radio I got at FDIM in Dayton and had my first SSB QSO with a good audio report using the built in mic and awkward touch screen TX/RX buttons. I think I'll wire up a hand speaker mic. However, I note a LONG delay (around 2-3 seconds) when going from TX to RX either using the buttons or just returning to RX after sending CW. I haven't tried FT8 except to observe it receives OK. The delay is from a slow AGC recovery as it is absent with AGC off and it looks like like perhaps a "pop" that is filtered in the Audio. I've seen that in uBitX radios after version 3 where the audio delay in the last stage is just enough to hide the "pop", but if AGC is tapped ahead of that it gets affected. My solution in my uBitX was to limit the max the AGC could clamp down. Is there something similar that could be done in software?
?
I did measure CW power out running on 7.6v batteries, drive 100, into a dummy load with less than 1.2:1 SWR.
80m?? 8.0w
40m ? 6.5w
30m?? 3.8w
20m?? 7.0w
17m?? 6.0w
15m?? 3.0w
12m?? 3.5w
10m?? 3.0w
?
These seem similar to the Vpp shown on the test card, but a bit more pronounced. Not sure why 20 and 17 are so good.
?
=Vic=
|
The zBitx display stops responding to the mouse scroll wheel or keyboard up/down keys after a few clicks.
?
Jim
N7IHQ
|
On Mon, May 12, 2025 at 05:15 AM, cam1aw wrote:
Mouse wheel doesn't work for me, sometimes a value will change by one when clicked on, not always.
?
I have the same problem with the scroll wheel on my mouse. The values changes intermittently. It works ok on OS windows.
?
Jim
N7IHQ
|
Re: Up and Running Fully Portable
Here's the keyboard/trackball I plan to use:?
I got this one because the keyboard isn't quite as tiny as the one I'm currently using. It's really for two thumbs, not two hands. But the bigger one is far more usable. Here's the small one, just for comparison:
If I run out of power in my 18650s, I carry a USB-C PD cable that terminates in the correct barrel connector and gives 9V. I've got PD power banks galore. I carry a 5000mAh bank in my EDC pouch in my pocket. I have not experienced any noise from it on a ham band. It's possible there are birdies elsewhere, but I've never noticed one. If I feel like bringing one of my 12v battery boxes, with a LiFePO4 battery in it (one is 12Ah, the other is 7Ah), I have them rigged with fused PowerPole connectors and have cables to match.
If my partner Maggie K3XS drives, her CR-V is all set up for ham radio use. There's a multiband antenna on the roof that is folded down while driving, and separate from the car's battery and alternator, we have a 100Ah battery and a hefty distribution system and BMS for keeping the battery charged from the engine. She figured out how to use the sBitX (the bigger one) with the Elecraft KXPA100 amplifier, and there's plenty of power for it. If it's not too taxing, I don't mind setting up for QRP away from the car. But having the setup for 100W in case of emergency is nice.
I'm in the process of setting up crash cases for all of my radios. They're really cheap on AliExpress if you keep a lookout for them and catch a sale. -=-=-=-=-=-=-=-=- 73, Gwen, NG3P
Having fun with mine as well.?? I did not bother to try the OTG
adapter at all, but I've been doing stuff with various RPis for a
while so I had several of the necessary USB and mini HDMI adapters
around.
For my keyboard and mouse I used a small multi port USB hub and
that's working fine.??? I do like the idea of a bluetooth keyboard
though, and one with a trackball would be awesome.? That would
keep from putting too much regular stress on that mini USB
connector.? Mind sharing which keyboard/trackball you're using?
As for batteries, I got a four pack of the 18650s and a charger,
but you can kill a set of those pretty quickly on an FT8 operating
session, so I have a 12.5-7V dc to dc converter on the way.? As
long as noise isn't an issue, I'll probably use that and run it
off my battery box.? I'm also an RC guy, so I thought I might make
an adapter for a Lipo 2S pack as well.? I have several of those
available.
73, Steve KF8KI
On 5/23/2025 12:11 AM, Gwen Patton via
wrote:
Yeah, I ran
into the same issue with that OTG adapter. I think it's not
so much an OTG adapter as a crisp, tasty waffle. I had one
around, and, like you, got my keyboard and mouse working.
Then I paired a handheld keyboard with a trackpad built in,
and I'll likely use that. I think I'm going to make a
technical bag to go along with the radio's go-bag, one with
an HDMI cable, a 14-inch portable screen, and a better
bluetooth keyboard I have that's got a built in trackball.
Each of those things is bigger than the repurposed
photographer's bag I'm using for the radio and its immediate
gear. I need to get the radio out and update the firmware
and UI anyway, so I'll do the pairing for the new keyboard
while I have it out.
I got an
8-pack of 18650 batteries and a USB-C powered 4-place
charger that has been a real workhorse. I like that it
doesn't need a mains plug. I'm about halfway through
assembly of a 5-band link dipole. Once I get the basic build
put together, I'll run it up a telescoping mast and tune it
up. Then I'll crimp on bullet lugs to make connecting the
links easy, and it'll be in service. I've got a 3-band link
dipole already, but I need to replace the connectors in the
links, and it doesn't have my favorite bands in it. Besides,
I made the new one out of Kev-Flex. Durable.
-=-=-=-=-=-=-=-=-
73,
Gwen, NG3P
BitXers,
??? Been playing with and
learning my new zBitx and I am having a blast.? I have
made a couple CW contacts almost immediately after
receiving my unit.? I had a couple unexpected hiccups in
that I wasn't prepared for the "Mini-HDMI".? The other
was the USB-OTG adapter that came with it was bad.?
Fortunately I had another stuffed in a drawer and
decided to try it.? Keyboard came right up and I was
able to get the unit up in Host mode.? I paired up my
bluetooth keyboard so I no longer have to deal with the
USB port for that.
I am running a portable
16inch (USB powered) monitor along with the wireless
keyboard and can run the rig on either 18650 batteries
or my Bioenno 1212Ah based battery box.? Got the clock
sync'd up and started up FT8 on 20m and made my first
contact within in a minute....
Great start so far.? It's
not a KX2 but this little unit is a lot of fun
especially with all the build in apps which I still need
to play with.
Attached is a couple of
photos.
--
"POWER, WE DON'T NEED NO STICKIN' POWER" - EVERY KX2 OWNER
Always respect the airwaves,
73 de KI7PBR, Gary
|
Re: Unreliable WIFI on zBitx?
One of the things going on, which may or may not impact overall performance, it that (as I understand it) the zBITX is using a 64 bit RasPi Zero 2.? But as released, there is a 32 bit RasPi OS running on it.? I have to wonder if it might be more efficient if it were running a 64 bit RasPi OS.? ?If anyone tries that, please post your results.
?
72,
Jody - K3JZD
?
|
Re: Unreliable WIFI on zBitx?
To clarify, I utilize RealVNC to remotely control the WSJT-X FT8 interface on the ZBITX system. Gridtracker, installed locally on my laptop, runs concurrently. The RealVNC client resides on my laptop, while WSJT-X is displayed on the external monitor connected to the ZBITX. This setup allows simultaneous use of VNC, WSJT-X, and Gridtracker, with Gridtracker typically prioritized as the foreground application. All connections are established with audio feedback and utilize Gridtracker's integrated logging mechanism.
Joe VE1BWV?
Thanks Joe.?? Desktop size change here was very obvious from with HDMI
monitor connected to without.??? RealVNC has been up and stable this
morning without the monitor attached for on the order of an hour and a
half now.?? Clearly more experimentation/troubleshooting is required.
That aside, mind sharing what your network setup is at remote locations
that you're doing RealVNC?
Steve KF8KI
On 5/23/2025 9:51 AM, Joe via wrote:
> I have been running the ZBITX for a month or so. My experience is no
> issues with wifi.
> I connect with real vnc both locally and at remote locations.
>
> I almost always have a 14 inch touch?screen ext monitor attached? and
> so far l have not seen any video changes.
>
> Also via web remote? so far it connects fine.
>
> Joe
> VE1BWV
> .
>
|
Re: Unreliable WIFI on zBitx?
Hi Steve,
?
I just had a WiFi RealVNC connection - I did not have any HDMI monitor, keyboard, or mouse connected.
?
72,
Jody - K3JZD?
|
Re: Unreliable WIFI on zBitx?
Steve,
For remote locations I installed Tailscale? on? the zbitz. Super simple Tailscale provides a free VPN access for nearly all devices, Linux, windows, snology nas? Android and iPhone. On your pc you just select your zbitx It automatically assigns an internet ip and a DNS, so you can choose
Enter that on your web browser and remote web from zbit appears from any place in the world... fun stuff.
Joe VE1BWV? No port fwd or opening settings on the router. It just makes it look like your connected on your home network.
I think they allow 100 devices. I run about 10 of my systems.
Thanks Joe.?? Desktop size change here was very obvious from with HDMI
monitor connected to without.??? RealVNC has been up and stable this
morning without the monitor attached for on the order of an hour and a
half now.?? Clearly more experimentation/troubleshooting is required.
That aside, mind sharing what your network setup is at remote locations
that you're doing RealVNC?
Steve KF8KI
On 5/23/2025 9:51 AM, Joe via wrote:
> I have been running the ZBITX for a month or so. My experience is no
> issues with wifi.
> I connect with real vnc both locally and at remote locations.
>
> I almost always have a 14 inch touch?screen ext monitor attached? and
> so far l have not seen any video changes.
>
> Also via web remote? so far it connects fine.
>
> Joe
> VE1BWV
> .
>
|
On Fri, May 23, 2025 at 09:39 AM, W1WCI wrote:
POTA a perspective, many activations with zBitx. ?Usually get five contacts in before taking a break for heat dissipation. ?My fingers are my gauge. Running the radio at 75% drive usually change in a second set of batteries after the first drops to 6 volts. ?Typically one and a half pairs of batteries per outing, the second set is usually still 7+ volts at the end. Did a five park outing on six batteries. ?A lot less gear than using my Sbitx.
Tim - W1WCI
?
Thanks for that input.? I think it helps set expectations quite well.?
?
As a sbitx owner I think I'd find needing a break after five QSOs for the rig to cool down a bit frustrating, yet the sbitx is three times the cost of zbitx, takes up more space and weighs more, etc.
?
--
Regards,
Dave, N1AI
|
Re: zBITX Crashes - Suspect Memory Leak
#zbitx
Hi John,
?
No, I did not try doing it that way.? ?I had no problem adding debounce to version 3.021, where only the falling edge needed it.? What I ran into while trying to debounce version 3.052 was that the rising edge now also needed to be debounced.? Changes made elsewhere in the CW processing chain now required that.? I was using paddles that are not often used, therefore having dirty contacts.? I was trying for a debounce time delay that would accommodate slow simple straight key use as well as 30 wpm electronic keying.? No timing that I tried seemed to work at both ends. Tiny delays were not effective because I still got bounces,? and longer delays resulted in my missing the rising edge causing the carrier to remain on.? It seems like disabling the IRQ (if the library being used allows it) would add another unpredictable timing element into the mix.? I'm already wondering how accurate the microsecond timer I have been using is with very low time delays.? The RasPi is running the full OS and GUI.? So it is busy. I still have my version that has the IRQs, and will give that a try whenever I have some time.??
?
72,
Jody - K3JZD
|
Re: Unreliable WIFI on zBitx?
For those following along at home on my WIFI issue:
- I discovered by using raspi-config and selecting a desktop size, I no longer got the difference in RealVNC in desktop between having an external monitor connected and not.? Both remained at the desktop size I chose.? I ended up digging into this because the original default size caused the sBitx program not to display completely on RealVNC without an external monitor being connected.
- After doing so, my RealVNC connection (and absence of wlan0 failures) reliability improved markedly.? I'm thinking this may be attributed to the fact that initially the resolution of the monitor I was using was much higher than the desktop size I ultimately chose (1024x768).? Intuitively I suspect this decreases the amount of traffic necessary for the RealVNC connection and thus reduces the load on the Pi.? Subsequently, I ran a RealVNC connection continuously for about four hours this morning, including running FT8 before a wlan failure occurred.? That I could see being attributed to a lack of available resources after having the connection open for so long.? Having executed many FT8 transmit cycles over that time, I was unable to attribute the wlan0 failure to anything other than the length of time the connection had been open.
73, Steve KF8KI
|
Jay Bremner
Ed K.
Chris (KQ4AET)
Vic WA4THR
Gwen Patton
Dave, N1AI