I got couple of emails how to power 7408 with 6V in QMX+.?
This is how I modified mine, lift pin 14 of 7408 from pcb carefully. Add a 7806 voltage regulator to 12V input and connect lifted up pin14 to that 6V regulator. That’s all. This will improve your high bands dramatically.
?
--
?
73, Barb
|
Re: Help with partially fried QMX
Hi Bruce,
I've verified that there is an air gap between the dc jack and encoder. I will put some electrical tape on the back of the DC jack to be safe.
I also did some additional testing and observed that the device still powers on immediately with draw when the LCD board and controls board are removed. Checked again for any solder bridges or solder flakes and was unable to see anything overtly wrong.
|
Re: Help with partially fried QMX
Could the body of the left encoder be shorting the DC jack?? You could put tape or thin plastic between them.
|
Chuck,
?
This ham is selling on Tindie:
?
?
?
--
?
73, Barb
|
That should have stated POWER MOD BOARD!
--
73 ChuckR
|
Hi all, is the circuit board availble to purchase and where to order? This would possibly solve a problem with my starion setup.
--
73 ChuckR
|
Re: Here's hoping for QMX+ v2
I think part of the problem with the Pi Zero is that it is a single core CPU. However, it's clock and memory pool are such that it can do the SDR transforms that are necessary. The T41 does it with a single Teensy 4.1 clocked at 600MHz. The Zero is cheap enough and stingy with power that you could dedicate one to the I/O functions and another to the number crunching. The Zero has the VFPv2 math co-processor which support both single and double precision floats. It might not be easy, but I think it would be doable.
Jack, W8TEE
On Saturday, April 12, 2025 at 10:24:43 AM EDT, Peter LaRue via groups.io <peter.f.larue@...> wrote:
I agree with Randy N4OPI that the Raspberry Pi (especially the Zero and Zero 2) is not an optimal choice for serious FT8 if your goal is to dig every possible signal out of the noise.? The FT8 software does some amazing things to find those weak and distorted signals, but it's brute force search methods are very CPU intensive.? I don't know for sure, but I assume that JS8 does similarly.?
For other modes, especially if only trying to decode signals at or near a specific center frequency, the Pi Zero 2W is sufficient and shares one of the advantages of the QMX: very low power consumption.? This means that, like the QMX, it can run for a long time on a small battery.? This makes it particularly useful for battery powered portable operations, or use where limited power is available from a battery system supported by solar.? For my interest in this sort of use, I also run my Pi Zero 2W with the lite version of Pi OS that does not include a desktop that is not needed for running Pat Winlink and ardopcf.
So, just like different radios are appropriate for different uses, so are different computers appropriate for different uses of digital modes.
Peter LaRue
AI7YN
|
Re: Here's hoping for QMX+ v2
I agree with Randy N4OPI that the Raspberry Pi (especially the Zero and Zero 2) is not an optimal choice for serious FT8 if your goal is to dig every possible signal out of the noise. The FT8 software does some amazing things to find those weak and distorted signals, but it's brute force search methods are very CPU intensive. I don't know for sure, but I assume that JS8 does similarly.
For other modes, especially if only trying to decode signals at or near a specific center frequency, the Pi Zero 2W is sufficient and shares one of the advantages of the QMX: very low power consumption. This means that, like the QMX, it can run for a long time on a small battery. This makes it particularly useful for battery powered portable operations, or use where limited power is available from a battery system supported by solar. For my interest in this sort of use, I also run my Pi Zero 2W with the lite version of Pi OS that does not include a desktop that is not needed for running Pat Winlink and ardopcf.
So, just like different radios are appropriate for different uses, so are different computers appropriate for different uses of digital modes.
Peter LaRue
AI7YN
|
Re: Here's hoping for QMX+ v2
|
Re: QMX+ with 100 amp/hour lithium iron battery
Jerry, thank you. That helps a lot. I live on a 20 acre horse ranch where there is absolutely no neighborhood RF noise, and even in that low-noise environment, I cannot detect any noise from the Drok. I can't take credit for choosing the Drok, someone on this forum suggested it as a good regulatory to use, and they were correct, apparently!
Thanks again and 73!
Braden Glett KD8ZM?
|
If you guys supply 74ACT08 with 6 Volts higher bands will improve
--
?
73, Barb
|
Hallo Volkert,
?
m?glicherweise hast Du meine Antwort weiter oben übersehen, ich hatte Dir einen Link zum Thema Treiber für Win7 geschickt. Im verlinkten Dokument steht auch ein Ablauf wie der Treiber zum Laufen gebracht werden kann, wenn er nach der Installation nicht funktioniert.
?
Falls Du lieber in Deutsch lesen m?chtest, schreibe mir bitte.
?
73 Ludwig
|
On Fri, 2025-04-11 at 22:30:30 +0300, Hans Summers via groups.io wrote: So I'm not currently working on a simplified CW filter. But I will...
And I wouldn't say it's a difficult thing to do. Just that the
cascaded filter thing was almost trivially easy at the time Reg
suggested it to me, I got intrigued and thought "why not". Hi all, To explain my thoughts, I attach a small C-program that I am running on my Linux PC. It takes the USB I and Q streams from the QMX and feeds them through a CW filteri, mixes then with the sidetone, and then sends them to the PC speaker. I have implemented four kind of lowpass filters: a Gaussian, a Butterworth, a Chebyshev and an Elliptic filter, all of order four (but any order will do). The filters can be adjusted for bandwidth and center frequency. The filters are defined in continuous time and then converted to discrete time. This conversion, which is only done when changing the parameters, involves the bandwidth setting. The adjustment of the center frequency is implemented by a mixer. Since the filter works in the IQ domain, the center frequency can be positive or negative. The filters are implemented on a balanced state-space form, which avoids many numerical problems, especially if the filter order is high. The implementation runs at 48 kHz but this can be modified to 12 kHz for the QMX. Best regards Anders
|
Hans,
??? No
hurry.?
??? You
solicited suggestions, I put one out.? If it never gets implemented, I
still have a very usable rig.
?
Bill
W2EB
East
Syracuse, NY
toggle quoted message
Show quoted text
Hello Bill
Thanks for the feedback. I would love to hear from Braden too. When not
liking the way something works, one should be able to propose an improved
alternative.?
?
From the user perspective,
I've found most other radios have you set your preferred CW tone/offset
(500hz, 600hz, 700hz, etc.), then choose a bandwidth around that tone.?
The QMX+ however presents you with a list of bandwidths that can be used with
several CW tones, then you must either live with the bandwidths that are
available for your chosen offset, or choose an offset based on the bandwidths
you'd like to use.?
Is it possible
to?change?the CW filter set up so that you can simply choose your
offset, then choose the bandwidth?? Or, stated differently,?can the
CW filters be made so that every bandwidth applies to every
offset/tone?
Yes this is exactly the?desired end goal. The reason it is the way
it is now, is because way back there was just the 300 Hz filter. In the
original way I had the code set up, there was just this single filter. It was
suggested that two offset overlapping versions of the same filter could be
used and that would provide a narrower bandwidth without causing ringing and
increases in delay.?
As it happened I already had on hand from way way back, 10 years ago, I
can't even remember where from, sets of filter coefficients for a bunch of
other filter centers and widths. It was a small amount of effort to code
this, and I released it in firmware 1_00_024 on 6-Aug-2024. At that time I
had already begun my main 9 month project of implementing SSB in QMX, which
so so many people have been waiting for, and I did not want to spend a large
amount of time redesigning audio code. I wanted to stick to minor changes
where possible. This implementation of the two overlapping filters was a low
effort thing but provided great benefit to CW operators. But it isn't ideal,
I know. It was intended as a temporary measure, with better ways to be
implemented post SSB.?
73 Hans G0UPL
|
Dr? Om Mel,
First of all, thanks for the quick reply.
Since this is the first time I've asked a question in a group, is it okay if I answer you directly?
I've installed the STM32102.
Installing the VCR driver: At the end, you're offered the option to update from the internet or cancel.
Although the ZOTAC is connected to the internet via WiFi and Ethernet via the Fritz!Box, it's not recognized. So the only option is Cancel!!!
Plugging in the QMX isn't recognized!
My USB stick is recognized.
?
To avoid any errors, I reinstalled Windows 7.
But then I had an error with the coprocessor.
Zotac Service gave me a link to download it.
Zotac ND22 Win 7 Home Premium 32-bit
https://www.zotac.com/de/support
ZOTAC DACH | SUPPORT | ZOTAC
here is the link of the site where you can download drovers for old devices
https://www.zotac.com/old_zotac/index.php?id=22&;L=0
https://www.zotac.com/old_zotac/index.php?id=170160&L=0%3Fshow%3Dglobal&tx_zoprodisp_pi1%5Bel%5D=10&tx_zoprodisp_pi1%5Bamounti%5D= 10&tx_zoprodisp_pi1%5Bproduid%5D=290&tx_zoprodisp_pi1%5Bsection%5D=5&tx_zoprodisp_pi1%5Bo rderi%5D=DESC&tx_zoprodisp_pi1%5Bsorti%5D=starttime&cHash=ab69fbdfe5460309d7e2a958e8b06169
?
After installation, the coprocessor error disappeared.
Tried installing the STM32102 again: Same result! Negative.
In 14 days, we plan to take a round trip in our motorhome for about 8 weeks, starting from the island of Rügen (EU057) via Denmark, Sweden, Norway, Finland, the Baltic States, and Poland, and back to the island of Rügen.
I bought the QMX for this purpose.
If I don't make it by then, I still have the 8 weeks!
Maybe you have another tip?
?
Vy73 de Volkert
dl6be
|
Re: Here's hoping for QMX+ v2
Hans, what are you doing up at this hour?? ?Oh, it's me who's up at a ridiculous hour, you're just having an easy Saturday morning.
Let's see, a simple wish list.? Dang, you're going to make me itemize the front panel interface and match it up against the CAT?+ Serial menu interface.? I'm too tired to do that now.??
But, if the CAT had a command to simulate the pressing of the four UI buttons and the turning of the two UI encoders, then the serial interface could drive the radio anywhere the front panel could go,?except perhaps for powering on.? ?But it wouldn't know where it was without the contents of the LCD panel, so CAT would also need to be able to pull those updates.? Note that this would open up a lot of automated testing possibilities that you don't currently have.? Like reading the bar graph?off the LCD and seeing if it agrees with the signal or mismatch presented to the RF input jack.
Then receiving IQ audio should allow for reproducing whatever the radio is putting out over its audio jack, because we know everything -- oh, no, we don't know what happened on the front panel interface, but maybe we could deduce it from the LCD panel contents log.? Ah, especially if the LCD panel log had a record of button presses and encoder twiddles.
Then if sending an IQ audio of the desired signal over the USB audio worked optimally for every mode, possibly?requiring the mode to be set as a hint to the firmware, then that would cover ASK, FSK, PSK, SSB, DSB, AM, and FM which are all the modulations I know about.
I think I like that list.? Allow the CAT/Serial/Audio USB host: to drive the UI pins; to read the LCD output and UI input log; to demodulate IQ audio to match the receiver output audio; to transmit anything that can be modulated as IQ on the USB audio channel.? It's all firmware, you don't need to rip up or reroute any traces, and you don't need to bring the front panel and serial menus?into congruence,?because the Serial/CAT host can see both of them,?which is another venue for testing.
-- 73 es gn -- rec -- ad5dz -- .
That's what I think, I see what I dream about?
On Sat, Apr 12, 2025 at 1:05?AM Hans Summers via <hans.summers= [email protected]> wrote: Hi Roger?
I'd like him to implement the few radio operations that can't currently be done with CAT over the serial line:? key down, key up, tune, and so on.? It's pretty obvious if you're trying to control all of the radio over the serial line.
I can extend the CAT commands easily. Put together a wish list for me :-)
There are things you can do, and things you can't do.? I don't know why the serial menu operations are a subset of the direct interface menu operations.?
It's all still in improvement phase. Lots more can be done!?
The tone detection to implement FSK is very clever, but the controller could simply specify a sequence of tones to send and skip all this zero crossing? detection nonsense.? Just because wsjtx?is wedded to sending USB audio doesn't mean everyone needs to do it that way.? But if he really wants USB audio as his transmit source, he should take CW as USB audio, too.
Every Digi mode software has audio in / out simply because it's the most easy and universal of "interfaces". Every multi mode transciever can send SSB. So it makes it as compatible with everything as possible. But again - I could add CAT commands along with the key down etc, to set the transmit tone frequency etc.? 73 Hans G0UPL
|
Re: Here's hoping for QMX+ v2
Hi Roger?
I'd like him to implement the few radio operations that can't currently be done with CAT over the serial line:? key down, key up, tune, and so on.? It's pretty obvious if you're trying to control all of the radio over the serial line.
I can extend the CAT commands easily. Put together a wish list for me :-)
There are things you can do, and things you can't do.? I don't know why the serial menu operations are a subset of the direct interface menu operations.?
It's all still in improvement phase. Lots more can be done!?
The tone detection to implement FSK is very clever, but the controller could simply specify a sequence of tones to send and skip all this zero crossing? detection nonsense.? Just because wsjtx?is wedded to sending USB audio doesn't mean everyone needs to do it that way.? But if he really wants USB audio as his transmit source, he should take CW as USB audio, too.
Every Digi mode software has audio in / out simply because it's the most easy and universal of "interfaces". Every multi mode transciever can send SSB. So it makes it as compatible with everything as possible. But again - I could add CAT commands along with the key down etc, to set the transmit tone frequency etc.? 73 Hans G0UPL
|
Re: Here's hoping for QMX+ v2
+1, Rpi is probably the most available and least capable linux microprocessor you can buy.? Well, you can buy worse, but you'd have to work at it.
I just bought a lenovo legion tab 3 tablet,? I'm going to use that to build an upgraded QMX*, starting with displays that my tired old eyes can read and buttons to push that don't?give me repetitive stress injuries and lots of UI shortcuts to replace the odysseys through the menus.? It's going to be fun, and there are lots of ways to do it if anyone else wants to get in on the fun.? You just can't expect Hans to reproduce the whole Android/Apple/Microsoft ecology of displays and user interfaces, he's just one guy, they're some of the most well heeled companies in the world.
I'd like him to implement the few radio operations that can't currently be done with CAT over the serial line:? key down, key up, tune, and so on.? It's pretty obvious if you're trying to control all of the radio over the serial line.? There are things you can do, and things you can't do.? I don't know why the serial menu operations are a subset of the direct interface menu operations.? The tone detection to implement FSK is very clever, but the controller could simply specify a sequence of tones to send and skip all this zero crossing? detection nonsense.? Just because wsjtx?is wedded to sending USB audio doesn't mean everyone needs to do it that way.? But if he really wants USB audio as his transmit source, he should take CW as USB audio, too.
-- 73 -- rec -- ad5dz --
I don't want to be a naysayer, but at one point a few years back I was excited about integrating a QDX and an 8GB RPi4 with a 7" monitor into one case with a small external kybd.
So I did some FT8 testing to compare it to my i7 windows 10 laptop. All else being made equal (without going into the details), the laptop consistently decoded ~10% more? stations than the RPi 4. On Field Day or any FTx competition, that matters. At least to me.
I would imagine based on other testing I've done with the RPi 2W that it would be considerably worse.
I really wanted this to work, but like Robin I'll be sticking with a small laptop that has extended battery life.
Maybe when they come out with an RPi6 I'll try again with the hope of stuffing it and a small monitor into the top half of a QMX+. :-)
--
Randy, N4OPI
|
Re: Here's hoping for QMX+ v2
I don't want to be a naysayer, but at one point a few years back I was excited about integrating a QDX and an 8GB RPi4 with a 7" monitor into one case with a small external kybd.
So I did some FT8 testing to compare it to my i7 windows 10 laptop. All else being made equal (without going into the details), the laptop consistently decoded ~10% more? stations than the RPi 4. On Field Day or any FTx competition, that matters. At least to me.
I would imagine based on other testing I've done with the RPi 2W that it would be considerably worse.
I really wanted this to work, but like Robin I'll be sticking with a small laptop that has extended battery life.
Maybe when they come out with an RPi6 I'll try again with the hope of stuffing it and a small monitor into the top half of a QMX+. :-)
--
Randy, N4OPI
|
Hey Randy, after doing some more testing I pretty much have the same results as you. On 80m 40m and 30m the difference was remarkable. 20m was mediocre. 17m and 15m were not great either. 10m for some reason was about the same. Although I was "cheating" a bit since I have my QMX+ wired for 9v and I was using 12v input. On the lower bands it was fantastic, but like stated by others the upper bands really suffer. It made the upper bands better but not by very much so in the end not really worth it, unless you like 80-40-30 only.? I just ordered a dozen TN0110 mosfets from Mouser to give them a try. Like I said, experimenting is part of the fun and learning.
?
Tim - K5DEZ
?
|
Barb WB2CBA
jjpurdum
Braden Glett
Volkert