Hello! I've finished assembling my new QMX, it only being my second such project after the Morserino-32 morse learning device. Once I plugged the QMX into a variable power supply, I was delighted to see it seemingly powering on and off as expected. After installing the firmware, however, I have stumbled upon a huge issue though: The display's backlight turns on, but there is not a single pixel appearing on the display. Using the QMX terminal, I was able to check that apart from this, everything seems to work for now.
Anyway, I have checked all joints and they seem to work fine. Which display pins and/or pin header connector do I have to focus on when the backlight works but the screen not at all? Which pins should I leave like that because they are already working?
|
Re: Hans , I can replace IC 203, STM32F446VET6 ,do I need the boot loader?
On Sat, Sep 30, 2023 at 08:41 PM, Jeffrey W Moore wrote:
I prefer the low temperature solder method
Thanks, I am glad to hear your opinion and your description.? I decided to try the low-temp desoldering alloy rather than the hot air gun for the same reasons.? Much gentler on the board, and there's no worry about capacitors dropping off the other side.? Since the STM32 is QFP (pins), not QFN (pinless), we have access to all the solder joints.? That's a big benefit ---- Thank You Hans for choosing QFP rather than QFN! Jeff, I am NOT at your level of skill and knowledge for diagnosing and fixing busted radios.? I just want to see if a monkey with a soldering iron can pull off this repair without damaging the chip. I need to know my QMX actually works before I start messing with it, so it's time to build the thing.? I'm a slow builder, but I'll eventually report back here. And THANK YOU again Hans, for hearing our plight.
|
1. I intentionally tripped the SWR protection--it seems to remain active until I enter the menu or restart the QMX. ?It remains active when tuning or changing bands. ?What's the intended way to "reset" it?
I watched the demo video and it answered this question--entering the menu or power cycling is the intended way to clear the SWR protection TX block. Jonathan
|
For anyone interested in the Zyyini 70cm LPF module, here¡¯s the VNA scan of the module I just received. As Barb previously indicated , it¡¯s a bit off meaning too low of a cutoff frequency as well as having only ~30 dB of return loss. Hopefully, playing with those tiny inductors I can barely see might make it usable in the 431-433 MHz range. How much effort before I toss it in the trash is unknown.
Tony AC9QY
On Mon, Sep 25, 2023 at 4:17?PM Tony Scaminaci via <tonyscam= [email protected]> wrote: Thanks for the info Barb. I ordered one today and will do a VNA scan when it arrives. Can¡¯t expect perfection at that price point.?
Tony On Mon, Sep 25, 2023 at 12:27 PM < wb2cba@...> wrote: On Mon, Sep 25, 2023 at 10:03 AM, Tony Scaminaci wrote:
Wow, you can't beat the price! Barb, have you scanned these for harmonic rejection? I'd like to pick up some of these if they meet FCC specs.
?
Tony
AC9QY
On Mon, Sep 25, 2023 at 11:58?AM < wb2cba@...> wrote:
I use these:
Zyyini Low Pass Filter Module, Power Distribution Control Equipment, Wide Operating Frequency Range£¬ for RF Signals
These have a pretty good response on 70m.
73
Barb WB2CBA
?
?
Tony,
Couple of years ago I built cubesat ground stations with these. They worked pretty good.
They are kinda hit and miss but not that off from FCC parameters. It¡¯s easy to tune them just by playing with inductors, sqeeze or extend, using a vna or spectrum analyser. I didn¡¯t even have to try cap replacement.?
73
Barb WB2CBA
?
|
Re: Hans , I can replace IC 203, STM32F446VET6 ,do I need the boot loader?
Here is what I do to replace the cpu. ?Keep in mind you need a programmed donor cpu for this. ?I always get very anxious when I use a heat gun with lots of small parts around the target. ?I¡¯ve used this technique successfully but I prefer the low temperature solder method. ?I put a bead of flux & a bead of low temp solder around the entire chip, I keep circling my iron around the chip until it is completely loose. ?I then pick up the chip with a vacuum pen. ?I clean up the board with solder wick. ?I clean up the chip by solder wick under the pins on a silicone heat proof mat. ?When installing a new cpu, I line it up carefully, tack solder one pin, straighten then tack solder another pin. ?Once all is lined up perfectly, I solder each pin by hand very quickly. ?I have a really small iron. ?Some pins (like GND) won¡¯t flow well. ?I then get a bigger tip and reflow those stubborn pins.
it¡¯s great when you turn it on and it comes to life.
good luck,
Jeff
W1NC
|
Re: QMX RX BPF troubleshooting, a new clue
That sounds pretty good, textbook DBM mixer.
|
QDX and safe battery operation ¨C do I need a buck converter?
I built a QDX (Rev 3a) for 9v operation, and got a Bioenno battery to power it in the field.
The Bioenno, although nominally 9V, actually puts out 10.5 volts when fully charged. I'm not sure if I should use a buck converter to limit the voltage to exactly 9V. Any thoughts, or words of wisdom from experience?
I'd be happy to do some calculations on this, but I'm not really sure which numbers to look at to determine what's a safe operating voltage to avoid blowing out the output transistors.
|
Re: Hans , I can replace IC 203, STM32F446VET6 ,do I need the boot loader?
Great.,? I can possibly repair two radios someday.? I won't trash them yet.?
I would be willing to help out.? Just send me a PM.
|
Hans,
Some very welcome features--thank you!
A question and a request:
1. I intentionally tripped the SWR protection--it seems to remain active until I enter the menu or restart the QMX. ?It remains active when tuning or changing bands. ?What's the intended way to "reset" it?
2. The on-LCD Tune SWR functionality would be much more useful if you could change the frequency, rather than just using the configured center-of-band (especially since you can't change that value without the terminal). ?When tuning a high-Q antenna you'd want to tune right near your desired operating frequency, which is likely to be variable.
Jonathan KN6LFB
|
Re: QMX RX BPF troubleshooting, a new clue
I might add that the diodes were selected for match and the mixer was properly terminated for both baseband and RF products.
JZ
On Sat, Sep 30, 2023, 7:41 PM John Zbrozek < jdzbrozek@...> wrote: No problem, Ryuji... 1N5711 Schottky diodes and 10-12 dbm LO injection.
JZ
On Sat, Sep 30, 2023, 7:32 PM Ryuji Suzuki AB1WX < ab1wx@...> wrote: On Sat, Sep 30, 2023 at 05:48 PM, John Z wrote:
I have had a diode ring DC receiver and a QSD Tayloe detector running
side by side. There is simply no comparison. The diode ring will serve
up plenty of news and music along with your CW or SSB. Not so much
with the Tayloe circuit.
I'm curious now which diodes you used, how many dBm of LO injected, and what filters used.... but don't bother answering these.
|
Re: QMX RX BPF troubleshooting, a new clue
No problem, Ryuji... 1N5711 Schottky diodes and 10-12 dbm LO injection.
JZ
On Sat, Sep 30, 2023, 7:32 PM Ryuji Suzuki AB1WX < ab1wx@...> wrote: On Sat, Sep 30, 2023 at 05:48 PM, John Z wrote:
I have had a diode ring DC receiver and a QSD Tayloe detector running
side by side. There is simply no comparison. The diode ring will serve
up plenty of news and music along with your CW or SSB. Not so much
with the Tayloe circuit.
I'm curious now which diodes you used, how many dBm of LO injected, and what filters used.... but don't bother answering these.
|
Thank you for your hard work and exciting update!
|
Re: -61 and -43 mix binocular cores and models for the RWTST and WTST
My previous comment got folded on the web view, so please click on Hide quoted text mark to see my response.
This is where you see what I was describing regarding the inductance vs freq. You want to look at the trace for mu'. The mu'' trace indicates the resistive loss component of the core.
|
Re: QMX RX BPF troubleshooting, a new clue
On Sat, Sep 30, 2023 at 05:48 PM, John Z wrote:
I have had a diode ring DC receiver and a QSD Tayloe detector running
side by side. There is simply no comparison. The diode ring will serve
up plenty of news and music along with your CW or SSB. Not so much
with the Tayloe circuit.
I'm curious now which diodes you used, how many dBm of LO injected, and what filters used.... but don't bother answering these.
|
Re: QMX RX BPF troubleshooting, a new clue
On Sat, Sep 30, 2023 at 05:34 PM, Hans Summers wrote:
Does it really though... this selectivity is POST the mixer. I still can't
convince myself that this makes it a pre-selector with a Q exceeding 3500.
I don't know exactly what mechanism that Q of 3500 is arising from. That number compares to a crystal ladder filter at IF. I assumed that the number itself might be from the marketing department but the input impedance would drop away from the LO freq, along the line of previous discussion with JZ. If the input impedance drops very fast, one could argue that is equivalent of a high Q, especially if the RF source impedance is very high to take advantage of deep off-band attenuation. Obviously, in QMX implementation, the Q cannot be 3500 while the receiver IF of 12kHz is working well with good sideband rejection.
I think you understand it but my point was that I would take any additional selectivity on top of a simple BPF but not in place of.
Imagine we had a diode ring mixer in a direct conversion receiver to
baseband, and a low pass filter implemented by some op-amps. It would be
easy to achieve a Q exceeding 3500 if we assume that this is the RF
operating frequency divided by the 3dB corner frequency of the low pass
filter.
Right. However, I didn't interpret that way because, as you expressed below, it is not a fair argument.
hence I find it hard to
introduce fundamentally new concepts such as the post-mixer baseband
roll-off magically becoming a built-in pre-selector with Q exceeding 3500.
In fact with that logic (that we disgree with) a CW mode receiver at 3500kHz would have a Q of 7000 (or more since I have a narrower AF BPF).
The genius of QSD/Tayloe is in the simple low cost high performance
implementation using common bus switches, which happen to provide a very
high IP3 performance because the switches are fast and low loss, much
closer to perfect switches than diodes are. This is the real reason why
it's so good. The switching action is much nearer to perfect and this gives
it a lower insertion loss, lower drive requirement and high IP3
performance. The high IP3 performance means you can get away with weaker
(or even NONE, though I consider that a step too far) filtering before the
mixer.
Agreed. I think a simple preselector is a prudent design choice, perhaps even if one is unnecessary in some situations. I think the controversy is an unnecessary one, but I guess one is largely out of unfortunate frustration with dealing with the interactions with the connected RF circuits such as LPF.
One real-world test may be to run a multi-operator station at a SOTA summit or in a POTA park with two QMX and two vertical antennas maybe 50 or 100m apart. Two stations operate on different bands simultaneously and see if the BPF makes any difference.
|
Some excellent new features. Thanks Hans!
I wonder if it would be better or at least more intuitive if the Tune % was expressed as Power rather than Voltage?
I suspect most people probably tend to think about RF output in power rather than voltage terms, so it would be easier than having to square the value.
Regards, Dav M0WDV
Hello all
I released firmware version 1_00_010 which is confined to mostly enhancement regarding the SWR meter implementation and bad SWR protection. You can download it here and the operating manual is also updated for this version. Here's a video demonstrating the changes:??
The changes are:
1. Single character battery voltage option instead of battery image
2. Initialization sequence on display so if stuck anywhere we'll know where
3. PA voltage is always on in diagnostics screen and green/red to indicate correct range
4. Power and SWR added to diagnostics screen
5. Power and SWR meter on QMX display; controlled by menu entry in Display/controls menu
6. Bug fix: Battery now shown correctly at power-up, and when CW decoding disabled
7. New protection menu, with SWR protection enable and threshold
8. Supply voltage min/max protection including warn or optional TX inhibit
9. SWR sweep added to terminal applications
10. SWR Tuneup screen on QMX LCD
Details:
1. Single character battery voltage option instead of battery image
There's now a choice of the usual battery voltage icon, or a new style character which makes a miniature number to indicate the volts, including the 1 of 10, 11, 12 for example, and a number of pixels lit to indicate the 0.1V increments.?
2. Initialization sequence on display so if stuck anywhere we'll know where
Now during power up the display shows (very briefly) status information; if the boot-up fails somewhere and gets stuck, the display will show where it got stuck and this could aid debugging.?
3. PA voltage is always on in diagnostics screen and green/red to indicate correct range
Previously the PA voltage was only down on the diagnostics screen when the T key was pressed (transmit); now it is displayed always and colored red/green to indicate whether or not it is in the correct range. On receive the PA voltage should be less than 0.5V.?
4. RF Power and SWR added to diagnostics screen
RF power and SWR measurements have also been added to the Transmit section of the diagnostics screen.?
5. Power and SWR meter on QMX display; controlled by menu entry in Display/controls menu
The RF power output and SWR measurements are now displayed on the main operating screen, on the top right in the location of the (future) S-meter. The RF power range is 0 to 6W and SWR is 1.0 to 4.0.?
6. Bug fix: Battery now shown correctly at power-up, and when CW decoding disabled
Self-explanatory
7. New protection menu, with SWR protection enable and threshold
We now have a new "Protection menu" containing various items that can help protect the QMX against adverse conditions. The first is a? configurable SWR protection, which defaults to enabled and 3. The SWR is measured every 1 millisecond during transmit and transmit is disabled if a high SWR condition is detected. Since the QMX contains RF envelope shaping and a reasonable enough SWR measurement is available even at low power as the envelope ramps up, the SWR protection acts before full power is reached and therefore provides good protection against both high voltage and high current types of SWR mismatch. In my tests at 12V supply the power output reached only 0.5W before it was shut down (gently).
8. Supply voltage min/max protection including warn or optional TX inhibit
Also in the new protection manu are parameters to control a minimum-maximum supply voltage range and the consequences of breaching it; which can be set to none, or warn only, or warn AND prevent transmit. The warning is indicated by flashing the battery icon/measurement.?
9. SWR sweep added to terminal applications
A new screen lets you sweep the SWR across the same range as the RF sweep screens. You can also press the T key to measure SWR and tune the antenna. The PA voltage is reduced according to the "Tune %" parameter, which is a new parameter also put in the "Protection menu". It defaults to 50%, remembering that voltage and power have a square law relationship, this is equivalent to 25% of full power output. See attached example, measurement on 20m band of my Off Center Fed Dipole which is not in a happy state at the moment, as the long leg has become detached from its far end fishing pole and is lying on the metal-framed tile roof. Still I've had some DX QSOs on it when I have a free moment ;-)?
10. SWR Tuneup screen on QMX LCD
A simplified version of the SWR screen (above) is also available on the QMX screen itself, in the "Hardware tests" menu; it doesn't provide a sweep, just a simple SWR measurement for checking your antenna or tuning up the antenna. SWR measurement takes place at the configured center frequency of the band, and you can use the rotary encoder to scroll through the bands. Again the PA voltage is reduced to the "Tune %" parameter so you don't have a risk of damaging the PA transistors.?
73 Hans G0UPL
|
Hans,
thanks for that. QMX going from good to better.
My first observations:
- SWR sweeps are having spikes, different ones when changing band up or down
- there is a brief SWR spike showing at the beginning of each TX (tune) on the QMX LCD
73 Bojan S53DZ
|
Re: -61 and -43 mix binocular cores and models for the RWTST and WTST
Thank you for sharing the report, Ross.
On Sat, Sep 30, 2023 at 01:18 PM, Afghan Kabulldust wrote:
I¡¯ve done a set of measurements of Inductance of the secondary winding of a
WTST on both -61 and -43 mix.
Is any of the QRP Labs transceivers using 61 mix or did you just try for comparison? (I previously said 43 ferrite is wonderfully lousy material for some but not all uses, or something to that effect... 61 is wonderfully less lousy material for some uses.)
Knowing the
resonant frequency and the C value allowed me to calculate inductance AT THAT
FREQUENCY and plot the values.
That is a very important point. Those L/C meters usually measure at a low frequency and in case of 43 ferrite often grossly overestimate the inductance at the HF operating frequencies.
The first attached plot, albeit very simple, is the 3t output of BN61-202 WTST
and you will note that the inductance INCRRASES with frequency, whereas in the
second plot of the 2t output of a BN43-202 RWTST, you¡¯ll note the inductance
DECREASES with an increase of frequency.
I would expect windings on 61 material to increase the inductance with freq up to about 20-30MHz and then decline. In case of 43, the increase is much shallower and only up to around 2MHz then decline. You can see that from Fair-Rite data sheets.
Also, the inductance versus frequency plot can be made very easily with a NanoVNA if you have one.
Now, my question is, what did you connect to the primary while measuring the secondary impedance?
Are the vertical axes in microhenries?
Now I¡¯ve done these plots to specifically complicate the task of those who
wish to model the QDC/QMX output transformers WTST and RWTST, of which I¡¯ve
included my functional model diagrams for LTSpice, noting that, in both cases,
L values for each segment, should be calculated from the either of 3 turn or 2
turn value from the charts.
I actually thought a bit more about possible transformer model after I called the simulation experts to join. I think these output transformers to be somewhat difficult to model in general circuit simulators, unless they have model framework for transmission line transformers.
|
Re: QMX RX BPF troubleshooting, a new clue
Roger that, Hans. It is an amazing circuit!
The question of "is filtering before-or-after the switches?" is sort of meaningless I think. The switches and integrating capacitors are married rather inseparably.?
The question might take on some merit if the input signal was so large that the channel resistance of the mux switches underwent modulation by that signal: a major nonlinearity.
That would require a rather *Yuge* input signal! Hundreds of millivolts, probably, to be meaningful.
We enjoy the benefits of this nearly unbreakable design every day as it has become a clear winner in cellphones. Kudos to the genius of Dan Tayloe!
JZ
Hello JZ
Yes but does it really move the filtering effect to BEFORE the multiplexer switch... or is it a filtering effect BEHIND the multiplexer switch, as would occur in the thought-experiment diode mixer?+ op-amp low pass filter (for example)... hmm... Anyway, regardless of the theoreticals, I love the QSD which is why I have used it so much for everything, fantastic circuit, unrivaled performance:price ratio.?
It's not magical, actually. It is a direct consequence of the switched-capacitor tracking filter origins of this circuit.
JZ
? I have had a diode ring DC receiver and a QSD Tayloe detector running
side by side. There is simply no comparison. The diode ring will serve
up plenty of news and music along with your CW or SSB. Not so much
with the Tayloe circuit.
Yes, I agree completely, it's an EXCELLENT mixer... I'm just not able to convince myself of the inbuilt pre-selector thing, and no benefit in any BPF at all ahead of it... I'm feeling the large IP3 is what makes it so good, not any magical pre-selection... could be I'm wrong though.
|
Re: QMX RX BPF troubleshooting, a new clue
Hello JZ
Yes but does it really move the filtering effect to BEFORE the multiplexer switch... or is it a filtering effect BEHIND the multiplexer switch, as would occur in the thought-experiment diode mixer?+ op-amp low pass filter (for example)... hmm... Anyway, regardless of the theoreticals, I love the QSD which is why I have used it so much for everything, fantastic circuit, unrivaled performance:price ratio.?
It's not magical, actually. It is a direct consequence of the switched-capacitor tracking filter origins of this circuit.
JZ
? I have had a diode ring DC receiver and a QSD Tayloe detector running
side by side. There is simply no comparison. The diode ring will serve
up plenty of news and music along with your CW or SSB. Not so much
with the Tayloe circuit.
Yes, I agree completely, it's an EXCELLENT mixer... I'm just not able to convince myself of the inbuilt pre-selector thing, and no benefit in any BPF at all ahead of it... I'm feeling the large IP3 is what makes it so good, not any magical pre-selection... could be I'm wrong though.
|
Haris
Ryuji Suzuki AB1WX
Leigh KG7WED
Dafydd Walters M0WDV