Re: QCX Challenge & QCX WARC Party -> New support web page :-)
Lol, It was the last one in the log, on QO-100, been working them /mm for the grids.
?
73
?
Ian G4GIR,
?
toggle quoted message
Show quoted text
----- Original Message -----
Sent: 06/01/2025 00:26:28
Subject: Re: [QRPLabs] QCX Challenge & QCX WARC Party -> New support web page :-)
Ian, funny that you used that call, as I have been tracking the "POLARSTERN" using WSPR for the past year....
?
Small world...
--
Rick
W5NR
|
Re: QCX Challenge & QCX WARC Party -> New support web page :-)
Nice work Tom, but please check the way you determine the dates.
The WARC Party is always two weeks before the Challenge, so for this month it's next week , January 13th, not the 20th
?
I know it's not possible to enter this in e.g. a Google calendar, so maybe we have to change this to "second monday of the month", which is easier for calendar programs.
Any supporters of changing this ?
?
73
Luc ON7DQ
|
Re: New QMX Rev 3 / High Band - Holiday Project failed after FW Install
It is also safe to use one of the 3V smps cards from either of your other two kits, now that you have shown the main board does not have a 3.3V short.
After doing that, and verifying no problems, you can also remove D109 from the bad 3.3V card, and go ahead and use it without D109 to make sure that is the problem.? It will function fine without D109, but that leaves you vulnerable to a new fault causing damage on the main board.? So don't operate with it that way for long periods.? But it would give you confidence that replacing D109 would make it serviceable again.
The most common causes of D109 failure (that I am aware of) is accidental momentary shorts of 12V or 5V rails to 3.3V during assembly or handling with power on - the minimal clearances within make this easy to do.? Also please check your power plug connector from your power supply, that it is 5.5/2.1, not the similar 5.5/2.5, which looks almost the same, but has a larger diameter hole for the power pin, and doesn't contact properly.? Some have reported D109 failures that seem to be traced to such a connector - which will work most of the time, but actually can give many very short power on/off spikes when it is moved, since it 'bounces' in the ill-fitting contacts.
Stan KC7XE
|
Re: QMX #008 Power Supply Search
The Bioenno has a full-charge voltage of 10.8V
This will be too high for a 9V QMX, unless you use at least 2 diodes to drop the voltage.? Or avoid transmitting until the voltage drops to near 9V.? IF you transmit only into a 1:1 SWR, and only transmit CW (low duty cycle), it MIGHT work just fine as-is at full charge, but there is some risk to damage the finals with over-voltage, especially on the lowest 2 bands.
Using this on a 12V QMX will work fine, with a small output power reduction (less than1 S-unit).
Stan
|
The left pushbutton goes through pin 5 of JP302.
The other three pushbuttons all go through pin 6 of JP302, so that is a good place to check.
Those three also require R211, a tiny resistor on the bottom of the main board very near several solder joints - you can check to see if it became dislodged.
The encoder rotations are independent, and go through pins 1-4 of JP302.? The most common problem with them is soldering as others have mentioned.
?
|
Hi Dan,
since several buttons are not working, I would start by checking carefully the connections between the control board and the main board. Measure for continuity and/or reflow the connector joints. Also check that the connectors don't have a short between them somewhere. Given the diagnostics, it seems rather unlikely that any individual encoder or button is the issue.
73, Peter
|
Tony,
?
I'm going for continuous mode as I feel that will keep the harmonics stable and predictable.? Ideal would be to choose regulators designed for the load current (I see about 130ma of 3.3v and 100ma of 5v with backlight on) so it stays in continuous mode anyway.
?
For a copper shield, cheapest and easiest seems to be a few L shaped header pins stuck in the board around the perimeter, the bend pointing toward the middle to hold up the "roof" of the copper foil.? Lightly tack the foil with solder around the edges where it meets the board.
?
For me at least, trying to keep 10m birdie free seems a fools errand.? I wouldn't use 10m much, and one birdie across a huge 2mhz wide band doesn't seem a burden.? May not hear it anyway if the supplies are well filtered and shielded.? If it becomes part of a real product, could sync it with a counter-timer pin from the processor which can jump around to avoid any birdies.? If I were to sync the SMPS to an oscillator, I'd shoot for 1.72mhz.
?
I'll just put the 3.3v and 5v SMPS's on this little board, keep it simple,? A second board at the coaxial power connector on the back panel.? Remove the board mount coaxial connector, replace it with a panel mount, have that connector also hold a PC board that has an LM2940CT-12 regulator.? Regulator tab has a bolt through the back panel to sink any heat.? Would be better to have it after Hans' power switch, but that's a more serious board hack (though still easy enough, just cut the short trace to the tab/drain of Q103 and insert the regulator between those two nodes).? Quiescent current of the LM2940CT-12 is single digit mA.
?
Jerry, KE7ER
?
|
Gwen. Yep Jerry is having fun.?
?
To the OM in Belgium, watch a video or two on qrp labs. Small enterprise, owner is from UK, does global commerce and shipping quite well. Sells them on all continents. So if you want the rig, check it out. Email the shop direct on any questions you might have. Or maybe someone else on list from Belgium.?
?
73 Curt wb8yyy?
?
|
Re: QMX #008 Power Supply Search
What about the 9v Bioenno battery.....should have enough safety margin to go directly to the QMX ???
|
I just checked those URLs, and they decidedly do not go to anything resembling a QDX.
-=-=-=-=-=-=-=-=- 73, Gwen, NG3P
|
Re: Wanted: low power 50 MHz. linear amplifier
Thanks Nate, very helpful.
|
Re: Help learning QMX CW filter settings?
?
p.s. ?
re... Not hearing some dx stations
?
This will become a moot point if you are the "activator" (calling CQ)
Now remember RIT is your friend?
?
BUT IF "tuning" while running "hunt es pounce" this is a BIG ISSUE... you will jump over (not hear) stations calling CQ
So CW BW must be wider than tuning rate
?
n3aaz
|
Hi Dan.?
I had a solder blob that hit the body of the encoder once.
Run a DVM check and make sure that one of the pins? is not hitting or have a solder blob to?
the metal body of the encoder. Also don't over-tighten the encoder nut.?
George
K3GK
|
Jerry,
There are a lot of factors at play in the simulation. I noticed efficiency varies with switching frequency, load current, input voltage, output voltage, burst mode vs continuous mode…. I settled on 5V @ 1A for the runs using the example circuit from ADI. I substituted real inductor and capacitor models for the critical output components to get a more real-world result than the ideal components in their example design. I’ve yet to play with burst vs continuous mode. More experiments need to be run but overall, I’ll be happy if I can stay above 85% efficiency over all the various factors. That’s a heck of a lot of runs though, especially considering how critical the selection of the inductor and output capacitor(s) are.
I thought it would be a good idea to a) minimize the number of interfering harmonics, ?b) limit the offending harmonics to the highest possible frequencies for ease of filtering, and c) avoid the need for an external oscillator at a specific frequency. The frequency range of 2.29-2.33 is easy to stay within, even though the switching frequency changes slightly as I vary other parameters. As you noted, other frequencies have more harmonics in lower frequency bands and staying locked to a specific “magic” frequency would require a stable external oscillator, maybe even a TCXO. The latter seems to be overkill for a power supply.
I also agree it’s important to place a copper shield around the circuit to minimize radiated emissions. No one likes birdies lol. The much more expensive silent switcher modules do a great job in this regard but geez, they’re expensive!
Tony AC9QY
Tony,
?
Curious that your LT8609 simulation shows it to be more? efficient operating from 15v than from 12v.
A smaller Vout/Vin ratio means shorter conduction times.
I always assumed that would hurt efficiency.
?
?
A few hits in 6m are inevitable, that band is 4mhz wide.
Most of us could live with one hit in the 30m band,
especially if we can fix it to an unused spot, or move it somehow.
?
Here's a scan of harmonics from 1.66 to 1.74, looks fine to me:
?
Switching at 1.66 MHz ? ? ? ? 28.22 ? ? 51.46 53.12
Switching at 1.67 MHz ? ? ? ? 28.39 ? ? 50.1 51.77 53.44?
Switching at 1.68 MHz ? ? ? ? 28.56 ? ? 50.4 52.08 53.76
Switching at 1.69 MHz? ? ? 10.14 ? 28.73? ?50.7 52.39
Switching at 1.70 MHz ? ? ? ? 28.9 ? ? ?51.0 52.7
Switching at 1.71 MHz ? ? ? ? 29.07 ? ? 51.3 53.01
Switching at 1.72 MHz ? ? ? ? 29.24 ? ? 51.6 53.32
Switching at 1.73 MHz ? ? ? ? 29.41 ? ? 50.17 51.9 53.63
Switching at 1.74 MHz ? ? ? ? 29.58 ? ? 50.46 52.2 53.94
?
The 6'th harmonic of 1.69mhz hits 30m:? ?6*1.69=10.14mhz.
The 5'th and 7'th are nowhere close:? ?5*1.69=8.45mhz,? 7*1.69=11.83mhz
?
It's the 17'th harmonic that walks through the 10m band.
?
The nearest neighbors start hitting a bunch of bands:
Switching at 1.65 MHz? ?18.15 21.45 28.05 29.7 51.15 52.8
Switching at 1.75 MHz? ? 3.5 7.0 14.0 21.0 28.0 50.75 52.5
?
Jerry, KE7ER
|
The encoders can be a bit tricky to get right with the way their leads are bent forward to reach the controls pcb. It can be a very tenuous connection.
?
One technique that I’ve used to repair a connection that looked good but had a hairline gap to the pcb was to solder a component cutoff to the hole in the pcb and then to the corresponding encoder lead.
|
Re: Help learning QMX CW filter settings?
?
CW filter bandwidth verses frequency tuning rate
?
Remember too that if CW bandwidth is more narrow than tuning rate you will jump over stations
?
for example...
CW BW = 100 Hz
and
Tuning = 14 ?04$ ?00x ?Hz
?
?"$" = one kHz QSY per click on Tune knob
?So you will not hear (jump over) several dx stations per click
?
72
John
N3AAZ
|
Tony,
?
Curious that your LT8609 simulation shows it to be more? efficient operating from 15v than from 12v.
A smaller Vout/Vin ratio means shorter conduction times.
I always assumed that would hurt efficiency.
?
?
A few hits in 6m are inevitable, that band is 4mhz wide.
Most of us could live with one hit in the 30m band,
especially if we can fix it to an unused spot, or move it somehow.
?
Here's a scan of harmonics from 1.66 to 1.74, looks fine to me:
?
Switching at 1.66 MHz ? ? ? ? 28.22 ? ? 51.46 53.12
Switching at 1.67 MHz ? ? ? ? 28.39 ? ? 50.1 51.77 53.44?
Switching at 1.68 MHz ? ? ? ? 28.56 ? ? 50.4 52.08 53.76
Switching at 1.69 MHz? ? ? 10.14 ? 28.73? ?50.7 52.39
Switching at 1.70 MHz ? ? ? ? 28.9 ? ? ?51.0 52.7
Switching at 1.71 MHz ? ? ? ? 29.07 ? ? 51.3 53.01
Switching at 1.72 MHz ? ? ? ? 29.24 ? ? 51.6 53.32
Switching at 1.73 MHz ? ? ? ? 29.41 ? ? 50.17 51.9 53.63
Switching at 1.74 MHz ? ? ? ? 29.58 ? ? 50.46 52.2 53.94
?
The 6'th harmonic of 1.69mhz hits 30m:? ?6*1.69=10.14mhz.
The 5'th and 7'th are nowhere close:? ?5*1.69=8.45mhz,? 7*1.69=11.83mhz
?
It's the 17'th harmonic that walks through the 10m band.
?
The nearest neighbors start hitting a bunch of bands:
Switching at 1.65 MHz? ?18.15 21.45 28.05 29.7 51.15 52.8
Switching at 1.75 MHz? ? 3.5 7.0 14.0 21.0 28.0 50.75 52.5
?
Jerry, KE7ER
|
Re: QCX Challenge & QCX WARC Party -> New support web page :-)
Ian, funny that you used that call, as I have been tracking the "POLARSTERN" using WSPR for the past year....
?
Small world...
--
Rick
W5NR
|
Re: Best QMX AGC settings?
On Sun, Jan 5, 2025 at 02:26 PM, Steve G4EDG wrote:
?
?
?For my QMX with firmware 027 the settings are as listed..
Why these settings...... I just liked the operation and performance..
Threshold S? 7
Slope dB per dB 60
Noise filter? 5
Hang time? 30
Smooth samples? 50
Recovery dB/s? 50
Sample blocks 2
S9 sounds like S9
AGC display ON
AGC dB per bar 3
?
--
Rick
W5NR
|
Jerry, ? Thanks for the python code. It looks like we have a number of good switcher frequencies: ? 2.15 MHz, 2.29 MHz?thru 2.33?MHz ? All of these have just two harmonic hits with both hits above 50 MHz. The sweet spot is most definitely?2.29 thru 2.33?MHz. It's great that there's a contiguous range since we can't control the switcher frequency that precisely without a separate oscillator. Further, all the 6 meter in-band hits are at least the 21st harmonic or higher making them easy to filter out, at least for conducted emissions. ? I also checked the extended range between 700 KHz and 1.8 MHz but there are no suitable lower frequencies. Finally, I changed the resolution from 10 KHz steps to 1 KHz steps to further detail the sweet-spot range of?2.29-2.33?MHz range but found no specific frequencies that are better or worse than with the 10 KHz steps. ? The LT8609 is spec'd to 2 MHz max so I'm a little wary of setting it to 2.3 MHz. However, it does work fine in simulation but I wonder if it'll work in real life. Efficiency in sims at 2.304 MHz with 5V out and a 1A load: ? 15V in = 92.7% 14V in = 93.2% 12V in = 86.6% ? The best thing about the LT switchers is no overshoot. They always produce a clean, fast, monotonically rising output voltage. Sim shows PowerGood rising at 360 us with a stable output voltage at 420 us. ? As for any devices from China, no thanks :-) ? Tony
Tony,
?
That Linear Tech stuff is good, we used a lot of it at work, especially the LT8610
which is very similar to your LT8609.? One very cool feature is the TR/SS pin,
which if below the feedback reference voltage can be used to set the output voltage
anywhere between 0 volts and the maximum you determined with the feedback resistors.
That may work fast enough to be useable as a modulator on the QMX (but do simulate first).
We can do about the same with a cheaper part plus a $0.10 op-amp used as an adder: Vfb=Vout*scaleA + Vadc*scaleB
?
Could drive the external sync using a divided down 25mhz from the TCXO.
If you can select one of two such divide values, you can avoid ever hearing a harmonic.
If you are Hans, you can drive the sync input from a processor counter-timer pin.
?
I started out assuming we needed to synchronize it somehow.
Decided to try instead to just keep it quiet and small and simple and cheap,
as most of us are happy with an SMPS for our 12v supply anyway, and this won't be much current.
You lose a bit of efficiency at 2.1 MHz, but keeps coils and caps especially small, harmonics few.
The LMR50410-Q1 is not exactly 2.1 MHz, datasheet says 1.8 to 2.4 MHz.
?
Here's python code to show where harmonics will hit the US ham bands
?
# Evaluate LMR50410-Q1 switching at 1.8 to 2.4mhz for harmonics in the ham bands
bp = ((1.8,2.0), ?(3.5,4), ?(5.330,5.407), ?(7.0,7.3), ?(10.1,10.15), ?(14.0,14.35), ?(18.068,18.168), ?(21.0,21.45), ?(24.89,24,99), ?(28,29.7), ?(50,54))
for fo in range(180,240,1): ? ?# 1.8mhz to 2.4mhz, steps of 0.01mhz
? print("Switching at", fo/100, "MHz")
? for n in range(60): ? ?# check each harmonic
? ? f = n * fo/100
? ? for b in bp:
? ? ? if f > (b[0]-0.01) and f < (b[1]+0.01): ?print(f)
?
The LMR50410-Q1 is $0.75, but there are similar switchers in a Sot23-6 that are far cheaper.
Mouser part 621-AP62201WU-7 is $0.116 each if you buy a reel, a Sot23-6 that runs at 740 KHz.
The Chinese MT2492 is available for $0.02 from if you buy a reel, a Sot23-6 that runs at 600 KHz.
A TPS54160 clone on is $0.26, can take an external sync.? (A real TI part on Mouser is $1.75)
The MAX42402 has external sync, is on Mouser at $0.87.
?
Lots of choices.?
The Sot23-6 package is relatively easy to work with, but there are a dozen different pinouts used.
?
Jerry, KE7ER
?
?
?
On Sat, Jan 4, 2025 at 01:15 PM, Tony Scaminaci wrote:
Jerry,
?
The LMR50410-Q1 runs at 2.1 MHz but somewhere around 2.2 MHz, switcher harmonics land only in one HF ham band (don’t know which one). Your chosen device looks pretty good though.
?
I’ve been looking at the LT8609 which offers more flexibility but at increased cost. That device allows the switching frequency to be adjusted to whatever exact frequency impacts the radio the least and I’m planning on using a pot for this purpose.
?
Tony
?
|
ON7DQ Luc
Peter HB9EBE
Gwen Patton
Robin Midgett
Rick Trommer