|
I just finished my QMX kit and it is working fine. Hans, thank you very much for this great kit and great documentation. I am eagerly looking forward to the formware development that will make use of ALL the HW capabilities, that you included into the design! These were my observations, when building the kit for 12V:
- Read the assembly manual carefully
- Wear protective glasses when cutting off wires with a wire cutter. This is particularly true for the pins of the connectors, that are really hard. They will fly off with high energy and I got a bloody finger, when holding it over the pin while cutting, to avoid? having the cut-off pin fly somewhere...
- The QMX will not forgive any assembly errors: I accidentally soldered a capacitor into the wrong place. Desoldering was impossible without destroying the capacitor. Luckily I still had in my junk box a 390pF C0G 100V SMD 0805 capacitor, that I could use as a replacement. Generally SMD 0805 capacitors fit nicely into the places targeted for the through-hole-capacitors. Just fill the hole with solder and then solder the capacitor.
- Carefully file the PCBs, particularly the controls PCB and display PCB, as during assembly into the enclosure, the controls PCB must move through the display PCB cutout.
- Look for the short between Q103 and Q104 on the buck converter board: My board HAD THE SHORT. I was able to fix it, by heating the drain of Q103 and moving it slightly away from Q104. To avoid tension of the gate and source connections to the PCB, I afterwards heated those up.
- When you put in the BNC connector X502 make sure that the flattened part sits on the PCB without a gap. I had a small gap, which did not allow, to put it into the enclosure. Luckily I had only soldered the centre pin and one of the ground pins, which finally allowed me, to heat the two up jointly and put the connector flush onto the PCB.
- Scrape off the enamel from the toroid wires with a sharp knife.In my kit the 0.6mm wires had a super temperature hardened enamel, that could not be burned off with the soldering iron.
- For the SWR bridge transformer BN43-1502 make sure, that you have ALL wire ends scraped off and the straight wires bent, before you insert it and start to solder. I completed first the 10 turns windings and then soldered the transformer onto the PCB. This will leave you almost NO space to fit in the straight wires and bent them 90 degrees. This took me a LOT of time.
- While soldering in the connectors, that will look through the enclosure, make enclosure assembly checks.
- The pins of the female connectors of the power supply boards were too long and had to be cut off. See #2 when doing that.
- I did the first smoke test on 40m with only the 40m LP filter L512/L508 populated:
- Lab power supply set to 7V and limited to 250mA.
- 0 mA current which went up after I pressed the left encoder.
- Connected to PC running Win 10 and the PC detected a new drive. Firmware copied onto the drive and the QMX restarted,
- Display lit up and showed 7,074,00 --> SUCCESS!
- Current consumption was 140mA
- PC now recognized a new device "QMX transceiver" and set it up.
- Also an additional COM port was created. The COM port number depends on?
- the history of COM ports already established on that PC
- the USB port of the PC, that is used for the connection --> if you connect the same QMX to another port of your PC it will most probably get a different COM port number. You will have to change that in WSJT-X / JTDX
- Connected an earphone to the QMX.
- Connected a TinySA as signal generator with 20+30dB attenatuators to the antena connector of the QMX and created a 7075kHz signal with the TinySA --> heard (too) loud a 1kHz tone in the earphones.
- Started WSJT-X, configured it for "Kenwood TS-480" and the correct COM-Port.?
- Pressed the "Test CAT" button and after several seconds (took longer than for other transceivers) the button went green.
- Changed the "Audio" settings in WSJT-X to "QMX Transceiver"
- Closed the config dialog and pressed "Tune" button in WSJT-X --> 30.8dBm output measured with my TinySA after 20dB+30dB attenuators
- Raised the supply power to 12V (after switching the QMX off)
- Same test, now with 12V resulted in 36.3dBm output power (4.3W, which were also shown by my ATU-100 tuner)
- Connected the QMX to an antenna:
- listened to some SSB stations on 40m after having found the place to switch to LSB
- switched back to USB and ran one FT8 QSO on 40m and one on 60m - GREAT
- After this success I wound and populated the remaining 4 LP filter toroids.
- Tested 30m and 20m with WSJT-X and they worked fine.
- Tested 80m: significant raise in noise in the attached earphone, when I touched the "Tune" encoder (no knob mounted)?
- switched QMX off
- looked at potential touch between encoder housing and other parts
- identified cathode of D513 as potential touch point
- moved it slightly to the left
- switched QMX on again
- no touch sensitivity of the encoder any longer
- "Tune" in WSJT-X resulted in 31.8dBm @ 7V and 36.3dBm @ 12V
I think overall it took me 10-12 hours to assemble the QMX kit and it was the most challenging build so far. The QMX is IMPRESSIVELY small compared e.g. to my DL2MAN-uSDX's and the uSDX-SOTA by WB2CBA and it has much more potential. Thanks to Hans for developing this masterpice of electronics, software and mechanics. Have fun building your kits! -- 73, DB2OO -Joerg
|
That was a joy to read, thanks Joerg
73
Rod
G0VKX
once was G8FJN
|
Somehow the Q103/Q104 short issue must become common knowledge for QMX builders. I found out only
after the smoke came out on power-up. (D101) I certainly hope that the rest of the radio survived this short -when
I replace Q103 later this week. I considered that It would likely destroy Q103 to heat it enough to remove it's fully
flowed tab. I was able to fit a regular 1N4148 in D101's space. Fingers crossed! Bob-AB5N
|
Joerg,
Thank you for the excellent assembly ?documentation and thanks also to the many other folks that have found issues and solutions in building the QRP-Lab kits.
To those that have yet built a kit, of any manufacture, the message is clear, ?read the manual, read the messages on a Social Media site such as this one AND follow instructions one step at a time…… check your every step.
Great job !
John KK4ITX? ?Each mistake is a learning opportunity.?
toggle quoted message
Show quoted text
On Aug 15, 2023, at 9:42 AM, Robert Nagy <7000mic@...> wrote:
?Somehow the Q103/Q104 short issue must become common knowledge for QMX builders. I found out only
after the smoke came out on power-up. (D101) I certainly hope that the rest of the radio survived this short -when
I replace Q103 later this week. I considered that It would likely destroy Q103 to heat it enough to remove it's fully
flowed tab. I was able to fit a regular 1N4148 in D101's space. Fingers crossed! Bob-AB5N
|
Hi,
I am very tempted to buy and build a QMX.
I read about it and I find this group very valuable and some post have excellent advices that should be fixed into a pdf as an addendum to the building instructions.
Or as a separate file easy to find somewhere in the File Section.
Cheers,
Thanks,
Adrian
|
Thanks for posting this Joerg. I saved your notes, so I can attach them to the manual to refer to when I build my QMX.
|
Can only agrea to what wrote Joerg DB2OO, I had also this Q103/Q104 short issue if you dont have a current limited PS I guess you fry the zener diodes and then the rest of the board!? you must unsolder Q103 shift is as far as possible to get rid of this.
Could be worth a warning to kit builders
may be in future would be worth putting SMD Caps to replace all through hole caps this would give a bit more space for he toroids
73s Hubert F6DUK
|
Can only agrea to what wrote Joerg DB2OO, I had also this Q103/Q104 short issue if you dont have a current limited PS I guess you fry the zener diodes and then the rest of the board!? you must unsolder Q103 shift is as far as possible to get rid of this.
If you have the Q103/Q104 short issue, you fry the 1N4148 diode D101. The rest of the radio (including the zeners) is not affected.?? ? Could be worth a warning to kit builders
Yes, I already added this to the assembly manual in the latest edit, this morning; and I have also put it on my new troubleshooting page??? ? may be in future would be worth putting SMD Caps to replace all through hole caps this would give a bit more space for he toroids
The reason for through-hole capacitors is that it allows a selection of different QMX versions, depending only on through-hole capacitors. Just like other QRP Labs transceivers (QCX+, QCX-mini, QDX) the band selection is dependent on the filter components. For example, there will soon be a high-bands QMX covering 10-20m. The main board doesn't change, only the toroids and capacitors change. No way could I afford the complexity and expense of manufacturing multiple versions of QMX boards, each with their own selection of SMD capacitors; without a large increase in price it would be a logistical and economic catastrophe!
|
Good write-up, Joerg! Spot on all! I tuned in Q103/Q104 magic smoke as I was getting close to putting them in and checked with DVM for the dreaded short and found none.
Mine is finally in the box and am in ops mode now! Whole new set of things to learn. Oh, yes! You may want to review proper shut-down of PuTTY affecting the operation. There was something about CAT control if one did only disconnect from the computer or just turn the QMX off. Well, you MUST exit PuTTY properly or the radio will NOT transmit but only a few milliwatts!
--
73
Karl
KI4ZUQ
|
Finished, in the box and working!
--
73
Karl
KI4ZUQ
|
Hi,
my QMX is running well at +12V. Had to repair the Q103/Q104 short issue first by moving a bit the bigger one. Fortunately I measured the short before using it.
I went the assembly my way:
- assembling the LCD and command PCB first,
- placing both SMPSs in,
- soldering the DC connector,
- using a +7V PSU current limited at 250mA for the first switch on,
- loading the FW,
Everything went well. Then I assembled the rest. Nice beast.
And yes Hans, if you don't know already, there is really really no place left on that board. What were you thinking? Haha.
73 Bojan S53DZ
|
Everything went well. Then I assembled the rest. Nice beast.
Awesome ? And yes Hans, if you don't know already, there is really really no place left on that board. What were you thinking? Haha.
Well it was all about romance. Marriage. Of QDX and QCX-mini. For whatever reason, I was determined to make it all fit in the same enclosure as QCX-mini. And use through-hole capacitors in the filters so we can have different band-coverage versions, like QDX, just by choice of supplied components. The mechanical design was going to be more like QCX-mini and the electronic design more like QDX (plus extras). Ask my XYL, she says I'm the most obstinate thing she's ever known (and as often as not, wishes she didn't). Of course, I disagree. But then again I did absolutely refuse to back down from my fixed PCB size, to fit in the QCX-mini enclosure. And when there still wasn't space in the building... well I just put in another floor (the SMPS floor!).?
|
Hi,
these are my first QMX HW test measurements.
73 Bojan S53DZ
|
|
|
Hi all - mine is also complete and working extremely well. Amazing kit and service from Hans so a huge thank you there. Built for 12V, 5W out on all bands except 20m where it's 4W. I made lots of mistakes along the way and the board survived my fixes so that's saying something.
WSJT-X frequency cal tool shows a df of less than 0.5 Hz - how is it so good??
Can't wait for the high bands version and further updates.
73 Steve MW0KST
|
Hi Steve
Hi all - mine is also complete and working extremely well. Amazing kit and service from Hans so a huge thank you there. Built for 12V, 5W out on all bands except 20m where it's 4W. I made lots of mistakes along the way and the board survived my fixes so that's saying something.
Thanks Steve! Nice to hear of your success.?? ? WSJT-X frequency cal tool shows a df of less than 0.5 Hz - how is it so good??
It's what you get when we put in a 25 MHz TCXO frequency reference which costs me almost as much as the CPU!
73 Hans G0UPL
|
Hi,
My +12V QMX draws from 89 mA to 95 mA on RX from 80 m to 20 m respectively.?
Without the audio signal (phones) present.
I think Hans has pushed the parameter a little bit once again (Hi) by saying the current was 80 mA.
Or, perhaps I am loosing that 15 mA somewhere.
For the High band QMX it will be even a bit more.
73 Bojan S53DZ
|
Hello Bojan
My measurement here is indeed about 80 mA without the backlight, and 87 mA when the display backlight is on.?
When?testing QMXs I did once see one which had an anomalous current consumption 10mA higher than I would expect. I traced this to a shorted GPIO pin (I forget which) due to a solder whisker; I think it was across the transmit/receive switch transistor gate if I recall?correctly (C508); so the processor was driving that output high into ground and maxing out at a certain number of mA, which by the time it was reduced (due to the buck converter) appeared as a 10mA increment on the 12V rail. Of course?I would have got as far as testing the receiver and it would fail at that point, because the signal never gets through the Tx/Rx switch. But I detected the problem ahead of that, just on the basis of the increased current consumption.?
I see very little variation between units so a 10mA increase is notable.?
So one of:
1) You could have a short somewhere? 2) Your current consumption measurement is inaccurate
3) MY current consumption measurement is inaccurate
4) Both 2) and 3)?
Hi,
My +12V QMX draws from 89 mA to 95 mA on RX from 80 m to 20 m respectively.?
Without the audio signal (phones) present.
I think Hans has pushed the parameter a little bit once again (Hi) by saying the current was 80 mA.
Or, perhaps I am loosing that 15 mA somewhere.
For the High band QMX it will be even a bit more.
73 Bojan S53DZ
|
Hi Hans,
Thanks for the tip. I will check for any small ball sticking illegally on my board...
My DC current measurements for this particular case are accurate to 1 mA.
So, now without the backlite I have 86 mA up to 92 mA from 80m to 20m.
I have changed the R306 previously from 33R to 200R.
73 Bojan S53DZ
|
Bob Nagy AB5N
KK4ITX John
F Adrian
GM0HUU Steve
Karl