I finished my QMX and it works. There were a few things that I learned on the way that I thought were worth reporting.
The TL;DR version is:
Do a fit check on the control and display board early.
Check where L507 is going to fit before putting in c514, 515, 517.
Mark the wires on the trifiliar T401, to make it easier to figure out which is which.
T507, I think it's easier to put the single wires in after the transformer is down.
I found what I think is an easier process for getting the power supply boards down.
I had trouble following the initial firmware flash instructions. After some trial and error, I wrote what I think is an easier to follow description of the process.
The full version.
Checked the board, it¡¯s Rev 4.
?
2.1 Board inspection. As I discovered at Step 2.27, a fit check between the control board and display board is a really good idea at this point. Ultimately, after I had put everything together I found they didnd¡¯t fit. I had to do a LOT of filing to get it to go together. It would have been much safer to do this on the bare board, rather than with the display installed. I filed off a lot more than just the ragged edges of the cut offs, I had to enlarge the display board hole and reduce the control board by quite a bit to get it to fit. This is all going to have to slide together later when putting it in the case, so it needs to be a loose fit. A set of small diamond files turned out to be very useful. Mine happens to be the 10 Piece Pittsburgh Needle File Set from Harbor Freight.
?
[Ignore this. This has been commented on elsewhere, that the USB-C connector doesn't have solder on the solder side.] Looks like the usb-c connector wasn¡¯t soldered at all. Maybe it was soldered where I couldn¡¯t see it, but there was no solder on the visible side of the board. My 2 mm soldering iron tip was way too big. I tried it with a 0.8 mm tip but couldn¡¯t heat things up enough to get it to melt solder. (It might just be that it is a cheap 0.8 mm tip that doesn¡¯t conduct heat from the element properly.) So I ended up going back to the 2 mm tip and fixing the solder bridges after.
?
2.7 My L507 did not fit as shown. That doesn¡¯t seem important until much later, when you¡¯re trying to install L508, at which point it¡¯s too late. My C514, 15, and 18 are fatter than shown, so there¡¯s no room for L507. Ir ended up getting pushed to the right, which produced other problems down the line. I don¡¯t know if it would have worked to leave longer leads, so L507 would be above the row of capacitors, or if that would have caused other problems. Maybe it would have been OK to let it stick up and run it horizontally.
?
2.9. I used the RWTST method. It is an 80-20 m version, so it probably would have worked the conventional way, but I did it this way for the practice. Some day I¡¯m going to re-wind the transformer on my low output 20-10 QDX.
?
On this one and on the trifiliar T401 I color the wires with a marker, so it¡¯s easy to figure out which wire is which. Even ¡°permanent¡± marker will rub off the varnish pretty quickly, so it¡¯s not enough just to mark a tiny bit, you have to mark a fairly long section to ensure that some of the color survives the handling. I used my drill to wind the wires together. I check it with an ohmmeter when it¡¯s done, just to be extra sure, but it¡¯s faster knowing which is which.
?
2.10 When installing L401, I just left the tail going into the ¡°23t (end)¡± hole a little long, and ran it back to ¡°23t¡± hole.
?
2.12 Install Low Pass Filter toroids. As a consequence of the L507 fit in 2.7, the LPF toroids would not go down as shown. Mine are all over the place, especially the 20 meter ones. We¡¯ll see if this works. It is really tough to get 28 turns neatly onto L512. In fact, I couldn¡¯t do it, some of the turns overlap at the inner diameter of the toroid.
?
2.14 Wind and install transformer T507. Getting everything to go down at once looked really hard. I cut and bent the single turn wires to fit, removed insulation, and set them aside. It looks like a staple. The instructions don¡¯t say, but I wound the ten turns in the same top-to-bottom, counterclockwise direction as everything else. That means one wire isn¡¯t against the ferrite when it goes to the hole on the other side. It might be possible to make a neater installation of one of the coils went the other way, but I¡¯m not sure that would work, so I didn¡¯t do it. I wound the ten turn coils, put the transformer in place, then threaded the single turn conductors through the core and into their holes on the board. Then solder it all down.
?
2.17 Installing power supply boards. I think this is easier. Execute steps 1 through 5. Skip step 6 (solder right angle to main board). Execute step 7 (solder top side to edge connector.) Solder the four pins on the 5 V board, then the pin closest to the BNC on the 3.3 V, then remove the 5 V board and finish the 3.3 V. The other two pins are very hard to get to with the 5 V board installed, but the board needs to be in there to get the alignment right. Hence get one pin soldered to lock the position, then remove the other board to finish soldering. Remove PS boards (step 8) and solder bottom pins. Then put the whole assembly back in, and execute step 6, solder right angle to main board.
?
?
2.18, figure 5. Installing the cutoffs in the LCD. Put all the wires in, then lift up the assembly and make sure that all the wires stay behind. That verifies that they are going all the way through the holes, and not hung up on top. Any that lift up when you lift the board should be checked and worked until they stay on the bench as you lift the assembly.
?
2.21 Install 20 k trimmer. I filed the nubs down using the same diamond file I used to enlarge the hole in the display board.
?
2.22 step 4a. Remove controls board by pulling it out of the 4x2 socket. It needs to be out before you can perform step 2.23, install the rotary encoders.
?
2.23 step 4a remove nut and washer
step 5. the pins barely reach. Install the nut and tighten so that the pins stay in the holes while you solder them. I didn¡¯t need to extend the center pin as described in step 7, but it¡¯s a reach.
?
2.27 It didn¡¯t go together. There should be a fit check for the control board in the hole in the display board back in step 2.1.
?
2.32 The firmware flash process is not written in a linear fashion.
This is the step by step procedure I followed.
?
?
0. Verify that the contrast pot R305 is fully counterclockwise. Fully counterclockwise is maximum.
1. Download firmware. Extract the firmware file from the .zip file.
2. Get a power supply, preferably 7 V. Verify that it¡¯s center positive, and the right voltage.
3. Get a USB-C cable, connect one end to your computer and the other to the QMX. Connect the power. If nothing catches on fire, continue.
4. Long press the left rotary encoder. The QMX should come up in boot loader mode. The display on the QMX will remain blank. Look for a new disk on your computer with the QDX as a USB flash drive. It should be empty. In Windows this will be a new drive showing up under the ¡°This PC¡±, so open this up before starting, so you will see the new drive when it connects.
5. Drop the extracted firmware file onto the QMX virtual USB flash drive.
6. Power cycle the QMX.
7. The QMX should boot up. Now the display will be active. It will ask for the hardware version. Adjust the right rotary encoder to select the correct version of hardware, then press the left button.
8. Adjust the contrast on the display of the QDX so it is at the desired level.
9. It should be ready to run at this point. Do whatever you would normally do. I ran mine into my QRP labs dummy load and measured the output voltage on each band. Then hook it up to WSJT-X and see if you can communicate with the device. Then get on the air.
?
2.33 Installing the case
0. Check photograph 9. Remove the indicated screws. That¡¯s not a QMX, but it¡¯s close enough. Take the display board off the main and control board. Now proceed to photograph 1.
?
And when I got to the end I realized that I didn¡¯t pay attention to the orientation of the screw back in step 2.8, and the case won¡¯t go on if it¡¯s backward. It has to be head under the board, nut on the washer. 2.8 should point that out as being important.
?
I also couldn¡¯t get the case on without partially filing down the head of the screw that holds the 5 V regulator. It interfered with the rail, so I pulled out my needle file and knocked off the edge of it near the outside of the board.
?
?
?
?
?
Donald S Brant Jr