What is the serial number of my QCX kit?
What is the tuning range of the VFO?
What is the current consumption of QCX?
Is there a version enforcing Japanese band limits?
When will my kit be sent, when will it arrive, etc etc.?
Will I get the latest firmware version when my kit ships?
Is there a mailing list for the QCX kit?
What technical support is available?
Is an enclosure available or planned for the QCX kit?
What are the dimensions of the PCB?
Are other bands possible?
Is a multi-band QCX possible?
Why do I have an MPS2907 transistor in my kit, but no MPS751?
The RF power output is lower than expected
I have 47uH inductors for L5 and L6, not 100uH...
I have a 10K or 22K trimmer pot, not a 100K trimmer pot...
My trimmer resistors are marked "203" not "503"...
Low Pass Filter
QCX kits are normally sent within 1 business day, as long as everything is in stock. If there is any shortage, it will be noted on the ordering page.
We have received a lot of questions regarding orders that are delayed, from people whose order includes the QCX enclosure, and who missed the notice on the ordering page about the enclosure availability. Please refer to http://shop.qrp-labs.com/enclosures/qcxcase to see the current estimate for your shipment.
For other questions regarding shipping, please refer to the main FAQ page.
QCX serial numbers were included in shipment notifications during the period that there was an order backlog. Since then, they are not in the shipment notification emails. You can find your serial number in the list of Order IDs vs Serial Number on the QCX Serial number page.
The QCX kit uses synthesized tuning using an Si5351A kit. Tuning covers the entire amateur band (not just the CW section). Technically, there is a limit imposed by the way the Si5351A chip is configured to produce quadrature (90-degrees phase shift) on two of its output pins. The tuning ranges are as follows:
80/60/40m: 3.2 MHz to 9.1 MHz
30m: 4.8 MHz to 13.7 MHz
20m: 6.3 MHz to 18.0 MHz
17m: 8.4 MHz to 24.0 MHz
15m: 10.1 MHz to 28.8 MHz
12m: 11.9 MHz to 33.8 MHz
10m: 14.5 MHz to 41.1 MHz
Practically speaking, the Band Pass Filter attenuates the signals outside the amateur radio bands. The -3dB points bandwidth of the BPF is sufficient to cover the entire amateur band which you build the kit for. There are example measurements of the BPF response in section 7.7 of the manual.
The current consumption varies with power supply voltage. There is a graph showing my measured values, in section 7.2 of the manual.
Yes. In Japan the regulations require that amateur radio transceiver CANNOT transmit outside the amateur band allocations. We have a version of the firmware which prevents transmission outside the Japanese band limits. Please email us, providing you order ID, if you require this special Japanese firmware version. The enforced band limits are:
3500000 - 3575000
3599000 - 3612000
3680000 - 3687000
3702000 - 3716000
3745000 - 3770000
3791000 - 3805000
7000000 - 7200000
10100000 - 10150000
14000000 - 14350000
18068000 - 18168000
21000000 - 21450000
24890000 - 24990000
28000000 - 29700000
Yes! All kits always use the latest firmware version available at the date/time of shipping.
If you think you need technical support, the following options are available - in order of preference: I mean, the first one in the list is to be done first! Do NOT start with 3...
- The vast majority of problems or questions are already answered by information in the documentation! The old adage "RTFM" generally applies. Are you sure you read the documentation thoroughly? Check again - also check elsewhere on this FAQ page and on the QCX kit page to see if the information you need is written there.
- Ask on the QRP Labs distribution list https://groups.io/g/QRPLabs. Not a member? You should be! Please join us!
- Study the QCX Troubleshooting page to learn how to debug your QCX
- View the QCX Troubleshooting video to see how to use simple inexpensive test equipment, or the built in test equipment, for debugging
- Email QRP Labs, CLICK HERE for details. We can decide how best to answer your question - by email, telephone, Skype, etc. In the very last resort you can send us the kit and we'll get it working for you (free of charge) and return it.
Yes! There is a German designed and produced aluminium enclosure with all accessories included. See this page http://qrp-labs.com/qcxcase .
The PCB size is 4.0 x 3.2 inches (101.6mm x 81.3mm). The 3.5mm jack connectors protrude on the left and right sides, and the BNC connector is the largest item sticking out on the right side at the rear (please refer to the diagrams in the manual). When the LCD module is plugged in, the top (front face) of the LCD is about 20.5mm above the surface of the PCB. The 12mm hex spacers on the lower side, plus the PCB thickness, bring the total thickness to 34mm. The two knobs stick out above this.
The transceiver should work on 10, 12 and 15m bands and kit versions for these bands may be made available in future, once we have had time to test operation and performance on these bands.
The transceiver as it stands, cannot work on 160m or LF bands. The reason for this is that the Si5351A Synthesiser chip can only produce a quadrature output (90-degree phase difference) on two of its outputs, down to a minimum frequency of 3.2MHz. Therefore the system can work for 80m (3.5MHz), but not for 160m (1.8MHz) band and below. Modifications would be required to generate the quadrature LO, probably involving a divide-by-4 circuit (e.g. with 74HC74 chip).
There are three circuits in the QCX which have band-dependent components:
- Low Pass Filter (LPF) which is for transmitter output harmonic filtering (inductors L1, L2, L3 and associated capacitors).
- Band Pass Filter (BPF) which filters unwanted signals and prevents them reaching the receiver input mixer (transformer T1, long secondary winding and capacitors).
- Class-E Power Amplifier resonant circuit (L4 and associated capacitor).
The assembly manual provides tables of component values per band, for each of these circuit blocks. Changing the band requires changing all three circuit blocks. In a multi-band QCX you would need to arrange switching between copies of these circuits, one copy per band. It is not impossible. But neither is it trivial...
The first 500 kits manufactured used the MPS2907 PNP transistor for key shaping. Some constructors reported that the transistor burnt out. It has a 600mA current rating. For the subsequent kit batches, the MPS751 transistor was used instead. The MPS751 is a PNP transistor with the same pinout as the MPS2907 but 2A current rating. So you should use the MPS751 for Q6, if the MPS751 is supplied instead of MPS2907 in your kit. The assembly manual does now refer to the MPS751, but you may have an older version of the assembly manual; or there may be some references still to the MPS2907.
Actually, the 600mA rating of the MPS2907 should never have been exceeded anyway. But some people reported a power amplifier instability problem, seemingly possible on 17m and 20m versions; I think that this is what caused the current to exceed 600mA and kill the MPS2907 in some cases. You can see a modification that resolves the power amplifier instability problem here: http://qrp-labs.com/qcx/qcxmods.html#pa . This simple modification involves adding a 10K resistor. It is recommended to carry out this modification even if you are using the MPS751. A spare 10K resistor is supplied in the kit from serial number 1500 and above.
First of all, make sure you understand that power output is related to supply voltage. You can check the curves shown in the final pages of the manual, to find out what to expect for your supply voltage.
Next, understand that a small increase in RF power is an even smaller increase in decibel terms, equating to a small increase in perceived signal strength at the other end. For example, an increase from 2.5W to 5W (Twice as much power in watts) is only a 3dB increased when expressed in decibels; this equates to only half an S-point of received signal strength at the other end of your QSO. Therefore, you have to ask yourself, if your power is a bit lower than expected, how much WORTH IT is it to fiddle with it, and perhaps risk breaking something, when it will only make a small difference?
Note also that it should be very easy to get the specified power output on the 40m band QCX kits. It gets progressively harder to get the target power output, as you go higher in frequency. And 80m is harder too.
But if you really are concerned that the power is low for your supply voltage, and if you think it is low enough that you really want to fix it, then this is what to look at.
1) First check the inductor wires are soldered properly (L1, L2, L3 and L4). If there is NO power output then this is very often due to no electrical connection at one of the wire enamel soldering joints. Failure to scrape away or burn away the enamel is a common fault.
2) Next check that the LPF capacitors are properly soldered too. This sounds silly. But be aware that one side of each of the capacitors C25, C26, C27, C28 and C30 is grounded. In this area under the LPF circuit, there is extensive ground-plane on both sides of the board. Even though, in accordance with good practice PCB layout, these ground pads are connected to the groundplane via "thermals", which means the capacitor pad isn't part of the groundplane directly, it is instead connected via thin tracks at 4 equal points (like North, South, East, West)... the fact there are 4 on top and 4 on the bottom still means a lot of heat transfer into the groundplane. So these ground joints at these capacitors do typically require more soldering iron heat, to solder properly.
3) Component tolerances are a possible cause of low RF output. Any variation in component value from the ideal theoretical value predicted by the Low Pass Filter design, results in a change to the filter passband characteristic. When the filter "cut-off" frequency becomes lower, it can soon start causing significant attenuation at the operating frequency. So, large differences between theoretical component value and actual component value, are likely to cause problems.
Of particular note, are the inductance values. The number of turns on a toroid is predicted theoretically. The model assumes that the turns are evenly spaced around the toroid. In practice this is not the case. We always have a larger gap between the ends of the winding. This results in the turns of the toroid being closer together than the model predicts; and this results in a higher inductance than we planned. You can minimise this by spreading out the turns as much as possible.
Another useful trick is to reduce the number of turns by 1 or 2 compared to what is prescribed in the manual. Fewer turns lowers the inductance. This can help counteract higher than intended inductance, caused by the turns being closer than the model predicts, or caused by the permeability of the powdered iron core being higher than it should be.
In the end, tweaking the inductance by stretching out and squeezing together the turns, can significantly alter the power output of the transmitter. Stretching out the turns around the toroid core, causes the inductance to be LOWER. Squeezing the turns together causes HIGHER inductance. So monitor the output RF and try stretching/squeezing the toroids one at a time. L1, L2, and L3 have the most effect. Usually it is not necessary to play with L4.
My method is to try stetching out the turns on L1 first. If you find that when fully spaced out, the power is at maximum (compared to when the turns are squeezed together a bit) then this is an indication that you have too many turns on there; you should try removing one turn then try again. Then repeat that with L2 and L3.
Often I find that L2 needs it turns SQUEEZED to increase the inductance, while L1 and L3 benefit from spreading out the turns (to reduce the inductance). This is a bit counter-intuitive, particularly given everything I said above, about inductance being too high (and so remove a turn or two). BUT, often does seem to be the case.
4) There have been a handful of occasions, nearly all were 20m kits, where the 390pF capacitors appeared to be lossy and replacing those with another NP0/C0G capacitor fixed the low power issue. A tell-tale symptom of bad capacitors is, if you key-down and the power appears to rise slightly after half a second or one second. This indicates the capacitors are lossy and are heating up, and in doing so their value is changing which alters the characteristics of the filter passband.
The inductors were supplied as standard in all kits from the second batch onwards (about serial number 500 and above). At the time of starting manufacture of the second batch of kits, it was very early days for the QCX transceiver; some people had noted that the microcontroller did not start up reliably every time and required several on/off switching cycles to get it to boot up properly (show text on the display). One possible solution for this was to use a smaller inductor L6 so that capacitor C47 charges faster at power-up. However this has NOT been found to be the preferred solution to this problem (it does not solve it in all cases); the proper modification is listed here. The 47uH inductors are still being supplied so please proceed and use them, do not worry about 47uH vs 100uH.
This is fine. The first batches were supplied with a 100K trimmer. Some people mentioned that a lower value is more typically used. Subsequent kit batches use 10K or 22K. It makes no differene at all, this component value is very non-criticial. It sets the LCD contrast with the potential divider formed by the trimmer - but the resistance value does not have a significant effect (within reason).
Some kits produced in the 4th/5th batches of kits have incorrect trimmer resistors supplied (serial numbers 1500-2500 approx, since there are 500 kits per batch). It appears that some tubes of 20K resistors (label "203") were mistakenly mixed in with the correct 50K resistors (label "503"). This mistake by the suppliers was not Most kits will have the correct 50K pair of resistors; however some may have one, or both supplied as 20K.
If you email us, we can send you the correct "503" resistors. However, it will take some weeks for the postal services to get this through to you (depending on your location, and many other factors out of anybody's control). The faster and better alternative is to use this simple modification (click here).
You do not need a PCB for the Low Pass Filter (LPF). The LPF components are installed on the main QCX PCB. Just follow the QCX assembly manual and everything will be fine. The LPF kit bag is the usual QRP Labs LPF kit but with the PCB removed, since it is not necessary. We recycle the PCB into LPF kits for non-QCX use.
Please see above. No LPF PCB is needed for assembly of the QCX kit. BUT, in some cases for logistical reasons, the LPF kit PCB originally in the LPF kit bag, was not removed. So you have it in your kit bag. You can just ignore this PCB! Throw it away, practice on it, buy another and make it into a pair of earrings... whatever you like!