开云体育

Hi,

I recently completed my QCX 80 kit, followed the instructions to the letter and even measured the resistor/capacitor values prior to installing them. The kit worked first time.? Unfortunately I did not measure the inductor values prior to fitting in the PCB. Because of the s-p-r-e-a-d in characteristics with the inductor cores my QCX suffered a lower than expected output power. It still worked OK but it "niggled" at me that the PA did not meet expectations so I decided the only way to get peace-of-mind was to fix it.

Reading some of the past posts on this forum I see the lower than expected power is quite a common problem and the fix (also described in the build manual) is to remove a couple of turns from each of the inductors.? I have no doubt this works but I did not want to risk damaging the PCB with re-soldering a number of times during the cut-n-try operation. So having a reasonably well stocked junk box I took a spare set of T37-2 cores and wound the number of turns specified for the QCX 80 (L1 25 turns, L2 27 turns, L3 25 turns and L4 24 turns) but this time around I used the QCX as a signal source with a kitchen table lash-up to measure resonance and therfore the value of each inductor with sufficient accuracy so that they matched the values quoted in the build manual.? The result was an instant increase in output power from the much less than 2 Watts I started with to 5 Watts as measured on the QCX. Now my power output-v-supply Voltage curve is close to that shown on page 129 of the manual plus I only had to desolder/replace the inductors ounce. So very happy :-)

While measuring the inductors, removing turns and re-checking the values a number of times the thought struck me that since the QCX already has an in-built signal source it would only take a small number of additional components (and software) to make some sort "peak detector" such that the resonance of each PA stage inductor can be checked while building. This (in conjunction with some simple math and a pocket calculator) would permit the measurement of the inductor values prior to fitting in the kit and also add a usefull bit of test gear to boot. The only "critical" component is a capacitor of reasonable close tollerance to resonate the inductors with (I used 820pF +/- 2%? for resonances around 3 to 3.5 MHz) but the value of the capacitor or the frequency are not that critical so long as both are known in order to work out the value of inductance. It might add a few 10's of cents to the cost of the kit but perhaps well worth it. So, what do you think?

73,s

Des (M0AYF)

I see the lower than expected power is quite a common problem and the fix (also described in the build manual) is to remove a couple of turns from each of the inductors.? I have no doubt this works but I did not want to risk damaging the PCB with re-soldering a number of times during the cut-n-try operation.

Des,

Hans has suggested that all that is needed is to re-solder the coils onto the board, not get the wire through the hole again.
I've not done this with a QCX but often only one end need be removed, cutting the wire is not necessary until enough turns have been removed. One could even rewind a turn. The inductance of the unwound wire will not be significant at at HF.

I used the QCX as a signal source with a kitchen table lash-up to measure resonance and therfore the value of each inductor with sufficient accuracy so that they matched the values quoted in the build manual.

After the last round of posts about this I started trying methods of using the QRP Labs VFO to do this.
I attach one of the working schematics. Resonance indicated by peak reading. And an idea of a crude test rig.
I was stumped however because many will not have a junk box with 2% components in it.
I was wondering if the actual LPF capacitors could be used, they are already good spec, would any variation be compensated while testing?

Your idea of using the QCX VFO sounds good.
There are various web pages like which save brain damage from the maths.

73 Alan G4ZFQ

The result was an instant increase in output power from

the much less than 2 Watts I started with to 5 Watts as measured on the QCX. Now my power output-v-supply Voltage curve is close to that shown on page 129 of the manual plus I only had to desolder/replace the inductors ounce. So very happy :-)
While measuring the inductors, removing turns and re-checking the values a number of times the thought struck me that since the QCX already has an in-built signal source it would only take a small number of additional components (and software) to make some sort "peak detector" such that the resonance of each PA stage inductor can be checked while building. This (in conjunction with some simple math and a pocket calculator) would permit the measurement of the inductor values prior to fitting in the kit and also add a usefull bit of test gear to boot. The only "critical" component is a capacitor of reasonable close tollerance to resonate the inductors with (I used 820pF +/- 2%? for resonances around 3 to 3.5 MHz) but the value of the capacitor or the frequency are not that critical so long as both are known in order to work out the value of inductance. It might add a few 10's of cents to the cost of the kit but perhaps well worth it. So, what do you think?

Hi Alan,

Thanks for posting the link to the on-line frequency calculator, very handy. That simplifies things a little more and speeds the job up a bit.

Your schematic is very nearly the same as I used except my input side resistor (from QCX CLK1) is 4K3 in series with 100 pF and on the output side I used 10K just to lighten the loading on the tank cct as much as possible while still leaving enough signal to be able to find the peak. I used an old 6 MHz CRO as a peak indicator because it has a nice high input-z just like a DVM and also gives easy visual feedback. But I did not use a detector since the signal level (in my case) was a little low due to the lighter coupling I used.

I imagine one of the capacitors included in the kit would be suitable as a reference. As the kit frequency goes up the capacitor values get smaller but so long as a suitable resonance frequency is chosen it should not matter.? BTW your test rig was much neater than mine Hi.

73,s

Des (M0AYF)

Des and Alan
I have just subscribed to the group and I wonder?if?either of you can help an old timer who has?only returned to amateur radio in the last couple of years after 30 years of no activity.
In fact, the last transceiver I built was a Heathkit HW8 which still works?- ?Hi Hi.

I have just completed building the QCX 80 and it worked first time thanks to the excellent Assembly Guide Manual.
In my opinion, I also have a slightly lower power output?than expected.
Using the online power meter, I ?have 5.2 watts on the drains of Q1, Q2 and Q3 (with both 9.5 and 11.3 V power source) and the following measurement across L1:
L1/L2 Junction = 3.3 watts (11.3V)? and?2.3 watts (9.5V)
L1/power output BNC socket = 2.0 watts (11.3V) and 1.5 watts (9.5V)

I understand these power measurements are not necessarily accurate values but serve for the purpose of comparison.

I note the possible removal of turns from the inductances L1, L2, L3 and L4 but before embarking on this I would be interested to know which inductance did you find was the most critical and turns removed.
I am fully aware that inductance values?can vary according to core material, tightness of turns, closeness of turns etc but would greatly appreciate your input as a quick fix until I can set myself up with the inductance measuring circuit.
Kind regards and 73s

Robin
G4DNP

In my opinion, I also have a slightly lower power output?than expected.
Using the online power meter, I ?have 5.2 watts on the drains of Q1, Q2 and Q3 (with both 9.5 and 11.3 V power source) and the following measurement across L1:
L1/L2 Junction = 3.3 watts (11.3V)? and?2.3 watts (9.5V)
L1/power output BNC socket = 2.0 watts (11.3V) and 1.5 watts (9.5V)
I understand these power measurements are not necessarily accurate values but serve for the purpose of comparison.

Robin,

The only one that has any chance of giving some sort of accurate indication is the final one.
The others are not into 50 ohms so really mean nothing.

I note the possible removal of turns from the inductances L1, L2, L3 and L4 but before embarking on this I would be interested to know which inductance did you find was the most critical and turns removed.

I think each has a small effect and all may require adjustment to make a significant difference.
Something somebody said off-list makes me wonder what current does the CQX take, RX and TX? Although perhaps that may mean nothing.

73 Alan G4ZFQ

Hi Robin,

Sadly I dont think I can help on this one. When I trimmed the number of turns on the inductors (L1, L2, L3 and L4) I decided through lack of time to avoid the cut-n-try approach in favour of the measuring the inductors and adjusting turns until they matched the specified values in the manual. It was more labour intensive but I felt it had a better chance of success given my currently limited spare time.

I performed the measure, adjust and re-measure on each inductor and then re-fitted all the inductors into the board in one operation so I did not notice which (if any) of the inductors made the biggest change to the power output. What I can say is that my inductors started off with the specified number of turns but ended up being way-off due (I suspect) to the well documented and notoriously poor sample-to-sample tollerances of this type of core.

If it helps I can tell you how many turns I had to remove from each core...

L1 manual says 25 turns (had to remove 5 turns)
L2 manual says 27 turns (had to remove 4 turns)
L3 manual syas 25 turns (had to remove 5 turns same as L1)
L4 manual says 24 turns (had to remove 5 turns)

As an aside, I rummaged around the shack and found some old (very old, 20 years +) T37-2 cores and wound the specified number of turns as a check to satisfy curiosity as to the s-p-r-e-a-d in core characteristics with cores from different suppliers and of different ages. The result was that in almost all cases the final number of turns required for a given inductance is subject to "tweaking" due to tollerances in the core. Conclusion: they have not got any better over the years Hi.

Hope you enjoy the QCX and look forward to hearing you on-air.

73,s

Des (M0AYF)

When I trimmed the number of turns on the inductors (L1, L2, L3 and L4) I decided through lack of time to avoid the cut-n-try approach in favour of the measuring the inductors and adjusting turns until they matched the specified values in the manual. It was more labour intensive but I felt it had a better chance of success given my currently limited spare time.

Robin,

Although I do not have a QCX it looks possible just to remove one end of an inductor, unwind one turn and resolder to the top of the board. The unwound length will have only a very small inductance. Then test, remove another turn, or even rewind that turn.
That is the way I have adjusted inductors on similar boards.

L1 manual says 25 turns (had to remove 5 turns)

Of course if you have an inductance meter you can trust then use that. Obviously if you remove more turns than you need then it will appear to work but harmonic rejection may not be so great.

73 Alan G4ZFQ

Hi Des

Thank you for documenting your results.? I found them to be very useful.??

When a good friend recently sent me a Version 2 QCX 80 kit to build, and being aware of this discussion, I decided to use your results as a starting point.? Accordingly, I wound as follows:

L1: 20 turns
L2: 23 turns
L3: 20 turns
L4: 22 turns

(In the case of L4, I went for 22 turns as a compromise between your results and the details given in the manual!)?

Anyway, with a supply of 13.5 volts, I got just over 4 watts out on 80 m.? This power output was initially measured on a Welz SP220 power meter, and confirmed by measuring Vpeak across a 50 dummy load with a Kenwood CS-6040 oscilloscope.? ?I was very happy with this result, and made no further changes.

Thank you, Des, for your contribution.

Best wishes, Steve

?

Robin

Hans posted a video rather recently of tuning up a qcx. The essence is to first remove turns at the LPF end closest to the bs170 PA devices. About 5 turns in my case. Then when it seems to be played out, do same for the middle inductor. I did not vary the inductor closest to the antenna, but maybe I should have. I boosted my power from 3 watts to a bit over 4 with 13.8 volt bias. Remember our power meters aren't perfect, I did optimize into a dummy load. Patiently adjust your rig. Nice to see several now reported on 80m.

73 curt wb8yyy