¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Topics

alternate build procedure for Mini? #qcxmini

 

I am thinking of deviating from the QCX mini build instructions this time around.
My previous mini build did not go so well (/g/QRPLabs/message/73018).
This time around I'd like to measure the BPF and LPF in place before completing assembly.
The idea is to validate the filters while it is still possible to measure them in isolation and correct any errors before proceeding.
But being very inexperienced at RF measurement, I'd like to run it past the group first.
Here is the idea (BPF first):
  1. Complete step 3.1:? Install T1 (but don't solder leads shown in red below)
  2. Complete step 3.10:? Install C5 & C8
  3. Complete step 3.18:? Install C1
  4. Inject signal from NanoVNA at primary 1 (shown below), measure at secondary 2 (shown below)
  5. Inject signal from NanoVNA at primary 1, measure at secondary 2


And for the LPF:
  1. Complete step 3.7:? install C25 & C26
  2. Complete step 3.8:? install C27 & C28
  3. Complete step 3.25:? install L2
  4. Complete step 3.26:? install L1 & L3
  5. Complete step 3.28:? install BNC connector
  6. Inject signal from NanoVNA at L3 (shown below) and measure at BNC connector


Please let me know your thoughts (be kind!).? Thanks!

73, Mike KK7ER

 

Mike,

I did that for the LPF's, see my blog

Note that I built an older version, and used manual V 1.05, so things may be different for your version.

I didn't bother for the BPF since this has quite a wide tuning range, and has the built-in adjustment procedure.

Even when you get the LPF's right this way, you will need some fine tuning, since the output class-E finals are NOT 50 Ohm, so the measurement/adjustment? with the NanoVNA will not give you the optimal point, but it should at least reduce the guesswork.

GL with the build
73
Luc ON7DQ
 

¿ªÔÆÌåÓý

Mike,

This is ham radio. Build the kit any way you desire and for any reason. Do tell us of your success or failure please. ?I teach all new hams that we learn more from our failures (and others failures) than we ever do from our successes. ?And we remember our failures far longer. Why do we still repeat some of them? ?Ha!

Dave K8WPE

David J. Wilcox¡¯s iPad

On Nov 13, 2022, at 7:08 PM, Mike KK7ER <groupio@...> wrote:

?I am thinking of deviating from the QCX mini build instructions this time around.
My previous mini build did not go so well (/g/QRPLabs/message/73018).
This time around I'd like to measure the BPF and LPF in place before completing assembly.
The idea is to validate the filters while it is still possible to measure them in isolation and correct any errors before proceeding.
But being very inexperienced at RF measurement, I'd like to run it past the group first.
Here is the idea (BPF first):
  1. Complete step 3.1:? Install T1 (but don't solder leads shown in red below)
  2. Complete step 3.10:? Install C5 & C8
  3. Complete step 3.18:? Install C1
  4. Inject signal from NanoVNA at primary 1 (shown below), measure at secondary 2 (shown below)
  5. Inject signal from NanoVNA at primary 1, measure at secondary 2
<dummyfile.0.part>


And for the LPF:
  1. Complete step 3.7:? install C25 & C26
  2. Complete step 3.8:? install C27 & C28
  3. Complete step 3.25:? install L2
  4. Complete step 3.26:? install L1 & L3
  5. Complete step 3.28:? install BNC connector
  6. Inject signal from NanoVNA at L3 (shown below) and measure at BNC connector
<dummyfile.1.part>


Please let me know your thoughts (be kind!).? Thanks!

73, Mike KK7ER

 

Hello Mike

Just building a 40m QCX mini.? Fitted T1 this morning along with LPF components out of sequence to ensure I could check out the completed filter before doing anything else,?having experienced an issue on a previous project caused by a wrongly marked capacitor.? Sequence is whatever is convenient for the builder - just check you haven't got too many bits left over at the end!

I fitted L1 to L3, then checked them out for continuity before fitting C25 to C29.? I didn't cut off the leads from C25 and C29 immediately, but used them to tag onto 2 flyleads made up from a couple of inches of RG174 with SMA and BNC connectors respectively. I keep them for that purpose.? Saves getting solder in any hole I need later.??

Plugging in the nanoVNA says the filter looks fine, as expected.??

Just finished the 1300Z QCX QSO party session on 30m, cut short by blown fuse, so having fixed that, soldering iron on and back to the QCX mini.

Good luck with yours.

Chris? G4CWS
 

Luc, thanks!
The blog looks very helpful.
My main concern was impedance mismatch.
I was hoping someone had some BKMs (best known methods).
I just planned to solder a piece of coax to the measurement points.

73, Mike KK7ER
 

Dave, I do tend learn a lot of things the hard way.
But life is short and sometimes it is worth asking a question.
I'll report back.
Thanks.

73, Mike KK7ER
 

Chris,

Thanks.? I am glad to know you followed a similar process and it worked for you.

73, Mike KK7ER
 

On Mon, Nov 14, 2022 at 03:43 AM, David Wilcox K8WPE wrote:
I teach all new hams that we learn more from our failures (and others failures) than we ever do from our successes.
I don't recall who the originator was but there is a quote that goes:
"Wise people learn from their mistakes.? VERY wise people learn from others mistakes as well." or something like that.
?
-------
Rob KB8RCO
 



On Mon, 14 Nov 2022 at 19:43, David Wilcox K8WPE via <Djwilcox01=[email protected]> wrote:
? And we remember our failures far longer. Why do we still repeat some of them?? Ha!

'Experience' is recognising a mistake the next time you make it.

--
MC
VK1MC

 

"Experience is what you get when you don't get what you want."
--
73, Dan? NM3A
 

OK, reporting back as requested.
Here are the results for the band pass filter.
Sorry it has taken so long but life is pretty hectic (won't bore you with the details).
The Mini is configured for 20m.
Only T1 and C1 are installed (C5 & C8 are not used for the 20m configuration).

The measurement setup is like this:

NanoVNA was calibrated and C1 was adjusted to put the peak in the CW part of the 20m band.
4.13 dB insertion loss seems high:

There is 10.22dB suppression w.r.t. peak at 13.48MHz.

And 4.8dB suppression w.r.t.? peak at 14.58MHz.

The shape is different than Hans' photo in the manual (I measure the notch below the passband and he measures the notch above the passband).

Seems reasonable to me but I would be glad to hear your opinions.
LPF is next.

73, Mike KK7ER

 

BTW, the measurement at secondary #2 was not nearly as nice.
The peak at secondary #2 was not in the same place as the peak on secondary #1.
So I adjusted C1 to move the peak on secondary #2 to the CW part of of the 20m band.




So only 2dB suppression w.r.t. peak at 13.58MHz.
Why should one 3-turn secondary measure so differently from another 3-turn secondary?

73, Mike KK7ER
 

Hi Mike,

remember that the secundary windings will have a 470 nF cap on them through IC4 (the sampling detector). Only one of four of C43-C46 will be connected at each instant.
I don't know how much detuning those caps will have on your final BPF curve, maybe you could redo your test with a 470 nF caps over the secondaries, and report back if that shifts the curve a lot.

Anyway, you will have to retune the BPF for real after you built the mini, with the built-in test procedure.

As you can see, the BPF is also very wide, I measured only 6 dB loss from the main band to the "one lower" band (from 17m tot 20 m in my case), when I tuned the BPF to the lowest end of 20m. So it's only 1 S-point less sensitive in reception.

Luc ON7DQ
 

Hi Luc,

That's a good point.
But wouldn't the 470nF capacitance on pins 2 and 9 of IC4 have the same effect on both secondary responses?
Could it really explain one secondary sweep looking correct and the other being pretty far off?

73, Mike KK7ER
 

Moving on to the LPF.
Here is the setup:


The LPF response looks reasonable:


NanoVNA-saver shows the following:




It does not look as square as Luc's, but it should do the job.

73, Mike KK7ER

1 - 15 of 15