On Thu, 2017-11-09 at 14:50 +0000, Alan G4ZFQ wrote: Red plots are wide spaced and green plots are tight spaced.
Peter,
Recently someone pointed out that the inductance does not change much,
it is the distributed capacitance that makes it look as if it does.
(Which would also mean that most, if not all, inductance meters do not
actually measure the inductance in isolation.)
The micrometals web site has this to say on the subject: "The cores are manufactured to the AL values listed; the permeability for each material is for reference only. Toroidal cores are tested with an evenly-spaced full single-layer winding in order to minimize leakage effects. Iron powder cores tested with a small number of turns which are not evenly distributed will produce higher inductance readings than expected." It says higher than expected, but is that just because of the change in distributed C or is there a real change in L ? It's hard to tell when all you can easily observe is 1/2¦Ð¡Ì(LC) :-) But you have shown how much the spacing might matter if the cutoff
frequency is close to the top of a band. I think that is the key thing to draw from these experiments. Your wide-spaced ones look about the recommended spacing, except not
quite as even as the finished product would be.
I wonder how much extra capacitance squeezing together adds? I would not
think it to be much, what happens if say 5-10pF were put across the
inductors? I'm not sure if the changing distributed capacitance or the changing inductance is the dominant effect, but at the very least I now have suitable coils wound for my 30m filter :-) PeterO G0DZB
|
Hello Hans! Thanks a lot for fast positive answer!??I had no any doubts concerning band pass filter!? I was just afraid that there is an artificial limitation of the synthesizer?frequency band? ... 73! Anatoly UT5ULX
Hello Anatoly
It covers the whole of 20m. The typical band pass filter response is shown near the end of the QCX manual. It is plenty wide enough to cover the whole of 20m.
The same applies also to the other bands.
73 Hans G0UPL? ?
|
Re: What is the effect of L4?
Looking at the schematic, L4 appears to be Jim, Look at the assembly manual section 5.10 73 Alan G4ZFQ
|
Re: What is the effect of L4?
This is explained in 5.10 of the assembly manual.
In the U3S and other class C designs the inductor at L4 is just a choke to feed DC power to the finals. The QCX works in class E and the L4 inductor is in resonance with the C30 capacitor and all the additional capacitance from the BS170 gates (and pc-board etc).
|
Re: What is the effect of L4?
Looking at the schematic, L4 appears to be a decoupling choke that works along with the .1uF capacitor C32 to keep RF off of the PA VCC voltage supply line coming from the transistors in the key shaping circuitry.? It needs to present a high impedance to the output frequency to keep RF flowing to the output filter, L1, L2 and L3 and not back into the 12 volt supply.? C32 shunts any RF that does make it through the choke (L4) to ground.? (This is a simplified explanation but basically that's what L4 is there for.)
Hans, correct me if I'm wrong please.
Jim - W0EB
|
Hello Anatoly
It covers the whole of 20m. The typical band pass filter response is shown near the end of the QCX manual. It is plenty wide enough to cover the whole of 20m.
The same applies also to the other bands.
73 Hans G0UPL? ?
On Nov 9, 2017 6:44 PM, "Anatoly G" < poselok@...> wrote: Dear Colleagues!
?I intend to purchase the device QCX-20m for CW operation. Unfortunately, I did not find the information about the width of the frequency range.?
Could the QCX works in the entire range of 20m (14.000-14.250kHz) or only in the CW section (14.000-14.100kHz)?
Thanks in advance! Anatoly, UT5ULX
|
Anatoly;
The receiver bandpass filter properties are shown on page 133 of the assembly manual.? I believe that QCX will cover 14.0 to 14.25 MHz.??? My experience with the 60m version indicates that the sideband suppression will also be more than adequate for the entire band.
Hope this helps
de va3rr
|
OM Anatoly, I am also interested in any answer to this, from those with QCX on 20m what difference between 14001 and 14349?
Is in Ukraine only allowed up to 14250? I thought all countries had 14000-14350? (in the very old days I think it was up to 14400, of course in the beginning, ALL frequencies on HF!)
|
Dear Colleagues!
?I intend to purchase the device QCX-20m for CW operation. Unfortunately, I did not find the information about the width of the frequency range.?
Could the QCX works in the entire range of 20m (14.000-14.250kHz) or only in the CW section (14.000-14.100kHz)?
Thanks in advance! Anatoly, UT5ULX
|
Re: U3S-19 in the air...struggling for altitude
|
Re: 40m - too much gain of whole RX chain
I built 40m version QCX. It Works fine, I was surprised how
this simple radio clear recieve.
But as radio has no AGC, I need use a volume knob a lot. I kept
it most of the time almost on zero. The total gain is just too
much for 40m (or 80m), I believe optimal for higher bands.
As a 'secure' solution I modified R37 from 1 K to 10K that
decrease gain by 20dB. I kept R38 and C23 as it defines cutoff
frequency (1326 Hz) for low pass filter.
It works fine, maybe sidetone is quiter. But I am still thinking
how to lower gain on first stage IC5.
I cannot touch 100 ohm resistors R5-R9 (input impedance of
amplifier) as they are tighten with quad. sampling detector.
What about R7-C4, R10-C7 in feedback?
2k2+47n, gain -13dB
3k3+30n, gain -9.7dB
cutoff freq. remains the same
What do you think? Could it affect IP3 / sensitivity of whole
in negative way?
73
Ondrej OM4DW
I reduced R38 (originally 120K) in my 80-meter QCX to 12K for a 20
dB reduction in audio gain.? It had way too much with the 120K.?
About right now and I turned up the sidetone level a bit to
compensate.? I'm think about ways to add AGC of some type to this
really fine rig.? Seems like one could take advantage of the audio
sampling that goes on with the audio1 line that is feeding one of
the ADCs in the CPU.? I think that is the source for the S-Meter
display data, so maybe use a JFET in the feedback where R38 is and
shut down the audio gain that way.? More experiments to do!? :-)
72,
Jim, K8IQY
|
What is the effect of L4?
Sorry for ignorance as an "advanced" class amateur, I'm naturally MORE amateurish ;-) but learning again...
L4 purpose is? Something to do with output transistors? Manual doesn't seem to make it clear in section 3.52 other than to say windings are not critical.
And what effects, if any, would a significantly higher or lower inductance than that recommended in the chart have, upon power output and/or durability of Q1-3?
For 30m, 14 turns are recommended. I measured (on my core, with my windings, thus it will NOT be the same for other people) for 14 turns 0.88uH, and removing 2 turns to 12 turns brought it to 0.79uH, very close to the recommended 0.78uH for 30m.
As I was, at the time at least, trying to make my QCX work (if repeaked) on both 20 and 30m, I removed a further two turns, not sure how low the inductance is now. Inductance recommended for 20m is 0.40uH. So, what effect would a too low inductance have on 30m operation? I plan to return it to get close to 0.78uH again.
I did blow my finals, but this was due to high VSWR when the wind (invisibly from ground level) transformed my doublet to an inverted L but being a manual balanced Antenna Matching Unit I was not on hand to rectify the intermittent and frequent swapping between low and high SWR until the smell of exploded Q(1-3) reached me.
|
Thanks Hans...would I reduce the 150.000.000 maximum frequency by 28.200.000?? ? On 11/09/17 10:54 AM, "Hans Summers" <hans.summers@...> wrote:
Hi Vernon
I suggest putting the sequence of the bands the other way around: start with the highest frequency and then step down.?
Park Mode 2 is useful in the gaps between transmissions but makes very little difference if the gap is only 8 seconds. So use Park Mode 2 for the long wait at the end of calibration and before the Mode 0] configuration starts. Have it sequence through starting at the highest frequencies then stepping lower.?
73 Hans G0UPL
?
|
40m - too much gain of whole RX chain
I built 40m version QCX. It Works fine, I was surprised how this simple radio clear recieve.
But as radio has no AGC, I need use a volume knob a lot. I kept it most of the time almost on zero. The total gain is just too much for 40m (or 80m), I believe optimal for higher bands.
As a 'secure' solution I modified R37 from 1 K to 10K that decrease gain by 20dB. I kept R38 and C23 as it defines cutoff frequency (1326 Hz) for low pass filter.
It works fine, maybe sidetone is quiter. But I am still thinking how to lower gain on first stage IC5.
I cannot touch 100 ohm resistors R5-R9 (input impedance of amplifier) as they are tighten with quad. sampling detector. What about R7-C4, R10-C7 in feedback?
2k2+47n, gain -13dB
3k3+30n, gain -9.7dB
cutoff freq. remains the same
What do you think? Could it affect IP3 / sensitivity of whole in negative way?
73
Ondrej OM4DW
|
The filters are really tight. I used Elsie to analyze the 30m filter.
3 dB cut-off frequency 11.4 MHz.
21dB attenuation at 14 MHz.
48 dB attenuation at 20 MHz.
VK5EEE had low power output and had to remove turns. Before removing turns someone measured his coils:
L1 = 1.22uH instead of 1,1uH
L2 = 1.6UH instead of 1.3uH
L3 = 1.32uH instead of 1.1uH.
Assuming the correct capacitors Elsie gives:
Cut-off at 10.35 MHz
1.9 dB loss at 10.100 MHz
2 dB loss at 10.150 MHz.
Jean-Claude
|
Hi Vernon
I suggest putting the sequence of the bands the other way around: start with the highest frequency and then step down.?
Park Mode 2 is useful in the gaps between transmissions but makes very little difference if the gap is only 8 seconds. So use Park Mode 2 for the long wait at the end of calibration and before the Mode 0] configuration starts. Have it sequence through starting at the highest frequencies then stepping lower.?
73 Hans G0UPL
On Thu, Nov 9, 2017 at 5:35 PM, Vernon Matheson <vmath@...> wrote: ?Hi all, ? I have been trying to reduce the drift on my wspr tx when running 5-6 bands. I seem to be able to get one band with zero drift and the next band in sequence has some drift. See the data for 20m versus 30m below ? I have attached a small heat sink on the 27mhz xtal as suggested but am wondering about park mode. U3S is also in a qrp labs case. ? If I set to mode 2 and say 149.000.000 as suggested it doesn't change much. ? Would I be better to use park mode #1 and then 149.000.000 for a frequency. ? Vernon ?
| Timestamp |
Call |
MHz |
SNR |
Drift |
Grid |
Pwr |
Reporter |
RGrid |
km |
az |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097148? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G4ZFQ? |
?IO90ir? |
?4509? |
?59? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097165? |
?-24? |
?-2? |
?FN85ij? |
?0.2? |
?F4GUK/SDR? |
?JN18et? |
?4828? |
?61? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097153? |
?-20? |
?-3? |
?FN85ij? |
?0.2? |
?K9AN? |
?EN50wc? |
?2102? |
?263? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097157? |
?-2? |
?-3? |
?FN85ij? |
?0.2? |
?W9HAM? |
?EN64? |
?1865? |
?275? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097156? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G0DLV? |
?IO70ru? |
?4286? |
?60? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097150? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G0LUJ? |
?IO83ls? |
?4305? |
?55? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097157? |
?-21? |
?-3? |
?FN85ij? |
?0.2? |
?KA4PKB? |
?EN41kf? |
?2287? |
?268? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140154? |
?-26? |
?0? |
?FN85ij? |
?0.2? |
?SWLKCL? |
?FN03ip? |
?1281? |
?267? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140156? |
?-9? |
?0? |
?FN85ij? |
?0.2? |
?SWLNXZ? |
?FN13co? |
?1165? |
?265? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140153? |
?-22? |
?0? |
?FN85ij? |
?0.2? |
?K9AN? |
?EN50wc? |
?2102? |
?263? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140152? |
?-18? |
?0? |
?FN85ij? |
?0.2? |
?VE3YX? |
?FN16fd? |
?1107? |
?279? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140160? |
?-22? |
?0? |
?FN85ij? |
?0.2? |
?VE3BSB? |
?FN14uv? |
?1020? |
?272? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140153? |
?-20? |
?0? |
?FN85ij? |
?0.2? |
?K4COD? |
?EM73sc? |
?2266? |
?240? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140154? |
?-18? |
?0? |
?FN85ij? |
?0.2? |
?K1BZ/3? |
?FM19ne? |
?1311? |
?243? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140150? |
?-21? |
?0? |
?FN85ij? |
?0.2? |
?N4CDN? |
?FM18iu? |
?1362? |
?243? |
| ?2017-11-09 14:20? |
?VA1VM? |
?7.040163? |
?-27? |
?0? |
?FN85ij? |
?0.2? |
?N4TVC? |
?FM18is? |
?1367? |
?242? |
|
Red plots are wide spaced and green plots are tight spaced.
Peter, Recently someone pointed out that the inductance does not change much, it is the distributed capacitance that makes it look as if it does. (Which would also mean that most, if not all, inductance meters do not actually measure the inductance in isolation.) But you have shown how much the spacing might matter if the cutoff frequency is close to the top of a band. Your wide-spaced ones look about the recommended spacing, except not quite as even as the finished product would be. I wonder how much extra capacitance squeezing together adds? I would not think it to be much, what happens if say 5-10pF were put across the inductors? 73 Alan G4ZFQ Frequency Red Green Gain
10103456 -0.55 -5.98 5.43
10125367 -0.56 -6.18 5.62
10147278 -0.57 -6.39 5.82
Changing the winding spacing had made nearly 6dB difference at the top
of the band !
|
Re: USBAsp question, voltage jumper setting?
Hello Alan,
I popped over to see him as he had trouble loading the driver, the
programmer looks like someone (probably over here...) has gone to the
trouble of going over all the solder joints, as there's fresh flux
deposits on it. But it worked straight off. I too have a few DOA
programmers, maybe if I find the enthusiasm going over their solder
points would revive them. Thanks for the help Alan.
Thursday, November 9, 2017
toggle quoted message
Show quoted text
So in theory
with the jumper on the programmer he has set to 5V he could programme
his chip without powering up the U3S itself with some U3S mods?
BUT I'll tell him to ignore it as it connects to an unused U3S pin
Chris, I'd say it could be used to power some devices but I'm not sure that
feeding the USB 5V to the 5V rail of something like a U3 is a good idea. I did mean to say that programmer looks good. Although it is more than
twice as expensive as my Chinese ones I had to buy two to get the fun of
updating the firmware! 73 Alan G4ZFQ -- 2E0ILY Best regards, Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)
|
Using QCX on WSPR I found great stability at home and got my frequency accurate to within I think one Hertz, but on the back seat of the car on a warm day in VK while driving around, some drift set in. I guess that would be normal, temperature affecting and thus heat sink may surely help?
|
?Hi all, ? I have been trying to reduce the drift on my wspr tx when running 5-6 bands. I seem to be able to get one band with zero drift and the next band in sequence has some drift. See the data for 20m versus 30m below ? I have attached a small heat sink on the 27mhz xtal as suggested but am wondering about park mode. U3S is also in a qrp labs case. ? If I set to mode 2 and say 149.000.000 as suggested it doesn't change much. ? Would I be better to use park mode #1 and then 149.000.000 for a frequency. ? Vernon ?
| Timestamp |
Call |
MHz |
SNR |
Drift |
Grid |
Pwr |
Reporter |
RGrid |
km |
az |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097148? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G4ZFQ? |
?IO90ir? |
?4509? |
?59? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097165? |
?-24? |
?-2? |
?FN85ij? |
?0.2? |
?F4GUK/SDR? |
?JN18et? |
?4828? |
?61? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097153? |
?-20? |
?-3? |
?FN85ij? |
?0.2? |
?K9AN? |
?EN50wc? |
?2102? |
?263? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097157? |
?-2? |
?-3? |
?FN85ij? |
?0.2? |
?W9HAM? |
?EN64? |
?1865? |
?275? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097156? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G0DLV? |
?IO70ru? |
?4286? |
?60? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097150? |
?-22? |
?-3? |
?FN85ij? |
?0.2? |
?G0LUJ? |
?IO83ls? |
?4305? |
?55? |
| ?2017-11-09 14:24? |
?VA1VM? |
?14.097157? |
?-21? |
?-3? |
?FN85ij? |
?0.2? |
?KA4PKB? |
?EN41kf? |
?2287? |
?268? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140154? |
?-26? |
?0? |
?FN85ij? |
?0.2? |
?SWLKCL? |
?FN03ip? |
?1281? |
?267? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140156? |
?-9? |
?0? |
?FN85ij? |
?0.2? |
?SWLNXZ? |
?FN13co? |
?1165? |
?265? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140153? |
?-22? |
?0? |
?FN85ij? |
?0.2? |
?K9AN? |
?EN50wc? |
?2102? |
?263? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140152? |
?-18? |
?0? |
?FN85ij? |
?0.2? |
?VE3YX? |
?FN16fd? |
?1107? |
?279? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140160? |
?-22? |
?0? |
?FN85ij? |
?0.2? |
?VE3BSB? |
?FN14uv? |
?1020? |
?272? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140153? |
?-20? |
?0? |
?FN85ij? |
?0.2? |
?K4COD? |
?EM73sc? |
?2266? |
?240? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140154? |
?-18? |
?0? |
?FN85ij? |
?0.2? |
?K1BZ/3? |
?FM19ne? |
?1311? |
?243? |
| ?2017-11-09 14:22? |
?VA1VM? |
?10.140150? |
?-21? |
?0? |
?FN85ij? |
?0.2? |
?N4CDN? |
?FM18iu? |
?1362? |
?243? |
| ?2017-11-09 14:20? |
?VA1VM? |
?7.040163? |
?-27? |
?0? |
?FN85ij? |
?0.2? |
?N4TVC? |
?FM18is? |
?1367? |
?242? |
|
Daniel Ekman SA2KNG
Jim Sheldon
va3rr
Jim Kortge, K8IQY