Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
QSX project retirement
|
Evan,
?
Ideally, a 50W amp for the QMX would replace everything on the final page of the QMX schematics.
The RF power amp final would be driven in switch mode from CLK2.
This would be far more efficient than just adding a linear amp to the QMX, and should have less distortion.
The new final amp and LPF's could be on a new QMX main board, a dev board, or in an external box.If in an external box, CLK2 should be transported as an LVDS pair, perhaps using a standard HDMI cable?
for all the required signals.
?
I'd go with a Class D audio amp modulating a Current Mode Class D RF power amp.
The AM transmitter guys have working examples of this (concepts good, build for the QMX could be simpler):
?
The newish GaN FET's look good for RF switching, those used for USB-C power bricks can be quite cheap.
? ? ? ?INN650D02? QFN-8
?
But first, the QMX needs to be fully featured and any bugs sorted out.
That will take some time.
?
Perhaps the QMX should stay at 5W.
Not sure the world needs more fully automated 50W FT8 contacts.
?
Jerry, KE7ER
?
?
?
On Wed, Mar 19, 2025 at 02:57 AM, Evan Hand wrote:
|
|
Ken,
?
Good point.
I'd pay another $10 for the QMX if the final was bomb proof.
I"d also like to see it move to standard SMPS IC's.
?
The BS170's are really cheap and have low capacitances, so hard to turn away from using them.
The QCX final also used BS170's and was fairly robust, there's probably a way to make the QMX final just as reliable.
Ideally, we would also be consistently getting 5W on all bands except maybe the really high bands.
Perhaps part of the problem is how the output networks must work for multiple bands.
I'd give up the huge band count for a more robust final.
?
Jerry, KE7ER
?
On Tue, Mar 18, 2025 at 11:52 PM, KEN G4APB wrote:
|
|
Hi Jerry,
Perhaps part of the problem is how the output networks must work for multiple bands.
That was really part of my original question if the 10w Linear could be used with the QMX and new SSB mode. I should have asked…does all the calibration, tweaking and pre distortion for QMX SSB work if the 10w linear is added, presumably with output feed back to the QMX lpfs, or is all that unique to the built in bf170 pa devices?
Again, I am not trying to boost the power to 10w or 50w, just trying to use something I have already have, (under thick dust now).
73 ken g4apb? |
|
The 10 watt amp may not be a good solution in conjunction with QMX+. It is my hope, Hans designs a new 50 watt linear amp with LPF multiband? switching and a new RF final transistor that is made for 160-6. It will bring the cost up but still be a very competitive offering. 25 watts would perfect for SSB. So to be conservative in rating, a 25 watts linear amp, and at least 50% efficiency seems ideal. 12 volts to match the 12 volts QMX+. Then we would have something suitable and integrated.? |
|
Hello Ken ?
It would match fairly well with a U3S. And since U3S is transmit only you wouldn't have worries about?Tx/Rx switching. Note that the 25-50mW input requirement of a 10W Linear is not so much about it being sensitive to damage, as about driving it totally into non-linearity, when its whole point was to be a linear amplifier. 73 Hans G0UPL |
|
Jerry,
?
I was thinking along the Elecraft KX line competition.? Elecraft offers a 100-watt amp for the KX2/3 radios:
?
Ideally, it would be 100 watts, but 50 at the right price would be good.
?
Currently, the QMX+ can be compared to the KX2.? The QMX is short on the number of bands.? At almost 1/10th the price, the QMX series is a steal.
?
73
Evan
AC9TU
? |
|
Ken,
?
The output of the QMX should be a clean SSB signal into 50 ohms, just like any other transmitter.
As such, with an appropriate attenuator, it should drive a linear amp such as the 10W amp from qrp-labs.
?
You wrote:
"I should have asked…does all the calibration, tweaking and pre distortion for QMX SSB work if the 10w linear is added, presumably with output feed back to the QMX lpfs, or is all that unique to the built in bf170 pa devices?" ?
Well, that's a thought.? I suppose you could adjust the 10W amp output transformer to give only 5W
so it doesn't overload the QMX LPF's.? Also adjust the T501 ratio to give much reduced power out from the BS170's.
Insert that 10W amp in between the output of T501 and the junction of C510,C511.??
The 10W amp should work fine with the output of T501 and the input of the roughly 50 ohm LPF's.
Might make a bullet proof QMX.
?
I think a better plan is to fix the QMX power amp to be more robust.
Removing L502 should reduce Vds voltage spikes on the BS170's, T501
would then be made 1:1 to bring the power back up.
However the increased current may be a problem for the BS170's.
Perhaps replacing the BS170's with TN0110's should be reconsidered.
The QCX has a fairly robust 5W output using BS170's, but must have band specific LPF's to operate in class E.
?
Jerry, KE7ER
?
On Wed, Mar 19, 2025 at 07:47 AM, KEN G4APB wrote:
|
|
Evan,
?
A linear amp for the QMX is a possibility, could work well.
Could just use that Elecraft 100W amp, at $1429 they have the budget for
a very nice amp.? But a major advantage of the QMX is EER transmission,?
which can use switch mode FET's in the final with no concern for linearity.
?
A beefed up QMX at 100 Watts might have the same design as the current QMX,
just beefier parts on the last page of the schematic.? The modulator should be made
a class D audio amp, not the linear amp we currently have, which would be a
bunch more parts.? Such a beefed up QMX should cost much less than
a QMX + linear, have a cleaner signal, and be extremely efficient with power.
A single board SSB transmitter at more than 5W often has trouble with RF from the final getting into various stages of the exciter.
That may not be an issue with the QMX, when transmitting all we have is CLK2
from the Si5351 feeding the final, all RF is at the same frequency.
Would still have to guard against RF getting into microphone cable and modulator,
but EER should make for a much simpler and cheaper 100W SSB transmitter.
?
Jerry, KE7ER
?
?
On Wed, Mar 19, 2025 at 08:19 AM, Evan Hand wrote:
|
|
Jerry,
?
We may be getting too far off-topic and need to start another thread.
?
A company tried an external EER linear amp (The Polar Explorer) highlighted in the March/April 2017 QEX (I have a subscription to QEX to read it.? I do not know if I can publish the article as the content may be copywritten).? The company never made a viable commercial product that could compete with the AB class amps.? Hans overcame the EER issues in the QMX by using the "calibrate" function to align the amplitude and phase components in the QMX.? That would need to be done for an external amp IF using the QMX EER approach.? Adding the circuitry to break down the input into amplitude and phase signals at RF frequency and power adds complexity and cost.? That is already done in the QMX at low signal levels, which makes it much easier to work.??
?
As you suggest, the best approach to keep with the EER design is to beef up the internal PA of the QMX by offering a QRP and QRO option.? That could change if Hans or another developer develops a low-cost way to manufacture an EER-based amp.
?
73
Evan
AC9TU |
|
Evan,
?
Fully agree that the ideal high power amp for the QMX would have about the same organization
as what we have now.? CLK2 feeds the gates of the final push-pull FET's.
Given that EER SSB works, there's little reason to deal with the complications and expense of a linear amp.
Leave that market to Elecraft.
?
Should probably be a new design in one box, perhaps call it the QMX++.
?
It would be possible to add a connector to the QMX+ (an HDMI cable so it is cheap and impedance controlled)
to carry CLK2, modulator control signals DAC and ADC_PA, tandem match outputs ADC_FWD and ADC_REV,
the band select signals, and the TX, RX, and RX_IN signals out to another box that has a high powered final
plus LPF's and tandem match in it.? Replaces everything on that final schematic page of the QMX.??
Kludgy, but if you want a cheap high volume 5W exciter and an optional amp, it could be made to work
and would be a whole lot cheaper and more efficient than Elecraft's 100W linear.
?
But there's plenty more to be done with the QMX before it's ready for higher power.
Higher power won't happen for quite some time.
?
The Polar Explorer article from QEX is publicly available from the ARRL website:
That's not just a linear amp replacement, it sends phase modulated RF to a switch mode final
and envelope information to a modulator, those two channels are generated digitally
just like the QMX.? There is no RF SSB signal till it comes out of the final.
?
There have been other products doing EER linear amps that take output from a
traditional SSB exciter and split it into RF phase and AF amplitude information
as shown in fig 2 of that polar explorer article.
These have been available for decades, they have been successful in?
commercial SW transmitters of perhaps 50kW or more where the
additional complication is worth the increased efficiency.
The QMX (and the uSDX) have far less complicated hardware, and thus are cheaper.
?
Evan said:
"We may be getting too far off-topic and need to start another thread."
?
I take considerable pride in being off-topic.
Shows I"m capable of independent thought.? ;-)
?
Jerry, KE7ER
?
?
?
On Wed, Mar 19, 2025 at 09:50 AM, Evan Hand wrote:
|
|
Hi Norm
The probability is certainly non-zero. But I don't have any concrete plans for a 10W QMX+ at the moment.? Bear in mind that the QSX would have had 10W output. But the QMX+ has CESSB, giving it effectively a 4 or 5 dB advantage. Which effectively makes its power a little more than 10W. So a QMX+ transmitting speech is at least comparable to a QSX transmitting speech at 10W SSB. Before anyone asks what about QSX with CESSB: I believe it is very unlikely that the QSX would have been able to do CESSB.? 73 Hans G0UPL |
|
On Wed, Mar 19, 2025 at 10:56 AM, KEN G4APB wrote:
Hi? With all that gain, you can use it for HackRF kind of miliWatt SDR transceivers. Even with U4B baloon Tracker 73 DE muhsin TA1MHS
|
|
Thanks, Hans for your reply. I have been 100% CW op for some time. I was looking for a potential Lab599 TX-500 or KX2/KX3 (or fill in the name of another "QRP" rig) killer. I know,... much different radios with different design criteria. I am not doing SSB (or Digital for that matter), so I guess the CESSB won't be of any use for me. No big deal, I was just curious.
--
73, Norm/KC1BMD |
|
On Fri, Mar 28, 2025 at 03:14 PM, Muhsin TA1MHS wrote:
I've wondered if the second stage of the 10W linear could be hooked into the QMX between the existing PA and the LPFs, sort of where T501 is.? T501 would become two transformers, one would provide the drive for the PA stage, the second would manage the impedance transform to the existing LPFs. |
|
On 29/03/2025 00:00, Chris via groups.io wrote:
I've wondered if the second stage of the 10W linear could be hooked into the QMX between the existing PA and the LPFs, sort of where T501 is.Chris, Anything is possible but is it worth the trouble? Often it has been remarked that 3dB improvement (1/2 S point) is no big deal. On the other hand an external 20+ watt amplifier would be neater, more effective and a lot easier to implement. 73 Alan G4ZFQ. |
Norm - KC1BMD
to navigate to use esc to dismiss