Date   
Re: SSB RX on QCX

James Daldry W4JED
 

Hi, Hans

I switched my QCX 40 to CW-R so the CW pitch would go down when I tuned the radio down-frequency. Just personal preference. Makes the radio feel nicer to my ears.

73

Jim W4JED

On 8/10/19 3:30 PM, Hans Summers wrote:
Hi Guido

Was not ware of this feature, and actually it has never came to mind that stations could come back on the CW-L band ...
I guess CW/CW-L switching is used when QRP stations are using direct conversion receiver and unaware of the side they use? is it that?

Transmit is ALWAYS on exactly the frequency shown on the display. The CW/CW-R (for CW-Reverse) setting only affects which sideband the receiver listens to. So, whether the LO is above (CW/USB) or below (CW-R/LSB) the operating frequency. 

I can't see how it could be related to the station at the other end being direct conversion and not knowing what side he is on - because the setting CW/CW-R only affects what you hear. Not what you transmit. So it can't affect him at his end. 

I've never ever switched on CW-R. I've never found a use for it. I implemented it because other radios do and because it is trivially easy to implement (being just a matter of the way the Si5351A is configured whether the phase shift is + or - 90 degrees). To change from one to the other just requires enabling the invert on one of the Si5351A quadrature outputs - thereby applying a 180-degree phase shift to it and bringing the total phase shift to 270 degrees. Or -90 degrees, equivalently. 

I suppose one conceivable use of it could be where there is a very very strong station on LSB which is interfering with your wanted weak signal CW station on USB. Maybe this could happen in a contest situation, I don't know. I am not good enough at this whole radio thing, to dare participate in contests. Anyway the unwanted sideband rejection in QCX is extremely good; so far I have never heard a station on the unwanted (rejected) sideband, let alone heard one at a strong enough level that it was annoying enough to go into the menu, switch to CW-R and tune 1.4kHz the other side of my wanted station! 

73 Hans G0UPL

Re: Hints for winding T1 in QCXs

dkwflight
 

Hi
After my limited experience winding T1 I think I will try to color code the different sections along with your suggestion.
Thanks
Dennis
AG4TD

Re: QCX-30

dmason@...
 

I will take that if it’s still available Jack.

Thanks,
Dale, W9ESE

Re: ProgRock not working #progrock

Robin Midgett
 

Great news..at least as far as working on frequency...I've finally had/taken time to build the QLG1 kit & connect it to my ProgRock. Once the QLG1 locked on GPS satellites & the 1PPS started, the ProgRock locked onto the desired frequency of 16.686666 MHz. and so far seems stable. 
Previously, using the Trimble T GPS module and a homebrew 555 based pulse stretcher to give a 116mS long 1PPS, I had a constant 18Hz. frequency discrepancy. In that case I fooled the ProgRock onto the desired frequency by programming register 8 18 Hz. higher than desired & setting register 3=0. Under those conditions the ProgRock would be stable and on frequency over a period of days. I"m still curious to experiment with the Trimble T and see if I can duplicate the reliability that comes with the QLG1, but for now I need to move forward with putting the propagation beacon, which is the basis for the project, on the air.
I'll see how reliable the ProgRock is over the next few days with the QLG1 supplying the 1PPS.

Robin Midgett K4IDC


On Sun, Jul 28, 2019 at 6:09 AM <mike@...> wrote:
I no longer have a trimble 't' so I can't test it, but yes, dats da sucker
afik.


On 27 Jul 2019 at 14:06, Robin Midgett wrote:

> Hey Mike,
> Is the software at the URL below the correct software for the Trimble T?
>
> Robin Midgett K4IDC
>
>
> On Mon, Jul 22, 2019 at 8:30 AM Robin Midgett <K4IDC@...> wrote:
>
> > Mike,
> > Is this the software you are recommending?
> >
> >
> http://trl.trimble.com/docushare/dsweb/Get/Document-484972/TrimbleStudio.exe
> >
> > Robin Midgett K4IDC
> >
> >
> > On Mon, Jul 22, 2019 at 8:16 AM Robin Midgett <K4IDC@...> wrote:
> >
> >> GM Mike,
> >> Ok, I'll check it out. Thanks much for that insight.
> >> My QLG-1 arrived Saturday, so that's 2 experiments I can perform soon.
> >>
> >> When you say the QLG-1 didn't change anything for you, are you saying
> >> your ProgRock worked properly with the QLG-1 and another GPS, or that it
> >> didn't work properly & by attaching a QLG-1 nothing changed, bringing
> the
> >> need for other action?
> >>
> >> Robin Midgett K4IDC
> >>
> >>
> >> On Mon, Jul 22, 2019 at 2:58 AM <mike@...> wrote:
> >>
> >>> Gm Robin..
> >>>
> >>> On 21 Jul 2019 at 22:56, Robin Midgett wrote:
> >>>
> >>> > How would triggering on an unrelated & unsynchronized 1PPS help?
> >>> It would confirm that the 1pps under test is not at fault. - some
> >>> Trimble T's
> >>> that are sourced in china are counterfeit and not actually Trimble T's
> >>> The same with some U-Blox engines.
> >>>
> >>> > I don't have Trimble's propriortized software, nor am I connected to
> >>> the
> >>> > GPS with a computer.
> >>> Then get it, from Trimbles website. Connect a computer , ttl serial, to
> >>> your
> >>> Trimble T and change the 1pps pulse duration to something appropriate
> >>> and
> >>> write it to the GPS's Eeprom. Using the software you can also check
> that
> >>> the
> >>> claimed defaults are set correctly.
> >>>
> >>> Remove the possible source of latency and see if the issue remains.
> >>>
> >>> I have seen many similar conversations on the list. The usual answer
> >>> 'get a
> >>> QLG-1' does'nt answer why one 5v TTL 1pps works and another,
> >>> functionally
> >>> identical 5v TTL 1pps does'nt - why not borrow or buy a QLG-1 and see
> if
> >>> it
> >>> actually changes anything. It did'nt for me.
> >>>
> >>>
> >>>
>


Re: ProgRock not working #progrock

Robin Midgett
 

Great news..at least as far as working on frequency...I've finally had/taken time to build the QLG1 kit & connect it to my ProgRock. Once the QLG1 locked on GPS satellites & the 1PPS started, the ProgRock locked onto the desired frequency of 16.686666 MHz. and so far seems stable. 
Previously, using the Trimble T GPS module and a homebrew 555 based pulse stretcher to give a 116mS long 1PPS, I had a constant 18Hz. frequency discrepancy. In that case I fooled the ProgRock onto the desired frequency by programming register 8 18 Hz. higher than desired & setting register 3=0. Under those conditions the ProgRock would be stable and on frequency over a period of days. I"m still curious to experiment with the Trimble T and see if I can duplicate the reliability that comes with the QLG1, but for now I need to move forward with putting the propagation beacon, which is the basis for the project, on the air.
I'll see how reliable the ProgRock is over the next few days with the QLG1 supplying the 1PPS.
Thanks,
Robin Midgett K4IDC


On Tue, Jul 30, 2019 at 10:38 AM Alan G4ZFQ <alan4alan@...> wrote:
> Is it easy enough for you to check the value of register 2 in your
> ProgRocks when they're on frequency, and then again when they're off
> frequency by whatever error you're seeing?

Robin,

I've never been too bothered about register 2, it's either set for
accuracy if there is no GPS or it sets itself. When setting manually
I've not noticed any non-linearity.
The attached picture taken some time ago shows my Progrock set at
14.996MHz from a cold start. As it starts about 2KHz off I guess
register 2 was about 4KHz away from correct.
As you see it rapidly corrects and settles 2Hz lower than the GPS
derived marker (Neo 7) at 14.996MHz. (Bottom spectrum, RX not quite
accurate) Looks like register 3 is 0?
If I set register 2 I can make it more accurate but if the GPS is
connected it will settle 2Hz low and stay there. I have done more
extended tests to confirm that within a few tenths of Hz.
A restart will then be close to the wanted frequency.
I suspect this test was with a 26MHz TCXO which might explain why the
frequency settles 2Hz low. All my available synth boards are now
non-standard, I cannot repeat this with a standard crystal but it does
show how the control works and that it settles as near as I can see at a
constant frequency.
I'm not sure what GPS I used, possibly a Neo-7, maybe an old Garmin.
I've not got a QLG1.

> I can set reg. 2 to a proper value to give the correct output frequency,
> set reg. 3 to something very high (30),

Why 30? I would think 5 was as much as anybody would want, zero if you
are looking for real precision.
I wish I could help further.

73 Alan G4ZFQ




Re: Hints for winding T1 in QCXs

Joe Cotton
 

Looking over my previous post for the 20th time, I think that I need to add this.

Winding the secondary 3 over 180 degrees of the toroid,  Point a short lead down as you hold the toroid and the wire, wind the long end of the wire up, over the top, into the toroid, and under.  Continue in this direction. When finished winding, the short end that you started with will go in hole 1, and the long end will go in hole 2.    

The purpose of this instruction is to get the winding direction correct. 

It turns out (no pun intended) that all windings start out with a short end pointing down, held close to the toroid on the outside of the toroid.  Then the windings go up, over the top, inside the toroid, and underneath, coming back up on the outside of the toroid, on the right side of the previous turn, (with the turn facing you).  The turns will continue in a counter-clockwise direction, looking from the top. 

Joe W3TTT  

QCX-30

Jack Hubbard <ni8njack@...>
 

Free to good home, non operating QCX-30.  You pay shipping.  Also a printed assembly/operating manual in a 3 ring binder.  ni8njack@...

IC10 short? No audio

Ashdee 9W8ASH
 

Greetings all,

Quick disclaimer. I'm by all means a non-technical person. I assembled only by the manual, and when in trouble (like wrongly assembled parts, which was half a dozen) I watched youtube build videos to help me out. I do not know how to test/troubleshoot that well as anything beyond the manual is out of my field.

I've been scouring this forum to find if anyone ever had a problem with audio and/or IC10. I had audio previously for maybe 5 minutes (which enabled me to align my QCX) but after that audio just went dead. Leaving it off for several hours will enable me to get audio when starting up, but again, for maybe 3-5 minutes.

For the past few days I've doing continuity checks, checking for any cold joints, bridges, etc. I've resoldered any dodgy looking pads.

I moved onto checking voltages of all the items as per the manual, since thats the only thing I have. Of all that was checked, only IC10 stood out, with measurements as per below using the built in QCX DVM feature:

IC10
Pin 1 - 11.18
Pin 2 - 11.18
Pin 3 - 11.18
Pin 4 - 0.00
Pin 5 - 3.03
Pin 6 - 2.26
Pin 7 - 11.16
Pin 8 - 11.96

Could this be the problem? I also did a quick check on C51 with an external multimeter, the readings weren't the same with C24, C37 and C38 (all are the same 10uf capacitor) and is constantly fluctuating (C24, C37 and C38 did not).

Any comments is highly appreciate.

73
Ashdee
9W8ASH

Re: Hints for winding T1 in QCXs

Joe Cotton
 

Hi Jim
Great hint for winding T1 for the QCX. The T1 winding step may be the hardest part of the whole megilla. 
The point that you make to wind the secondary winding 3 first and to mount the T1 as the first step are great ideas. 
However, your description of winding the primary and the secondary 1 and secondary 2 are not clear to me.  Allow me to say it in my own words and see if I understand you correctly. 

Step 1.  Wind secondary 3 over half of the toroid.  Solder it in holes 1 and 2, leaving room between the toroid and the board for wires (to come). 
Step 2.  Take about two inches of wire, tin one end, solder to hole 5.  This wire is standing straight up.
Step 3.  Bend the wire over the top of the toroid, then down into the toroid, and then underneath the toroid, coming out between hole 5 and hole 1. 
Step 4   Repeat step 3, two more times.
Step 5.  You should have a short length of wire left.  Bend this over the top of the toroid, and down into hole 6. Solder hole 6.

Step 6,  Repeat steps 2 through 5, for holes 3 and 4.  The first winding should come out between holes 3 and 5 in step 3.
Step 7.  Repeat steps 2 through 5, for holes 7 and 8.  The first winding should come out between holes 7 and 3 in step 3.

It's the step 3 which I want to get correct, and leave no doubt exactly what to do!  Do I have step 3 correct? I think that the windings all need to be in the correct direction. Do I have it correct?

thanks
Joe W3TTT

Re: SSB RX on QCX

ajparent1/KB1GMX
 

CW-R is a good feature for SSB RX on 40m...

The best mod for the filter is leave it stock and duplicate it for SSB bandwidth
as its only two opamp packages and a few parts.  Then a switch selecting the
output of one or the other gives excellent CW and SSB.  Would not take much
of a board to make one or just dead bug it.

Allison

Re: Using an SDR with QCX

Hans Summers
 

Hi Bob

Gain before selectivity reduces intermodulation performance. IP3, dynamic range, linearity. 

In the old days mixers were lossy and it was necessary to have gain ahead of the mixer (the RF amp stage), otherwise overall receiver sensitivity would not be good enough. I think RF amps ahead of the selectivity were a necessary evil, not a desirable feature. 

Nowadays the switching type mixers e.g. Quadrature Sampling Detector (as used in QCX and the Receiver module) have very low insertion loss and one of the benefits to come out of that, is no need for the RF amplifier. The QSD also has very high IP3 and not putting an RF amp in front means the high IP3 of the mixer (and therefore good signal handling of the whole radio) is preserved. 

VHF still usually needs RF amps because band noise levels are so low. But anyway at VHF demands on receiver are less than on crowded HF bands.

Superhet receivers have other additional problems, like mixing products (birdies) requiring good attention to shielding and filtering throughout the radio. To get general coverage HF and to avoid the difficulties of choosing an IF which doesn't adversely impact some amateur band or another, dual conversion is necessary, using a first IF of something some way above HF e.g. 45MHz. But double conversion multiplies up on all the problems...

Everything is tradeoffs but I think generally with modern components, the ability to get high performance at HF using simple direct conversion designs, easy to align and use, has been a very beneficial development. I'm not a superhet fan. Not that I'd never build one... but if doing so I think lots if filtering and shielding are in order... and probably much harder to get the same high level of performance. 

73 Hans G0UPL 






On Sun, Aug 11, 2019, 01:55 Bob Macklin <macklinbob@...> wrote:
I come from the time where the superhet was the design of choice.
 
Good ones had an RF amp. Better ones had 2 RF amps.
 
A single IF amp was most common. But better receivers has two IF amps.
 
Then there were double conversion receivers.
 
Why not a superhet followed by a quadreature IF second stage. Why do we want all the gain after the quaderature detector?
 
BTW:
There are 1836KHz crystals and oscillaord that would allow building a Taylor detector behind a 455KHz IF.
 
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"
----- Original Message -----
Sent: Saturday, August 10, 2019 3:32 PM
Subject: Re: [QRPLabs] Using an SDR with QCX

OK , its a problem with all mixers, there are not that are immune totally.

IF you run a switching mixer and the control lines and the control source (this case 5351)
are totally shielded.... guess what you still get RF at the input port from the switching
action and incidental leakage inside the IC.  Happens with DBMs too.  Even the best
mixers are rarely more than 50db below the RF drive. considering most have a drive
level in the 7-20dBm range there will be some signal in the microwatt region getting
to the antenna.

What the solution?

There are two common ones.

Use a common base of common gate RF amplifier before the mixer (between
mixer and antenna) as this type of amplifier has a fairly high reverse transfer
(S21) meaning signals at the output have a hard time getting to the input.

The other is superhet where the input filter attenuates the signal from the mixer
at the RF+IF (or RF-if).  The filter passes 7mhz but not the other possible
frequencies.

So most of the DC RX (be they SDR or analog) radio have this, unless they
have an RF amp of low gain but high isolation.  Even then some amount of
LO is radiated unless one has a well laid out enclosure that provides good
shielding.

With all that its touch to have a radio near another radio and not expect
to find the first radios receivers emissions.

However If one knows where to look (frequency) you can easily track 
where the RX is (both location and frequency!).

IF you want to use a SDR for RX build a TX  to match that radio.
For CW that is simple.

Allison

Re: Hints for winding T1 in QCXs

terryhugheskirkcudbrifght@...
 

hi

did you have any difficulty in passing the wire for s1,2 and pri under the core

terry gm4dso

Re: Clock: Assemble in conjunction with the operation manual??

Don--AE4DW
 

What you did is the same as I did..scan thru the operations manual, get a general awareness of how to set up the clock once its operating, and proceed. Its an easy fun build, a few jumper decisions you'll need to make, and they're all very simplistic.Once operating, it took a little practice to get used to the configuration button sequences, but its still all pretty straightforward.

Not sure if you're using a gps for timing (definitely recommend it, makes for one very cool shack clock), but be aware of the "date bug" where your local date will be incorrect by a day. You'll find it discussed a bit here in the group, and on the facebook qrp labs group.





On Sun, Aug 11, 2019 at 01:47 AM, Mike wrote:

I'm going to begin assembling my clock but the assembly instructions state I should be consulting with the operation manual. I've scanned through the operation manual and if I assemble the clock as it comes from the factory with no variations, I don't see a need for the operating instructions. Am I missing something?

Thanks,
Mike

Re: Video about U4B balloon tracker development

Alan de G1FXB
 

COLD.....
I guess that is what prompted the experiment with super caps in the last flight, (rather than batteries?)
Not sure if you flown those before?
(And the U4B prototype, having the TCXO glued thermally bonded to the micro for warmth. :-[)

I'm sure you have exhausted the obvious, and done the lateral thinking......
In a previous life I did instrumentation.

(Before the advent of Lithium types.)
?Stuff destined for extreme cold were specified to use the oldest, simplest primary cell type.......
(And against all common advertising about how a "bunny" can go on and on and on.)
"They MUST be Zinc / Carbon," because they were the best chemistry to work at low temperatures.

(I'm not sure if they be "re-charged" like primary alkaline types using dirty DC, perhaps a buck convertor with no smoothing cap?
and even less so, when whatever is the chosen chemistry is frozen solid......)


If only it was as simple as using an old thermos flask to keep everything in a chamber, and "just tie another balloon to it" to overcome the weight penalty.....


Alan

On 06/08/2019 11:24, Hans Summers wrote:
Hi Alan
?
Good to hear about the power saving possibilities,
It will be interesting to see if more balloons could fly with some sort of battery to perhaps operate at a reduced service during darkness,
maybe a single " I'm alive " once per hour / two hours rather than getting lost at night and not knowing if it came down or the GPS etc had locked up.
(Will still have the weight & temperature / discharge, limits.)

Dave and I have experimented with that quite a bit. The issue is COLD... it gets so cold at night that the battery freezes. Seemingly you have to burn some power, just to keep warming things up. So far the record Dave got was 4 hours after sunset.?

But U4B will be very good for experimenting with such things because the BASIC allows you to change the behaviour based on battery voltage, temperature, whatever you want. All those parameters are available in BASIC. So you can easily do an IF statement on them.??
?
I don't know what the true LOFERS look for in their beacons, but maybe the U4B could find a home there to?
They can then bake in a desert rather than freeze....
OK on band switching & the existing filter PCB.

Maybe have to get the daughter to learn CW, she will know whether the horses need their rugs or not in the field.
Not used BASIC since a RM380Z.
?
Hi hi. Well U4B is designed to be easy, cheap and flexible! You will not find it difficult. Absolute beginners won't find it difficult either.?

If I keep it reading in degK, if it ever reaches absolute zero......

...then you'll know the entire universe just stopped.

73 Hans G0UPL

Re: Clock: Assemble in conjunction with the operation manual??

Alan de G1FXB
 

The kit was launched early 2015, little will be "new" discoveries.....
Be Brave, Go for it.?


If there is anything not covered in either of the manuals,
or the forum you can always ask.


Good luck!
Alan

On 11/08/2019 06:47, Mike wrote:
I'm going to begin assembling my clock but the assembly instructions state I should be consulting with the operation manual. I've scanned through the operation manual and if I assemble the clock as it comes from the factory with no variations, I don't see a need for the operating instructions. Am I missing something?

Thanks,
Mike

Re: #u3s Relocating R5 #u3s

Nik
 

I’ve not tried the RD16HHF1 in a 5W PA kit yet, the one I have been talking about is fitted directly into the U3S, and under test at 5V. If all is successful then I will use this U3S to drive a 5W PA kit.  As the 5W PA kit has provision for the RD16HHF1 I may try this device in this kit too at some time. 

My goal is a system that outputs 5W fairly linearly across the HF bands. Then I will lower the power to 2W and use it as my home WSPR system. 

Clock: Assemble in conjunction with the operation manual??

Mike
 

I'm going to begin assembling my clock but the assembly instructions state I should be consulting with the operation manual. I've scanned through the operation manual and if I assemble the clock as it comes from the factory with no variations, I don't see a need for the operating instructions. Am I missing something?

Thanks,
Mike

Re: #u3s Relocating R5 #u3s

J68HZ
 

Well let’s not kid ourselves here.  The IRF parts and the RD parts are very (very very) different animals.  The IRF510’s dies are hard-on switches, meaning they saturate/ switch on hard and abruptly…they are designed that way because of their actual intended service as motor switches and power supply parts…. and that, in an RF world can cause harmonics, ringing, and trash.  It means you need to work extra hard to filter that stuff out of your output signal when you push them… and it’s frequency dependent.  The RD devices are made to be RF switches with a smoother approach on the curve at full on.  Much less trash generated.  You already know about the capacitance issues.

 

 

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton – J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com

Like us on Facebook!

 

Moderator – North American QRO Group at Groups.IO.

 

email:  bill@...

 

 

From: QRPLabs@groups.io [mailto:QRPLabs@groups.io] On Behalf Of ajparent1/KB1GMX
Sent: Saturday, August 10, 2019 5:59 PM
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] #u3s Relocating R5

 

for a 5W amplifier the RD15HHF or RD16VHF is a over sized device.
There are also RD06 parts that might be a better match.  Or better
yet MRF136.

As to bias, this is to be a class C or maybe E amplifier.  Two things
have to exist for that enough drive to saturate the device and also 
drive it to cutoff.  The yabut is that Most mosfets have a gain vs
bias curve that forces some current flow to get reasonable operation.
So the gate bias voltage (NOT CURRENT) needs to be sufficient for
a small amount of drain current (under 50ma!) or you moving from
class C to class B or even class A.   That move impacts gain,
power out and output and input circuits for stability and power.

So dumping the supposedly better RD part usually is more illusion
that actual performance.  Those that claim improvement often had
other issues in the works or due to the significantly different
characteristics they changed the amp in ways they do not realize.
At RF there are few (nearly none) drop in improvements.

Bottom line is to get 5W out you need enough power in and about
200-500mw is enough depending on circuit and frequency.

FYI I use IRF510s and have RD15 and RD16 parts and generally I get
more out of the IRF510s as I can run higher voltages (up to 28V) that
would destroy the RD part (designed for 13.8V nominal, 16V max use).
As to the claim the IRF is not a good HF part the QRPlabs amp runs
to 70mhz at acceptably reduced gain (could be optimized) and solid
to 30mhz.   I have two amps of my design that are for 6M use one is
20W class and the other using 8 of them does 220W at 30V.  I also
have refitted a 2006 6M SSB radio that was then designed for 1W
based on junkbox final and that is now a nice 2.5W using the
IRF510 and of course the right circuit to extract that power.
Could it do more yes, but the existing driver is the limiting factor
not the output and I didn't want to change that.

The only place the RD parts have been better for me than IRF510
is at 144mhz  and up but I can get more power per $ using MRF
series parts.  But then we are talking 100++W levels and 30-50$
transistors.

Allison


Virus-free. www.avg.com

Re: #u3s Relocating R5 #u3s

ajparent1/KB1GMX
 

for a 5W amplifier the RD15HHF or RD16VHF is a over sized device.
There are also RD06 parts that might be a better match.  Or better
yet MRF136.

As to bias, this is to be a class C or maybe E amplifier.  Two things
have to exist for that enough drive to saturate the device and also 
drive it to cutoff.  The yabut is that Most mosfets have a gain vs
bias curve that forces some current flow to get reasonable operation.
So the gate bias voltage (NOT CURRENT) needs to be sufficient for
a small amount of drain current (under 50ma!) or you moving from
class C to class B or even class A.   That move impacts gain,
power out and output and input circuits for stability and power.

So dumping the supposedly better RD part usually is more illusion
that actual performance.  Those that claim improvement often had
other issues in the works or due to the significantly different
characteristics they changed the amp in ways they do not realize.
At RF there are few (nearly none) drop in improvements.

Bottom line is to get 5W out you need enough power in and about
200-500mw is enough depending on circuit and frequency.

FYI I use IRF510s and have RD15 and RD16 parts and generally I get
more out of the IRF510s as I can run higher voltages (up to 28V) that
would destroy the RD part (designed for 13.8V nominal, 16V max use).
As to the claim the IRF is not a good HF part the QRPlabs amp runs
to 70mhz at acceptably reduced gain (could be optimized) and solid
to 30mhz.   I have two amps of my design that are for 6M use one is
20W class and the other using 8 of them does 220W at 30V.  I also
have refitted a 2006 6M SSB radio that was then designed for 1W
based on junkbox final and that is now a nice 2.5W using the
IRF510 and of course the right circuit to extract that power.
Could it do more yes, but the existing driver is the limiting factor
not the output and I didn't want to change that.

The only place the RD parts have been better for me than IRF510
is at 144mhz  and up but I can get more power per $ using MRF
series parts.  But then we are talking 100++W levels and 30-50$
transistors.

Allison

Re: Using an SDR with QCX

Bob Macklin <macklinbob@...>
 

I come from the time where the superhet was the design of choice.
 
Good ones had an RF amp. Better ones had 2 RF amps.
 
A single IF amp was most common. But better receivers has two IF amps.
 
Then there were double conversion receivers.
 
Why not a superhet followed by a quadreature IF second stage. Why do we want all the gain after the quaderature detector?
 
BTW:
There are 1836KHz crystals and oscillaord that would allow building a Taylor detector behind a 455KHz IF.
 
Bob Macklin
K5MYJ
Seattle, Wa.
"Real Radios Glow In The Dark"

----- Original Message -----
Sent: Saturday, August 10, 2019 3:32 PM
Subject: Re: [QRPLabs] Using an SDR with QCX

OK , its a problem with all mixers, there are not that are immune totally.

IF you run a switching mixer and the control lines and the control source (this case 5351)
are totally shielded.... guess what you still get RF at the input port from the switching
action and incidental leakage inside the IC.  Happens with DBMs too.  Even the best
mixers are rarely more than 50db below the RF drive. considering most have a drive
level in the 7-20dBm range there will be some signal in the microwatt region getting
to the antenna.

What the solution?

There are two common ones.

Use a common base of common gate RF amplifier before the mixer (between
mixer and antenna) as this type of amplifier has a fairly high reverse transfer
(S21) meaning signals at the output have a hard time getting to the input.

The other is superhet where the input filter attenuates the signal from the mixer
at the RF+IF (or RF-if).  The filter passes 7mhz but not the other possible
frequencies.

So most of the DC RX (be they SDR or analog) radio have this, unless they
have an RF amp of low gain but high isolation.  Even then some amount of
LO is radiated unless one has a well laid out enclosure that provides good
shielding.

With all that its touch to have a radio near another radio and not expect
to find the first radios receivers emissions.

However If one knows where to look (frequency) you can easily track 
where the RX is (both location and frequency!).

IF you want to use a SDR for RX build a TX  to match that radio.
For CW that is simple.

Allison