Date   

Re: 50W amplifier: power rating of TR switch

John Seboldt K0JD
 

Thanks, Hans, for nicely pointing to all of the stuff that was right there in the documentation! Wow, back biasing the diode at 550 volts! There's some insight into the total ratings issues involved. I once bought a number of true surface mount PIN diodes, I will have to check the junk box and see if their ratings might be greater than a 1N4007. If so, I'm in business.

John K0JD
Milwaukee

On 1/27/2020 14:56, Hans Summers wrote:
Hi John

See the references in the theoretical section of my 50W PA manual to the document by W6JL. And his QRZ page https://www.qrz.com/db/w6jl 

Don W6JL uses a single 1N5408 diode for switching 600W RF, forward biased with 100mA. 

The requirements on Don's switch are not the same as mine in the 50W PA. And it's different again in QSX, and in QCX. Often T/R switches need to be considered at the system level. A T/R switch design can't necessarily just be transplanted from one place to everywhere else. 

Note that sufficient forward bias current is very important, as I found out in another project. All the more so at lower frequencies, where eventually the 1N4007 loses its PIN characteristics and starts to return to rectifier behaviour; then you suddenly end up with a whole bunch of unwelcome harmonics starting with the 2nd. 

Solid state T/R switching is not, in my opinion, when done properly, a reasonable way to reduce current consumption of the switch, compared to relays. But solid state does have a lot of benefits compared to relays including smaller physical size, lower cost, silent operation, very fast switching and longer lifetime. 

73 Hans G0UPL 

On Mon, Jan 27, 2020, 16:39 John Seboldt K0JD <k0jd-l@...> wrote:
Glad to see the 1N4007-based diode TR switching in the 50W amp kit. I
also have a 120W linear amp kit, long ago completed but has been waiting
for final packaging with TR switching and filtering. Do you think this
switch will work at the 100W or so level? If not, who can guide me to a
nice polished design for a switch at the 100-200W level?

John K0JD
Milwaukee, WI






Re: Experiences with 50 watt Amp

geoff M0ORE
 

An efficient way of providing reverse power protection is to run the supply through a relay with the relay switched via a diode such that the relay will not operate with the supply reversed. This has several benefits. 1) You can position the relay near to the supply entry point to reduce voltage drop on long leads. 2) the supply can be switched on/ off using a switch rated at much less than the current drawn by the rig, it only has to operate the relay. 3) no voltage drop through the diode, it only supplies the relay.

A suitable relay is the automotive type used to supply fans, heated rear windows etc at currents up to around 30 Amps. Designed to work at 13.8 volts.

On 26/01/2020 23:34, ajparent1/KB1GMX wrote:
Hint there is none.
Did you consider a fuse?

Even the QCX, I use a 1A fuse, because wires cross.

FYI 30w at likley less than 17V is doing good.  Oh, why less than 17V, at 4a you
get voltage drop in your wires to the battery and unless you measured the battery
while transmitting you do not know your getting the same voltage as open circuit.

High current reverse polarity protection is not a 1$ as a series diode is cheap but costs 
about .7V (sometimes more at high currents) voltage loss and would have to be a
heavier 5A part.  Works at QRP power levels, but not at all at 80W input.

Power out for a fixed amount of drive will vary based on voltage applied so .7V
(minimum some diodes are more) so loosing that voltage to the diode would
be costly in terms of power out and energy wasted.

The people put in diodes across the power leads.  A parallel diode to
"blow the fuse" on reverse polarity.  About some percentage  of the time
the transistors go first.  Or the user puts in a bigger fuse to "correct that"
weak fuse resulting in a more disastrous  failure.

If you have room about 1sq inch of board and more than 1$ ts possible to do
relay based protection.  Note the relay can't be some cheap thing as reasonable
contact rating is 5A or more.  Since in a kit the user is also testing if it was done
right there is again a percentage of time when smoke is the only course.

Which brings the point.  Kits are assembled and final tested by the builder.

As a pilot you make sure there are wheels with air in them and the oil in the
engine, adequate quantity and correct fuel, control surfaces that operate properly,
plus a long list of other checks before climbing into the flier as once in the air it
can be terribly unforgiving of error or inattention.     Kit built airplanes have a
very rigorous test program that does not even allow for builder saying its right,
he has to prove it with  third party checks during assembly and beyond.

Electronics is the same way.  At high currents parts will evaporate if 
subjected to current well past reasonable limits, and same for excess
voltage.  There are few if any  commercial amps with reverse polarity
protection as well.   Most errors are preventable by not proceeding
without double checking.  

I've built more than a few amps, some using devices with 200$ price tags,
you just don't power up and pray.  There are a list of things to do before
attaching a 50V at 30A power supply.   So when that stage is reached 
there is no surprise, or smoke.   Why?  Not only would the expensive
LD-MOSFET vaporize but likely half the traces on the board because
1500W of energy will do that.  That's a lot of electronic and mechanical
work to "fix" it.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: Bias setting 50W. PA #qcx #pa

ajparent1/KB1GMX
 

The bit of data sheet figure 4 is the define the limitations of the IRF510
safe operating area.  Safe in this case means no failure if on a infinite
heatsink.

Its current vs voltage and the result is how much heat it can stand.

-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: Bias setting 50W. PA #qcx #pa

ajparent1/KB1GMX
 

Don,

It was heat or the device was bad initially from carpet lightning.

Keep in mind you can get a IRF510 hot enough to die without
even warming the heatsink  Simple way is make it dissipate 50W 
(heat it hard and fast its only a 43 W max device) or exceed the
max pulsed drain current of 20 A (blow the bond wires off it).
Usually in those cases you hear a pop and the case ruptures.
Reason the thermal resistance is high enough you can heat it
hard and fast before the heat transfers.

I would verify that nothing else is not quite right.  For example a
intermittent bias pot (or soldering) can easily bounce the bias
to 5V and the device will try to draw maybe 10A (and melt).

I don't doubt they failed but the cause is not conclusive and
heat is still the leading killer. 

The bias setting is per manual just below the threshold. I used a
Tripplite 630 set for 120ma to see that small increase and then
back off from it.

Actually doing the bias setting at 20V is procedure in the manual.

The transistor is not a big old power mosfet at 43W dissipation or
5.6A max current (at 25C).   If you want big ole, check IRF520
at 14A, but the gate capacitance at 670pf makes it tougher to
drive.

Also I've run my 'EBY amp at 26V for years (since '06) and it
can do a mere 55W at 20M with 2W drive. Its heatsink is
not the limiting factor (its larger) its the max current and its
power supply has a current limiter to 4A (that is 26V at 4A
or 104W DC input) and pushed to that limit its 75W
without blowing up.  Gets mighty hot if sustained.   During
testing 90W out (at 30V and 5A) was the smoke test, 
devices got to max die temp well before the heatsink was
warm.   Not bad for two IRF510s in my book.

Testing on my bench a relay driver using a IRF510, shorted coil
meant the moment the gate was driven the mosfet expired
instantly the tab never got warm. Peak current about 14A
at 28V.   Why?  The only limiting fact was the power supply
and leads and the on resistance of the IRF510 is about .6 ohms.
So the bond wires to the die evaporated.  The relay was defective!

FYI I have a MSC 100W 2M repeater amp that has an unregulated
DC supply nominal voltage for the MRF174 is 28V.  However the
Transformer-rectifier-capacitor idles (key up, bias off) at 35V
and at full power the DC droops to 27V.   Not a great design
and the PS is being replaced with a 12v to 28V switcher as it
had too much hum (rough audio) for SSB.

The average nominal 12V radio in the car will see as low as 11.7V
(most stop working right there) and as high as 16V  non -trivial swing.
The spec is + or - 15% centers at 13.8V or a 30% swing in operating
voltage.  

Allison 
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: 50W PA success on 40M, mostly Europe and Scandinavia

Mike Hoddy
 

With 22v from a SMPS and 1.8V I’m getting about 45W on 40M, it’s ideal for what I want.  Just ordered another one for 20M so fingers crossed!


73 de Mike, G0JXX


Re: QRP Monthly Challenge

Steve in Okinawa
 

Wow, I must stay awake next time for that time segment. I was on 40 shortly before 1300z and experienced the most rapid propagation collapse ever, from 599/599 with a Tokyo friend to unreadable the next minute. 20 was also dead of course, so i just shut down for the duration. JS6TMW 


Re: QRP Monthly Challenge

neil G4BRK
 

Tried the 1300z session on 20m, as I have several times in the last few months.
Still no QSOs at all despite numerous RBN spots for my CQs all including US.
At least I heard some stations this time :-)  Antenna a 1/4 wave vertical.

Maybe 20m isn't the best band for QRP in the current radio conditions.

73,  Neil  G4BRK


Re: Bias setting 50W. PA #qcx #pa

Don, ND6T
 

Wasn't heat. It was only keyed about 15 seconds. Immediately upon failure I felt the heat sink and it was still stone cold (the lab was unheated and 10 C.).
I originally thought that there had to be some other cause. Turned out to be the MOSFETs.
Yep!  Dummy load was a 250W thick film mounted on an enormous heat sink and better than 35 dB return loss. Input  during bias setting was BNC  mount with 32 dB RL. Power supply was a well filtered and regulated analog Harrison Labs. A separate 4.5 volt supply applied PTT signal.

Looks like a little bit of DIBL to me. Not something to be expected on a big 'ol power MOSFET. But sure feels like.

The amplifier works perfectly, just exactly like it is supposed to. Up to the point of failure (which is my fault entirely) the original transistors did, too. My point in posting was that it might be prudent to perform the bias setting at the highest value of supply voltage anticipated. Can't hurt.

I would not assume that there are many amplifiers built where such large excursions of supply voltage are encountered when considering bias settings. Most of those are in linear service and not as subject to variations in threshold. This is actually the first time that I have ever blown an IRF510, come to think about it. I've used them in a lot of rigs. An incredible device and dirt cheap. Then again, this is the first time that I've pushed them this hard. My bad.


Re: 50W amplifier: power rating of TR switch

Hans Summers
 

Hi John

See the references in the theoretical section of my 50W PA manual to the document by W6JL. And his QRZ page https://www.qrz.com/db/w6jl 

Don W6JL uses a single 1N5408 diode for switching 600W RF, forward biased with 100mA. 

The requirements on Don's switch are not the same as mine in the 50W PA. And it's different again in QSX, and in QCX. Often T/R switches need to be considered at the system level. A T/R switch design can't necessarily just be transplanted from one place to everywhere else. 

Note that sufficient forward bias current is very important, as I found out in another project. All the more so at lower frequencies, where eventually the 1N4007 loses its PIN characteristics and starts to return to rectifier behaviour; then you suddenly end up with a whole bunch of unwelcome harmonics starting with the 2nd. 

Solid state T/R switching is not, in my opinion, when done properly, a reasonable way to reduce current consumption of the switch, compared to relays. But solid state does have a lot of benefits compared to relays including smaller physical size, lower cost, silent operation, very fast switching and longer lifetime. 

73 Hans G0UPL 

On Mon, Jan 27, 2020, 16:39 John Seboldt K0JD <k0jd-l@...> wrote:
Glad to see the 1N4007-based diode TR switching in the 50W amp kit. I
also have a 120W linear amp kit, long ago completed but has been waiting
for final packaging with TR switching and filtering. Do you think this
switch will work at the 100W or so level? If not, who can guide me to a
nice polished design for a switch at the 100-200W level?

John K0JD
Milwaukee, WI






Re: si5351a issue spectral output above 90Mhz

ajparent1/KB1GMX
 

The biggest issue is phase jitter is essentially noise in the frequency
domain ie: FM.  All of the papers that address this try to translate that
from to amplitude where noise is measured (usually).  Its very system
dependent.

Since the SI5351 is really a UHF PLL followed by integer or fractional
(user decides) counters its easy and hard to evaluate noise and side
bands.  Every divide lowers the phase noise of the UHF VCO and the
multiply from 27mhz to PLL VCO frequency multiplys the the phase noise.

So on a 1:1 basis at 27mhz or below its likely to be as good or only slightly
poorer than the 27mhz reference.  At higher than 90-100mhz you in the
path problem if even pushing the internal PLL higher to 1ghz you max
divisor is only 10 for 100mhz!

Using a SA to evaluate spurs be very careful to not introduce spurs
or worse from incorrect technique.

For those that need better there is he SI570 (also a pain to program).
Or go to a crystal and multipliers.

For those that want cheap and easy above 100mhz, sorry you need to
look at other parts if it isn't good enough for you.  They exist but not
cheap and likely harder to use.

As is complaining about a device that was designed as a clock source
data systems is fortuitous for us RF weenies that it works.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: QRP Monthly Challenge

N3MNT
 

Two 1900hr contact both non QCX. First contact was impressed any plans to order QCX. 


Re: 50W amplifier: power rating of TR switch

ajparent1/KB1GMX
 

The 1N4007 is a 1A diode.

That is a magic intersection for power and current 1A into 50 ohms is 50W.
So for higher power you need a heftier diode with good RF switching and
they are both scarce or your forced to a RF PIN device at higher cost
as those for HF use do not go cheap.

At 120W (out I presume) that's well over 1.4A and you will need a heaftier diode.

Look here for another design for low power http://www.4sqrp.com/MagicBox.php
thats the Jim Kortge MagicBox.

That said I've run 1N4007 in parallel for a 100W amp but you need a few changes
to get that to work.  Also the reverse bias derived from the RF has to be higher so
the diodes used there also need higher voltage rating while sill rectifying at RF.
Scaling up to higher power is not trivial or easy.


Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: QRP Monthly Challenge

Hans Summers
 

Correction... JR1BLX just after 11pm local time (2000Z)... 

On Mon, Jan 27, 2020 at 11:16 PM Hans Summers via Groups.Io <hans.summers=gmail.com@groups.io> wrote:
Hi all

Family life intervened and I only managed the last 20 minutes... I did work on 40m, 7029.5:

YU7AE Kara QRP, but not QCX
OM5VS Vlado (not QRP)

I was very pleased to be called, after the OM5VS QSO and just after 10pm local time here, by Tada JR1BLX... yes I QSO'ed Japan 9,000km with my 5W QCX 40m... very nice!

73 Hans G0UPL

On Mon, Jan 27, 2020 at 11:04 PM Brian N7BKV <cl@...> wrote:
With my QCX 40 I clearly heard KJ4BQS at about 1945z on 7030.  No joy on contact though.  I am grid CM97vu in California near Yosemite NP.  Will be back on for 0300 round.

GL

Brian N
N7BKV


Re: QRP Monthly Challenge

Hans Summers
 

Hi all

Family life intervened and I only managed the last 20 minutes... I did work on 40m, 7029.5:

YU7AE Kara QRP, but not QCX
OM5VS Vlado (not QRP)

I was very pleased to be called, after the OM5VS QSO and just after 10pm local time here, by Tada JR1BLX... yes I QSO'ed Japan 9,000km with my 5W QCX 40m... very nice!

73 Hans G0UPL

On Mon, Jan 27, 2020 at 11:04 PM Brian N7BKV <cl@...> wrote:
With my QCX 40 I clearly heard KJ4BQS at about 1945z on 7030.  No joy on contact though.  I am grid CM97vu in California near Yosemite NP.  Will be back on for 0300 round.

GL

Brian N
N7BKV


Re: Reverse Polarity Protection [Was: Experiences with 50 watt Amp]

ajparent1/KB1GMX
 

John,

If at 100W your IC725 is only pulling 15A somethings wrong.  It will be 18-20A.
Typical SSB is about 220 too 240 W input power at nominal 13.8V.  The finals
run about about 45-to 55% so for 100W out they alone require 200W DC input
and the rest of the radio another 20-40W (VFO system, a processing and
transmit driver chain).  Measured radio is about 2A RX and 3 to 4.5A TX
no signal.

Parallel devices work if done correctly.  Incorrectly they can be unstable.
Obviously RF layout is everything and most RF MOSFETs are more
efficieint at 28V and higher.   I have an amp for 6M using eight total IRF510s
in parallel push pull (4x4) for a mere 220W at 30V.  Since the device s are
current limited to get more power you have to run more voltage (and deal
with the impedance). 

The 50W amp is class C at 70% (no worse than 60%) efficiency and the power out
means power in of about 72-83W average at 20V or about 4A.
Further the IRF510 has a max current of 5.6A!

The current draw is NOT twice 3.6A, it is 3.6A average as only one is on (conducting)
at any one point in time.  Class C push pull.

So during normal ops the dissipated heat is 22 to 30W and the heat sinks are adequate.
The yabut is that the IRF510 die has a 2.5 degree C rise for every watt of heat dissipated
due to the thermal resistance of the package (die to flange) is the limiting issue.

At 30W (80w DC in for 50W RF out)  between the two devices is 15W of heat each
and the die will heat and additional 37.5C to 62.5C (room temp is 25C).  That's at
60% efficient.  If we degrade that to 50% we are at 25W each device and we are at
a whopping 87.5C die temp.  Since the heatsink is at 25C it has thermal mass
(intertia) and will not instantly heat up and will take a while to reach temp while
the die is getting warmer.  As the heatsink heats the die is hotter still.

So despite the large enough heatsink you can still heat the device to destruction before
it warm the heat sink much over 50C.  Long key down is obviously a big problem.
IT was designed for class C and CW use.

FT8 has proven to the commercial transceiver users and builders that heat is an
issue and the only solution is big heatsinks and fans.  or a corresponding reduction
in power.  This was true for the old school AM and RTTY users heating 3-500Zs
to a nice orange glow.  Now people are frying their finals never noticing the
heatsink is very HOT.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: QRP Monthly Challenge

Brian N7BKV
 

With my QCX 40 I clearly heard KJ4BQS at about 1945z on 7030.  No joy on contact though.  I am grid CM97vu in California near Yosemite NP.  Will be back on for 0300 round.

GL

Brian N
N7BKV


Re: QRP Monthly Challenge

Martin DK3UW
 

heard a lot of stn but could not ge a single one despite good RBN feedback on 7 mhz
OK2BQN at the beginning but very week.

I give up for 2day
73s
Martin
DK3 UW


Re: Bias setting 50W. PA #qcx #pa

J68HZ
 

You beat me to this and it is the salient point of most in service part failures:

 

I've mentioned to people in the past (and likely forgotten) running the amp
with bias move s it from fairly efficient class C to less efficeint class AB or
AB1.  What you get for that is linearity (not required for CW) and HEAT. 
That heat is an enemy as the IRF510 has a thermal resistance from the
DIE to the flange that limits its ability to dissipate heat.   That leads to
a calculation where to keep the die under 150C you have to reduce
power for every degree C heating you have to reduce power by .29W
and we start at room temp of 25C.    The other ways to say that is for
every watt of power not put to the load is heating the device and it
increases temperature 2.5 degrees C for every watt.  So when you
run the bias up you get heat, if you key down for long periods you
get heat.  There is even with the generous heatsink there is a
finite limit.

Let’s remember that IRF510’s were designed to run in switching power supplies and as motor current switches with a duty cycle of 50% or less at  their listed full ratings.  The thermo-conductivity of the die with adequate heat transfer is what limits this part, and operating it at greater than 50% duty…. in a higher class… say AB2 or greater will cause more heat than design and possible part failure.  While this is mostly a design concern, one can surely get into trouble by cranking the on bias without understanding all of the limiting parameters involved.

 

 

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: Monday, January 27, 2020 1:35 PM
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] Bias setting 50W. PA #qcx #pa

 

All was well until the current dropped and the power meter did likewise. Quite the surprise.

Don,

How long was key down?  Reason I ask is what you thought you were seeing
was likely not what you thought.

IRF510 like most Hex, Tench, VMOS, and other LDmos FETs  exhibit a bias
point shift not with voltage, but temperature.   That is with a fixed bias and
fixed drain voltage increasing die temperature will see an increasing current,
which will heat the die further and increase the current, rinse later and
repeat to failure..

What that is the Gate threshold decreases with increasing temperature.
The other reason is that with high power out the device also heats, more
power more heat.  So what appears to be bias related and if set right
its very low should not cause that.  However is set high without understanding 
how that can head to failure.

See figure 7 of datasheet page supplies not gate threshold for 175C(very hot)
and 25C (room temp).  Tj means the temperature of the junction (die).

I've mentioned to people in the past (and likely forgotten) running the amp
with bias move s it from fairly efficient class C to less efficeint class AB or
AB1.  What you get for that is linearity (not required for CW) and HEAT. 
That heat is an enemy as the IRF510 has a thermal resistance from the
DIE to the flange that limits its ability to dissipate heat.   That leads to
a calculation where to keep the die under 150C you have to reduce
power for every degree C heating you have to reduce power by .29W
and we start at room temp of 25C.    The other ways to say that is for
every watt of power not put to the load is heating the device and it
increases temperature 2.5 degrees C for every watt.  So when you
run the bias up you get heat, if you key down for long periods you
get heat.  There is even with the generous heatsink there is a
finite limit.

For those the feel the IRF510 is insufficient for a inexpensive amp a
pair of MRF137, or MRFE101 would be far more robust at more than 
60$ for a pair.  Then again they need more supporting parts as well 
so a cheap amp is out of the question.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: Bias setting 50W. PA #qcx #pa

ajparent1/KB1GMX
 

All was well until the current dropped and the power meter did likewise. Quite the surprise.

Don,

How long was key down?  Reason I ask is what you thought you were seeing
was likely not what you thought.

IRF510 like most Hex, Tench, VMOS, and other LDmos FETs  exhibit a bias
point shift not with voltage, but temperature.   That is with a fixed bias and
fixed drain voltage increasing die temperature will see an increasing current,
which will heat the die further and increase the current, rinse later and
repeat to failure..

What that is the Gate threshold decreases with increasing temperature.
The other reason is that with high power out the device also heats, more
power more heat.  So what appears to be bias related and if set right
its very low should not cause that.  However is set high without understanding 
how that can head to failure.

See figure 7 of datasheet page supplies not gate threshold for 175C(very hot)
and 25C (room temp).  Tj means the temperature of the junction (die).

I've mentioned to people in the past (and likely forgotten) running the amp
with bias move s it from fairly efficient class C to less efficeint class AB or
AB1.  What you get for that is linearity (not required for CW) and HEAT. 
That heat is an enemy as the IRF510 has a thermal resistance from the
DIE to the flange that limits its ability to dissipate heat.   That leads to
a calculation where to keep the die under 150C you have to reduce
power for every degree C heating you have to reduce power by .29W
and we start at room temp of 25C.    The other ways to say that is for
every watt of power not put to the load is heating the device and it
increases temperature 2.5 degrees C for every watt.  So when you
run the bias up you get heat, if you key down for long periods you
get heat.  There is even with the generous heatsink there is a
finite limit.

For those the feel the IRF510 is insufficient for a inexpensive amp a
pair of MRF137, or MRFE101 would be far more robust at more than 
60$ for a pair.  Then again they need more supporting parts as well 
so a cheap amp is out of the question.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO


Re: si5351a issue spectral output above 90Mhz

Glen Leinweber
 

Hans,
Thank you for adding phase noise measurements for SI5351. Just another
value-added QRP_Labs feature for those of us who build-to-learn. Noise
floor measurements seem prone to mis-interpretation and measurement
errors. Including your noise-measurement setup progression is particularly
enlightening.

Hans didn't test fractional MultiSynth dividers for the very good reason that
QCX uses integer MultiSynth division only. Don't confuse the SI5351
MultiSynth PLL with its MultiSynth divider. I believe that QCX uses
fractional division in the PLL, but integer division in the divider. Seems very
reasonable that QCX spurious response is quite good with this
arrangement: SI5351 configuration (187 internal registers!) can be set
up many different ways, impacting spurious response.

Hans' measurements of phase noise would include spurii, but I'm wondering if all the
noise averaging would hide sharp spurious responses of jittery PLL, mushing them into
the measured noise floor of about -135dBc/Hz? I see that some phase noise is
measured not including spurs, i.e. between spurious peaks.
Perhaps the internal SI5351 fractional PLL that takes 27 MHz up to nearly 900 MHz
is mostly contributing to the -135dBc/Hz measured noise floor? All guesses, since
the SI5351 data sheet is so sketchy. Silicon Labs could learn a thing or two
about documentation from QRP-Labs ;-)