moderated Class C Amplifier Application #transmit


Bob McGraw - K4TAX
 
Edited

I've not seen any clear or definitive answer to the question "can one successfully run FT-8/4 or other WSJT-X modes via a Class C solid-state VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX


alanbh@...
 

The answer would be NO. FT8 and FT4 is a tone modulate SSB type signal. You cannot put an SSB or modulated signal through class C amplifier without distorting the signal. The maximum class you can put a modulated of SSB type signal through without or absolute minimal distortion is a class B Amplifier.

73, alan WA3EKL

----- Original Message -----
From: "Bob McGraw - K4TAX" <rmcgraw@benlomand.net>
To: WSJTX@groups.io
Sent: Tuesday, July 27, 2021 7:05:24 PM
Subject: [WSJTX] Class C Amplifier Application #transmit


I've not seen any clear or definitive answer to the question "can one
suddessfully run FT-8/4 or other WSJT-X modes via a Class C solid state
VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX


Bill Somerville
 

On 27/07/2021 20:05, Bob McGraw - K4TAX wrote:

I've not seen any clear or definitive answer to the question "can one suddessfully run FT-8/4 or other WSJT-X modes via a Class C solid state VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX
Bob,

sure, so long as the PA has sufficient harmonic suppression for Amateur Radio use. All the WSJT-X modes use constant envelope output and linear amplification is not required for that. Note that MSK144 is a little wider than most typical SSB transmit filters so truncation of the modulation frequency extremes may introduce enough AM content that could make your output, amplified by a non-linear PA, wider than that allowed in the sub-bands where WSJT-X modes are normally used. If you intend to use MSK144 with that PA you should probably check your Tx signal for IMD products before connecting to a high gain aerial.

73
Bill
G4WJS.


Bill Somerville
 

On 27/07/2021 20:14, alanbh@cablespeed.com wrote:
FT8 and FT4 is a tone modulate SSB type signal.
Alan,

that is not a description of any modulation type used by WSJT-X. Frequency shift keyed signals are frequency modulated, they have no AM content so long as the whole bandwidth of the signal is transmitted as intended.

73
Bill
G4WJS.


Rik Strobbe
 

Although SSB modulated FT8 and FT4 are FSK signals and thus don't require linear amplification.
So a class C (or D or E) amplifier can be used wthout distorting the signal.
On LF and MF it is common practice to use high efficient (90% or more) class D or E amplifiers for FT8, JT9 etc.

73, Rik OR7T
________________________________________
Van: main@WSJTX.groups.io <main@WSJTX.groups.io> namens alanbh@cablespeed.com <alanbh@cablespeed.com>
Verzonden: dinsdag 27 juli 2021 21:14
Aan: main@WSJTX.groups.io
Onderwerp: Re: [WSJTX] Class C Amplifier Application #transmit

The answer would be NO. FT8 and FT4 is a tone modulate SSB type signal. You cannot put an SSB or modulated signal through class C amplifier without distorting the signal. The maximum class you can put a modulated of SSB type signal through without or absolute minimal distortion is a class B Amplifier.

73, alan WA3EKL

----- Original Message -----
From: "Bob McGraw - K4TAX" <rmcgraw@benlomand.net>
To: WSJTX@groups.io
Sent: Tuesday, July 27, 2021 7:05:24 PM
Subject: [WSJTX] Class C Amplifier Application #transmit


I've not seen any clear or definitive answer to the question "can one
suddessfully run FT-8/4 or other WSJT-X modes via a Class C solid state
VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX


Michel Bernard VE2BJG
 
Edited

Everyone is blinded by the word "FSK tones" and assume a class C amplifier will work... It won't.

We are broadcasting 8 different discrete tones simultaneously (8-PSK) and that won't simply pass through a class C amplifier without excessive intermodulation products.

You need strict class B, AB1 or A amplifiers. And good ones at that. Simply observe any bands when some hams try to do FT8 with older kilowatt junk amplifiers. They are painting the entire band with junk. They are the one claiming there is no need for an ALC feedback line by the way...

Just read some of the PDF documents available on a search engine (try "8PSK modulation") and keep reading keep reading and again keep reading...

Alan WA3EKL is correct.


Bill Somerville
 

On 28/07/2021 17:15, Michel Bernard VE2BJG wrote:
Everyone is blinded by the word "FSK tones" and assume a class C amplifier will work... It won't.

We are broadcasting 8 different discrete tones simultaneously (8-PSK) and that won't simply pass through a class C amplifier without excessive intermodulation products.

You need strict class AB1 or A amplifiers. And good ones at that. Simply observe any bands when some hams try to do FT8 with older kilowatt junk amplifiers. They are painting the entire band with junk. They are the one claiming there is no need for an ALC feedback line by the way...

Just read some of the PDF documents available on a search engine (try "8PSK modulation") and keep reading keep reading and again keep reading...

Alan WA3EKL is correct.
Michel,

I don't know where you are getting your information on FT8 modulation, but you are wrong. The modulation is 8-FSK, i.e. only one tone transmitted at any time. Further to that, the tone transitions are Gaussian filtered in the frequency domain so even the frequency transitions are smoothed. There are no amplitude variations, perhaps you need to look carefully at the waterfall in WSJT-X where it should be quite obvious how FT8 is modulated!

73
Bill
G4WJS.


Kai-KE4PT
 

Hi Bob,
FT4 and FT8 are each constant envelope signals with just one signal at a time (one"tone") in the FT4 or FT8 bandwidth. They are purely single-tone FSK signals, and as such there is no reason why they cannnot be passed through a class C amplifier, since there is no "amplitude" content to contend with. Just think of "SSB" mode as the lowest frequency IF in your transceiver chain and FT4 or FT8 are sequential single tones in the full SSB bandwidth. Note that this is not true for modes like PSK31 and QPSK family of protocols, because they DO have an amplitude variation component.
Kindest regards,
Kai, KE4PT



On 7/27/2021 15:14, alanbh@... wrote:
The answer would be NO.  FT8 and FT4 is a tone modulate SSB type signal.  You cannot put an SSB or modulated signal through class C amplifier without distorting the signal. The maximum class you can put a modulated of SSB type signal through without or absolute minimal distortion is a class B Amplifier. 

73, alan WA3EKL

----- Original Message -----
From: "Bob McGraw - K4TAX" <rmcgraw@...>
To: WSJTX@groups.io
Sent: Tuesday, July 27, 2021 7:05:24 PM
Subject: [WSJTX] Class C Amplifier Application #transmit


I've not seen any clear or definitive answer to the question "can one 
suddessfully run FT-8/4 or other WSJT-X modes via a Class C solid state 
VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX












Ben Nardi
 

Hi All,

If you refer to the WSJT-X manual under Protocol Specifications section 17.1. Overview, the last sentence "Generated waveforms for all of the WSJT-X modes have continuous phase and constant envelope." states all WSJT-X modes are constant envelope which means no amplitude variations so class C amplifiers are suitable.

Regards,
Ben W3ZUP


On 7/28/2021 1:29 PM, Kai-KE4PT via groups.io wrote:
Hi Bob,
FT4 and FT8 are each constant envelope signals with just one signal at a time (one"tone") in the FT4 or FT8 bandwidth. They are purely single-tone FSK signals, and as such there is no reason why they cannnot be passed through a class C amplifier, since there is no "amplitude" content to contend with. Just think of "SSB" mode as the lowest frequency IF in your transceiver chain and FT4 or FT8 are sequential single tones in the full SSB bandwidth. Note that this is not true for modes like PSK31 and QPSK family of protocols, because they DO have an amplitude variation component.
Kindest regards,
Kai, KE4PT



On 7/27/2021 15:14, alanbh@... wrote:
The answer would be NO.  FT8 and FT4 is a tone modulate SSB type signal.  You cannot put an SSB or modulated signal through class C amplifier without distorting the signal. The maximum class you can put a modulated of SSB type signal through without or absolute minimal distortion is a class B Amplifier. 

73, alan WA3EKL

----- Original Message -----
From: "Bob McGraw - K4TAX" <rmcgraw@...>
To: WSJTX@groups.io
Sent: Tuesday, July 27, 2021 7:05:24 PM
Subject: [WSJTX] Class C Amplifier Application #transmit


I've not seen any clear or definitive answer to the question "can one 
suddessfully run FT-8/4 or other WSJT-X modes via a Class C solid state 
VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX














Hank Pfizenmayer
 

 Everyone who is interested in exactly how  the FT4 and FT8 signal is generated and the extreme care taken to limit bandwidth by careful treatment of the transistions between tones should read  -

https://physics.princeton.edu/pulsar/k1jt/FT4_FT8_QEX.pdf
Paragraph 5 discusses this
page ten  figure 3 shows a theoretical  FT4 spectrum --  I have to admit I have not ground thru the math to calculate a FT8 spectrum - and I have not been able to con anyone into doing that !.

If you go to WSJT home page

https://www.physics.princeton.edu/pulsar/K1JT/

and "references" its the last one in the list .  

Now while we are on the subject however -- I see a few VERY FEW  signals that the close in "distortion" or whatever it is -  is down some 55 dB - The vast majority are more like 40 -  45 dB - I am NOT talking about 120hz hum products . 

So I am curious what the FT8 close in  levels are. -I suspect one of the Gurus knows. 

Hank K7HP 


Bill Somerville
 

On 28/07/2021 20:00, Hank Pfizenmayer via groups.io wrote:
 Everyone who is interested in exactly how  the FT4 and FT8 signal is generated and the extreme care taken to limit bandwidth by careful treatment of the transistions between tones should read  -

https://physics.princeton.edu/pulsar/k1jt/FT4_FT8_QEX.pdf <https://physics.princeton.edu/pulsar/k1jt/FT4_FT8_QEX.pdf>
Paragraph 5 discusses this
page ten  figure 3 shows a theoretical  FT4 spectrum --  I have to admit I have not ground thru the math to calculate a FT8 spectrum - and I have not been able to con anyone into doing that !.

If you go to WSJT home page

https://www.physics.princeton.edu/pulsar/K1JT/

and "references" its the last one in the list .

Now while we are on the subject however -- I see a few VERY FEW signals that the close in "distortion" or whatever it is -  is down some 55 dB - The vast majority are more like 40 -  45 dB - I am NOT talking about 120hz hum products .

So I am curious what the FT8 close in  levels are. -I suspect one of the Gurus knows.

Hank K7HP
Hi Hank,

side bands of FSK modulation, using continuous phase, are normally expected to be >60 dB down from the carrier amplitude, OTOH PSK side bands would be of the order of 33 dB down from the carrier amplitude and therefore rather higher than would wise to use a non-linear class of amplifier with on a crowded Amateur band.

73
Bill
G4WJS.


Michel Bernard VE2BJG
 

Bill G4WJS you were right and I was wrong. Indeed constant enveloppe signals may be amplified with class C amplifiers. There are no amplitude components in the FT8 modulation.

There is a nice short article on "Advantages of Constant Envelope Modulation" on the sss-mag.com website.

Nice reading.

Thanks for private remarks.

73
Michel
VE2BJG


Evan Hand
 

Hi Hank,

I do not remember exactly the values, but I did some signal purity tests for two low power FT8 transceivers:
QRPGuys AFP-FSK Digital Transceiver III by QRPGuys and The Phaser by Midnight Design Solutions
https://qrpguys.com/qrpguys-digital-fsk-transceiver-iii
https://midnightdesignsolutions.com/phaser/

The results were drastically different driven by how the transmitted signal was generated. 
- In the case of the QRPGuys rig, it is measuring the audio signal input frequency and then adjusting a Si5351 DDS directly.  Since this is at audio frequencies the PIC controller is more than enough for this application.  This one had the best signal, and I would feel very comfortable feeding it into a class C amp.  The drive setting only had a threshold where the VOX would detect and turn on the transmitter.  Going past that point had no effect on the signal as far as I could tell
- The Phaser was not so good and could be over and under driven resulting in low power or a very "splattered" signal.  Evey with careful adjustment of this rig I would hesitate to feed it into any amp, let alone a class C amp.  The Phaser is taking the audio signal and generating an I/Q pair to then feed a pair of Tayloe mixers that are then summed together to get the RF frequency.  Since the phase shift is done using audio op-amps I am not sure how accurate the result could be.  Any phase or amplitude variations will show up as signal distortions.

The QRPGuys transceiver had very low close in generated noise, and the 3rd harmonic was down by over 45db.  I did not do the harmonic measures on the Phaser as the signal was not good enough for me and I have abandoned it.  Close-in measurements were made with an RSP1a using the SDRPlay Spectrum Analyzer software using a dummy load and a -50db tap.  Harmonics were measured with the same dummy load and tap with a TinySA spectrum analyzer.

Note:  I have found that I need both devices to get close to what a Rigol or Siglent SA can do.  The total cost for both is under $200 compared to close to $1000 for the lower cost Rigol.

Since both transmitters are under 5watts, I did not add a step attenuator between the -50db tap and the test instruments, as the power level is well below the input threshold of either device.  When I test my 100watt rig I add that into the chain.

Above are my observations, and could be in error.  I am always open to corrections or expansions on them.
73
Evan
AC9TU




Kai-KE4PT
 

All that said, let's not forget DXpedition mode where the DX station might send multiple FT8 signals simultaneously. THEN a clean linear amp is necessary.
73,
Kai. KE4PT


On 7/28/2021 15:34, Michel Bernard VE2BJG wrote:
Bill G4WJS you were right and I was wrong. Indeed constant enveloppe signals may be amplified with class C amplifiers. There are no amplitude components in the FT8 modulation.

There is a nice short article on "Advantages of Constant Envelope Modulation" on the sss-mag.com website.

Nice reading.

Thanks for private remarks.

73
Michel
VE2BJG




KC2WTG <kc2wtg@...>
 

Why use a amp mostly low power 50watts and less


On Wed, Jul 28, 2021 at 12:15 PM, Michel Bernard VE2BJG
<ve2bjg_qbc@...> wrote:




Michel Bernard VE2BJG
 

I welcome my 200 watts of RF out to punch through poor 6m conditions. That's when brute power is needed. Otherwise on HF I am amazed at QRP stations.


Bob McGraw - K4TAX
 

In my application, I am working 2M meteor scatter.   I find that power and antenna gain are both needed.    I just happen to have a solid-state Class C commercial amplifier capable of about 500 watts output. 

There are two very good meteor showers in August.  

73
Bob, K4TAX


anthony.simons@...
 

Why don't a couple of you try it and see?

On Thu, Jul 29, 2021 at 10:57 AM Bob McGraw - K4TAX <rmcgraw@...> wrote:
[Edited Message Follows]

I've not seen any clear or definitive answer to the question "can one
successfully run FT-8/4 or other WSJT-X modes via a Class C solid-state
VHF amplifier?"  Asking because I have a VHF amplifier which is not linear.

73

Bob, K4TAX








--
73, de W8AF  ~  Anthony  


Andy TALBOT
 

FST4 goes very nicely through my Class D 137kHz PA
and 400W class E for 475kHz

As does WSPR, and JT9
We don't use FT8 ion those bands


Joe
 

As has been written here and elsewhere repeatedly, the WSJT-X modes all use constant-envelope waveforms. 

YES, you can use a Class C amplifier.

YES, there are good meteor showers in August.  But with MSK144 you don't really need a shower.  You can make meteor-scatter QSOs any day of the year.

  -- 73, Joe, K1JT