Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 [2] 3 Next   Go Down
  Print  
Author Topic: Software Pecker  (Read 8728 times)
K4DPK
Member

Posts: 1077


WWW

Ignore
« Reply #15 on: September 01, 2012, 08:52:32 PM »

G7EIX: " the only difference I have implemented in S/W is the ability to adjust the frequency of the tone and the amplitude."

There's the problem.....changing the amplitude of the pulse varies the drive and consequently, the load settings.

Phil C. Sr.
k4dpk

Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #16 on: September 01, 2012, 10:00:50 PM »

There's the problem.....changing the amplitude of the pulse varies the drive and consequently, the load settings.

Logical.  By default I set it at 20% of the max.  So if that's the case...then cranking the amplitude to 100% on the slider would emulate the h/w version fully.
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #17 on: September 02, 2012, 04:30:32 AM »

I'm not saying you can't make your system work and be useful. A pulser can be useful, but to build and sell something, and especially to write instructions, we have to know how it really works.


 
If someone has tried your device and found it results in lower load settings, that very clearly means your device has a problem. The fact someone got grossly different tuning results, resulting in a more meshed load control from normal methods, means you system has an operational defect.

If this is the case, then I have simply replicated the defect that exists in the H/W version.

That's probably true. The basic idea, other than the porn name, complexity, and how it connects, is no different than a CW pulser.  :-)

Quote
The output tone is the same - the only difference I have implemented in S/W is the ability to adjust the frequency of the tone and the amplitude.

The ampliftude should always be at 100% or slightly more of the expected absolute maximum PEP levels for voice or CW.

As for tone and frequency, apparently there is a common misunderstanding of what an RF single-tone signal is. I've seen PE's be totally lost at what a single tone or multi-tone SSB signal really is. Look at the fellow on the west coast who evaluated a radio frequency electronic bias system using an audio tone! He never has gotten his head around the difference between 30 Hz and 30 MHz.

A single tone SSB signal is equal to a steady CW carrier. A pulsed audio tone, other than passband characteristics of any IF filters or audio stages following the tone injection point, is exactly the same as a CW carrier. The only difference between injecting a CW pulse and a SSB single-tone pulse is the frequency response of the system before the signal reaches the mixer that converts it to the final radio frequency.

If I have a 3.80003 MHz signal from a 30 Hz tone that makes it through the entire system, it would be exactly the same as a 30,000 Hz tone that makes it though the system, except for a radio frequency shift in frequency to 3.830000 MHz.  If the CW signal rise and fall time is in the passband of the modulation path and filters, and a 1 kHz audio tone is in the passband of the modulation path and filters, they are exactly identical for the same pulse rate and duration.

With that in mind, the only thing changing the tone frequency does, if the filters pass it, is change the output frequency the same as turning the VFO dial does.

Collins, in early 30S1 and KWM2 rigs, injected an audio tone into the SSB input to make CW. The only difference was the audio tone method had wider bandwidth and troubles with spurious signals, so they had to use a better system.

Quote
Like I said above, I don't own an AMP, so this was created for an existing owner of the H/W Pecker.   So as to whether it is flawed or not - I cannot comment further.   I shall bow to those with experience and knowledge far greater than mine in these areas.

It was a fun project to play with - but from this point on I have other projects to occupy my mind.

Well, I hope it is a learning experience in understanding how these systems work. My general view is we should always understand what we are designing or emulating as much as possible, otherwise we can accidentally degenerate a design or an idea. Unfortunately it appears the basic idea of using a specific pulse duty rate and audio was flawed from the very concept, and probably went unchallenged.

I'm not saying a tuning pulser is a bad idea. I am saying the notion an audio injected pulsed single tone, other than rise and fall rates (which creates sidebands and make it a multiple tone signal that occupies wide bandwidth), is not any different than just running a keyer on dots once it is inside the bandwidth.

I'm just puzzled what changing the "tone" actually does in the system you have. Changing pulse width is a good idea because it can be tailored to the meters used  and the bandwidth of the injection and filtering system beyond the pulse generator.

However, this this "audio tone" and "speech duty cycle" idea got off on the wrong foot the minute it was proposed. An off-and-on audio tone through a SSB rig is exactly the same as a single tone CW off and on signal, except for a slight shift in output frequency and limits of the modulation system frequency response.

If I had a 5 Hz audio tone that made it though the audio and filters, it would be an RF signal absolutely unnoticably different than a 30 kHz audio tone that made it through the audio and filters, and the same as a CW carrier, if the VFO was adjusted slightly (30 kHz) to make them appear on the same frequency. The only way to emulate voice is with a multiple tone signal, so AG6K (the originator of the duty cycle theory, so far as I know) and the people who followed that idea are really emulating CW dits injected into a mic jack.

The critical thing is staying within limits of power detector meter response and the modulation system, otherwise the audio or CW systems are the same.

Also, claims a pulser will prevent arcs are just not true. A pulser reduces heat by reducing duty cycle, but it should not alter peak voltages or reduce chances of arcing at all. Duty cycle affects heat, not peak voltage.

If you wanted to add features, they could have been rise and fall time adjustment, pulse repetition rate, and pulse duty cycle so the pulser could match the metering and any ALC system. One thing is critical, the peak power from your pulser should be slightly greater than the maximum PEP power ever expected on SSB, and the meters have to respond to those peaks.

With a steady carrier, the meters will always work. :-)


73 Tom
« Last Edit: September 02, 2012, 04:47:14 AM by W8JI » Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #18 on: September 02, 2012, 08:13:38 AM »

Tom,

I appreciate the time and effort you have put into your responses.  My understanding increases each and every time I read them.   Taking your points on board and I may revisit the whole thing in the coming weeks to see if I can produce something more beneficial than a reproduction of something that already exists.

Regards and 73's

Richard
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #19 on: September 02, 2012, 08:34:16 AM »

Tom,

I appreciate the time and effort you have put into your responses.  My understanding increases each and every time I read them.   Taking your points on board and I may revisit the whole thing in the coming weeks to see if I can produce something more beneficial than a reproduction of something that already exists.

Regards and 73's

Richard

It's all about making things better for fellow Hams.  I hope this helps all of us, myself included, better understand how this all works and solutions.

What surprises me more than anything is how people fell into the trap of thinking a pulsed single-tone emulates voice, while another pulsed single-tone (CW) does not.

Also there seems to be a lack of understanding of failures and stresses. The only thing pulsing does, for a fixed peak level, is reduce heat. It cannot reduce arcing, although it can make an arc occur in pulses.

One problem is bandwidth of the tune signal and offset frequency, which affect annoyance of others on adjacent SSB channels. That is determined by the actual RF frequency and any audio harmonics, plus the shape of rise and fall. The audio pulse is limited by the bandwidth of SSB filters and audio stages. The CW pulse is limited by the bandwidth of the keying system and rise and fall characteristics, which is generally narrower than SSB.

The end results in tuning are identical with identical potential problems. Those problems include meter response and making sure full peak power is reached by the exciter in the tuning process, and that the amplifier is not saturating or non-linear. Neither method is any different for that, and all methods are equally critical for pulse shape, pulse level, and pulse rate.

No matter what method is used, grid current should be checked at maximum drive. Both methods, when applied correctly, give exactly the same result regardless of injection method. It may be necessary to use 100% duty, or someone might get away with 2% duty cycle, but any duration and injection method will have the same result if transmitter limits are not hit, and if the meters track the pulses correctly.

The two critical adjustment parameters are pulse repetition rate and duty cycle, which are defined by the response of the system. They are NOT defined by some imagined voice duty cycle emulation. If we want to emulate voice, we have to use at least three tones in a test...unless someone speaks in a single one-frequency pulsed tone.  :-)

73 Tom
Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #20 on: September 02, 2012, 10:31:28 AM »

I did some quick audio captures using the Flex and Audacity to look at the Spectral Analysis of what is generated using the Pecker set at 35hz and also 150hz.

If anyone is interested in what the Pecker generates, the images are here:

Audacity Frequency Analysis (35hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/35hz.jpg

Audacity Audio Capture (150hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/audacity.jpg

Audacity Frequency Analysis (150hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/150hz.jpg

Audacity Frequency Analysis (SSB from MD-200 Mic) (post processed audio from Flex)
http://pecker.g7eix.com/images/voice.jpg
« Last Edit: September 02, 2012, 10:40:03 AM by G7EIX » Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #21 on: September 02, 2012, 11:05:48 AM »

I did some quick audio captures using the Flex and Audacity to look at the Spectral Analysis of what is generated using the Pecker set at 35hz and also 150hz.

If anyone is interested in what the Pecker generates, the images are here:

Audacity Frequency Analysis (35hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/35hz.jpg

Audacity Audio Capture (150hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/audacity.jpg

Audacity Frequency Analysis (150hz) (post processed audio from Flex)
http://pecker.g7eix.com/images/150hz.jpg

Audacity Frequency Analysis (SSB from MD-200 Mic) (post processed audio from Flex)
http://pecker.g7eix.com/images/voice.jpg

 

I'm not sure what you are implying, what that means, or why that is important. Can you explain what you think it all means?
Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #22 on: September 02, 2012, 11:16:54 AM »

I'm not sure what you are implying, what that means, or why that is important. Can you explain what you think it all means?

Not implying anything.   I got a few emails from people asking me what the Pecker does / produces.   So I have added these to help people visualize.   
Logged
KM3F
Member

Posts: 501




Ignore
« Reply #23 on: September 02, 2012, 11:44:40 AM »

Problem I have with the wave shapes is they are pointed and not anywhere near a square tops.
To use time and width pulses, they should have some dwell time at peak value to be sure the amplifier gets to it's design limit  or your intended drive limit and the readout is able to see this dwell time long enough to register the peak correctly.
Still, driving the amp at less than full on is still not correct even if the meter is able to read correctly.
Just my view about the dynamics.
« Last Edit: September 02, 2012, 12:16:53 PM by KM3F » Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #24 on: September 02, 2012, 01:08:21 PM »

As of today I have frozen this project.   But as a result of a request from a couple of people who want to experiment I have added the ability to change the Waveform from a Sine-Wave to a Square-Wave.

I will revisit the whole tuning aid project again later in time when I acquire myself my very own Amplifier.

Thanks to all for the input, be it good, bad or indifferent.   

Regards

Richard
Logged
K5USF
Member

Posts: 83




Ignore
« Reply #25 on: September 02, 2012, 02:38:12 PM »

Good luck Richard.  73s Jim
Logged
K0IZ
Member

Posts: 737




Ignore
« Reply #26 on: September 02, 2012, 05:43:17 PM »

Tom, there is a difference between a pulsed single tone and a pulsed CW:  The tone and CW pulses can be equivalent but the radio may be operating with different parameters on SSB vs CW.  If the point is to tune up an SSB amp at full PEP, but much less average (which the pulsed tones do), then the radio should be running in SSB.  Else no assurance that the radio drive in CW will be same as the SSB drive.   I have used the pecker for a number of years, and adjusted the average to about 20% of normal peak.  Very easy on amp.   John.
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #27 on: September 03, 2012, 05:53:01 AM »

As of today I have frozen this project.   But as a result of a request from a couple of people who want to experiment I have added the ability to change the Waveform from a Sine-Wave to a Square-Wave.

I will revisit the whole tuning aid project again later in time when I acquire myself my very own Amplifier.

Thanks to all for the input, be it good, bad or indifferent.   

Regards

Richard

Hi Richard,

I think you are missing the point of how the system works. The distortion in a SSB transmitter is not because of a time-frozen audio spectrum, which is what the square wave tells, but envelope distortion. It is a transfer function issue.

If the square wave was the shape of the RF envelope, and if it was perfectly flat, then you would be adjusting the amplifier at the crest of that square-shaped envelope. That could easily be done with a sinewave at audio. The envelope shape is the rise and fall times and amplitude level of the sine wave going into the TX audio.

The only thing making a square waveform instead of sine waveform at audio does is add odd-harmonics to the frequency content.

The issues with your system are the 20% power thing, and the tone duration and tone rise and fall times. For proper tuning the meters have to be able to follow and indicate the PEAK level of the envelope, and that relates to duration, slope, and repetition rate. The exciter power also has to be at, or slightly above, the maximum PEP power of the exciter that will ever be applied. Peak envelope exciter pulse power should be directly tied to the maximum power setting or limit of the exciter that will ever drive the PA. The meters must be able to track the envelope peaks.

So again, if you had a single tone (which is either through the mic jack or through CW keying) and had the envelope TOP  reach and hold maximum possible peaks with enough duration and proper rates for meters to follow, the system would work. It isn't the shape of the cycles in the tone, but the shape and duration of the tone that matters.

John K0IZ wrote:
 
Quote
Tom, there is a difference between a pulsed single tone and a pulsed CW:  The tone and CW pulses can be equivalent but the radio may be operating with different parameters on SSB vs CW.  If the point is to tune up an SSB amp at full PEP, but much less average (which the pulsed tones do), then the radio should be running in SSB.  Else no assurance that the radio drive in CW will be same as the SSB drive.   I have used the pecker for a number of years, and adjusted the average to about 20% of normal peak.  Very easy on amp.   John.

John,

Virtually all modern radios with SSB work the same way. I'd appreciate you pointing out any exception to this, other than something like a T4XC Drake that uses grid current detection on the PA tube that limits SSB peak but not CW peak. 

They have a power detector on the output that samples antenna port power, usually combined with reflected power. That detector feeds a comparator. The comparator has a DC bias that sets the limit of peak detector voltage, which effectively limits peak envelope power. The comparator output voltage is used to gain control an early stage, just like AGC in a receiver.

It does not matter if we run CW or SSB, the peak envelope power is limited to the same value by the gain control loop.

If you run CW dots at an effective weight of 20% duty, you have exactly the same results as running an off-on audio tone at a duty cycle of 20%. There is no difference at all. If the transmitter has no distortion and perfect carrier and sideband suppression, and if the tone is a pure sine wave, the RF output is exactly identical for CW or SSB. Peak power is identical, and so is average power.

This is just how a SSB transmitter works.

I'm not sure where the misconceptions come from, but my best guess is they come from people not understanding a tone going in is really just mixed to a new frequency in the "radio" frequency range. Once at the radio range, it is all the same. For example, if I inject 30 Hz into an audio chain of a distortion-free ESSB transmitter, the output is a radio frequency carrier absolutely indistinguishable from a 10,000 Hz tone.

http://www.w8ji.com/electronic_bias.htm#The_Signal_Detected

My guess is the misconception comes because unregulated power supplies can have higher voltage with lower duty cycle. When someone locks a tone or a carrier (each are the same) on for a period of time, the power supply can sag. When it sags peak power is reduced. The cause of this is absolutely not the mode. The problem is the duty cycle. If you were on SSB and whistled steadily, the voltage would sag exactly like holding a key for the same duration at the same power.

For the amplifier, if voltage sags, you can tune exactly like you would with full voltage and get the same results. This is because the optimum operating impedance of the output device stays about the same. A modest reduction in voltage results in a reduction of current so impedance remains the same. This is why Heathkit told people to tune the SB220 in the CW/TUNE voltage position, and then just flip to SSB voltage. That was a substantial anode supply voltage change, but the impedance tracked for the same drive level.

All that aside, even if tuning at a different voltage was an issue, there is absolutely no difference between a "CW ditter" and a pulsed audio tone. BOTH have to match the meters so we can read the meters, and both have to be at full exciter drive.

Anyone having a good feel for how SSB transmitters and amplifiers work should realize all this, but recently some people have ran with some theoretical misinformation from one source without considering how the entire system works, and assumed a certain duty cycle pulsed-tone results in "better tuning".

There is no argument about reduced heat from pulsed tone (which absolutely includes CW dits because they are identical). The danger or problem is in getting people to understand the tuning must be done at absolute full peak envelope power and the meters must accurately respond.

By the way, because we use something for a number of years does not mean we have been using it correctly, or we know what it does. It doesn't mean it works like we think, or doesn't work. It doesn't mean it is right or wrong.  It simply means we used it.    Grin

What is more important is understanding how the system works, rather than if someone is content.

73 Tom


 
Logged
G7EIX
Member

Posts: 19


WWW

Ignore
« Reply #28 on: September 03, 2012, 09:00:59 AM »

The issues with your system are the 20% power thing, and the tone duration and tone rise and fall times.

I really do wish I could take the credit for this whole thing (or maybe I do not), but I simply created an EMULATOR (with a few adjustable options) of an original idea.     Grin

http://www.qsl.net/ei7ba/Woodpecker.htm

Also, let us look on the bright side.  You don't have to spend $60 to try out a Pecker, you can download one, try it, and then delete it if you want.   No poorer for the whole experience.


« Last Edit: September 03, 2012, 09:10:28 AM by G7EIX » Logged
G3RZP
Member

Posts: 4479




Ignore
« Reply #29 on: September 03, 2012, 09:24:43 AM »

They were popular for a while in the 1970s, when sweep tube amplifiers were all the vogue. I really cannot see the point of it, though - the length of time it SHOULD take to tune a PA to full output on a carrier is short enough to avoid problems.

Unless you stop half way through to stuff more tobacco in your pipe and then light it - which I've seen done!
Logged
Pages: Prev 1 [2] 3 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!