1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

SSB Carrier Supression

Discussion in 'General Electronics Chat' started by Space Varmint, Dec 3, 2008.

  1. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA

    Hey thanks there flat5. You sound like a ham...?? Scrooge just likes ducks...lol. Maybe I'll throw up another recording. ..Wait here's some AM: http://www.mediafire.com/file/z23m12zmxwm/Dave&Joyce.MOV

    But I think he heard a unfinished version of the receiver with SSB recordings. They did have a quacking sound on strong signals because the free running oscillators were being loaded down. I put diode clamps on all three oscillators and completely eliminated it.
     
  2. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,445
    Likes:
    937
    Location:
    Canada, of course!
    The Dave and Joyce movie starts with a cow (?) sound. Then a man and woman talk and laugh with horrible sound quality. No high audio frequencies except high distortion.

    I saw a blue LED flashing with the sounds. Maybe it causes the distortion.
     
  3. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    Sorry Varmint, it does sound much worse than necessary.
    Lots of distortion. Do you have access to an RF signal generator?
    Audio generator? Scope?

    "The problem is, I want to adjust for upper and lower sideband using one crystal but don't think I can warp it that far.

    What is the recommended difference frequency again? It's like 15K apt it?"

    If I understand what you are asking, you want the carrier 2k from the filter or a little less.

    I don't remember your IF freq. but below 1mhz, you won't be able to pull the xtal even 1kHz, I think. I have wondered if a ceramic resonator would be good for this but have not researched it.

    Edit: not sure you could call this research, but I found this right away.
    Ceramic Resonators

    more info:
    Ceramic Resonators

    data sheet:
    http://www.electro-tech-online.com/custompdfs/2008/12/p16e9.pdf

    Once more:
    VFO with Ceramic Resonator
    Oscillators

    Simple Modular DSB TX:
    A Simple Modular DSB TX
     
    Last edited: Dec 6, 2008
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA


    I think they should be seperated about 3KHz so that you get approx 1500 Hz either side of the Xtal filter for upper 5Kz of audio frequency. I believe that is the general rule of thumb. That is pushing an Xtal into the wobbly region.

    That wasn't the best reception I got from them by far. In fact it was the first reception I got of them and it was raining at the time. Turned out to be about the worst reception but I got it recorded.
     
    Last edited: Dec 8, 2008
  6. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    If you can understand speech filtered to 3k to 5k your ears are much different than my human ears. 300hz to 2.5khz is about right. Your first estimate of 1.5khz (not 15khz) is a good offset.

    Like others here, I believe you should scrap your audio section and start over.
    Want to try an experiment? Connect a hi-fi amp to the output of the detector and see what it sounds like. Turn the tone controls to minimum and see if it sounds better than what you have now, esp. distortion. If it not less distorted then you may not be using enough bfo injection.
     
  7. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,445
    Likes:
    937
    Location:
    Canada, of course!
    Good speech requires a bandwidth from 80Hz to 14khz. The consonants are very important in speech and they extend to 14kHz.

    If you remove the frequencies with consonants then listeners will be saying, "What did you say?" over and over and over and over and ...
     
  8. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    Oh that's just silly. AM radio often has a bandwidth of 6khz.
    Can you still hear out to 14khz? I don't think I can.
    (What did you say?)

    He is designing a ham band receiver. It has a special purpose.
    The transmitters he is listening to are not wasting power on high and low audio frequencies. Bell labs decided many years ago that 300-3k would get the job done and they were right.
     
    Last edited: Dec 8, 2008
  9. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,445
    Likes:
    937
    Location:
    Canada, of course!
    High speech frequencies don't use much power. They use bandwidth.
    AM radios have a high frequency limit of only 2.5kHz to 3kHz so that the 10kHz beat frequency is not heard. AM radio stations peak the audio at 4kHz to 5kHz so it isn't too muffled on good radios.

    Bell decided on an upper frequency for telephones of only 3kHz because in those days that was as high as their tin can transducers would go and the distortion was so high that they used a lowpass filter to reduce it.
    Modern phones with electret mics and dynamic earphones sound much better than those old ones.

    If a ham receiver produced 300Hz to 3kHz with low distortion then it would sound a lot better than the extremely narrow bandwidth and high distortion from The Varmint's radio.
     
  10. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    Ok, you get debate and technical points.
    However, to not listen to more noise than necessary it's nice to be able to restrict the audio bandwidth to what is REQUIRED.
    I presume that you don't enjoy listening to the ham bands but I would if my environment would let me. Noise on the hf bands at this location is very bad.

    10k beat note? On the ham bands we should be so lucky :)
    Do you know how much extra circuitry is employed in a hf receiver to combat hetrodynes and other noise?

    A Drake or Collins receiver sounded good by ham standards using a bandpass to 2.7khz.
     
  11. Nigel Goodwin

    Nigel Goodwin Super Moderator Most Helpful Member

    Joined:
    Nov 17, 2003
    Messages:
    39,228
    Likes:
    641
    Location:
    Derbyshire, UK
    The radio under discussion is so badly designed it's going to perform badly and sound terrible - but good 'quality' 300Hz to 3KHz is all you need for good speech communication. It's all about low bandwidth and communicating across large distances, under poor conditions, on relatively little power.
     
  12. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA
    The radio in question has picked up German ham operators in the middle of the day as well as Italy and many others with perfect clarity and if the designer says so himself, the only spec that could be improved improved in the audio would be the AGC. The dynamic range being so great in HF communications, it falls short of the mark in reducing the very very strong signals but is superior in the weak ones.

    The radio in question's design has never been posted so the above statement is speculative and with bias & prejudice.

    The radio in question has never been declared a finished product and is modularized to have the advantage of removing slices to be upgraded such as the AGC which may be changed to rf, AGC instead of the current two amplifier 1st IF & 2nd IF gain control.

    The thread was about transmitter modulators. Who will stick their neck out? Is this all we have is a bunch of critics?

    The thread starter my have a little time to re-do the SSB modulator and would appreciate if he could discuss just that for the time being. The thread starter has been a ham for 36 years and has had and used many radios and likes his receiver just fine thank you.

    .
     
  13. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    Here is an excerpt from Audels Radiomans Guide 1939 reprint 1945. It talks about speech bandwidth. I hope some find it interesting.

    BTW, quite a few old radio texts, mostly books on tube circuits and design but also including the ARRL Handbooks for 1936 and 1940 and a few of Orr's Handbooks are downloadable at the site. Show interest and I'll post the link. Copyright long gone on these books.

    Sorry Spaceman, not too many of the circuits will be useful in your current project :)
     

    Attached Files:

  14. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,445
    Likes:
    937
    Location:
    Canada, of course!
    A hearing test was designed in 1951. The transducers in those days couldn't go higher than 8kHz and many deaf people couldn't hear frequencies that high so that is still the highest frequency that is tested.

    I think that a bandwidth to 15kHz makes speech sound natural.
    If the cutoff is at 8kHz then speech is a little muffled.
    A cutoff at 4kHz is very muffled because most of the consonants are gone.
    A cutoff at only 2kHz is "Wad ya say? Wad ya say?"
    A cutoff at 1kHz is ducks quacking.
     
  15. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA
    So what I think you are driving at, is if I limit the audio bandwidth, it will reduce the noise that can leak through the balanced modulator?
     
    Last edited: Dec 11, 2008
  16. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    I just found the info interesting. Speech bandwidth is one of the topics this thread is about, indirectly...as in filter bandwidth and bfo/filter freq. offset. I also thought you might like to check out the page. Lots of old radio books in PDF.

    Most important in the bal. modulator is matching of components in port impedances, and injection level...your talking about transmitter not receiver. I guess it might help to shape the audio before the modulator, yes. Depends how you shape it, your choices.
     
  17. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA
    Yeah yeah, that's what I'm talking about. You say you got a page? I will be very anxious to see what you got. I started building one and there was too much leakage/carrier. I got a small window of opportunity to do it right. So, trying to collect as much info as possible on the subject matter.

    Did you understand my question about using an attenuator after the crystal filter? What few schematics I could muster up, I found a guy using that technique, I believe to reduce any signal being passed to the amplifiers that isn't actual audio derived energy.

    As a side, if your interested flat. I built this PLL for 40 meters. I'm just using VFO for full coverage but my antenna is cut for 40 so I will be swapping the VFO out when operating: Aaron's Homepage Forum - Building a PLL Synthesizer
     
    Last edited: Dec 12, 2008
  18. JimB

    JimB Super Moderator Most Helpful Member

    Joined:
    Sep 11, 2004
    Messages:
    6,329
    Likes:
    585
    Location:
    Peterhead, Scotland
    Putting an attenuator after a crystal filter or a mixer, will not necessarily improve the carrier rejection or unwanted sideband supression.
    What it will do is to improve the impedance matching.
    Suppose a filter requires a terminating impedance of 50ohm.
    In an ideal world, the following stage will have an input impedance of 50ohm - perfection.
    However in the real world, the input impedance of the following stage may vary across the frequency range of interest, say it varies from 30 to 70 ohms. I we put a 6dB attenuator between the two stages the impedance seen at the output of the filter will be 44 to 54 ohms.

    As a result of this there may be in improvement in the performance of the filter or mixer.

    But adding a 20dB attenuator does not mean an automatic 20dB improvement in (say) carrier suppression, what it does mean is a 20dB reduction in EVERYTHING, both wanted and unwanted signals.

    JimB
     
  19. flat5

    flat5 Member

    Joined:
    Oct 26, 2008
    Messages:
    866
    Likes:
    8
    Location:
    Amsterdam
    Aaron's site is interesting. I have plenty to learn before I tackle a PLL design. I'm trying to get practical knowledge of toroids for now. 30 years ago I built a good 80 & 40 meter ssb receiver.
    I'm starting the hobby again. Just starting to collect parts.
    I want to build a receiver again.

    The site with the old books won't help you with what you're doing. It would only distract you, or not.
     
  20. Space Varmint

    Space Varmint New Member

    Joined:
    Jul 10, 2008
    Messages:
    1,487
    Likes:
    11
    Location:
    Tucker GA...USA
    Great! Cool even! I just built a receiver again and I put some recordings up but it really doesn't do it justice. In other words, I absolutely love this one. It does the job. I got some help with it right here. There are a few guys that know something like Mikebits and some others, but watch out, they will take shots at you in the process.

    The things I did with this receiver were just careful, careful, careful design. I biased everything class A and if I saw any loading I buffered the the stages,

    At any rate I got to cover my butt because they will hurrel the insults, but I can give you some good advice. I've done my share of building and I am licensed ham 36 years.

    The receiver's doing a bang up job. Considering skip and other propagational phenomena, I'm hearing almost all ends of all conversations or if you are a ham, the QSO's.

    Really I have had a great deal of difficulty doing the recordings. I get some kind of white noise that just does not exist in real life, but this one was recorded even before I had finished the radio: http://www.mediafire.com/file/l2ywknexcnn/80MetersB.wma

    Anyway it is a communications sweetheart. I put 1st IF & 2nd IF gain control on it and a 6 position preselector. I got a two position band pass selection on the Xtal filter. It's a home brew filter. Nothing fancy on the transistors. Just a hand full of 3904's and a couple of 2N5486 JFETs. Audioguru showed me a great audio chip which I'm using...TDA2822 for speaker drive. And I got a couple of 741s, one goes in the AGC.

    At first I was just gonna build a receiver, but the dern thing turned out so good I need to get me a transmitter going again.

    As far as the toroids, that's about all I use. I built a tranceiver back a while ago and since lost it in moving, but one of my very first QSOs was to the Cayman Is. and the guy gave me a 5,9 signal report.

    What I suggest is that you go to Amidon and get the pamphlet. It is free. They give the AL values for all their cores and the frequency response and anything else you need to know. OK, they call it "The Flyer". It's right here:https://www.amidoncorp.com/items/55

    I'm kinda of an old hand with the toroids, so if you need any help don't hesitate to ask. In HF I use allot of Amidon FT-37-43's in the tank circuits and transformers. For oscillators and filters I usually use the number 2 & number 6 mix such as T-37-02 or T-80-06 where (T) is toroid and (80) would be the size parameter for .8 inches diamter and (6) is the powdered iron mix. The designator (FT) means ferrite toroid.

    I can show you some tricks for making cheap crystal filters with just two off the shelf crystals and a specially wound transformer to achieve a very narrow band pass < 5KHz. You can stack them too, to make nice binocular core transformers for rf power amps.
     
  21. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,445
    Likes:
    937
    Location:
    Canada, of course!
    Your 80m recording sounds awful like ducks quacking beside a loud waterfall. The BFO sounds like its frequency changes with the incoming signal level.

    My audio spectrum analyser is not calibrated but it shows the voices have no low frequencies and no high frequencies. The noise is wideband and continuous so I think it comes from your audio amplifier circuit.

    Post the schematic of your audio amp from the detector diode to the speaker and we will show you where the noise comes from.
     

Share This Page