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

Very simple Audio compressor for FM transmitter

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by Willen, Dec 6, 2013.

  1. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    How?
     
  2. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    Why not use the simple IC in post #82?
    I spiced around with the circuit in spice. The MOSFET I used year ago, made a good voltage to resistance converter. The 2n7002 has a sharp knee. It goes form little attenuation to great attenuation in a small number of mV. I did not get it to work well.
     
  3. Willen

    Willen Well-Known Member

    Joined:
    May 22, 2012
    Messages:
    1,293
    Likes:
    115
    Location:
    Nepal (GMT +5:45)
    Hi Ron,
    The original one has 2N7000 mosfet. Will you try using this one?

    Which mosfet had you used in your limiter in past days?
     
  4. dave

    Dave New Member

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


     
  5. Willen

    Willen Well-Known Member

    Joined:
    May 22, 2012
    Messages:
    1,293
    Likes:
    115
    Location:
    Nepal (GMT +5:45)

    Hi Ron, I have few THAT2252 (RMS Detector) and THAT2155 (VCA) chips, though working principle of simpler circuit is visiable and interesting. Also chips are limited with me but have many mosfets.
    The original one has 2N7000 mosfet. Will you try using this one?

    Which mosfet had you used in your limiter in past days?
     
  6. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    Too many years ago. I do remember getting a MOSFET with out a internal diode. I don't remember if it was better.

    One thing I hate about the circuit is that the "AGC voltage" (more or less gate voltage) starts out a 0 volts and must get up to 1.25 volts before any limiting can be seen. Then from 1.25 to 1.4 is the entire range. Before I added a pot that set the "DC" voltage on the gate with no audio. So I would set the voltage at 1.25 to start out. That way the "attack delay time" was faster. Then to get the distortion out of the picture, no audio frequency noise should be on the 10uF cap. (bottom line the gate voltage should stay in the just turned on linear region)
    upload_2015-11-7_17-38-16.png
     
  7. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ
    Hello,

    Well the devices i was referring to were made specifically for audio applications such as compressors.
     
  8. Nigel Goodwin

    Nigel Goodwin Super Moderator Most Helpful Member

    Joined:
    Nov 17, 2003
    Messages:
    39,199
    Likes:
    640
    Location:
    Derbyshire, UK
    There have been lot's of them used, perfectly fine for audio as that was their intended purpose.
     
  9. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ
    Hi there Nigel,

    That's what i thought too, but we can check the spec's if anyone has any doubts i guess. I've have to find a data sheet, but not sure if i have the part number anymore.
    I guess i could test one also, just to get some real life data in.
     
  10. Nigel Goodwin

    Nigel Goodwin Super Moderator Most Helpful Member

    Joined:
    Nov 17, 2003
    Messages:
    39,199
    Likes:
    640
    Location:
    Derbyshire, UK
    They were generally a combination of an LDR and a small incandescent lamp, this combination gave an excellent attack/decay shape - something which is lacking if you use an LED instead of the lamp (and you have to add extra electronics to compensate).
     
  11. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,430
    Likes:
    933
    Location:
    Canada, of course!
    I used an opto-Jfet as a good compressor. I think the attack time should be fast like Rod Elliot says at Elliot Sound Products. He also agrees that an LDR is too slow.
     
  12. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ
    Hi,

    What do you consider to be "too slow" ?
    I could test one at some point to see how fast/slow it really is.
    If will also depend to some extent on any compensation network, but i could test the basic functionality.
     
  13. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,430
    Likes:
    933
    Location:
    Canada, of course!
    Rod Elliot says most LDR compressor circuits have an attack time too slow at 15ms. Most Jfet circuits use a simple distortion cancelling capacitor that increases the attack time to 100ms and attempting to reduce it to 20ms results in clicks and overshoots.
    He shows a Jfet circuit with low distortion and a 5ms attack time.
     
  14. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ

    Hi again AG,


    I think it would be very hard to adopt a blanket policy concerning LDR's as to the working or non working status for a given application.
    For one thing, there is quite a bit of variance from model to model. There are units with 3.5ms to 5ms turn on, with turn off ranging from 10ms to 500ms. Which ones were tested? That's one question here.

    Second and equally as important as the innate spec's themselves, part of the application functionality is by design. We can make a slower device faster by using positive feedback, for just one example. We also are not really limited to using only one device in the same application.

    So the conclusion here seems to be that although for a given circuit or typical circuit we may find a device that works better than an LDR, given the engineering freedom to create any reasonably simple circuit we wish the LDR should be a likely candidate.

    What we'd have to do to investigate the statement made by your favorite audio guy is find out what units he tested, and what circuits he used to test them with.

    Another small point brings us to the word, "most". The word "most" can be translated into "many, but not all". We want to know about "all" of them, not really just "most" of them, if we really want to know for sure if this can ever be done or not.
     
  15. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    When I used LDRs, I went to the store and got one of every kind they had. Then I mail ordered more kinds and they all were slow. I used an op-amp to speed up the response. I frequency sweep the LED-LDR circuit and added high frequency gain to make it faster.

    If you can find LDRs with 3mS response I would use these. They have very good distortion.

    Using the 2N7000 MOSFET in spice I have distortion. I used a linearizing circuit and it gets better but I just can't make it good. It could just be SPICE. When I built these things in real life I used MOSFETs with very high gate turn on voltages.
    upload_2015-11-10_7-35-14.png
     
  16. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,430
    Likes:
    933
    Location:
    Canada, of course!
    I tried using an LDR as a compressor but it was obviously too slow, "HELLo". The hell part of hello was finished before the LDR reduced the level. It was a no-name-brand Chinese one. I used a little light bulb then a fairly bright LED but I did not try an Xenon flash tube to make the LDR conduct. I do not see how positive feedback can speed up the physical part.

    Claims for The Batterizer voltage booster product also use the words many, most then all then says disposable AA cells are thrown away when their voltage drops a little. They make The Batterizer and sometimes it drains a battery cell instead of making it last longer. Here is a video about its false claims:
     
  17. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ
    Hi there Ron,

    I guess you would have to check into what exactly causes that distortion. It looks like the response of the compression is too fast for the frequency in question, unless that is across the board distortion that happens even when the compression is not changing. Do you know which it is?
     
  18. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,026
    Likes:
    951
    Location:
    NJ
    Hi AG,


    We can always find false claims around but that certainly does not prove that every claim is false.

    Positive feedback basically works by measuring the error in the output and knowing what it is supposed to be and that it is not yet that, it increases the drive. Of course this means everything else has to be right too, such as the bias, because obviously if the bias is already set high then there's nothing higher left in order to drive harder.
    The audio couplers are made with fast attack, and i guess that is because that is the most wanted feature. But that's not the end of it either, because we also have to choose the range of operation carefully. For example, trying to use the entire dynamic range would lead to something even a snail would not want to listen to :)
    But limiting the range by using say a parallel resistor (for a 10 to 1 dynamic range) will allow the overall action to speed up already. For example, 1ms to 50 percent audio attenuation. It seems to me that ought to be good enough. What do you think, does 1ms to 50 percent attenuation sound good enough to you or no?
    In other words, if the device resistance spread is from 1k to 100k, we dont have to use that entire range we can use 1k to 10k (or about 900 ohms to 10k). If it takes 10ms to get from 100k to 1k then it probably takes around 1ms to get from 10k to 1k.

    But i am working from theory and the spec sheets that i knew about. For the devices i have it could be totally different. As Ron pointed out, if the devices are not chosen carefully with respect to their KNOWN spec's, we can not hope to get anything useful out of it except by pure chance.

    I will try to test the devices i happen to have on hand, but as noted this can not be considered an all encompassing test as these were probably the cheapie brand. The good ones are probably four or five bucks each (USD).
     
  19. Gasboss775

    Gasboss775 Member

    Joined:
    Nov 9, 2015
    Messages:
    247
    Likes:
    13
    Location:
    Scotland
    These ones can be and have been used in pro audio equipment:

    http://uk.farnell.com/webapp/wcs/st...001&langId=44&storeId=10151&gs=true&st=Vtl5c3

    They are known as vactrols ( actually a trade name ) and contain an encapsulated led facing a fast good quality LDR, if you look at the data sheet. It can be found on the Farnell product listing, or just Google it. The VTL5C3 has a very fast ( 3mS ) attack time and is eminently suited to this sort of application.
     
    Last edited: Nov 10, 2015
  20. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,430
    Likes:
    933
    Location:
    Canada, of course!
    Mr Al, 50% audio attenuation is only -6dB which is a small drop in level. Maybe you mean 50% of the range of resistance change? A fast Vactrol has a range of 50dB so you mean limit it to 25dB?

    Gordon, the Vactrols (never heard that word before) are very good.
     
  21. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    Here is the MOSFET-drain and Vin, with 1.605 volts on the gate. (compressor shut down and a "battery" on the gate) Vin is 20mV pk. It only takes mVs to make huge changes in gain. I think this is the wrong FET! I know how to remove much of the distortion but.....
    upload_2015-11-10_18-42-41.png
     

Share This Page