Callsign
ad: giga-rw
Page 1 of 4 1 2 3 4 LastLast
Results 1 to 10 of 34

Thread: Building my own mobile hf Amplifier....Filter questions

ad: L-HROutlet
ad: l-assoc
ad: l-tentec
ad: l-rl
ad: l-Waters
ad: l-innov
ad: l-sarc
ad: l-WarrenG
ad: l-gcopper
  1. #1
    Join Date
    Dec 2005
    Location
    Ocean City MD
    Posts
    265

    Default Building my own mobile hf Amplifier....Filter questions

    Ok, I am planing a build for a 1kw mobile amplifier. I have looked at the Ameritron amp and SGC schematics. From what I can tell, they both have tuned inputs which are nothing more than Band pass filters for each band which in turn goes through a combiner network and to the PA's Then there seems to be nothing more than a Low pass filter on the output. ....I was under the impresion there were tuned outputs for each band...But it looks to me like a standard low pass filter......What do you guys say?????

  2. #2
    Join Date
    Dec 2005
    Location
    Ocean City MD
    Posts
    265

    Default

    Here is the band pass filter for the input I was thinking about http://site.hfprojectsyahoo.com/low_pass_filters

  3. #3

    Post

    I am planing a build for a 1kw mobile amplifier.
    You will need metal fabrication skills as well as electronic assembly skills. Material costs will likely be > $600 plus your time.
    In addition, you will need to know what you are doing (HF amplifier theory), since what exists on the Internet (DIY builds) is incomplete and many sites are hacked Helge Granberg (sk) Application Notes (Motorola) notes without the proper information on transistor bias; IMD and output filtering.

    AVOID the mis-information from the pirate/illegal CB web sites.
    Here is the band pass filter for the input I was thinking about
    At least you care about RFI generated in a mobile.
    It is just a matter of time for the first lawsuit related to operations of a dirty HF amplifier -- that causes an automobile's electronic computers (and CAN bus) on an expressway to malfunction (e.g. Toyota acceleration) -- resulting in injuries or even loss of life. Even though very unlikely -- illegal operations give lawyers a reason to "go after you"

    READ VIRGIL's web site more carefully.
    Start with reading his construction manual (HF Projects)
    http://site.hfprojectsyahoo.com/low_pass_filters
    and a good Radio Handbook (Bill Orr's; ARRL or RSGB)

    This is the LOW PASS filter found on the OUTPUT of the RF amplifier (you are thinking of building).
    The capacitors and wound toroids are sized for the output power -- in this case (HF Projects amp) -- 100 watts! --
    You would burn up this unit with 1 kW RF output !!

    K6IF also makes PC boards for Low Pass Filters (OUTPUT)
    http://n6eaj.com/K6IF_PCB_graphic.jpg

    The various types of filters – Butterworth filter, Chebyshev filter, Bessel filter, etc. –
    all have different-looking "knee curves".

    I prefer 5 or 7-pole Chebyshev filter designs for soild-state HF amplifiers, which are sharper than Butterworth
    http://en.wikipedia.org/wiki/Chebyshev_filter

    Matt, N6EAJ is building a 600 Watt HF amplifier (base not mobile) based on the CCI EB104 board,
    which was covered in one of Helge Granberg's MOSFET Application Notes for Motorola.
    http://n6eaj.com/EB104.html

    WHAT MAKE'S A GOOD SOLID-STATE AMPLIFIER - Adam Farson VA7OJ/AB4OJ
    http://www.ab4oj.com/quadra/sshfamp.html

    w9gb
    Last edited by W9GB; 05-08-2010 at 01:10 PM.
    Nullius in verba

  4. #4

    Default

    Quote Originally Posted by KB3HLK View Post
    Ok, I am planing a build for a 1kw mobile amplifier. I have looked at the Ameritron amp and SGC schematics. From what I can tell, they both have tuned inputs which are nothing more than Band pass filters for each band
    Yep, often they perform an impedance transformation as well, so the amp can "present" a 50 ohm load to the transceiver.

    Then there seems to be nothing more than a Low pass filter on the output. ....I was under the impresion there were tuned outputs for each band...But it looks to me like a standard low pass filter......What do you guys say?????
    Nope all you need is the low pass with a cuttoff at the appropriate point. Often designers will press one filter into service for two bands, like 30 and 20m, 12 and 10 etc.

    The folks at http://www.communication-concepts.com/hf_amplifiers.htm will have a lot of the stuff that you are looking for.

    Good luck on your project, building stuff is fun.

    73 m/4

  5. #5
    Join Date
    Apr 2001
    Location
    Colorado
    Posts
    1,719

    Default

    Quote Originally Posted by KB3HLK View Post
    Ok, I am planing a build for a 1kw mobile amplifier.
    That's not an easy thing to do! Electrically the challenge is getting the 2KW or so that you'll need outta your car's electrical system (almost 150A at 14V - think big alternator, high idle switch, BIG copper wires, etc.), but probably the biggest challenge is dissipating a KW or so of heat - not an easy task! This is assuming you're planning on an AB configuration, which is what the other amps you're looking at are.

    Getting the thermal design right is going to be critical. The small surface area of solid state devices and the (relatively) small temperature delta is why tubes are still so popular in power applications where cost is an issue. I know hams who have built 1.5KW RF solid state amps and had to go with water cooling!

    Now, a class E mobile amp might be the way to go...

    Steve
    KV6O
    Signature line removed for your viewing pleasure.

  6. #6
    Join Date
    Dec 2005
    Location
    Ocean City MD
    Posts
    265

    Default

    Quote Originally Posted by W9GB View Post
    You will need metal fabrication skills as well as electronic assembly skills. Material costs will likely be > $600 plus your time.
    In addition, you will need to know what you are doing (HF amplifier theory), since what exists on the Internet (DIY builds) is incomplete and many sites are hacked Helge Granberg (sk) Application Notes (Motorola) notes without the proper information on transistor bias; IMD and output filtering.

    AVOID the mis-information from the pirate/illegal CB web sites.

    At least you care about RFI generated in a mobile.
    It is just a matter of time for the first lawsuit related to operations of a dirty HF amplifier -- that causes an automobile's electronic computers (and CAN bus) on an expressway to malfunction (e.g. Toyota acceleration) -- resulting in injuries or even loss of life. Even though very unlikely -- illegal operations give lawyers a reason to "go after you"

    READ VIRGIL's web site more carefully.
    Start with reading his construction manual (HF Projects)
    http://site.hfprojectsyahoo.com/low_pass_filters
    and a good Radio Handbook (Bill Orr's; ARRL or RSGB)

    This is the LOW PASS filter found on the OUTPUT of the RF amplifier (you are thinking of building).
    The capacitors and wound toroids are sized for the output power -- in this case (HF Projects amp) -- 100 watts! --
    You would burn up this unit with 1 kW RF output !!

    K6IF also makes PC boards for Low Pass Filters (OUTPUT)
    http://n6eaj.com/K6IF_PCB_graphic.jpg

    The various types of filters Butterworth filter, Chebyshev filter, Bessel filter, etc.
    all have different-looking "knee curves".

    I prefer 5 or 7-pole Chebyshev filter designs for soild-state HF amplifiers, which are sharper than Butterworth
    http://en.wikipedia.org/wiki/Chebyshev_filter

    Matt, N6EAJ is building a 600 Watt HF amplifier (base not mobile) based on the CCI EB104 board,
    which was covered in one of Helge Granberg's MOSFET Application Notes for Motorola.
    http://n6eaj.com/EB104.html

    WHAT MAKE'S A GOOD SOLID-STATE AMPLIFIER - Adam Farson VA7OJ/AB4OJ
    http://www.ab4oj.com/quadra/sshfamp.html

    w9gb

    THANKS!!! That is some good info!!! I was planning on using the HF packer Filter for a tuned input and just using a LARGE Low pass filter in the output. I did plan on using 8 2sc2879 transistors and the heat sink from a cb amplifier (Texas Star dx1600)

    I plan on the amplifier being AB biased, I think thats .62 volts emitter to base at 13.8 volts B+

  7. #7

    Default

    Quote Originally Posted by KB3HLK View Post
    THANKS!!! That is some good info!!! I was planning on using the HF packer Filter for a tuned input and just using a LARGE Low pass filter in the output. I did plan on using 8 2sc2879 transistors and the heat sink from a cb amplifier (Texas Star dx1600)

    I plan on the amplifier being AB biased, I think thats .62 volts emitter to base at 13.8 volts B+
    You need to used bandswitched low pass filters on the output, not just one.

    A 30 MHz LPF passes 8 harmonics of 80m, 4 harmonics of 40m, the second harmonic of 20m, etc. That's important to remember. A single LPF is really only effective on 15, 12 and 10 meters. If won't help with the frequencies lower than that.

    BTW there was an earlier comment about the LPFs possibly performing impedance transformation; that's usually not the case. The impedance transformation is normally done using a wideband transformer and the filters are usually 50 Ohms in and out.

  8. #8
    Join Date
    Jan 2000
    Posts
    2,956

    Default

    Quote Originally Posted by KB3HLK View Post
    THANKS!!! That is some good info!!! I was planning on using the HF packer Filter for a tuned input and just using a LARGE Low pass filter in the output. I did plan on using 8 2sc2879 transistors and the heat sink from a cb amplifier (Texas Star dx1600)

    I plan on the amplifier being AB biased, I think thats .62 volts emitter to base at 13.8 volts B+
    The tuned filter input will not be a significant benefit to you for the cost and space involved. Most modern transceivers provide a signal clean of harmonics, so there will be nothing to filter out on your amplifier's input. Since the amplifier is "broadband," it IS prone to generating/amplifying harmonics and does need an output filter.

    A low pass filter on the amplifier output will minimize any signal beyond the filter cutoff frequency from reaching your antenna. Lets say you select a single low pass filter with a 30 MHz cutoff. That means that if you are transmitting on 40M (7MHz) that you may also be passing a signal of twice that (14MHz), three times that (21MHz) and four times that (28MHz). So what good did your low pass filter do for you if it allowed you to transmit on four amateur bands at the same time?

    You will need low pass or band pass filters designed for each band of interest.

    Bias is not a specific voltage, rather one that is adjusted for the best linearity for your particular amplifier. It will be somewhere in the range that you mentioned, about +0.6 to 1.0V DC. The difference lies in the variables of the components themselves. Each has a tolerance within a specification range, thus you will need to adjust the bias voltage for your particular amplifier's collection of components. I suggest an IC voltage regulator/ driver transistor combination mounted on the same heat sink to aid in preventing thermal runaway. Each 2SC2879 pair (modules) should have their own bias supply and adjustment, so you will need four supplies built into the amplifier housing. Start off with the bias set for 200MA of collector current for each pair of 2SC2879 transistors. Each module's linearity should be tested and bias set separate from the other modules. When all modules have their bias properly set, their RF inputs and RF outputs may be combined in the final configuration.

    Eight 2SC2879's will provide about 800 watts output PEP with an intermodulation distortion of about -24dB. I would not run them any higher than that. Too much drive power will short the emitter base junction of the finals. Build a 3.5 to 4dB input attenuator in the amplifier box to prevent your 100 watt transceiver from over-driving the amplifier. 40 -45 watts of drive should do it. The input attenuator will also present a nice 50 ohm load for your transceiver.

    If the amplifier dos not have a fan cooling the internal components and external heatsink, add one......or two.

    This project will be a good use of parts from a former CB amplifier.
    73,
    Terry
    Last edited by K7FE; 05-08-2010 at 11:43 PM.
    Terry Graves, K7FE
    Chief Editor, QRZ.COM

    "Some people call CW a MODE but in
    reality it is an autonomous LANGUAGE."

  9. #9

    Default

    Quote Originally Posted by K7FE View Post
    The tuned filter input will not be a significant benefit to you for the cost and space involved. Most modern transceivers provide a signal clean of harmonics, so there will be nothing to filter out on your amplifier's input. Since the amplifier is "broadband," it IS prone to generating/amplifying harmonics and does need an output filter.

    A low pass filter on the amplifier output will minimize any signal beyond the filter cutoff frequency from reaching your antenna. Lets say you select a single low pass filter with a 30 MHz cutoff. That means that if you are transmitting on 40M (7MHz) that you may also be passing a signal of twice that (14MHz), three times that (21MHz) and four times that (28MHz). So what good did your low pass filter do for you if it allowed you to transmit on four amateur bands at the same time?

    You will need low pass or band pass filters designed for each band of interest.

    Bias is not a specific voltage, rather one that is adjusted for the best linearity for your particular amplifier. It will be somewhere in the range that you mentioned, about +0.6 to 1.0V DC. The difference lies in the variables of the components themselves. Each has a tolerance within a specification range, thus you will need to adjust the bias voltage for your particular amplifier's collection of components. I suggest an IC voltage regulator/ driver transistor combination mounted on the same heat sink to aid in preventing thermal runaway. Each 2SC2879 pair should have their own bias supply and adjustment, so you will need four supplies built into the amplifier housing. Start off with the bias set for 200MA of collector current for each pair of 2SC2879 transistors.

    Eight 2SC2879's will provide about 800 watts output PEP with an intermodulation distortion of about -24dB. I would not run them any higher than that. Too much drive power will short the emitter base junction of the finals. Build a 3.5 to 4dB input attenuator in the amplifier box to prevent your 100 watt transceiver from over-driving the amplifier. 40 -45 watts of drive should do it. The input attenuator will also present a nice 50 ohm load for your transceiver.

    If the amplifier dos not have a fan cooling the internal components and external heatsink, add one......or two.

    This project will be a good use of parts from a former CB amplifier.
    73,
    Terry
    Very well written, and I couldn't agree more.

    Keeping transistors linear over a wide temperature range is a challenge. Tubes are way easier: They're always hot.

  10. #10
    Join Date
    Dec 2005
    Location
    Ocean City MD
    Posts
    265

    Default

    Quote Originally Posted by K7FE View Post
    The tuned filter input will not be a significant benefit to you for the cost and space involved. Most modern transceivers provide a signal clean of harmonics, so there will be nothing to filter out on your amplifier's input. Since the amplifier is "broadband," it IS prone to generating/amplifying harmonics and does need an output filter.

    A low pass filter on the amplifier output will minimize any signal beyond the filter cutoff frequency from reaching your antenna. Lets say you select a single low pass filter with a 30 MHz cutoff. That means that if you are transmitting on 40M (7MHz) that you may also be passing a signal of twice that (14MHz), three times that (21MHz) and four times that (28MHz). So what good did your low pass filter do for you if it allowed you to transmit on four amateur bands at the same time?

    You will need low pass or band pass filters designed for each band of interest.

    Bias is not a specific voltage, rather one that is adjusted for the best linearity for your particular amplifier. It will be somewhere in the range that you mentioned, about +0.6 to 1.0V DC. The difference lies in the variables of the components themselves. Each has a tolerance within a specification range, thus you will need to adjust the bias voltage for your particular amplifier's collection of components. I suggest an IC voltage regulator/ driver transistor combination mounted on the same heat sink to aid in preventing thermal runaway. Each 2SC2879 pair (modules) should have their own bias supply and adjustment, so you will need four supplies built into the amplifier housing. Start off with the bias set for 200MA of collector current for each pair of 2SC2879 transistors. Each module's linearity should be tested and bias set separate from the other modules. When all modules have their bias properly set, their RF inputs and RF outputs may be combined in the final configuration.

    Eight 2SC2879's will provide about 800 watts output PEP with an intermodulation distortion of about -24dB. I would not run them any higher than that. Too much drive power will short the emitter base junction of the finals. Build a 3.5 to 4dB input attenuator in the amplifier box to prevent your 100 watt transceiver from over-driving the amplifier. 40 -45 watts of drive should do it. The input attenuator will also present a nice 50 ohm load for your transceiver.

    If the amplifier dos not have a fan cooling the internal components and external heatsink, add one......or two.

    This project will be a good use of parts from a former CB amplifier.
    73,
    Terry
    I hope you all know, Im copy and pasting this good info to a document for later reference! This is great!!! I have four 3 inch fans for the heat sink and was also going to put two blowers side by side and blow in one end across the transistors and vent the other side, make a positive pressure ventilation system in the amp.

    My next question is, Why do the "BIG" builders use input low pass filters? I would think output filters to withstand 800w would be massive!!!

    My next question, is where can I find a schematics for active biasing for the 4 boards of 2 2879's?

Page 1 of 4 1 2 3 4 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •