Topics

Amplifier Talk- Communications Concepts

Jeffrey Benedict
 

I see that there has been a little bit of talk about amplifiers for ubitx.  Motorola published amplifier designs which used their transistors a few years ago and a company named Communications Concepts packaged many of them into kits with boards and parts- as much or as little as you want.  These are not "kits" as in Elecraft or Heathkits or even ubitx.  You really have to be able to go it alone with no one to hold your hand.

One that might fit the needs of the ubitx user is:

http://www.communication-concepts.com/eb63a-140w-hf-amplifier/

It takes 1-5 watts and makes as much as 140 watts out.  They have the parts listed and you can buy everything except things like cases from them.  The website is kind of cryptic I think to keep the "good buddies" from finding and building high power amps for their See Bees.  But, poke around and you will find amps of varying power and type as well as the parts to build them.  The Motorola documentation is there also.

eHam has some reviews of amplifiers built with information and parts from this site.

73

Jeff
--
Jeff, K7AIL  CN87

raoul@...

 

Hi Jeff,

Well that one really is just a CB amp -- the biasing not linear and uses only a diode (a design for CW and FM use) and the output transformer design dated (leading to low efficiency).  The AN762-140 also low power in/140W out but is a much better design (with OK IMD specs) and $4 cheaper! 

73,


Mark.

Razvan (M0HZH)
 

The Motorola Helge Granberg designs are almost 30 years old now. Keep in mind they are only engineered to the level that demonstrates the functionality of a part and since then there have been important developments in materials and components. Once you understand that, designt he rest of the required circuitry and the right output filters, you might end up with a decent amp.

Alternatively, for about the same cost, less work and much more predictable results, you can find on eBay PA Units and Filter Units from old transceivers.

Cheers,
Razvan M0HZH

MadRadioModder
 

I know Helge and what you say is not true.  His RF designs are outstanding still today.  His work pioneered matching transformer designs for transistor devices, taking into account many parameters (saturation, capacitance, stray inductances, etc. over frequency).  His designs incorporated temperature compensated bias for thermal creep, and some later used RF chokes to offload current flux and pre-saturation from it in the RF transformers… something that few designs do even today.

 

Now, if you mean that his amplifier boards didn’t incorporate filtering to meet FCC regs (etc, name your country)… yes that is true.  This is because as you say his designs were demonstration designs, and it was left to the designer to add such filters… But that is true of almost everything you buy off eBay today except for complete amplifiers.

 

MRM

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of Razvan Fatu
Sent: Friday, September 13, 2019 3:52 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Amplifier Talk- Communications Concepts

 

The Motorola Helge Granberg designs are almost 30 years old now. Keep in mind they are only engineered to the level that demonstrates the functionality of a part and since then there have been important developments in materials and components. Once you understand that, designt he rest of the required circuitry and the right output filters, you might end up with a decent amp.

Alternatively, for about the same cost, less work and much more predictable results, you can find on eBay PA Units and Filter Units from old transceivers.

Cheers,
Razvan M0HZH


Virus-free. www.avg.com

--

…_. _._

iz oos
 

The AN762-140 should be a good amplifier, I think it was described so in the ARRL 2010 Handbook. It needs lowpass filters like those Communication Concepts sells. It seems not overly complex to build.


Il 13/set/2019 07:03, "Mark - N7EKU" <n7eku@...> ha scritto:
Hi Jeff,

Well that one really is just a CB amp -- the biasing not linear and uses only a diode (a design for CW and FM use) and the output transformer design dated (leading to low efficiency).  The AN762-140 also low power in/140W out but is a much better design (with OK IMD specs) and $4 cheaper! 

73,


Mark.


Ashhar Farhan
 

The WA2EBY amp has been the staple around the bitx community. I am bored with the IRF510s now. I just bought a triplet of 811a tubes. I am mortified at the thought of 1000v dc in the shack...


On Fri 13 Sep, 2019, 10:03 PM iz oos, <and2oosiz2@...> wrote:

The AN762-140 should be a good amplifier, I think it was described so in the ARRL 2010 Handbook. It needs lowpass filters like those Communication Concepts sells. It seems not overly complex to build.


Il 13/set/2019 07:03, "Mark - N7EKU" <n7eku@...> ha scritto:
Hi Jeff,

Well that one really is just a CB amp -- the biasing not linear and uses only a diode (a design for CW and FM use) and the output transformer design dated (leading to low efficiency).  The AN762-140 also low power in/140W out but is a much better design (with OK IMD specs) and $4 cheaper! 

73,


Mark.

Sajid Rahum
 

Well Ashar

I plugged KL203 and not bad.  Not the best but works.  I need to mod KL203 to expand the power range; on 20m, it was good no issues.

Sajid

iz oos
 

The KL203 is a tabu in the ham community. Don't even think to add in parallel or replace C14 with a 500V 1000pf.


Il 13/set/2019 19:18, "Sajid Rahum via Groups.Io" <zs735=yahoo.com@groups.io> ha scritto:
Well Ashar

I plugged KL203 and not bad.  Not the best but works.  I need to mod KL203 to expand the power range; on 20m, it was good no issues.

Sajid

Gordon Gibby
 

Asher, the 811A amp that I built when I was a kid used more like 2000 V;   In the coming year I hope to refurbish it. 


On Sep 13, 2019, at 13:18, Sajid Rahum via Groups.Io <zs735@...> wrote:

EXTERNAL EMAIL: Exercise caution with links and attachments.


Well Ashar

I plugged KL203 and not bad.  Not the best but works.  I need to mod KL203 to expand the power range; on 20m, it was good no issues.

Sajid

Martin KM6TCD
 

Thanks for the feedback. What about the kenwood 440s? What's its reputation?
 
Sent: Saturday, September 14, 2019 at 12:14 PM
From: "iz oos" <and2oosiz2@...>
To: BITX20@groups.io
Subject: Re: [BITX20] Amplifier Talk- Communications Concepts

The KL203 is a tabu in the ham community. Don't even think to add in parallel or replace C14 with a 500V 1000pf.

 
Il 13/set/2019 19:18, "Sajid Rahum via Groups.Io" <zs735=yahoo.com@groups.io> ha scritto:
Well Ashar

I plugged KL203 and not bad.  Not the best but works.  I need to mod KL203 to expand the power range; on 20m, it was good no issues.

Sajid

 

 

John Seboldt K0JD
 

I have the other similar amp, AN762, that I got in the late 90's. http://www.communication-concepts.com/an762-140/ I have the amp built up and mounted on a monster heat sink, with a supporting case ready, and just have to craft filters and bandswitching. A kit for filters is almost done - https://hfprojects2.com/shop?olsPage=products%2Flpf-140-r1-low-pass-filter .

As usual, takes a while to finish multiple things....
John K0JD
Milwaukee, WI

On 9/12/2019 21:14, Jeffrey Benedict wrote:

I see that there has been a little bit of talk about amplifiers for ubitx.  Motorola published amplifier designs which used their transistors a few years ago and a company named Communications Concepts packaged many of them into kits with boards and parts- as much or as little as you want.  These are not "kits" as in Elecraft or Heathkits or even ubitx.  You really have to be able to go it alone with no one to hold your hand.

One that might fit the needs of the ubitx user is:

http://www.communication-concepts.com/eb63a-140w-hf-amplifier/

It takes 1-5 watts and makes as much as 140 watts out.  They have the parts listed and you can buy everything except things like cases from them.  The website is kind of cryptic I think to keep the "good buddies" from finding and building high power amps for their See Bees.  But, poke around and you will find amps of varying power and type as well as the parts to build them.  The Motorola documentation is there also.

eHam has some reviews of amplifiers built with information and parts from this site.

73

Jeff
--
Jeff, K7AIL  CN87

raoul@...

ajparent1/KB1GMX
 

OK, staring from a point of experience...

First serious and decent amps in the 100W and up class even if home built will
cost not less that 2-5$/watt.  There are big metal parts and expensive transistors
in many.

Most of the cheap "70W" ebay amps have not been seen as working, for long. 
They are still in need of box, likely a real heatsink (adequately sized), low pass
filters and plenty of spare Mosfets and some do not work at all wall above 20M. 
Caveat emptor.

I have built many of the EB series of amplifiers and they are good stable designs that 
generally perform to the limits of the devices used.  Comments like "CB amps" are
generally meaningless and better avoided.  I have seen those apnotes applied
to CB and some are junk and a few were well done.  Quality is often a builder
thing rather than application.  All of the amp modules require a Low pass filter(s) 
and will be likely switched with relays or mechanically.

One last thing what you put in had better be good as 10x that will reflect
all of the flaws in you signal and more people will hear them and comment.
if that is not clear, garbage in, means garbage out.

The biggest issue for those amps are the transistors for them are scarce and often 
very expensive.  I built them as I already had the devices and last look MRF454s
are over 100$ a pair for real ones.

EB63, basic simple amp, the down side is the diode current alone will be high
and that establishes the bias.  Efficiency when built right is over 50% however
at 160M or over 10M the design is not optimal, that's hard to do over a decade
of frequency range.  However with a good low pass filter and clean input the
result is excellent but the transistors used are costly and it can be intolerent
[unstable] of random substitutions. 

AN762  is a later design that uses a 723 voltage regulator to do a precision bias 
and that is an improvement.  The deal is its excellent and not a simple build
plus the devices are near extinct, unless treasure is expended.  Most will
choke at the cost of the machined copper heat spreader alone.

An779 20W amplifer for a SSB transceiver, excellent amp with good efficiency
and low IMD  output.

AN593 amplifier for 160W (28V use) and a version for 12V that does about 80W
SSB and its a good design but one has to pay attention to the mechanical details
or cooling suffers and for 12V use the power is limited.  Since it has the driver is
part of the design its not suited for more than a fraction of a watt input.  For the
80w (13.8V) version max power in would be around 125mW.  Both versions when
correctly adjusted offer very good IMD.  I've built the 28V version.

Others built include AN791 using MRF247 for 80W at 2m. and the datasheet  MRF492
circuit with bias for 6M and near 100W output.  MRF140 28V 160W amp for 6M,
MRF174 for 2M at 150W, and a utv8100B amp for 60W (with 5W FT817 drive) at 432.

For VHF and UHF I've built far more but there I do use more power but at HF 
I tend to build to the 4-5W level and rely on either my 100W radio if more is
needed or one of several amps that include the EB63, Wa2EBY, and a
design that is similar to the EB63 with active bias.


Allison

Ashhar Farhan
 

Allison,
Glad to see you pitch in. How has your experience with the paralleled IRF510s been? You managed to push them to 50 Mhz.

On Sun 15 Sep, 2019, 3:43 AM ajparent1/KB1GMX, <kb1gmx@...> wrote:
OK, staring from a point of experience...

First serious and decent amps in the 100W and up class even if home built will
cost not less that 2-5$/watt.  There are big metal parts and expensive transistors
in many.

Most of the cheap "70W" ebay amps have not been seen as working, for long. 
They are still in need of box, likely a real heatsink (adequately sized), low pass
filters and plenty of spare Mosfets and some do not work at all wall above 20M. 
Caveat emptor.

I have built many of the EB series of amplifiers and they are good stable designs that 
generally perform to the limits of the devices used.  Comments like "CB amps" are
generally meaningless and better avoided.  I have seen those apnotes applied
to CB and some are junk and a few were well done.  Quality is often a builder
thing rather than application.  All of the amp modules require a Low pass filter(s) 
and will be likely switched with relays or mechanically.

One last thing what you put in had better be good as 10x that will reflect
all of the flaws in you signal and more people will hear them and comment.
if that is not clear, garbage in, means garbage out.

The biggest issue for those amps are the transistors for them are scarce and often 
very expensive.  I built them as I already had the devices and last look MRF454s
are over 100$ a pair for real ones.

EB63, basic simple amp, the down side is the diode current alone will be high
and that establishes the bias.  Efficiency when built right is over 50% however
at 160M or over 10M the design is not optimal, that's hard to do over a decade
of frequency range.  However with a good low pass filter and clean input the
result is excellent but the transistors used are costly and it can be intolerent
[unstable] of random substitutions. 

AN762  is a later design that uses a 723 voltage regulator to do a precision bias 
and that is an improvement.  The deal is its excellent and not a simple build
plus the devices are near extinct, unless treasure is expended.  Most will
choke at the cost of the machined copper heat spreader alone.

An779 20W amplifer for a SSB transceiver, excellent amp with good efficiency
and low IMD  output.

AN593 amplifier for 160W (28V use) and a version for 12V that does about 80W
SSB and its a good design but one has to pay attention to the mechanical details
or cooling suffers and for 12V use the power is limited.  Since it has the driver is
part of the design its not suited for more than a fraction of a watt input.  For the
80w (13.8V) version max power in would be around 125mW.  Both versions when
correctly adjusted offer very good IMD.  I've built the 28V version.

Others built include AN791 using MRF247 for 80W at 2m. and the datasheet  MRF492
circuit with bias for 6M and near 100W output.  MRF140 28V 160W amp for 6M,
MRF174 for 2M at 150W, and a utv8100B amp for 60W (with 5W FT817 drive) at 432.

For VHF and UHF I've built far more but there I do use more power but at HF 
I tend to build to the 4-5W level and rely on either my 100W radio if more is
needed or one of several amps that include the EB63, Wa2EBY, and a
design that is similar to the EB63 with active bias.


Allison

ajparent1/KB1GMX
 

Farhan,

Still around and busy.

Check out OZ1PIFs work.  I've run one section of that a 4x4 parallel push pull
on 6M to over 200W.  The yabut is you need 28V power at 15A!  At 12V that
doesn't work well.  The up side is to get that power not much drive is needed
and its easy to over drive it 5-6W was enough.  Its remarkably tough as I
haven't fried it.  The IMD is only OK.  However a really good and heavy built
low pass is needed as the first one with #18 air wound coils heated too much,
#14 wire solved that along with 500V metal clamp mica caps.

A pair of IRF510s can do 40-50w easy at 28V with 1-2W of drive. 
Parallel parts are harder as then you have parallel paths for
RF current to balance.  Cheat on the layout it becomes unstable
or poor power out.  Getting them to as high as 70mhz was not all
that hard if you pay attention to input and output impedance
they must be low for stability and reasonable power.

The 8 IRF510s are cheap but the total system is not cheap as you still
need a big heatsink (130mm by 130mm with 40mm fins  and a  fan) , TR
switching that can handle the power, and a package to put it all in. 
Cheap 28 or 48 volt switching  power supplies are not a solution as
often they are way to noisy on receive.

Things to consider 100W at 12V is about the limit using the right parts.
However the relays now have to be good for at least 5A and 100V
at the output and the input they can be fairly light.  With higher power
higher currents and voltage are the expected thing and parts like
#12 wire, 500V silver mica or ATC 800B are the required parts.

High power (more than 20 or so watts) tends to be expensive and
getting a reliable amp is never trivial task.

Allison

Gordon Gibby
 

Allison—

Can I ask your advice on a boat anchor project?  Soon I will start refurbishing more and more  HW 100s & SB 100s.   I read that only fractions of a white are needed to drive a 6146 amplifier (driving the grid).  It’s a very high input impedance.   We just learned how to make 1 to 49 baluns easily for End Fed half wave resonant antennas...(fantastic results!)

. would some 1:100 or so ferrite Baluns work well to feed the output of a micro Bitx  into the final amplifier stage one of these older boat anchor radios??

Gordon KX4Z




On Sep 14, 2019, at 23:48, ajparent1/KB1GMX <kb1gmx@...> wrote:

EXTERNAL EMAIL: Exercise caution with links and attachments.


Farhan,

Still around and busy.

Check out OZ1PIFs work.  I've run one section of that a 4x4 parallel push pull
on 6M to over 200W.  The yabut is you need 28V power at 15A!  At 12V that
doesn't work well.  The up side is to get that power not much drive is needed
and its easy to over drive it 5-6W was enough.  Its remarkably tough as I
haven't fried it.  The IMD is only OK.  However a really good and heavy built
low pass is needed as the first one with #18 air wound coils heated too much,
#14 wire solved that along with 500V metal clamp mica caps.

A pair of IRF510s can do 40-50w easy at 28V with 1-2W of drive. 
Parallel parts are harder as then you have parallel paths for
RF current to balance.  Cheat on the layout it becomes unstable
or poor power out.  Getting them to as high as 70mhz was not all
that hard if you pay attention to input and output impedance
they must be low for stability and reasonable power.

The 8 IRF510s are cheap but the total system is not cheap as you still
need a big heatsink (130mm by 130mm with 40mm fins  and a  fan) , TR
switching that can handle the power, and a package to put it all in. 
Cheap 28 or 48 volt switching  power supplies are not a solution as
often they are way to noisy on receive.

Things to consider 100W at 12V is about the limit using the right parts.
However the relays now have to be good for at least 5A and 100V
at the output and the input they can be fairly light.  With higher power
higher currents and voltage are the expected thing and parts like
#12 wire, 500V silver mica or ATC 800B are the required parts.

High power (more than 20 or so watts) tends to be expensive and
getting a reliable amp is never trivial task.

Allison

barry halterman
 

I have a complete EB63A that has not been built with a heat sink 165mm x 120mm x 40, 15 fins.
Anyone interested let me know.
Barry
K3BO

On Sun, Sep 15, 2019, 7:31 AM Gordon Gibby <ggibby@...> wrote:
Allison—

Can I ask your advice on a boat anchor project?  Soon I will start refurbishing more and more  HW 100s & SB 100s.   I read that only fractions of a white are needed to drive a 6146 amplifier (driving the grid).  It’s a very high input impedance.   We just learned how to make 1 to 49 baluns easily for End Fed half wave resonant antennas...(fantastic results!)

. would some 1:100 or so ferrite Baluns work well to feed the output of a micro Bitx  into the final amplifier stage one of these older boat anchor radios??

Gordon KX4Z




On Sep 14, 2019, at 23:48, ajparent1/KB1GMX <kb1gmx@...> wrote:

EXTERNAL EMAIL: Exercise caution with links and attachments.


Farhan,

Still around and busy.

Check out OZ1PIFs work.  I've run one section of that a 4x4 parallel push pull
on 6M to over 200W.  The yabut is you need 28V power at 15A!  At 12V that
doesn't work well.  The up side is to get that power not much drive is needed
and its easy to over drive it 5-6W was enough.  Its remarkably tough as I
haven't fried it.  The IMD is only OK.  However a really good and heavy built
low pass is needed as the first one with #18 air wound coils heated too much,
#14 wire solved that along with 500V metal clamp mica caps.

A pair of IRF510s can do 40-50w easy at 28V with 1-2W of drive. 
Parallel parts are harder as then you have parallel paths for
RF current to balance.  Cheat on the layout it becomes unstable
or poor power out.  Getting them to as high as 70mhz was not all
that hard if you pay attention to input and output impedance
they must be low for stability and reasonable power.

The 8 IRF510s are cheap but the total system is not cheap as you still
need a big heatsink (130mm by 130mm with 40mm fins  and a  fan) , TR
switching that can handle the power, and a package to put it all in. 
Cheap 28 or 48 volt switching  power supplies are not a solution as
often they are way to noisy on receive.

Things to consider 100W at 12V is about the limit using the right parts.
However the relays now have to be good for at least 5A and 100V
at the output and the input they can be fairly light.  With higher power
higher currents and voltage are the expected thing and parts like
#12 wire, 500V silver mica or ATC 800B are the required parts.

High power (more than 20 or so watts) tends to be expensive and
getting a reliable amp is never trivial task.

Allison

ajparent1/KB1GMX
 

Gordon,

Short form no.  The 6146 does require power to the grid to run though small.
the key thing is you need 25-50V p-p of RF and to do that across the grid resistor
and the 10-20pf of input capacitance.

The other part is that the bitx needs a stable load and the grid of 6146 is not that.

I'd think at several watts a 4:1 or maybe 9:1 would do.  However if yu want stability 
you have to load that transformer with resistance as the 6146 just like IRF510
is very intolerant of sloppy RF design and will oscillate to destruction.

Allison

Gordon Gibby
 

Thank you very much for that information, I will try either 450 ohms or 900 ohms, Little noninductive dummy load and a step up balun to match, that should work in the 1 to 3 watt range if I my calculations are correct;  you’ve given me excellent guidance.




On Sep 15, 2019, at 09:21, ajparent1/KB1GMX <kb1gmx@...> wrote:

EXTERNAL EMAIL: Exercise caution with links and attachments.


Gordon,

Short form no.  The 6146 does require power to the grid to run though small.
the key thing is you need 25-50V p-p of RF and to do that across the grid resistor
and the 10-20pf of input capacitance.

The other part is that the bitx needs a stable load and the grid of 6146 is not that.

I'd think at several watts a 4:1 or maybe 9:1 would do.  However if yu want stability 
you have to load that transformer with resistance as the 6146 just like IRF510
is very intolerant of sloppy RF design and will oscillate to destruction.

Allison

 

Sorry for using "CB amp" as a description,

It isn't a real specification anyway.  I just meant a type of amplifier that was gererally of a simple design that doesn't have a lot of features.  As Alison says, e.g. the EB63 uses basic diode biasing which needs high current in order to maintain a stable bias.  This results in low efficiency and higher battery drain.

Such types might also might be considered a single band design, as there is often only one (or none) lowpass filter.  They also generally have no gain compensation, which means uneven power output has you change bands, which would require transmitter power adjustment to keep from overdriving and burning out the final transistors which are not cheap.

The latter point really makes it not a good match for the ubitx since the transciever already has quite uneven power ouput with good power on the low bands and less and less as you go up.  An uncompensated amplifier will also do the same thing so that can be a troublesome thing to handle.

The AN763 would be a better match up as it has gain compensation built into the design already (no need to figure it out on your own, or extra parts to buy).  It also has a more efficient design for both biasing and for the final transformer (less heat and lower current draw).

Alison, I don't quite get what you said about the copper heat spreader and the AN762?  Reading the construction notes for both the EB63 and the AN763, I couldn't find a mention of that, or that the AN763 would have any more rare parts than the EB63?  Both seemed to be complete kits that only required a large aluminum heatsink and a heatsink fan (besides a case and lowpass filters and switches).

73,


Mark.