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: [1] 2 Next   Go Down
  Print  
Author Topic: AN758 300W AMPLIFIER HELP  (Read 4946 times)
N2XZS
Member

Posts: 8




Ignore
« on: January 23, 2009, 06:23:49 PM »

Hello Everyone,

I am in the process of building communications-concepts an758 300w hf amplifier. I have a few questions. This is my fist hf amp. So I ordered all the components and the 80m, 40m, 20m and 10m low pass filters too.

1)   Do I need any kind of spacing between the filters as I arrange them next to each other? Each comes with its own board, and I was thinking to arrange them next to each other on a top of another pcb since I have to connect some relays for switching.
2)   What amperage should I use on the switching relays? I was thinking in around 10amp DPDT
3)   If anyone built this amp before, is there a way to adjust the output power on the amp, or just changing the input will change the out as well?
4)   My biggest problem so far is, where to find a 36v/AC 20A transformer for the power supply? Since it requires 45-50vDC 15A power source?

This is it for now and I appreciate all the help the community can give me.

PS: perhaps someone has a web site or some building notes besides the Motorola bulletins.

Thank you
73 de n2xzs
Robert
Logged
K0BG
Member

Posts: 9839


WWW

Ignore
« Reply #1 on: January 24, 2009, 11:38:38 AM »

If it is like the other kits from CC, there are circuit boards available. Are you building your own PCBs?

Alan, KØBG
www.k0bg.com
Logged

N3JBH
Member

Posts: 2358




Ignore
« Reply #2 on: January 24, 2009, 02:54:03 PM »

1) Do I need any kind of spacing between the filters as I arrange them next to each other?

" Each comes with its own board, "

and I was thinking to arrange them next to each other on a top of another pcb since I have to connect some relays for switching.

I think he knows that. One of his questions is what kind a spacing is needed between the filters sounds like he wishes to build this as compact as possible.,
Logged
WB2WIK
Member

Posts: 20542




Ignore
« Reply #3 on: January 24, 2009, 06:19:10 PM »

A very good power supply providing 48V at 20A+ is readily available as a Telco "rectifier."

I have several, they normally provide 48V at 40-50Adc and they're pretty inexpensive: Hundreds of sources are available, since Telco (telephone company) central offices all run on 48Vdc (at hundreds of amps).

This is a really common voltage/current for inexpensive surplus power supplies.

WB2WIK/6
Logged
N2XZS
Member

Posts: 8




Ignore
« Reply #4 on: January 24, 2009, 08:12:28 PM »

First off I like to thank you all for the reply.

I am getting the amp and the filters from CC. I guess what im asking is when im constructing the amp:

i was thinking to place the filters on a separate board, since i have to add some relays, so i wanted to know how close they can be to each other. i m thinking of interference between them.

Hope that clarifies some questions.

Robert
Logged
AI4NS
Member

Posts: 320


WWW

Ignore
« Reply #5 on: January 25, 2009, 09:32:56 AM »

If you look at the LPF filter board for almost anything, the band components are very close to each other. Simply line them up in an eye-pleasing manner, put a relay on each end for your switching, and wire it up. What kind of rig are you connecting it to for a driver? You could use the band data out from the rig to autoswitch your LPF board.

Mike
AI4NS
Logged
N2XZS
Member

Posts: 8




Ignore
« Reply #6 on: January 26, 2009, 12:25:40 PM »

Thank you Mike. have to see if the FT847 has such a band data out but thats a good idea.

robert
n2xzs
------------------------------------------------------
If you look at the LPF filter board for almost anything, the band components are very close to each other. Simply line them up in an eye-pleasing manner, put a relay on each end for your switching, and wire it up. What kind of rig are you connecting it to for a driver? You could use the band data out from the rig to autoswitch your LPF board.

Mike
AI4NS
Logged
W9GB
Member

Posts: 2597




Ignore
« Reply #7 on: February 07, 2009, 07:10:16 AM »

Robert -

The problem with the Motorola applicaiton notes (AN758) and CC kits of parts -- is that it is NOT a Heathkit or Elecraft style kit.

As for your Low Pass Filter quesiton, my best advise is sit down and do some more research ont he Internet and LOOK at the photographs and construction of some very good builders.  LEARN from their earlier mistakes and Apprentice under their final achievements

HF Projects
Their catalog of LPF boards
http://www.hfprojectsyahoo.com/lowpassfilters.html
(look near bottom of web page)
This LPF is not rated for usage at 300 watts!
http://site.hfprojectsyahoo.com/hf_packer-amp

SuperPacker LPF board
http://www.somerset.net/arm/SuperPacker/SuperPacker.html
http://www.falara.org/tektalk/VALDEZ2.pdf

Bill Sabin's great Sept/Oct 2000 QEX article
http://www.arrl.org/tis/info/pdf/001009qex013.pdf

G3VPX 100 watt amplifier (see 7 band LPF using relays)
http://www.g3vpx.net/pa100w/pa100w_photos.htm
http://www.g3vpx.net/pa100w/pa100w.htm

Logged
N2XZS
Member

Posts: 8




Ignore
« Reply #8 on: February 02, 2010, 09:40:37 PM »

Hello Everyone,

So i built this amplifier but im not having much luck. i already blow out one transistor.

Did anybody built this thing before? i can use some help.

the problem is:

it seams that its working ok but i can only key up for about 3 seconds before both of the transistors get SUPER hot.

I checked the bias voltage. it seems to be ok.
i connected a separate bias power supply and it does the same thing.

Here is what i noticed. the bias power supply that i connected i have volt meters (its a variable power supply that i have home).
so every time i key up i see the volt meter jumps from 0.5 volts to about 2V which is too much for the bias. it does the same thing with the one on board bias circuit. I don’t know why and I don’t know if it should do that.

if anyone has any suggestions please email me or post here. i ran out of ideas.

Thanks
Robert
N2XZS
Logged
KM3F
Member

Posts: 494




Ignore
« Reply #9 on: February 02, 2010, 10:54:13 PM »

Without seeing the bias circuit it is hard to tell what problem you have.
Since you see the same action with an external supply, my guess is that the amplifier is breaking into oscillation with the bias turning the amplifier on even more.
That may be a "run-away" condition so the devices will over dissapate very quickly.
Try the following; with a power meter in the output, key up the radio without any driving signal.
Watch the power meter for any sign of output with no drive.
If you see anything, the amp is oscillating either at the intended driving frequiency or at some other frequency determined by the circuit elements.
If you see any tendency like this, you have your answer then need to find out what is causing it.
One area that can be critical is the 'negitive' feedback circuit between collector and base of each device.
I replaced devices in my 2 meter amp with more modern matched pair and discovered there was a small amount of oscillation on dead key.
A grid dip meter 'dipped' in the 280 mhz range at the feedback circuit that has an inductor.
The fix was to increase the series resistance from 15 to 47 ohms. No further oscillation after that change.
The thing with my amp was it has the series inductance that most other amps don't have but I left it in as long as the issue was resolved with a resistance change.
Good luck.
Logged
KM3F
Member

Posts: 494




Ignore
« Reply #10 on: February 02, 2010, 10:54:51 PM »

Without seeing the bias circuit it is hard to tell what problem you have.
Since you see the same action with an external supply, my guess is that the amplifier is breaking into oscillation with the bias turning the amplifier on even more.
That may be a "run-away" condition so the devices will over dissapate very quickly.
Try the following; with a power meter in the output, key up the radio without any driving signal.
Watch the power meter for any sign of output with no drive.
If you see anything, the amp is oscillating either at the intended driving frequiency or at some other frequency determined by the circuit elements.
If you see any tendency like this, you have your answer then need to find out what is causing it.
One area that can be critical is the 'negitive' feedback circuit between collector and base of each device.
I replaced devices in my 2 meter amp with more modern matched pair and discovered there was a small amount of oscillation on dead key.
A grid dip meter 'dipped' in the 280 mhz range at the feedback circuit that has an inductor.
The fix was to increase the series resistance from 15 to 47 ohms. No further oscillation after that change.
The thing with my amp was it has the series inductance that most other amps don't have but I left it in as long as the issue was resolved with a resistance change.
Good luck.
Logged
N2XZS
Member

Posts: 8




Ignore
« Reply #11 on: February 03, 2010, 06:11:24 AM »

Thanks for the reply.
the circuit is basically the Motorola application notes an-758 you can see the diagram on cc website. http://www.communication-concepts.com/appnotes/AN758300Sharp.pdf

it does sound like what you describing. although i don’t have any negative feedback at collector to base. i will try what you suggested with the no rive signal. at this point ill try anything. i just dont want to fry the transistors again. they are about $100 a pair. :-( and i alrady fryed one.

Thanks for all you help

Robert
Logged
WB2WIK
Member

Posts: 20542




Ignore
« Reply #12 on: February 03, 2010, 08:22:57 AM »

What size heat sink do you have on the transistors?  Did you use thermal grease between the transistor cases and the heat sink?  Are they bolted down very flat with no gaps or bubbles?

This amp cannot handle even one second of transmitting without an appropriate heat sink.  You probably know that, but I didn't see you mention it, and you did mention that they get very hot.
Logged
N2XZS
Member

Posts: 8




Ignore
« Reply #13 on: February 03, 2010, 10:05:55 PM »

WELL I DID NOTICED A SLIGHT GAP ON ONE THAT IM TAKING CARE OF. BUT THE OTHER IS PERFECTLY FLAT ON THE HEAT SINK. BUT THEY BOTH GET VERY HOT. EVEN THE ONE THATS ON FLAT. I DONT HEVE GEL ON IT I HAD BUT IT DIDNT MAKE A DIFFERENCE.

THE INTERESTING THING IS THAT THE TRANSMITTER WORKS WITHOUT THE BIAS VOLTAGE SET TO .5V. I TURNED THE ADJ. POWER SUPPLY ALL THE WAY DOWN AND IT STILL WORKED. ONLY ABOUT 40MV GOING INTO BIAS?Huh?
Logged
W8JI
Member

Posts: 9304


WWW

Ignore
« Reply #14 on: February 03, 2010, 10:58:39 PM »

The boards you are using are intended to be experimental boards that are starting points for engineers. They are not finished perfect products. They might or might not work when built exactly as described, and many of them are not prefect.

It sounds to be like you are either getting RF into the bias system either through ground loops or inadequate bypassing, or the transistors are getting so hot they are leaking from the collector to the base.

The transistors have to dissipate well over 100 watts each. This means the underside of transistors must mate against a very flat very smooth clean surface area of a large and must have a very thin layer of special heat conducting grease. You have to do that, and you MUST torque the transistors down correctly. Any air pockets, gaps, ridges, scratches, ripples, or dirt below a transistor can cause catastrophic problems.

RF stability problems will generally show on a spectrum analyzer, or as collector current that suddenly "pops up" and stays high after drive is removed. You might hear buzzing birdies or noise in a receiver. It does not really sound like you have this problem.


As for the filter question, the answer is "it depends". You cannot always just copy another layout and assume it will work. Filters that are not being used will have resonant effects, and if the open circuit or short circuit (depending on how you switch the filters) condition causes a resonance to move into the operating frequency or harmonic frequency range, it can upset operation of the filter. The best idea is to use a low capacitance relay and keep the filters some small distance apart while avoiding ground loops. This is something you can work out later.

When you shut the bias down the gain goes down, but the linearity changes. The amplifier, with improperly low bias, will certainly still amplify but will be class C. This can cause signal bandwidth problems on amplitude modulated modes like AM, SSB, and even CW (off-on keying).

If I were you, I would stop where I was at and go back to transistor mounting. I would make sure the heatsink was very flat and very thick or conductive under the transistors. The metal has to be so smooth and flat you should not see any gaps or thick spots if you check it with a thin layer of heat conductive grease. Bolt the transistors down with a few small dots of grease on the transistor bottom, and then remove the transistors. The bottoms of the transistors and the heatsink should wind up with an even uniformly thin (VERY thin) layer of grease.

Normally a 150 watt dissipation RF transistor requires a copper heat spreader, although a very thick lower thermal resistance aluminum heatsink will work. I designed so special extrusions that were thick enough, with the right airflow, to dissipate the power of four 150 watt transistors without using copper, but in most cases with normal heatsinks a copper spreader is required. The copper spreader should be very flat and thinly greased with compound.

It might help you to think of the transistors and heatsink as cylinder heads on an engine. you really want the same surface contact characteristics, where the deck is perfectly flat to match the mating surface. You can NOT have any air gaps, and you MUST use some very light coating of thermally conductive grease. Since even the very best grease is hundreds of times more resistive to heat flow than aluminum or copper, you MUST minimize the thickness of the grease. Do not expect it to do more than fill in microscopic gaps that are microns thick. Grease will NOT make up for a scratched or warped surface.

I generally make a pass over the mounting area with a four-flute end mill using a lot of clean lube spraying on the bit. The surface has to be almost mirror-like when using 150 watt dissipation transistors.

Tom












Logged
Pages: [1] 2 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!