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

donate to eham
   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: SSB rated bipolars versus FM class C rated power transistors.  (Read 13037 times)
N3OQD
Member

Posts: 104




Ignore
« on: May 01, 2010, 07:13:59 AM »

In my 2 meter RF Concepts 160w linear, it has a pair of FM rated MRF-247 transistors rated 75w each.  In order to get the FM rated output, I have to drive it with about 30w as printed on the front of the amp.  Yet it does have the capability of SSB operation.  Most believe that the 160w is also the SSB rating.  Now what I did was a simple test some time ago.  When driven between 5 and 10 watts, its gain is steep and consistant.  Appearing to be the most linear.  10w in gives about 100w or so out.  After this point and increasing drive, percentage of gain starts noticably tappering down.  So in this, when operating SSB, always I drive it with no more that 10w for that 100w.  Then I remembered the posting about the HLA-150.  How Tom got about 90w out of it and its saturation point was just above the total ratings of the two power transistors.  So I looked up the SD1446 transistors that are in it and discovered that they too are class "C" FM rated 70w each.  Looking at the drive ratio to output power that I got from my 2 meter amplifier I pretty much came up with roughly the same ratio that Tom got with the HLA-150.  Is my results coincidental or more than that?
Logged
N3OQD
Member

Posts: 104




Ignore
« Reply #1 on: May 01, 2010, 06:23:06 PM »

Driving an amp with bipolars into the gain compression region SSB mode is ok?  Hmmm.   Many years ago I tried that and I got scores of splatter complaints.  Pissed a person or two here and there as well.  Yes, it was more raw power out but it was annoying to others on close in ajacent frequencies.  I experimented and found that by driving the amp SSB mode in the region just before compression starts solved the issues.  Question is, why did this solve my problem if doing the latter is ok?
Logged
KM3F
Member

Posts: 902




Ignore
« Reply #2 on: May 01, 2010, 07:53:28 PM »

Your gain compression observation is corrent except for actual numbers.
The simpler way to look at this is the output power vs input power.
On the curve, the straightest portion of the curve is the most linear.
Driving the peak power to just the region where the compression begins to get very pronounced is the max you may do and still keep it in the "practical" linear portion of the curve.
Many use these amps without the sleightest thought of how much drive they should limit to and keep the signal relatively clean.
The term FM rated is more for how much heat they will dissipate without loss 'of life'.
That is largely determined by the size of the heat sink and how much cooling as well as key up time.
Otherwise the same device will work SSB with out special issue as long as the bias point is proper for this mode.
Also remember that the so called linear function is of little consequence on FM where gain compression makes little to no difference because the gain is not changing as you transmitt, only a small frequency swing, so linearity is not an issue on FM.
To go one step farther, the use of these amps on AM mode requires a different thought approach.
Remember that a carrier modulated 100% = a power 4 time the carrier at peak output.
Using your example of 100 watts for low distortion; 100 watts divided by 4 = 25 watts max carrier, then 100% modulation will drive the peak power to 100 watts and should sound fine as long as the driving radio has no distortion in the driving signal.
Trying to drive the amp to what many consider 160 watts  will only result in severe distortion and splatter because it would drive the amp into gain compression and a form of clipping and you should know what effects that would have at RF frequencies.
Just because the mfger has rated the amp for a certain power does not make it so in every mode. You must know the differences and how to use it properly in each mode
It's not really difficult to understand if one sits down and really thinks about what is going on with each mode and the requirements that must be met for each.
Lastly there is a large difference between IMD and harmonic output in the usual sense but one can add to the other if the equipment is just plain abused in use.
Usually IMD cannot be filtered out like harmonics are with harmonic filters.
IMD is the generation of 'mixing products' in the amplifier operation and usually do not respond to any filtering in the output or input.
Logged
N3OQD
Member

Posts: 104




Ignore
« Reply #3 on: May 01, 2010, 08:48:01 PM »

KM3F:  As you have stated I did and still do compare input to output SSB power for a basic comparison.  When I find the beginning of the gradual "knee" point then I mark down my drive and not exceed that.  This knee point does move upward or downward in preportion to the raising or lowering of the main B+ supply (that is within its supply operating limits).  As unorthadox as this might sound as "documentation" none the less this is my refrence.  As far as FM (usually like unto class "C") the amp can be driven pretty much near saturation without too much problems.  This would explain my RF Concepts 2-315 rating as a total of 160w.  As far as AM, I almost never use this mode but I do thank you for the information.  One part RF carrier to four parts peak AM modulation.  This 4 to 1 ratio within the linearity of the amp just like SSB. 
Logged
G3RZP
Member

Posts: 1090




Ignore
« Reply #4 on: May 01, 2010, 10:58:27 PM »

In the days when I did a lot of this sort of stuff, if I recall correctly, a lot of the bipolar 'SSB' devices had internal emitter ballasting resistors. This gave negative feedback, thus improving IMD. For FM, where efficiency is wanted, the ballast resistor was removed so that the RF saturation volatge was as low as possible.
Logged
N3OQD
Member

Posts: 104




Ignore
« Reply #5 on: May 02, 2010, 06:07:28 AM »

I did some looking around for a user manual for my amp but was unable to find info partaining to this subject.  I did though find the user manual for a similar amp the Mirage B5018-G.  It is a 2 meter amp with drive capability of 50w.  There is a section dedicated to linear operation.  It says not to drive the amp any higher than 50w FM or 20w AM or PEP.  It further explains the importance of keeping AM and SSB drive levels reduced in order to keep the amp linear and reduce splatter and IMD.  So, I had no idea that my question partaning to two different bipolar transistor power ratings would carry this far but this is very good.  As far as SSB rated transistors being ballasted and the latter not, that is an interesting thing and I will try to look further on it.  Thanks to all the guys for their time and efforts in posting information here.
Logged
KM3F
Member

Posts: 902




Ignore
« Reply #6 on: May 02, 2010, 04:20:09 PM »

Here is my experience with an older Brown faced M 3016 amplifier.
Rebuilt  it with different power transistors and set the bias for SSB use according to the bias value used in my KW TS 480 that uses the same devices in it's final stage.
Results without lab type measurements look very good and side monitoring by other stations reports a clean operation to either side of the operating frequency.
A test of relitive gain for input vs output shows no notable gain compression within the drive levels I use. I was very suprised at seeing this result.
The transverter seems very well matched to the amplifier as it is now.
On AM, I use a 10 db pad to reduce the drive power to the transverter so it only drives the TV  amplifier to 7 watts carrier for about 28 watts output (at 80% AM modulation) to drive the the amplifier resulting in about 150 watts PEP in the AM mode with clean results.
On SSB the drive is about 28 watts in for about 150 out into a 12 element beam.
Using this setup with speech processing results in a very heavey, loud, clean signal, according to all reports. I get comments about hearing me even with no signal meter reading on the recieving end. The problem I have with this is I get out better than I can hear the other stations, very often.
FM operation is equally clean.
Granted the system has not been looked at for IMD or even harmonic output but the results of actual operation have never been in question as I have repeaedly ask to be alertedi by local ops if there is ever any issue I need to know about..
I'm very happy with the results and pride on a clean signal with fine audio.
As an extra bit of info, I use a switching power supply loaded up with a lot of capacitance external to the supply to feed peak currents from, keeping the supply quite 'hard'.
The amp and transverter are also fan cooled.
I beat and beat on it with high average drive and it never gives up.
With speech processing, the supply voltage has nearly no drop in voltage more that about a tenth of a volt on voice peaks as measured with a Fluke digital meter powereing the amp at 15 volts..
Ok guys bang away at it if you like but it don't get much better than this.
I have no doubt that lab measurement will show the IMD levels I cannot see.
It's the way I intended it to be and about the best quality hard punching 2m signal on the band.
Logged
KM3F
Member

Posts: 902




Ignore
« Reply #7 on: May 02, 2010, 05:01:00 PM »

Owen, thanks. That's good to know it may be possible to do with software analysis.


Ken.
Logged
EA2FL
Member

Posts: 4




Ignore
« Reply #8 on: May 08, 2010, 09:02:01 AM »

Hi good people,
Just my tests with the AN758 from Communication Concepts, using new HF50-250 bipolar RF power transistors using two of them in this push-pull from master Helge, W7ES (sk).
These transistors I chose are rated 250w each, instead of the 150w out from the old Motorola, resulting in a 66% higher nominal power.
Besides power, this new design has 60% improvement in breakdown voltage increasing an always needed robustness, and get up to 300w output each according to manufacturer specs:

http://www.spectrumdevices.com/products

I only had to make the holes at PCB a little bigger to case 550 to fit them in, replacing  the 30 year old, 150w, Motorola transistors.

Even if, eventually, any amplifier device will saturate at max output, we only have to take care of not over-driving. Is easy to check output versus input power simply dividing OP/IP and finding gain factor changes. At some point, near max output, gain gets too low indicating that amplifier is not linear anymore even if it looks like linearity is pretty good at lower levels. See an almost straight line at Output vs Input diagram here:

http://www.spectrumdevices.com/datasheet/HF50-250_Aug08.pdf

We can check IMD (InterModulationDistorsion) by looking at a spectrum analyzer, of course. If analyzer is not at hand, let's get a simple DTMF microphone and transmit into the load by pressing any keypad "number".
It is so easy to tune a receiver, as someone said, to check at which power levels our intermodulation products are only 30, or as high as 20dB below the two tone transmitted signal (i.e. the DTMF two tones show a 20dB over S9 signal, while other products are as high as an S9 signal).

If we keep increasing input driving power, our output watt meter will read more and more power, but increasing IMD adding interfering power to the sides almost as high as our desired signals.

Do not confuse this with harmonics. When transmitting on 1.8MHz you generate a little power on 3.6MHz (2nd) and 7.2MHz (4th), 10.8MHz and so on (all the even order harmonics) PLUS a "lot of power on odd harmonics" like 5.4 (3rd), 9 (5th), 12.6(7th), 16.2 (9th) and so on up to the maximum frequency the amplifier will work, well over 30MHz.

Low pass filter is mandatory. If you check power before and after a well made low pass filter you MIGHT think that filter is bad since is robbing so much power, when is not.
Watt meter will read and add all harmonics also!

Not that difficult to see ...

I still have to check different VDC to find best voltage to choose to get best output transformers efficiency, since output transformer has a fixed ratio of 1:4 and output impedance from transistors changes with output power level AND supplied voltage.
In the meantime let me tell you what I found out, be wise and think twice:

With a strong power supply of around 50V and 25A (1250w PEP from P.S.), you can end up thinking that efficiency is much higher than rated.
You will get a dirty 800w (64% apparent efficiency) including all harmonics, about 650w (52% nice efficiency) after the low pass filter but over-driving and generating too many IMD's, since efficiency gets higher the higher the output power over driving, while gain factor drops down.

You only loose 0.7dB (oh-my-god, oh no, signal will DROP form s9+20.7 to ONLY 9+20, hi, hi) or 100w less from 650 to get a good+clean 550w PEP output (at rated 44% efficiency) with just a small, easy to get, push-pull PCB and pair of these strong new devices.
Logged
KB1GMX
Member

Posts: 1768




Ignore
« Reply #9 on: May 15, 2010, 08:19:37 PM »

I use a pair of MRF247s at 2M and one at 6M (not rated but works well).

The biggest item to pay attention to is the more linear devices are both
costly and  generally under 100W at 12V.  Most are designed for less power
or higher power at higher voltages.  Also devices designed for 12V (nominal 13.Cool
really want 13.8 and at 12V the power does drop a lot.  that seems trivial
bit 75W out at 13.8V means something near 150W in or about 11A and pair
will easily suck down 22-23A so maintaining 13.8V means heafty power
supply and wires.

The device is rated for 75W FM and will do that well for SSB but don't push it
high as it will do that but above 75W it does get into gain compression.
forr FM class C thats fine, for SSB not good. Note the 2SC2782 is the current
replacement for the MRF247 and the specs are a bit better.

Running pair of them is clean (better than -30imd) at 140W, push them to 160w
(they will do it!) it drops to -20imd.  For the price of the device and design
it's decent.  Now if you want more power, gain or cleaner run a MRF141G
or better BLF248(28V), BLF278(48V) both will want more than 12V power
but will  give more power 300-350W and a whole lot cleaner too.

The big difference is with the MRF247s is you need at least 18W drive for
full power the FETs (MRF141G or BLF) will need less drive for 300W.  The con
is most of the fets want 28 or 50V power and for mobile work that means a
more expensive power source.  At home the power issue is a wash as there
are sources for cheap 28 and 50V suppies.

Note those BLF2xx parts are not cheap but 2 MRF247s run about 70$, MRF141g
about  $150 and the BLF parts under $150 currently.  Prices vary so shop.

All depends on how much power you want, How much drive you have
and how many bucks.


Allison
Logged
Pages: [1]   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!