hundreds of watts per 2879 ???????????????????????

[W5LZ]

I made a mistaken assumption to start with, so ignore any post I've made in this thread. Next time I'll pay more attention (I hope).
- 'Doc

 

[Captain Kilowatt]

Captain, the topic wasn't how to design a quality final power amplifier for the next HF rig to hit the market. I won't begin to dispute the fact the spectral purity sucks when you run every spec right to the maximum. I believe the basic topic of the thread was if it's possible to obtain hundreds of watts from a 2SC2879 on the fundamental frequency. The transistor is rated for 250 watts of dissipation. Anyone should understand it's quite possible to obtain 250 watts of output all the way down to 50% efficiency without exceeding the parts real specifications.

I know what the topic was and in regards to the parts real specifications, IMD characteristics IS part of that specification. As for big power out of a transistor, I once got nearly a kilowatt out of a 2N2222A. It didn't last very long but the numbers indicated what the total power output was before the device failed. BTW I'm still finding little pieces of it.

 

[Shockwave]

I know what the topic was and in regards to the parts real specifications, IMD characteristics IS part of that specification. As for big power out of a transistor, I once got nearly a kilowatt out of a 2N2222A. It didn't last very long but the numbers indicated what the total power output was before the device failed. BTW I'm still finding little pieces of it.

I'm not saying that IMD is not an important specification. I am saying it's not one of the specifications that determine what the maximum power output capabilities are once you're willing to sacrifice spectral purity for power output. LOL @ the KW from the 2N2222A. Switch over to the older metal cased 2N2222A and the pieces won't be so small next time

 

[dxhound]

To suggest a transistor with a collector dissipation of 250 watts can't make over 160 watts output running class C indicates a lack of understanding. The fact that Toshiba has published a minimum output of 100 watts has far less to do with what the part can actually make then things like collector dissipation and maximum collector current. The guys that do get hundreds of watts from each 2SC2879 have 18 volt battery banks and charge at 21 volts. These are competition amplifiers that were built to max out the transistors. They don't care if it blows up as long as they get the "five". The amps aren't clean and they aren't running linear either.

Lets apply ohms law here rather then BS. If you raise voltage by 50% power goes up by 100%. Raising voltage causes current to rise proportionally. Since current goes up by 50% when voltages is raised 50%, the power goes up by 100%. Volts times amps give watts. Realizing that the 2SC2879 will blow up at about 23 amps of collector current means that you must reduce drive so as not to exceed this limitation. When you have 18 volts at 20 amps on a 2SC2879 that's 360 watts of input power. I'll be conservative and say it's running 70% efficient in class C. That's 252 watts of output with only 108 watts being dissipated by the transistor.

It becomes clear to see that even at 252 watts of output the transistor hasn't even come close to it's maximum collector dissipation however, it is riding the maximum current close to the limit. Even if you were to bias it up in the linear mode you would still not exceed the maximum collector dissipation at 250 watts output. This is certainly not the best way to run an amplifier but the topic here was "hundred of watts from a 2SC2879" as though it's not possible when it certainly is.

Wow this thread exploded!!! Just like a 250watt 2879 will!! You are refering to input power, not output!! Take an avg class C effeciency rating of 70% and use that against your 250 watt input...and guess what ..............175 watts. The rest is deisipated in heat!! That 250 watt dispation on the specs is an all inclusive rating.......not an out put rating.

Think about it....Do you really believe you would get a 150% increase in power out with only a 50% increase in Vcc??????? The only way to truly test this would be to build a jig and look at it on a Spectrum analyzer at the target frequecy.

Your lack of understanding on how watt/swr meters work is proven in your comments about, if the harmonic content was that high, it would show as reflective power (SWR). I can tell you that on a fatboy 2 2sc2879 the 2nd order harmonic was almost 4 watts output. I know this from not only the Specan showing me, but also because for the hell of it, I hooked up a directional coupler tied to a 6 meter vertical and managed to produce an S5 on a buddy on 6 meters 5 miles away.......and he could understand me!!!!

There are two ways to calibrate an SWR meter. W/ a carrier (best method) recheck with same carrier applied for reflective. Well unless you are saturating the transistors with a high carrier, the IMD will be low. So you will get a more accurate reading. But when you talk, the Calibration is now lost because the power level has changed.

The other method is to calibrate the meter off the peak output you are using. the best would be to use a 2 tone 100% modulated signal. But everyone usually whistles. Guess what, you are now producing all those harmonics.....and now you are calibrating your meter to a false output.

So all those Harmonics are now coming back from the non resonant antenna(at the harmonic freqs) and causing exsessive current on the zener diode inside your meter. Well, that diode, and the current across it, is what makes the meter dance!!!!

 

[dxhound]

Oh and I dont know if this is what you meant or not, But class C does not mean more output. It just means the tramnsistor conducts for less than 180 deg of the signal and draws less current. truth is, Class AB will produce more power with the same Vcc and same drive level.

 

[Shockwave]

Wow this thread exploded!!! Just like a 250watt 2879 will!! You are refering to input power, not output!! Take an avg class C effeciency rating of 70% and use that against your 250 watt input...and guess what ..............175 watts. The rest is deisipated in heat!! That 250 watt dispation on the specs is an all inclusive rating.......not an out put rating.

Think about it....Do you really believe you would get a 150% increase in power out with only a 50% increase in Vcc??????? The only way to truly test this would be to build a jig and look at it on a Spectrum analyzer at the target frequecy.

Your lack of understanding on how watt/swr meters work is proven in your comments about, if the harmonic content was that high, it would show as reflective power (SWR). I can tell you that on a fatboy 2 2sc2879 the 2nd order harmonic was almost 4 watts output. I know this from not only the Specan showing me, but also because for the hell of it, I hooked up a directional coupler tied to a 6 meter vertical and managed to produce an S5 on a buddy on 6 meters 5 miles away.......and he could understand me!!!!

There are two ways to calibrate an SWR meter. W/ a carrier (best method) recheck with same carrier applied for reflective. Well unless you are saturating the transistors with a high carrier, the IMD will be low. So you will get a more accurate reading. But when you talk, the Calibration is now lost because the power level has changed.

The other method is to calibrate the meter off the peak output you are using. the best would be to use a 2 tone 100% modulated signal. But everyone usually whistles. Guess what, you are now producing all those harmonics.....and now you are calibrating your meter to a false output.

So all those Harmonics are now coming back from the non resonant antenna(at the harmonic freqs) and causing exsessive current on the zener diode inside your meter. Well, that diode, and the current across it, is what makes the meter dance!!!!

Think about it.... Go back and read my post to let me know where I said 50% increase in voltage is 150% increase in power. It looks like you're not reading or you're ignoring facts. Ohms law says a 50% increase of voltage into the same load causes a simultaneous 50% increase in current. The end result is a doubling of power into that load.

Your lack of understanding on how the VSWR or power meter works couldn't be any more evident then when you referred to the "excessive current on the zener diode inside your meter". Some of the first things you learn about in electronics school is that a zener diode is used as a constant voltage source and has no purpose in the rectification of RF.

The story about using a Fatboy 2 pill to talk to someone on six meters 5 miles away with a decent signal from the second harmonic generated from a CB had me so LOL. You really found an amp that makes about 5% of it's output on the second harmonic? You think 4 watts of VHF power won't show any reflected power into a CB antenna? Try connecting your 6 meter radio to your CB antenna with 4 watts forward and measure the reflected power!

We can keep going with this nonsense but no one here has yet been able to give a single valid reason why a transistor with 250 watts of collector dissipation can't possibly make 250 watts of output power on the fundamental frequency. There isn't one. No one claimed it would be reliable or clean either.

PS. Class C does mean you can use more of the transistors dissipation for RF and less for the heat the reduced efficiency the DC bias would add in the linear mode. Class AB will blow up at these power levels in many cases.

 

[Shockwave]

Wow this thread exploded!!! Just like a 250watt 2879 will!! You are refering to input power, not output!! Take an avg class C effeciency rating of 70% and use that against your 250 watt input...and guess what ..............175 watts. The rest is deisipated in heat!! That 250 watt dispation on the specs is an all inclusive rating.......not an out put rating.

Think about it....Do you really believe you would get a 150% increase in power out with only a 50% increase in Vcc??????? The only way to truly test this would be to build a jig and look at it on a Spectrum analyzer at the target frequecy.

Your lack of understanding on how watt/swr meters work is proven in your comments about, if the harmonic content was that high, it would show as reflective power (SWR). I can tell you that on a fatboy 2 2sc2879 the 2nd order harmonic was almost 4 watts output. I know this from not only the Specan showing me, but also because for the hell of it, I hooked up a directional coupler tied to a 6 meter vertical and managed to produce an S5 on a buddy on 6 meters 5 miles away.......and he could understand me!!!!

There are two ways to calibrate an SWR meter. W/ a carrier (best method) recheck with same carrier applied for reflective. Well unless you are saturating the transistors with a high carrier, the IMD will be low. So you will get a more accurate reading. But when you talk, the Calibration is now lost because the power level has changed.

The other method is to calibrate the meter off the peak output you are using. the best would be to use a 2 tone 100% modulated signal. But everyone usually whistles. Guess what, you are now producing all those harmonics.....and now you are calibrating your meter to a false output.

So all those Harmonics are now coming back from the non resonant antenna(at the harmonic freqs) and causing exsessive current on the zener diode inside your meter. Well, that diode, and the current across it, is what makes the meter dance!!!!

Just to add a little more info here, the 250 watts I was talking about is not input power. It's the power the transistor can dissipate in heat. That 250 watt collector dissipation translates into the capacity to produce 250 watts of output in a system that has an efficiency as low as 50%. You didn't even completely calculate out that 250 watts of input power at 70% efficiency is only still only 75 watts of collector dissipation. Long way to go before you hit 250.

You're also confusing the characteristics of IMD and harmonic distortion. IMD creates in band spurs in your sidebands when you modulate. Right along side of the fundamental frequency and well within the meters frequency range for accurate measurements. Not at all what the original post was suggesting. That talked about harmonic frequencies several times the line section ratings. It is excessive harmonic component that has the capacity to exaggerate power readings because of how it's much higher frequency relates to the sampling within the line section.

 

[dudmuck]

We can keep going with this nonsense but no one here has yet been able to give a single valid reason why a transistor with 250 watts of collector dissipation can't possibly make 250 watts of output power on the fundamental frequency. There isn't one. No one claimed it would be reliable or clean either.

Its certainly possible, but not practical.
Its sort-of like cpu overclocking with nitrogen cooler. Sure you can get 6GHz, but your average joe cant do that.

Trying to get 250 watts out of a 2sc2879 is like trying to overclock a pentium2 to run as fast as todays off-the-shelf computer.
Today there are pills with Rth(j-c) better than 0.3 Celsius/Watt. They can dissipate more than 600watts without problem, and put out half that in RF without breaking a sweat.
Yet its hard to beat $20 per pill.

 

[BOOTY MONSTER]

I made a mistaken assumption to start with, so ignore any post I've made in this thread. Next time I'll pay more attention (I hope).
- 'Doc

its all good Doc . being perfect isnt all its cracked up to be .... look what happened to jesus christ .

 

[mackmobile43]

its all good Doc . being perfect isnt all its cracked up to be .... look what happened to jesus christ .

He wasn't perfect, just a sucker for a good cause.

 

[Shockwave]

Its certainly possible, but not practical.
Its sort-of like cpu overclocking with nitrogen cooler. Sure you can get 6GHz, but your average joe cant do that.

Trying to get 250 watts out of a 2sc2879 is like trying to overclock a pentium2 to run as fast as todays off-the-shelf computer.
Today there are pills with Rth(j-c) better than 0.3 Celsius/Watt. They can dissipate more than 600watts without problem, and put out half that in RF without breaking a sweat.
Yet its hard to beat $20 per pill.

How practical 250 watts per 2SC2879 would be depends on what you're trying to do. If you're looking to make under 1 KW in your vehicle then it's probably not practical. It's not economical for someone running anything between a "two pill" and a "five pill" to purchase the 18 volt battery bank and install the Leece Neville alternator with a modified regulator. It just makes more sense to buy a "ten pill"

Now let say you have a 10 transistor amp and still want more power. You have to at least double your power to see a change. That means at least 16 finals and a bigger driver. Now it makes more sense to buy the batteries and alternator to get the extra 3 db because you're going to need them anyhow to power this size amp properly.

As far as the comparison of overclocking a CPU or getting 250 watts output from a 2SC2879, when you overclock the CPU you have exceeded it's specifications. When you run a 2SC2879 at 250 watts output in class C you haven't exceeded a single maximum published specification on this part except perhaps the drive level. It's still dissipating less then 108 watts out of 250 maximum. The 100 watt output rating is a MINIMUM rating at 12.5 volts VCC. Gain (HFE), and output will be much higher at 18 volts.

The new 600 watt FET's they have today are amazing and so is their price! They make a matched pair in a single package now too. They are commonly used in FM broadcast transmitters final power stage. Often many are combined together to form outputs well into the kilowatts. They also need 50 volts VDD to operate and that makes them not practical for mobile use.

 

[BOOTY MONSTER]

He wasn't perfect, just a sucker for a good cause.

class c or ab biased ........
What Would Jesus Do ??????

 

[BOOTY MONSTER]

"The 100 watt output rating is a MINIMUM rating at 12.5 volts"


hey shockwave , are you saying a 2sc2879 can do more CLEAN power on 28MHz. using 12.5 volts of power than its 100 watt PEP rating ? . if it can wouldnt the extra watts have so little effect on a real world signal (none) that it wouldnt be worth the extra heat and stress put on it and potientally shorten its life ?

 

[Shockwave]

"The 100 watt output rating is a MINIMUM rating at 12.5 volts"


hey shockwave , are you saying a 2sc2879 can do more CLEAN power on 28MHz. using 12.5 volts of power than its 100 watt PEP rating ? . if it can wouldnt the extra watts have so little effect on a real world signal (none) that it wouldnt be worth the extra heat and stress put on it and potientally shorten its life ?

No, I sure am not saying it can do more "clean" watts. As has already been pointed out the IMD will begin to suffer as output is increased. You might get about 120 watts of clean power in unmodulated CW use. Of course this won't make a noticeable increase in signal and wouldn't warrant doing. Especially on the ham bands using SSB where it would be quickly noticed and whined about.

When you talk about CBer's that run kilowatts from transistors, convincing them that the IMD created by running 18 volts at 250 watts output per transistor makes it not worth doing is.........just not happening. They have one valid point too because they have gained over 3 db in signal and that is very noticeable if you're competing with others side by side. You are absolutely right that this does increase the risk of transistor failure.

 

[Shockwave]

Here is an option for those on AM who want to experiment with getting about twice the output from 18 volts the easy way. First understand that you are risking your transistors if you make a mistake with drive or carrier level and your amplifier has to have a few key things in place. Many can do this without modification. You are also likely to splash more because of IMD.

The amplifier has to be class C. That means the bases of the transistors are at DC ground trough the input transformers. The amp should have 2SC2879 transistors. The capacitors across the output transformers and transistor collectors should be no higher then 680 pf. If your amp fits this description you can probably run 18 volts. Some 16 volt electrolitics may get warm.

You need a 6 volt battery. The bigger it is the longer you can talk. Place this battery in series with the positive 12 volt line feeding your amp. You can use a 6 volt battery charger powered from a cigarette lighter 110 volt AC inverter to charge the 6 volt battery. You could also just use the battery if you're not talking long and take it home to charge.

The most important thing to consider in order to prevent transistor failure is drive level. The amp will require LESS drive then it did on 12 volts. Your radio should have a variable RF power control on it. Drive should never be turned up past the point where the carrier exceeds 25% of the max PEP. If you use a true RMS meter this point is when the meter still moves the slightest bit forward with modulation. Not backwards.

Lower drive will increase reliability. One quick mistake with too much drive will burn open the transistor junctions! The amp will get hotter and may need more air flow. I don't offer this information here for anyone to try it. You need to know what you're doing and what the tradeoffs are. I offer the info for the guy that "needs" 18 volts the easy way but sees mounting alternators under the hood as not an option.