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: Prev 1 [2] 3 4 5 Next   Go Down
  Print  
Author Topic: Will New Power Transistors End the Need for Tube Amplifiers in the Future?  (Read 21796 times)
K2ACB
Member

Posts: 64




Ignore
« Reply #15 on: January 23, 2012, 06:12:48 PM »

I am not an expert in semiconductor parts and technology. Doing some research  the the transistors I originally mentioned in my first message are part of the MRFE6VP power transistors that Freescale semiconductor of Austin, Texas manufactures.

The part number for 1250 watt power transistor is MRFE6VP61K25HRS. It is a 1,8 -600MHz 1250 w 50v lateral n-channel broadband RFPower mosfet,Its DS breakdown voltage-min 125v. Its gain is 24 DB. it can be purchased from Future Electronics for $245:00

The part number for the 600 Watt power mosfet is MRFE6VP5600HRS. Its price is $158:00

I understand in 2004 motorola divested itself from manufacturing and designing semiconductors. That division was spun off as a seperate company under the name Freescale Semiconductors. One study has said that Freescale has 60 percent of RF semiconductor market. The company went public in 2011.
73
Alan-K2ACB
Logged
W8JI
Member

Posts: 9296


WWW

Ignore
« Reply #16 on: January 23, 2012, 06:23:21 PM »

I am not an expert in semiconductor parts and technology. Doing some research  the the transistors I originally mentioned in my first message are part of the MRFE6VP power transistors that Freescale semiconductor of Austin, Texas manufactures.

The part number for 1250 watt power transistor is MRFE6VP61K25HRS. It is a 1,8 -600MHz 1250 w 50v lateral n-channel broadband RFPower mosfet,Its DS breakdown voltage-min 125v. Its gain is 24 DB. it can be purchased from Future Electronics for $245:00

It is really a 700-800 watt maximum linear transistor. The 1250 watts is class C pulse with a short duty cycle.
 
Quote
The part number for the 600 Watt power mosfet is MRFE6VP5600HRS. Its price is $158:00

I never looked at that device.

Thanks for the info Alan. I was wondering if it was a new device.
Logged
K2ACB
Member

Posts: 64




Ignore
« Reply #17 on: January 23, 2012, 06:31:19 PM »

According to the specs of this power transistor the MRFE6VP61K25HR6 is an extremely rugged 1250 watt LDMOS transistor capable of withstanding a near 100 percent load mismatch of 65 to 1 VSWR at full CW power.According to the literature these devices offer 50 volt operation ,high gain,high efficiency,extremely low thermal resistance and are very stable down to 2MHz. This makes them ideal transistors for very high stress applications. They can be used for co2 lasers,plasma genrators,MRI Power Amplifiers and Broadcasting ,lLghting and Avionics.

I have not heard of any other power transistors that can be used with a commercial or amateur radio amplifier that can withstand nearly a 100 p-ercentload mismatch of 65 to 1. i do;t know what tubes can do this. but again i am not a specialist on power transistors.

I agree with those that wrote that solid state hf amplifiers are much more complex than tube amplifiers and require band filters. They also have been  much less tolerant to changes in SWR. Isthis new power transistor revolutionary when it comes to SWR matching?
73
Alan-K2ACB
Logged
W8JI
Member

Posts: 9296


WWW

Ignore
« Reply #18 on: January 24, 2012, 05:42:19 AM »

According to the specs of this power transistor the MRFE6VP61K25HR6 is an extremely rugged 1250 watt LDMOS transistor capable of withstanding a near 100 percent load mismatch of 65 to 1 VSWR at full CW power.According to the literature these devices offer 50 volt operation ,high gain,high efficiency,extremely low thermal resistance and are very stable down to 2MHz. This makes them ideal transistors for very high stress applications. They can be used for co2 lasers,plasma genrators,MRI Power Amplifiers and Broadcasting ,lLghting and Avionics.

I have not heard of any other power transistors that can be used with a commercial or amateur radio amplifier that can withstand nearly a 100 p-ercentload mismatch of 65 to 1. i do;t know what tubes can do this. but again i am not a specialist on power transistors.

Be really careful of that. Those types of spec's are extremely misleading, to the point where we really should just ignore them.

To be blunt, they are really just marketing hyperbole.

If you ran that transistor into a 65:1 VSWR at full power and rotated phase angle, you would either destroy the junction from excessive voltage (instantly) or damage it from excessive heat.

As a matter of fact, even a 2:1 VSWR at full ratings could pop the transistor.

Any experienced designer will likely confirm what I am saying here.
 
Now if it is a short duty cycle pulse operation, and the circuit limits voltage or the power supply limits current, the device certainly will tolerate a 65:1 SWR.

For example, in the 1990's, I designed a two-transistor MRF150 PA stage for a medical device that ran 1200 watts peak power with a 65 uS pulse. At 1200 watts output, I could short the load, open the load, or run any phase angle impedance between those conditions and not hurt the transistors. I could make a big spark, and not hurt the transistors, and I had no SWR shutdown circuit at all.

I could have used that as a case that MRF150 transistors will take a 65:1 VSWR at all phase angles at 600 watts per device, but the truth really is otherwise. The short pulse time and power supply limitations prevented damaging the FET's.

If I took the very same FET's I could fry them with a 1.5:1 SWR  in another more conventional circuit.

It is more than a little silly to think we could build a linear amplifier and run those FET's at 1000 watts CW or SSB, and expect them to last with even a 2:1 SWR in a higher duty cycle mode.

Quote
I agree with those that wrote that solid state hf amplifiers are much more complex than tube amplifiers and require band filters. They also have been  much less tolerant to changes in SWR. Isthis new power transistor revolutionary when it comes to SWR matching?

No.

FET's have been good for a long time, as have tubes. If duty cycle is short, and we don't exceed peak voltage ratings, they all can make impressive SWR tolerance numbers.

The reality is, when the device is in a real amplifier with some significant duty cycle, it will be easy to damage them unless the system somehow limits voltage, current, and heat. The same is true with a tube.

There isn't anything there that excites me.

73 Tom
Logged
KC8Y
Member

Posts: 247




Ignore
« Reply #19 on: January 24, 2012, 06:42:39 AM »

W8JI---

I agree with all comments...

Have an EE degree and a Master in Industrial Computer Technology.  I seem to recall studying the power transistors.

Now handicapped and retired, now; BUT still remember which is why I've decided to hold off buying either a tube or transistorized amplifier.  (In no hurry)
Logged
NO9E
Member

Posts: 416




Ignore
« Reply #20 on: January 24, 2012, 12:16:04 PM »

I guess that all depends on the imagination.

Assume that we drive MRFE6VP61K25HR6 with a 5 W class A amplifier and SDR via a D/A converter. The amplifier can apply forward correction, separately for each push-pull transistor, constantly self-adjusting for any combination of power and frequency in a milisecond.  Higher efficiency would bring lower dissipation and separate feeding of each transistor would reduce even harmonics. Flattopping would be detected in miliseconds and gain readjusted, for a perfect ALC. If a change in gain is detected (e.g., sparcing), the amp could shut itself in microseconds. Instead of band-pass filters, perhaps one use switching networks as used in antenna tuners. ALl of these would mean fewer parts, less heating, automatic matching of not only antennas but also transistor outputs, and matching for high efficiency at reduced power.

Most of the complexity would be in software. May be a dream or it may be already used in a military design...
Logged
DF3KV
Member

Posts: 1




Ignore
« Reply #21 on: January 24, 2012, 01:04:34 PM »

OK, the eff on those SS amps is lousy at best.  Like 50% max. 

Modern LDMOS do well above 70%

For the same $540.00  you can still buy a chinese 8877  that will blow the doors off your SS devices.   I can't see  SS taking over any time soon. 

In Germany on VHF/UHF they have taken over a couple of years ago

http://www.beko-elektronik.de/index.php?do=03,01,01,02&lang=en

73
Peter, DJ7WW
Logged
W8JI
Member

Posts: 9296


WWW

Ignore
« Reply #22 on: January 24, 2012, 04:25:55 PM »

How does a modern LDMOS do well over 70% when in linear class AB? That would imply a voltage swing from rail to nearly zero volts, while remaining linear in transfer characteristics, even if matching was lossless.

What are the IM specs while well over 70%?
Logged
KB1GMX
Member

Posts: 780




Ignore
« Reply #23 on: January 24, 2012, 07:11:20 PM »

The modern LDmos parts are very clean, at full power we are talking better than -30DB or better.  Keep in mid they are run as Class AB1 push pull (a few hundred mA bias per device).

The closest of the MRF150 era to this is the MRF151G dual PP device.

Those new generation LDmos devices are a different class and  if you can cool them
they put out big power.

Me I find 350W from a BLF278 (improved cousin to the MRF151) enough and it seems tough enough.

I also have a single MRF150 6M amp and it sees all manor of abuses and still runs without failure and
there is NO protection save for the B+ is only 48 and there is a thermal cutout at 60C, never seen it open though.  It's used for a CW/SSB and FM service and I've never seen the heat sink over 40C.  I bought two
in case I fried the MRF150 during initial testing.  I built that three years ago and never a burp.  Still have
the unopened spare.

Oddly I have several homebrew bipolar (MRF247 and MRF492) amps that seem to be unbreakable as well.
I can't explain why I haven't fried them but they continue.  I use both for SSB and FM service.

Yet I know commercial gear that has failed and I suspect due to bias system failures and cooling issues.

Is it fragility or poor assembly?  Maybe design, too small or not enough air?


Allison

Logged
VE7RF
Member

Posts: 212




Ignore
« Reply #24 on: January 25, 2012, 01:03:02 AM »

How does a modern LDMOS do well over 70% when in linear class AB? That would imply a voltage swing from rail to nearly zero volts, while remaining linear in transfer characteristics, even if matching was lossless.

What are the IM specs while well over 70%?

##  You beat me to it..I was gonna  ask the same thing.  The best I see on the yaesu's  is 50%..when run at 200w out.   I back the power down to 100w...and eff is down to 35%.  [ class AB].  The only way  to get the eff back up..when running low power is to reduce the B+ a bit.   Problem with that is... then the IMD goes to hell. The yaesu 767GX has superb TX IMD...like -40db IMD3.  It uses a pair of MRF-422's..with 24 vdc on em... and operates  at 100w pep out.    Now my yaesu  FT-1000-D  uses the same pair of MRF-422's.... but runs at 200w pep out... and uses 30 vdc.   The imd on the 1000-D is worse at 200w  vs the 767GX..[ that uses the same MRF-422's, but is optimized for 100w].

##  to kill several birds with one stone..what's really required is more devices [ or bigger]  to be used  for a given power level.  Then the heat is spread out between more devices..and also the TX imd would improve by a huge amount.   Every SS amp I have looked at has sub standard TX imd..with imd that goes on forever.   The KPA-500 is no better.  too bad, scratch that one.  The new 800 W SS amp from Acom is not much better,so scratch it too.   The 400w yaesu FTDX-9000MP is a joke..with it's  -22db pep IMD3 on 80 + 10m.  I'd get run outa town with that kind of poor imd. Our local RI would shut us down asap.

##  what's really required is min TX IMD standards for ham gear.

Later... Jim  VE7RF
Logged
EI9JU
Member

Posts: 61




Ignore
« Reply #25 on: January 25, 2012, 06:09:00 AM »

This is the company and these are the parts.
http://www.youtube.com/watch?v=KZF-FR8b71s
Logged
W6RMK
Member

Posts: 656




Ignore
« Reply #26 on: January 25, 2012, 06:38:01 AM »

That's sort of like the NXP video with the 1200W FET that was mentioned on this forum some months ago (in that one, they shorted the connector with a screwdriver, etc.)

A few comments:
- You don't know how many of those parts the engineer blew up while designing and building that circuit. I know some folks who build high power solid state tesla coils, and they've all commented that it's not a design activity for those who are using free sample parts in small quantities.

 - You don't know how picky the design is (i.e. is it a reproducible design, or a one off).

- Pulling arcs isn't actually a very demanding test. Flashy, showy, and yes a "talking/singing spark" is way cool (which is why people do the solid state tesla coil thing)

- You can't tell what waveform is being used.  Are they doing 1.25kW peak with a low duty cycle pulsed waveform? That makes the thermal management much easier.  You can hear a whine during the demo.. is that the PRF? 

- Hmm, is he doing this at work?  And the RF safety folks at work let him do that?  Drawing arcs to a metal rod held in your hand from a 200+ Watt RF source?  Well, it's not very long in duration, so they get to average over some minutes.
Logged
W8JI
Member

Posts: 9296


WWW

Ignore
« Reply #27 on: January 25, 2012, 06:45:08 AM »

The modern LDmos parts are very clean, at full power we are talking better than -30DB or better.  Keep in mid they are run as Class AB1 push pull (a few hundred mA bias per device).

"Very clean", by almost anyone's definition,  is certainly not -30dB PEP! Acceptably clean for amateur use is about -35 dB PEP. Very clean would probably be -45 dB PEP.

Also, I doubt anyone is getting 70% efficiency in class AB.

Where is the two-tone or multiple-tone IM data published?

Quote
Those new generation LDmos devices are a different class and  if you can cool them
they put out big power.

All FET's will. A single MRF150 can make 600 watts in pulsed applications.
 
Quote
Oddly I have several homebrew bipolar (MRF247 and MRF492) amps that seem to be unbreakable as well.
I can't explain why I haven't fried them but they continue.  I use both for SSB and FM service.

One amplifier in one station does not mean unbreakable in another amplifier or another station.

73 Tom
 
Logged
M0HCN
Member

Posts: 473




Ignore
« Reply #28 on: January 25, 2012, 01:06:30 PM »

Also those tricks with dynamic bias and drain modulation mentioned upthread can only really be done in a rig not a standalone AMPLIFIER....

I should know, I have just such a thing on the bench, but mine uses old school bipolar devices (Because I had a load of BLW96 available).
A pair of FETs should sub with only minor changes and a new PA board however.

Efficiency can be good to excellent, but it does need a fair amount of DSP support to pull it off, and I would never try it except integrated into a radio (It needs far too much integration with the baseband processing to make it work well), also the drain power supply is a complicated piece of switchmode power engineering (Making the supply swing from ~12V up to 50V at envelope tracking speeds is a hard problem, as is keeping switching noise out of the RF stage).

I would reiterate that the high power sand is really not the problem, heatsinks, and capacitors are, ATC100B (Which is about the minimum I would consider at a few hundred watts for the output filter) are ~$1 each, and you will need 3 - 5 per filter, plus the toroids, with probably 6 or 7 filters to cover 160 - 6M. Moving up to something man enough for US legal limit just makes things really expensive.

Building a truly repeatable radio is hard, my prototype makes -60dB ref one tone, but coming up with suitable BITE to make the setup for that automatic is a headache, and dialling it in manually takes tweaking a lot of parameters in the software.

Dropping the drain voltage a modest amount does NOT automatically make the IMD go to pot, as long as you keep the voltage above the level needed for whatever PEP is being demanded given the drain impedance, modulating the drain voltage does change the gain, but there are ways around that as long as you do not drop it too far (LDMOS Cdg skyrockets at low drain voltage).

Regards, Dan.

Logged
KD8MJR
Member

Posts: 2499




Ignore
« Reply #29 on: January 25, 2012, 02:37:42 PM »

"One day" solid state devices will completely replace vacuum tubes in ALL applications.  The only valid question is "when" that will happen.

Like previously posted, getting rid of heat is the issue. The heat generated in a typical very small PN junction is much harder to dissipate than the same amount of heat generated inside a vacuum tube.  
Dick  AD4U

I agree and IMO one of the other stumbling blocks are Hams themselves, most of them feel much more comfortable around tubes because they grew up with this stuff and know how to fix it, also many Hams have a lot of stray RF in the shack and while this is not a problem with Tube amps but is a real problem with SS amps. IMO SS will take over completely in probably 15 years and it could take over right now, I don't see heat as being the deal breaker but I do see the adoption of the technology by older Hams as a hard hill to get over.
Logged
Pages: Prev 1 [2] 3 4 5 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!