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] 2 3 4 5 6 ... 13 Next   Go Down
  Print  
Author Topic: Goodbye tubes.  (Read 6715 times)
AH7I
Member

Posts: 128


WWW

Ignore
« on: July 06, 2019, 07:15:18 AM »

NXP has 300W 50V with 28 dB gain for $35/each (Mouser) and $25/each in qty suitable to a small manufacturer. https://www.mouser.com/datasheet/2/302/MRF300AN-1381586.pdf
Four of these make 1200W output HF amp.
It looks very attractive for single band/amp per antenna application that is competitive with anything having a tube.


Logged
K6AER
Member

Posts: 5721




Ignore
« Reply #1 on: July 06, 2019, 08:16:36 AM »

Stated wattage of a transistor is its maximum dissipation. The transistor will not last long at maximum wattage. You will find that in order to have a good IMD level these devices generally have to be run at 70% of maximum wattage. A solid-state amplifier with enough head room for a 1500-watt output and good IMD numbers (-35dB two tone) will cost almost twice that of an equivalent tube amplifier.
Logged
DL8OV
Member

Posts: 1054




Ignore
« Reply #2 on: July 06, 2019, 09:45:01 AM »

Read the datasheet carefully, the output figure you see may be for pulse operation.

Having said that, a pair of these ran at 45V and generating 100W would be a nice clean P.A. providing you got the output transformer right.

Peter DL8OV
Logged
AC2RY
Member

Posts: 739




Ignore
« Reply #3 on: July 06, 2019, 11:48:05 AM »

Read the datasheet carefully, the output figure you see may be for pulse operation.

Having said that, a pair of these ran at 45V and generating 100W would be a nice clean P.A. providing you got the output transformer right.

Peter DL8OV

This device won't be very efficient in real life operation. Miller capacitance shoots up below 20 volts D-S. Thus for wide band stability, you need to keep drain voltage above that. Which means significant loss of efficiency. Just compare it with BLF188XR in which capacitance jumps up below 5 volts D-S, thus allowing for more useful drain voltage range. When you design amplifier for a single frequency you can compensate  that capacity, but this is not possible when you need to cover the whole range between 1.8 and 55 MHz.

« Last Edit: July 06, 2019, 11:52:45 AM by AC2RY » Logged
N8FVJ
Member

Posts: 883




Ignore
« Reply #4 on: July 07, 2019, 05:32:30 AM »

IMO tubes have been obsolete for years with the exception of high power tube amplifiers. Tubes easily tune into a 2 to 1 SWR whereas solid-state cannot tolerate more than 1.5 to 1 SWR. I use the MyAntennas EFHW8010 antenna that has a 2 to 1 SWR on 75 meters and about 1.5 to 1 on 40 to 10 meters. You need an amp on 75 meters at night anyways. No expensive high power antenna tuner required.

Now, I read some solid-state amps have a built-in tuner, but very expensive. Only new tube left that lasts 10 years or longer is the Chinese 3-500Z. You can buy NOS Cetron 572B that are known to last 20+ years, but expensive and rare.
Logged
AC2RY
Member

Posts: 739




Ignore
« Reply #5 on: July 07, 2019, 08:29:38 AM »


Now, I read some solid-state amps have a built-in tuner, but very expensive.

Everything is relative. Good mid-range transceiver is around $3000. Good 500-600W solid state amplifier WITH tuner can be purchased for similar amount of money.

Things get more expensive when you go over 1000W, but then everything in your transmission chain allowing effective use of that power becomes expensive. Just like high-end transceivers go up in price to $10K and above.
Logged
K6AER
Member

Posts: 5721




Ignore
« Reply #6 on: July 07, 2019, 10:09:51 AM »

IMO tubes have been obsolete for years with the exception of high power tube amplifiers. Tubes easily tune into a 2 to 1 SWR whereas solid-state cannot tolerate more than 1.5 to 1 SWR. I use the MyAntennas EFHW8010 antenna that has a 2 to 1 SWR on 75 meters and about 1.5 to 1 on 40 to 10 meters. You need an amp on 75 meters at night anyways. No expensive high power antenna tuner required.

Now, I read some solid-state amps have a built-in tuner, but very expensive. Only new tube left that lasts 10 years or longer is the Chinese 3-500Z. You can buy NOS Cetron 572B that are known to last 20+ years, but expensive and rare.

The Chinese FU728F Tetrode is rated for 1800 watts continuously and only cost about $400. Gain is about 15-16 dB depending on screen current. I have seen these tubes put out easily 2.3 KW in SSB mode with 60 watts drive. This tube was modeled after the 4CX1500B.
Logged
WB8VLC
Member

Posts: 635




Ignore
« Reply #7 on: July 08, 2019, 01:23:31 PM »

The  Motorola/Freescale/NXP group (sorry to the NXP guys but I'm ex Moto from the 1st days of LDMOS  development so it's still Moto to me)  down in Tempe already have these operating I believe across HF/6 meters and they were heard during field day  on 6 meters using a P-P MRF300 A/B amp at around 600 watts on FT8.

Also the NXP guys discovered that with the MRF300's and the newer LDMOS 65 volt parts that running them off of 50  or 65 volts and not backed off in Po actually resulted in the best overall efficiency which wasn't the case with the P-P MRFE6VP5300 amp that I made. nor with a similar single BLF184 amp.

I'm using 2 X $69 dollar 15 amp mean-well 48 volt  supplies (because I'm cheap) on my Mouser bought MRF300A/b P-P parts from 14 to 54 MHz with good results,  14 to 54 Meg are the only bands which I presently have antennas for at this time.

So far the 300A/B's are more rugged than the 2x MRFE6VP5300 P-P amp (500 watts out) that I made 3 years ago similar to Lionel Montgin's fine design of several years ago and the 300 A/B is on par with a BLF184 single amp that I made although the BLF184 is more suitable for ~650 watts whereas the 300A/B P-P amp is best around 575 watts max in my case.

So far using mine under extended FT8 mode on 20 somewhat  and many hours  on 10 and 6 meters CW/SSB /FT8 and even 50.3 and 29.6 FM, even for one 3 hour US to JA opening on 6 meters FT8 the 300 A/B P-P amp has worked just fine.

As mentioned I also use it on 50.3 FM  and 29.6 FM and my nearest neighbor W7EW, 1 1/2 miles from me, even with his massive array of antennas pointed at me, hasn't had any issues when we operate within 15 KHz of each other  he on 50.313 and me on 50.3 and even when we are operating different sequences on the 50.313 so he hears me there haven't been any issues.

The NXP guys in Az  also have a design challenge running thru I believe Arrow for a home brew challenge for builders using the MRF100 A/B 100 watt and the MRF300 A/B 300 watt devices.

There are already a couple of russian guys who are working on commercial amps using the 300 A/b's and one may already have one on the air using the MRF300's  on HF/6 meters.

As long as 50 Vdd and a 3 db over drive are not exceeded they seem to be very rugged parts and at around 40 dollars even if I kill one the pain is minimal compared to the MRFE6VP5300 P-p and a BLF184 which are considerably more expensive and I have already killed some.

My success at not killing anymore of my LDMOS devices may because I'm now using an input overdrive Pin limiter circuit similar to ON7EQ's fine design to keep from hitting above a 1 db or whatever I set it at over-drive level and so far no dead 300 A/B's.

In the very early days of LDMOS during the development of the  2nd gen parts in the late 1990's at Moto in Phoenix I was know for killing more LDMOS devices than cockroaches in Phoenix's Arcadia neighborhood where our plant was located.

On the other end of the extreme my company just bought 2  new, 1-2 GHz/4 KW Solid State  amps for use in  reverb chamber and these amps also are based on several combined NXP parts and presently being used in our EMC lab.

Although being a commercial lab grade EMC amp they are considerably more expensive than my home brew amp, still these newer LDMOS devices are very nice.
« Last Edit: July 08, 2019, 01:34:45 PM by WB8VLC » Logged
AH7I
Member

Posts: 128


WWW

Ignore
« Reply #8 on: July 09, 2019, 07:21:58 AM »


...

So far the 300A/B's are more rugged than the 2x MRFE6VP5300 P-P amp (500 watts out) that I made 3 years ago similar to Lionel Montgin's fine design of several years ago and the 300 A/B is on par with a BLF184 single amp that I made although the BLF184 is more suitable for ~650 watts whereas the 300A/B P-P amp is best around 575 watts max in my case.

...


Thank you for commenting and references. I am interested in constructing a pair of amplifiers to drive phased 40/80 verticals with the phasing done on the input side. I was looking at a pair of SD2933 per each but these Motorola parts are considerably less expensive.

73, -Bob ah7i
Logged
WB8VLC
Member

Posts: 635




Ignore
« Reply #9 on: July 09, 2019, 11:53:26 AM »

Mine is far from complete right now from what I would eventually like to have which would be an auto switched LPF/swr detector design that one day is controlled by my main HF/6meter rig.

 Right now I have  individual coax cabled in low pass filters for the main 3 bands that I use only because I didn't have the extra time nor the extra money to spend to roll a multi-relay or multi-Pin diode switched low pass filter board which would be in the hundreds of dollars.

Possibly, some day I'll sit at the PC and roll a relay or Pin switched LPF board in PCB-Express but only if I have the free time and spare money to spend other wise it will be a hillbilly design using cabled in one at a time LPfilter boards for the band that's being used at any particular time. 

DC switching for Vdd and Vgg were cheap but very good designed W6PQL boards along with a W6PQL 4 event sequencer that does input/output switching and Vdd/Vgg switching at the proper times.

The various boards available from W6PQL let me get the amp on the air quickly.

Also everything is pretty much open frame design on a large freescale provided copper heat-spreader/ bare board with attached 5 inch long beautifully milled aluminum finned heatsink, this is a real Freescale provided A1 heatsink design that lets me operate without a fan.

In typical cheap-ham-dot-com fashion, which I rigorously adhere to, I used anything from my junk box or anything that I had donated to keep cost down.

I have everything crammed in an old Palomar CB amp case that was donated by a ham friend so overall looks-wise it's not very pretty but at least it's clean RF wise and lets me operate with considerably more than my exciter power.



Logged
OZ8AGB
Member

Posts: 566




Ignore
« Reply #10 on: July 11, 2019, 01:31:52 AM »

Goodbye tubes??

Oh.. I just bought an ACOM 1010 tube amp...
Logged
M0HCN
Member

Posts: 566




Ignore
« Reply #11 on: July 11, 2019, 07:16:22 AM »

As ever be a little careful with front of datasheet ratings, especially with TO series plastic packages as getting heat out can be a pain.

I would consider a pair of those to make a decent 300W or so stage while allowing the heatsink to run at a reasonable temperature, 600W in AB from those packages would have me really nervous (It could certainly be done, but solder mounting and some very careful figuring of Tj would I think be required to make a reliable design).

TBH, at this point I take the view that the cost of the power sand is low enough compared to the incidentals that running it balls out is generally a poor design decision (Really it always was), much better to over spec the power devices and save on the heatsinks, most of your cost is not in the power transistors anyway (LPF bank, power supplies, case, cooling, matching will generally cost more then the power devices). It used to be the case that the power semiconductors were the elephant in the room, but I would argue that this has not been true since the modern LDMOS stuff hit the streets.

 
73 Dan.
Logged
G3RZP
Member

Posts: 1265




Ignore
« Reply #12 on: July 11, 2019, 02:33:39 PM »

So just how good are the power LDMOS in terms of high order IMD? The last generation of tube PA gear had relatively little 7th order IMD and negligible 9th order and higher....

A friend of mine who works on designing microwave PAs with LDMOS thinks they are pretty bad from his experience....
Logged
AH7I
Member

Posts: 128


WWW

Ignore
« Reply #13 on: July 12, 2019, 04:06:55 AM »

So just how good are the power LDMOS in terms of high order IMD? The last generation of tube PA gear had relatively little 7th order IMD and negligible 9th order and higher....

A friend of mine who works on designing microwave PAs with LDMOS thinks they are pretty bad from his experience....


HF IMD is good enough to meet code. My application, CW, is much less a concern. LID is by far(orders of magnitude) the greatest contributor to crap on HF.

73, -Bob ah7i/w4
Logged
AH7I
Member

Posts: 128


WWW

Ignore
« Reply #14 on: July 12, 2019, 04:15:29 AM »

As ever be a little careful with front of datasheet ratings, especially with TO series plastic packages as getting heat out can be a pain.

I would consider a pair of those to make a decent 300W or so stage while allowing the heatsink to run at a reasonable temperature, 600W in AB from those packages would have me really nervous (It could certainly be done, but solder mounting and some very careful figuring of Tj would I think be required to make a reliable design).

TBH, at this point I take the view that the cost of the power sand is low enough compared to the incidentals that running it balls out is generally a poor design decision (Really it always was), much better to over spec the power devices and save on the heatsinks, most of your cost is not in the power transistors anyway (LPF bank, power supplies, case, cooling, matching will generally cost more then the power devices). It used to be the case that the power semiconductors were the elephant in the room, but I would argue that this has not been true since the modern LDMOS stuff hit the streets.

 
73 Dan.


Thanks. My application is one or two bands and morse. I expect a pair of 300W/each continuous rated devices running at morse duty cycle to have plenty of thermal overhead when reasonable consideration has been given to heat transfer.



Logged
Pages: [1] 2 3 4 5 6 ... 13 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!