Merv touched on a very important point. These SS devices are built for a
specific purpose and industry market that will purchase enough to cover
the high cost of development and show a profit. Few were designed for
"linear" operation as we define it. OK, so with unprocessed SSB the
duty cycle is usually figured at 20%, but those peaks may be much wider
than those pulses.
The only way we can know for sure is when someone who can afford enough
of those devices to test them running in class AB1.
The greatest problem with these devices is HEAT and it takes care and
knowledge to transfer a thousand watts of heat from that little device
through its tiny footprint with a high enough efficiency to keep the
junctions or gate temperatures low enough to prevent thermal runaway, or
a drastically shortened life span.
I'm becoming attached to the idea that the word "linear" should be
banned from the advertising Ham Amps, unless they (the seller) can show
proof that the amp is truly linear, neglecting HOW Linear an amp must be
before we can call it linear. How often has an add for a 2-meter Amp
stated, "160 Watt 2-meter FM linear amp? A linear class C amp is quite
a trick.
Often, or more likely with "most" new devices, the yields are far from
being a viably financial product. Even after a temporary shutdown of
production, yields will be down on the re-startup. This is generally
true from creating the Silicon from sand, the process for creating
HSiCl3 (Trichlorosilane from which the poly crystal Silicon is made),
creating the polished wafers and all the steps for creating the finished
devices. I'll skip describing those steps as they could more than double
the size of this post. Google is your friend if you really want to
know! I've also skipped the hazards associated with these steps.
More and more new devices are developed for what we see as exotic nodes.
We re-purpose these multi kilowatt devices into (hopefully) linear devices.
Well the literature did say 100% duty cycle, BUT that was with some
exotic wave form (just call it pulse for simplicity), or biased into
DEEP class C. If the efficiency is much over 60%, like 73% to 75% your
AB1 or 2 is going to reach thermal overload FAR SOONER than the power
ratings in the data sheets.
HOWEVER high power "linear" devices are available and still being
developed. A pair of 1200 W transistors is 2400W. Two pallets with a
pair of these devices run at around half their ratings of 4800W "can"
give good life, efficiency, and low IM products. Add "dynamic
pre-distortion" and the resulting signal can beat the quality of the
best tube rigs. The key is the pallets and combiners need to be designed
for the power being run
Development of these devices is more mature than earlier devices and
accurate modeling and speed and more accurately design these new devices.
I would "guess" that instead of linear, those devices were run in deep
class C, while the modulator was run at half that power PP class B.
80 to 90 percent efficient tells me they were either using one of the
more exotic forms of amplification, operating in deep class C, or a
combination. They were definitely not running class AB, or B in PP.
90% tells me they only had to dissipate around 5,560 watts of heat for
50,000 W output. Class AB1 or 2 (60%) would require getting rid of at
least 33,333 Watts of heat and circuits for these frequencies are much
simpler than a 160 through 10 meter band switched amp.
73, Roger (K8RI)
On 6/6/2017 Tuesday 6:42 PM, Merv Schweigert via Amps wrote:
Cant quote you the device numbers, but have seen KW plus MOSFETS
around since the
middle to late 1990s. When I purchased this AM station property
there was a Harris
50KW chugging along, it used 50 MOSFET modules, 50 for RF and 50 for
modulation.
plug in boards and would auto adjust if a module went down, it kept
running fine at
lower power.
Commercial use has existed for some time. perhaps price and class of
service was not
conducive to amateur use, but they have been around.
By the way the Harris 50KW ran 85 to 90 percent efficient, thats
power line in to power out.
73 Merv K9FD/KH6
OK, name some 1400W solid state devices and they're first date of
manufacture. Seems to me they didn't exist until a few years ago.
You don't like solid state....we get it.
People didn't like cars and tractors much when they drove horses out
to pasture either.
On 6/5/2017 9:54 PM, w8hw@comcast.net wrote:
Larry,
You made and proved my point for me. The transistor you mention is
NOT an 1400 watt class (as I specified in the email) in fact your
transistor is not even close, it is closer to 140 watt out (if
memory serves) That makes your statement off by a factor of 10. You
asked me to be careful, I was very carful to specify 1400 watt class
in my email.
Thank you , Bruce, W8HW
====================
-----Original Message----- From: Larry Benko
Sent: Monday, June 05, 2017 20:40
To: amps@contesting.com
Subject: Re: [Amps] LDMOS availability
Bruce,
The MRF-150 came out in 1983/84 and Helge Granberg produced the
Motorola
app not EB-104 using the XRF-150 in the same year. It is still readily
available. Be careful with your FACTS!
Larry, W0QE
On 6/5/2017 5:12 PM, w8hw@comcast.net wrote:
Actually you do NOT have your crystal ball to predict the future so
your statement is NOT based in fact, only an opinion. I prefer to
go with facts, so hear we go.
FACT... Because we can not see into the future, we only have
history to look at.
FACT... History has shown us that RF power devices have NOT
survived 15 years of manufacture very well... IN FACT... I
challenge you to show me the facts. Show me a solid state RF
transistor in that power class (RF 1400 watts out) that has been in
production for 15 years and is still in production today. I am open
minded and would love to see your report.
FACT... I noticed that you did not mention if they agree to the
$100k Guarantee. Why is that important, see below.
FACT... If they are sure that they will keep production going until
the year 2032 (15 years), then that guarantee should BE NO PROBLEM
for them. Talk is cheep
FACT... If you feel that is an unfair request, then that tells us
that (Deep down) you do not believe them either. Please do not tell
me that you DO have a crystal ball.
FACT... Many high power tubes (8877 and 3-500) do have a rich 40+
year history with no end in sight.
FACT... I have purchased the BLF188 because I CAN afford both the
cost and the risk. However people who can not afford the cost and
risk may wish to take this into consideration.
FACT... I have been in the RF repair industry since 1971 and have
seen and faced this issue happen over and over again. How long have
you been in it???
73, Bruce, W8HW
============================
-----Original Message----- From: Per Molund
Sent: Monday, June 05, 2017 14:55
To: w8hw@comcast.net ; amps@contesting.com ; Tom Thompson
Subject: Re: [Amps] LDMOS availability
Actually they have put it in writing, on the data sheet (although
prelimenary at the moment) they state 'Included in NXP product
longevity program with assured supply for a minimum of 15 years after
launch' which I see as a quite firm commitment. Or do You have an
example that NXP has not fullfilled their 'longevity program'
commitment for some components?
73 de LA9XKA - Per
At 19:08 05.06.2017, w8hw@comcast.net wrote:
Talk is cheep. Ask them if they will put that in writing? Also ask
if they will put a penalty clause stating they will give you a
$100k payment should they fail the 15 year test. After all if it
was true, they should have no problems with that. On the other
hand if they baulk , that tells you that they are just talk with
no truth. Additionally, Please be honest... If you are reluctant
to ask them for that guarantee, that shows that you KNOW they are
just talk. Someone or Anyone let me know if anyone gets that
guarantee.
Remember company plans do change with new management or when they
get a bad P&L. This is especially true if they are listed on any
exchange such as NASDAQ. Like the dip in early 2016 (NXPI at $65)
and now back on a good clime this year ($109.00) now. companies
tend to make strong claims on the peek and pull claims back claims
on the stock dip. Often they have no choice. The device market
history is the only good indicator. Solid State RF power devices
still have a very poor history. Translation... the market speaks
with authority, companies just talk.
I have a solid state amp (dual BLF188) homebrew amp, and have two
(2) tube amps one of which is AL-1500 with 8877 tube. I love them
all, but understand the risk of SS replacement as I have been in
broadcasting or the RF repair business since 1971 and a Ham since
1962. I have lived through everything that I have said above.
73, Bruce, W8HW
============================
-----Original Message----- From: Tom Thompson
Sent: Monday, June 05, 2017 11:18
To: amps@contesting.com
Subject: [Amps] LDMOS availability
I recently took part in a webinar on the new, 1800 watt, 65 volt,
LDMOS
transistors made by NXP. It is the MRFX1K80, and NXP plans to
produce
it for at least 15 years.
Tom W0IVJ
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
---
This email has been checked for viruses by AVG.
http://www.avg.com
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|