This has been an interesting discussion, but still not sure which
might be the best way to insure a long and uneventful life for

MOSFET power amplifiers.? Several ideas have been presented
by myself and other individuals.? All seem to have been rejected
by others for various reasons.? Makes me wonder if rejections
are based on reality or just a fear of using any switching device
as an RF amplifier.?

The alternative of using real RF transistors or RF MOSFETs seems
to defeat the objective of designing and building low cost transceiver
from components that can be readily obtained at relatively low cost.

Are the IRF510 failures because they are not suitable for the way

we are attempting to use them, or because we are not being careful

enough in how we use them?? Are IRF510 failures a real problem,

or maybe being exaggerated?? Is there an acceptable way to make

switching MOSFETs more robust in RF applications.

Given the relatively high cost of RF rated devices it may be time to

revisit something like Farhan's (Just a Bunch Of Transistors)

design to see if that can be scaled up to the 5 to 10 watt range and

cover up to 30 MHz.

Arv? K7HKL
_._

On Mon, Jan 29, 2018 at 4:48 PM, K9HZ <bill@...> wrote:

ALC.? What a ghost from the past.? I’m not aware of anyone who designs RF amplifiers today that thinks ALC is a good idea today specifically it generates nasty products in the PA when it cuts in and makes a linear amplifier…nonlinear.? People used it years ago so that they didn’t over drive (read that blow the grids out of) their tube amplifiers.? Today we use more sophisticated algorithms, head room in over design, ?and compression to avoid the issue.?

?

Of course current limited power supplies are not linear RF amplifiers.

?

I think you should present your design, build it, and try it on the air… and have a look through a spectrum analyzer.? Maybe all of the theory I had in graduate school is bunk.

?

?

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

?

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

?

Owner – Operator

Villa Grand Piton – J68HZ

Soufriere, St. Lucia W.I.

Rent it:

Like us on Facebook!

?

?

email:? bill@...

?

From: [email protected] [mailto:[email protected]] On Behalf Of Arv Evans
Sent: Monday, January 29, 2018 12:49 PM
To: [email protected]
Subject: [SPAM] Re: [BITX20] IRF510 amplifier failures

?

Maybe part of the point has been missed at several levels.? The proposed current

limiter is not much different from current limiters in many commercial power

supplies.? It is not active until some current value has been reached.? Up to then

there is no current limiting or "modulation" or "compression".? Yes, the in-series

current sensing resistance does allow an output voltage drop until voltage across

this resistance reaches about 0.65 volts to forward bias the sensing transistor.

If your power supply output is not stiff enough to hold a steady voltage this could

add to the sensing resistance and cause voltage droop and in really bad cases
compression of RF output.?

Question...is RF compression all that bad?? It is what most transmit ALC circuits

do.? It would also be possible to make the proposed circuit driven by reflected

power instead of current.? Then we could call it a transmit ALC to protect the

IRF510(s).?

?

It would be fairly simple to add a latching function to the limiter design so that it

totally shuts down output when Imax has been reached.? This is possible but would

then require a RESET button or power cycling to restore power to the RF PA stage.

It is also possible to change the proposed design to put the current sensing ahead

of the control MOSFET and make that MOSFET into a voltage control circuit.? This

was my initial approach but the component count became just too high to be an

internal modification for transceivers.? It would be more suitable for converting

an unregulated power supply to a regulated one.?

?

For those who are using modified PC power supplies it is possible to change the

current sensing so that the switching regulator would shut down if Imax of 2 to 3
amperes is exceeded.? This approach would require power cycling to restore
operation.?

Problem with using fuses is that they have a short but measurable heating period
before they go open.? When they do open they spark internally for another short
period.? Both the delay and the spark are potential problems for circuit components.

Poly-fuse? or resettable fuse devices are intended to address a different issue.?

Some are equivalent of FAST-BLOW fuses, and some are just conventional

resettable circuit breakers.

Suggestion here is that "if you can come up with something else that is adequate,

please submit a circuit drawing and explanation of how it works"? We could all

benefit from that.

Arv? K7HKL
_._

?

On Fri, Jan 26, 2018 at 11:27 AM, Tim Gorman <tgorman2@...> wrote:

A current "limiter" which modulates the current, i.e. limiting the
current no matter the drive level, is like applying an ALC to the
exciter. Improper design can result in compression of the signal which
can then generate spurious responses in the output.

If the current limiter is of a "drop-out" style which requires
resetting then this won't happen, it will just shut down the exciter.
In that case, however, it doesn't provide much efficiency over a plain
old fuse!

tim ab0wr

On Fri, 26 Jan 2018 05:32:08 -0800
"n5ib_2" <n5ib@...> wrote:

> One other possibility to consider to protect against overcurrent
> situations is to use a current limiting power supply, or add a
> current limiter to an existing power supply or battery.
>
> The current limiter circuit whose link is at this site:
> <>
> will go up to 2.2 amps, and includes a low-dropout voltage regulator
> that also provides reverse polarity protection.
>
> There is a version, in a different package, of that same regulator
> chip that has an added "tracking" feature - being able to parallel
> multiple chips to increase the available current. Details are on the
> datasheet in the document.
>
> Jim, N5IB
>
>
>

?