Owen,
You have obviously have extensive knowledge and familiarity with some
equipment on this subject. I haven't studied this aspect of the radios I
have yet, but offer the following from my design experience.

There are several relatively independent considerations which have been
mentioned in this thread and I hope this clears up any confusion about which
is doing what.

My understanding of the true purpose of (what we call) ALC is to prevent
distortion in the final amplifier and the resulting "splatter" it causes,
while allowing the maximum power out. As you so correctly point out, this
originated in the linear amplifiers of old. Distortion at this point causes
increased intermodulation (mixing) of the desired signal components and the
resulting products appear out of the normal bandwidth = "splatter". There
is always SOME of this going on, but as long as it is below about -30 dBc
this is considered acceptable.
Because there is no Grid current to sense in solid state amplifiers, some
other method of detecting the onset of distorting is required. Not knowing
what _IS_ done, I can only speculate that knowing the highest level of
output power which causes only a small, allowable increase in distortion, an
output level detector can be designed which begins to provide feed back
which cuts drive when this point is reached; thus paralleling the function
of the grid current detection method of old. Because this is an
instantaneous feedback system with a relatively high gain, the feed back
voltage will exhibit a considerable rise as the device is driven closer and
closer to the undesirable point. This must be what we are looking at that
is called ALC on the front of the radio. As long as the level of this
signal is within the manufacturer's limits, the distortion is limited to the
specification and you are assured of the highest power with in-spec
distortion.

With solid state comes other considerations which are actually separate
from the distortion issue although from a circuitry standpoint they may seem
identical. Because these devices have been and are still somewhat sensitive
to mis-treatment and can be destroyed instantaneously, various methods are
available to help save the devices. However, had they been easily available
in tube days, they would have been equally applicable.

Sensing and placing a limit on collector/drain current is one method.
Some load conditions can cause excessive current resulting in device
destruction, so sensing over-current electronically and reducing drive can
prevent this.

With the advent of simple reverse power sensing circuitry, another fault
detection scheme can be used to reduce drive and protect the final power
device. Sensing SWR and reducing drive can prevent high device current and
destruction, as well.

Heat rise can also destroy the device, so sensing device temp and
reducing drive controls this also.

Because a design has one or more of these, the identity of each could be
confused by combining the feedback signal and calling it only "ALC" when in
actuality, there are two or more protection schemes present along with the
(distortion limiting) ALC.

73, Steve, K'9;D.C,I



snipping quite a bit:

"Owen Duffy" wrote in message
news
On Sun, 9 Oct 2005 20:11:09 +0000 (UTC), "Reg Edwards"
wrote:


"Owen Duffy" wrote various ways and means of automatically
controlling the power output of transmitters. Thanks very much.

Presumably control is needed to limit the power dissipated ...

Another method of PA protection is thermal protection of the PA...


In my Icom 735 transceiver the first control requirement appears to be
met by by placing a limit on the DC collector current or power input.

You are right, some radios do incorporate over-current protection ...

Whilst probably intended for PA protection, it has the great benefit
that the radio is less likely to draw excessive current while
adjusting an ATU, ...

As I described, and for example, from the IC706IIG service manual:
"The reflected wave signal appears ...the antenna is mismatched. The
...signal level is detected ...and
applied to the ALC circuit as the reference voltage." ....

Owen
--

On Tue, 11 Oct 2005 16:32:59 -0500, "Steve Nosko"
wrote:


Because a design has one or more of these, the identity of each could be
confused by combining the feedback signal and calling it only "ALC" when in
actuality, there are two or more protection schemes present along with the
(distortion limiting) ALC.

Yes Steve, once you have the gain control mechanism, it becomes the
obvious control point for various over-x closed loop control systems.
But you are right that the highest priority function is for limitation
of distortion due to exessive signal (for the current conditions), and
it is that loop that has the most onerous dynamic performance
requirements (eg fast attack) for it to perform the intended fuction
well.

Unfortunately, there is a growing common belief that unless the ALC
meter is high upscale, then the rig isn't being talked up enough (the
dumbing down of ham radio). There seems a common ignorance that most
ALC detectors have a threshold (usually the rated PEP for solid state
PAs), and that the slightest deflection of the indicator means that
some peak signal has reached that threshold, further deflection causes
greater gain reduction in the IF stages, causing the ALC to act more
and more as a compressor where overdrive of the PA is likely to occur
on transients during ALC attack time. Exceeding the red line, even for
occasional transients, is to exceed the capability of the ALC to
contain PA distortion due to overload. My thought is that if you want
audio compression, use a speech processor, not the ALC.

Mentioning speech processors. A correctly adjusted speech processor is
proably better protection against overdrive than depending on ALC
alone. The peaks are contained (clipped) and the distortion products
filtered off, before getting near the PA which does not have effective
post filtering for clipping distorion.

Owen
--

"Owen Duffy" wrote in message
...
On Tue, 11 Oct 2005 16:32:59 -0500, "Steve Nosko"
wrote:

[ clippers active...]

Owen,
Though this is not antenna talk...

Thanks for the confirmation of my half experience, half
judgment-based-on-knowledge based understanding.


Unfortunately, there is a growing common belief that unless the ALC
meter is high upscale, then the rig isn't being talked up enough (the
dumbing down of ham radio). snip... My thought is that if you want
audio compression, use a speech processor, not the ALC.

Interesting misconception. I'll remember this when I find that some
clarification is needed on ALC.
A good explanation "tool" is to say that the ALC "meter" deflection is an
indication of how much over drive you are trying to give due to too much
audio. I understand that it is an indication of the gain reduction being
applied to keep the PEP at the design max. If it is done correctly, it
should provide no compression effect at all.



Mentioning speech processors. A correctly adjusted speech processor is
proably better protection against overdrive than depending on ALC
alone. The peaks are contained (clipped) and the distortion products
filtered off, before getting near the PA which does not have effective
post filtering for clipping distorion.
Owen
--

I'd be careful in saying it this way because it appears to connect the
compression concept with the over drive concept and they are independent
concepts. The purpose of ALC is to set the proper PEP level (highest, but
not "too much") and the purpose of compression is to improve "talk power /
intelligibility" by increasing the average envelope power ( AEP ?) while NOT
changing the peak (improve peak-to-average-ratio of speech).

73, Steve, K,9.D;C'I

My thought is that if you want
audio compression, use a speech processor, not the ALC....

For sure...Another problem these days is some radios
deal with ALC indication in reverse from each other.
IE: I *think* most kenwoods will show more indication
the closer to the edge you get, and if you were to disable
the ALC. the meter would show very high peaks, or maybe
even peg. But many Icoms are reverse. The closer to the "edge"
you set them, the lower the indication. Or at least on my older
730. If you see no indication, Houston, we have a nasty problem.
When you reduce power on an icom, the ALC indication increases,
and I think pegs, if you dial down to QRP range.
One thing that strikes me as real silly, are the ones that try to get
more power by disabling the ALC function. All you gotta do is
turn a trimmer...But it makes the radio nasty as all get out, and
is silly. Not to mention the extra power amounts to little in the
real world, and then you add on the extra wear and tear on the
poor radio...Silly... I hear stories of people getting 140-150-160
watts outa lil IC -706's and I just roll my eyeballs... Boy, I bet
it sounds great...Not...MK