[Amps] FCC Denies Expert Linears' Request for Waiver of 15 dB Rule

[Manfred Mornhinweg]

Jim,

> You are confusing speech processing, which is generally done at base 
> band (that is, on the audio signal before it is applied as modulation),
> with ALC, which is control feedback between an RF amplifier and its 
> driver.

No, I'm not confusing them. Maybe I wasn't clear enough, by using the 
term "ALC" in its generic meaning (automatic level control), rather than 
specifically over the RF stages. In this sense, speech processing is a 
combination of ALC (often with two different time constants) and filtering.

To avoid further misunderstanding in this regard, from here on I will 
use "ALC" in the sense you use it (and radio manufacturers too), as an 
automatic gain control loop sensing the RF output and acting typically 
on one or a few low-level TX IF stages; and I will treat all gain 
control loops at AF without using the "ALC" acronym.

 > Most modern transceivers include speech processing. In the pro
> world, we use both peak limiting and compression. Peak limiting being a 
> short time constant that simply reduces gain on speech peaks, and 
> compression being more of a dynamic gain-riding. Good signal processing 
> can sound very good with up to 10 dB of gain reduction, and some systems 
> are good for more than that.

The problem with that is that 10dB of control range is far too small to 
accommodate the variations in the audio level coming from the 
microphone, as the operator moves closer or farther away, and speaks up 
or speaks softly. And to maintain the 10dB compression you first need to 
have a stable audio signal. So, in order to achieve that 10dB 
compression, you need to place this compressor after an automatic gain 
control system, that has a larger control range, and has such a long 
decay time (1 second or longer) that it doesn't cause significant 
distortion. Good speech processors usually do this.

After such an audio processor, indeed you don't strictly need ALC, as 
long as the operator always sets up the TX gain in a correct way, so 
that all amplifier stages are kept out of saturation. But with any band 
change this gain setting will be different. Often it will also change 
with frequency changes inside the same band, and what's worse, the gain 
of most amplifiers changes with temperature, so the operator will have 
to watch the output and readjust the TX gain rather frequently, to stay 
at the optimum output level. That's quite inconvenient, and so we use 
ALC to perform that task automatically.

The manufacturers build ALC into the transmitters, as a 
non-user-defeatable feature, because they have good reason to suspect 
that most hams will not properly set the TX gain by hand all the time. 
Even more so in case of radio operators in other services, who don't 
have any technical knowledge at all!

> Audio processing done entirely at baseband creates artifacts at 
> baseband, but those baseband components won't get past the TX passband 
> filter.

Yes. Very true. But the same is true for RF speech processing, as long 
as it's done before that filter!

> W4TV has noted, however, that some rigs, notably Yaesu and ICOM, 
> do part of their processing at RF, and can splatter pretty badly.

If they splatter badly, it's because of some other reason. RF speech 
processing, done before a good filter, cannot create more splatter than 
audio processing. A good RF speech processing scheme needs a first 
sideband filter, then the clipping and compressing, and then a second 
sideband filter to remove the out-of-band artifacts. The advantage of RF 
speech processing, relative to audio speech processing, is that fewer 
artifacts fall inside the passband. So, RF speech processing should 
result in a cleaner signal, having less distortion within the passband, 
and no more crud outside the passband than AF speech processing causes.

Now if the radios you mean happen to use low quality second sideband 
filters, with slopes that aren't steep enough, and with poor stopband 
rejection, then of course there will be more splatter. But that's a 
problem of cheap implementation, not of the principle.

With current DSP technology pretty sophisticated and clean speech 
processing can be implemented, at very low cost. Good speech processors 
in DSP often shift the audio signal to some low RF (which is the same as 
passing it through a balanced modulator and sideband filter in the 
analog world) and then apply the compression and limiting, precisely to 
move most of the artifacts out of the passband and then filter them away.

>> ALC doesn't have to cause splatter and key clicks. It's _bad_ ALC, or 
>> improperly used ALC, that does. Typically setting the mic gain far too 
>> high and having the ALC throttle back the gain by 30dB or so, and 
>> doing this on a poorly designed radio that has an ALC with a very slow 
>> attack time.
> 
> Most amplifier manufacturers disagree with you. As long ago as 1980, Ten 
> Tec's manual for their Titan amp advised against using ALC, and the 
> manual for most power amps includes that advice. Perhaps it's the need 
> for that careful matching of time constants that forms part of the basis 
> for their advice.

Yes, I think that's the reason. Unfortunately there is no established 
standard for the ALC interface between transceivers and power 
amplifiers, except that ALC voltage is usually negative relative to 
ground. For the ALC loop to work correctly, the entire loop needs to be 
optimized. If a radio takes ALC input and then processed it with some 
frequency/gain/delay curve, and the amplifier also has such a circuit 
between its RF output sensor and its ALC output, we are in trouble.

Since most of the gain is inside the radio, it would be logical for an 
amplifier to provide a fast, linear sample of the amplitude at its 
active devices to the ALC connection, without any further processing. 
The radio should do the rest, but that woud require the operator to 
configure an ALC threshold in the radio. The other way around is that 
this threshold is set in the amplifier, and then radios have to provide 
a linear gain reduction according to ALC voltage. The problem here is 
how fast a radio can react, which depends on its design.

"Fast" means that its attack speed must be fast enough to react well to 
the whole significant spectrum of an SSB envelope, and that extends up 
to many times the highest audio frequency. A bandwidth of 50kHz or so is 
probably adequate for an ALC interface. Decay time instead has to be 
slow enough to avoid significant distortion.

If the amplifier has a peak detector with a slow decay time feeding the 
ALC ouput, and the radio's ALC input has its own slow decay time, then 
indeed it's better not using that signal, as the loop might easily 
become unstable!  Again, this is a problem of improper implementation, 
not of the principle.

I'm just checking the schematics of some amps. For example, the ALC 
output of the Ameritron AL-811 is linear, has a fast attack, and a decay 
time constant of 1ms, which is pretty slow. It's taken from the RF 
output, so that it doesn't really reflect whether or not the tubes are 
being driven into clipping - it depends on exactly how the amplifier was 
tuned! So I would say that this ALC output is pretty useless, as it 
gives no information about saturation and clipping.

The Drake L4B provides an ALC signal derived from its grounded-grid 
tubes' cathodes. The signal provided is non-linear, having a threshold 
at which it starts. It has a fast attack time, but its decay time 
depends totally on the input impedance of the radio's ALC input, varying 
from 0.3ms to infinite! Such a response is usable only with radios 
designed specifically to use this signal.

The Heathkit SB200 uses a similar setup, with a threshold, but being 
grid-driven it takes the sample from the tubes' grids. Its ALC decay 
time constant also depends totally on the radio's ALC input impedance, 
and starts from a shorter value than the L4B. It has two-stage low pass 
filtering, which adds phase delay of fast signals. So its usability 
again depends on the radio used to drive it.

The National NCL-2000 has a cleverly thought out system, that's derived 
from the grid bias and basically starts delivering an output when the 
grids are being driven a certain amount into conduction. So it does have 
a threshold like the two above. Its attack is fast, it's decay is 
totally dependent on the radio's ALC input impedance - it might vary 
from zero to infinity! So, again, this isn't usable with every radio.

The Yaesu FL-2100 uses basically the same setup as the SB-200, just with 
single stage low pass filtering and with a time constant varying from 
zero to infinite depending on the radio's input impedance.

Those are the well-known amps for which I happen to have the schematics 
on file, and clearly none of it has a really well implemented ALC 
output. And the style implemented by Ameritron is totally different to 
that of the other models cited here: The Ameritron has linear ALC 
output, while the others have an internal threshold. This means that 
radios that are ALC-compatible with the AL-811 will not be 
ALC-compatible with the others, and vice-versa!

No wonder that hams find it hard to use that sort of ALC. But still, the 
concept of overall ALC from the final amp's active devices to the 
low-level RF stages is a good one - if properly implemented. We need a 
standard in that regard, so that manufacturers of transceivers and 
amplifiers can provide ALC inputs and outputs that are truly compatible.

 > Most amp manufacturers also provide ALC output ONLY so
> that its omission won't be a buying obstacle for those customers who 
> think they need it.

When I started writing this post, I intended to question that remark - 
but now, after having looked closely at all those ALC implementations, I 
see your point. Let's be kind and think that amplifier manufacturers 
provide ALC outputs tailored to specific radios. Obviously those 
manufacturers that make the amps as accessories for their own radios 
will tailor them to these radios, but I wonder what the amplifier-only 
manufacturers have in mind, when defining their ALC outputs... Maybe 
some specific, widely used radio? Or the radio the company's boss 
happens to use?

Maybe I should investigate the ALC input behaviour of a few common HF 
radios... Maybe in a future post...

> The pot that Vic was talking about sets the MAX output level.

Yes, and that was the one I meant when I wrote that many hams around 
here tweak that pot to get 150W from their 100W radios, and don't even 
start to understand that this causes intense splatter and is a nuisance 
to everyone else on the band.

> There's a 
> front panel control to set the operating power.  I set it for about 50W 
> of drive for my Titan.

Actually I do the same with my TS-450SAT and my NCL-2000. This one 
requires around 30W drive for full output. The TS-450 has a 
slow-attacking ALC, so it has severe power overshoot at the attacks. So 
this isn't a really good solution, but I have no better one, short of 
modifying the radio. I just try to avoid excessive splattering by 
keeping the mic gain low enough to just start activating the radio's ALC 
on voice peaks.

>> And this adjustment simply sets the trigger level for ALC. Nothing 
>> else.  Let's face it: SSB transmitters control the output power by 
>> means of ALC. It's the best method found to date, as far as I know, at 
>> least for voice transmission. Instead in digimodes, including CW, it's 
>> better to adjust the drive to stay just below the ALC activation 
>> level. With voice you can't really do this, because the voice level 
>> changes too much.
> 
> I completely disagree with you, Manfred.  Audio processing has been 
> widely used in broadcasting for at least 60 years. The AM station where 
> I worked as a student was using it in the early '60s. That is done 
> entirely at AF, NOT at RF.

But there you are talking about a very different animal: AM 
broadcasting! In AM you cannot use ALC to control the peak output of a 
linear amplifier string, because that would modulate the carrier level 
too. Instead the normal approach is to drive the final amplifier into 
saturation, and use automatic gain control for the audio signal so as to 
keep the modulation at 100%, or whatever level the station uses. Indeed 
this has to be done at AF, as there is no RF stage in an AM transmitter 
where you could control the modulation level separately from the carrier 
level! In SSB, instead, it's just as valid to apply automatic gain 
control to audio stages, as it is doing so to IF or RF stages. Except 
for possible dynamic range limitations in mixers and amplifier stages, 
the end result will be the same. Note that here I'm referring to slowly 
changing automatic gain adjustment, no faster than the syllabic rate. 
Any faster changes of gain must be done before the last (or only) narrow 
filter, to keep distortion products confined to the passband.

And broadcasting is very different from ham operation, too. In 
broadcasting, you set up the transmitter on one frequency, then usually 
run it there, 24/7, for years. Or at least for hours, in shortwave 
broadcasting. You have no frequency changes, and no thermal drifts from 
circuits heating up and cooling down. In ham radio instead you change 
frequency often, you also might change bands every now and then, and 
your transmitter stages are all the time changing temperature, due to 
the RX/TX cycling. So there is far more need for ALC than in broadcasting.

 > VE7RF does extensive audio processing in his station.

Audio processing in ham stations, using external consoles with 
compressors, AGC, equalizers, etc, seems to be all the rage at present. 
Some hams indeed produce excellent transmission in that way. Others not 
so much. I keep hearing hams with lousy signal quality, bragging about 
their studio mikes and all the audio equipment they are running to 
process their audio signal.  Some even add cathedral-style echo effects, 
like those CB operators of 30-40 years ago! :-)

The funny thing is that so many hams use this sophisticated audio 
processing on HF SSB, where HiFi audio makes little sense, due to all 
the usual QRM and QRN and the intrinsic limitations of SSB. It would be 
far more logical to strive for excellent audio quality on VHF and UHF in 
FM, but that's something no ham in my area does.

> I've adjusted my K3 to roll off everything below 500 Hz and 
> provide about 10 dB of compression on audio peaks.

That's the same I did in my PQD-5 QRP transceiver, which I built for 
backpacking and mountaineering when I was 21 years old. The compression 
was rather crude, being implemented in a saturating balanced modulator! 
This combination of low frequency roll-off and compression provides 
about the best intelligibility per power, and that's what QRP is all 
about. And it's valid too at QRO, when conditions are rough.

>> It's not an error, Jim! While I agree with you on the fact that many 
>> radios have rather high phase noise and poor ALC by design, in the 
>> case I described the phase noise was far higher than normal even for a 
>> rather bad radio, so it must have have been a fault, not a "feature".
> 
> Yes, it IS a design error. Any rig that is broader than necessary for 
> the mode being transmitted is badly designed or badly operated. 

I have to insist: When a radio works correctly, with acceptable phase 
noise, for 20 years or so, and then suddenly the phase noise shoots 
through the roof and becomes totally unacceptable, then that's a fault, 
not a design error. And for all I know, that was what happened in the 
particular case I described.

> Quite 
> common with the el-cheapo "do everything from 160M to 440 MHz for $1K US 
> boxes,

Those are built to a price, obviously. As long as enough hams prefer to 
buy radios loaded with multiple features, while not understanding their 
technical specs, the manufacturers will serve this market and make their 
business. They will make radios full of features, bells and whistles, 
and they will make radios that try to cover DC to daylight in all modes, 
and they will sacrifice performance as much as they have to in order to 
meet a pre-determined price limit. And many hams will keep buying them.

Take the example of the country where I happen to live, Chile. The 
average ham here is maybe one tenth as wealthy as the average ham in the 
US, or in western Europe. So these hams buy only the cheapest radios 
available. Add to this that radios are more expensive here than in the 
US - the cheapest way to get a new radio is still to order it from a 
store in the US, paying the US price plus international shipping plus 
around 30% in Chilean taxes, so that your $1000 radio turns into a $1500 
radio when a Chilean ham gets it. If you buy that same radio in a 
Chilean store, it costs $1800. Buying a new HF transceiver is thus a big 
effort for many hams here, even when it's the absolutely cheapest HF 
transceiver available on the world market. That explains why the TS-120 
was so popular here in 1980's, and remains popular even today, and why 
the FT-450D and FT-857D are the most popular new transceivers at this 
time. And Chile isn't really bad off in this regard, compared to many 
other countries! You know that a country has a bad ratio between average 
income and radio prices, when you notice that it has a very small ham 
population and that the few hams there are mostly either foreigners or 
adept homebrewers!

So we shouldn't criticize inexpensive HF radios for having marginal 
specs. Only those hams who can build and sell better radios, at that 
same price or lower, have a right to criticize the manufacturers who are 
currently building those "cheap" radios!

> but also with the high priced rigs from some mfrs.

That's indeed more objectionable. It seems that there is a significant 
market for extra big and expensive radios that look very fancy, have 
some added bells and whistles, but inside have the same technology, at 
the same quality, as much cheaper radios. I don't like such rigs - but 
as long as there is a sufficently large market, there will be companies 
serving it.

At the end it's "to each his own", and that's what makes this world so 
colorful and interesting!

Manfred

========================
Visit my hobby homepage!
http://ludens.cl
========================