Linear Amplifier ALC interfacing

User avatar
Posts: 8
Joined: Sun Apr 09, 2017 3:14 pm
Location: Naperville, IL USA

Linear Amplifier ALC interfacing

Postby n9vv » Thu Mar 14, 2019 10:48 pm

Linear Amplifier ALC interfacing
Some Linear Amplifiers require an ALC interface from the Linear to the Transceiver. The Apache rigs do not provide a negative voltage "ALC" interface for Linear Amp equipment. This seems to be especially important for older Tube type Linear Amplifiers.

Have you come across an interface circuit or device that will work with our Apache gear?

Is there a document or discussion somewhere that speaks about the special feature of an Apache TRx that prevents it from OVERSHOOT or OVERDRIVE of a Tube type Linear Amplifier?

Can the new "CESSB" software cause any ALC problem for a Linear Amp?
For example W1AEX says "see a check box for CESSB Overshoot Control" on this webpage:

I thought that Warren NR0V wrote about the careful attention he paid to ALC inside the PowerSDR software and Apache firmware:

The PowerSDR Developers are very proud about how you can NOT exceed ALC = 0.0 when driving the Audio chain with MIC or VAC Audio. This certainly is an essential feature for proper operation on digital modes like FT8 or PSK31.

W1AEX has presented some excellent guidance here:

[Way back in March 2011 K6JCA surprised the Flex community with scope pictures to prove that overshoot destroyed the PA in a Flex-5000: I assume Flex solved that problem and it has not come back to haunt them.]

Is there any evidence that an Apache rig could cause a overshoot or overdrive of a Linear Amp that is relying on an ALC control voltage to ensure that the Linear operates within it's intended output range?


thanks for any info, opinions or comments,
72/73 de Ken N9VV
User avatar
Posts: 2332
Joined: Fri Mar 10, 2017 1:47 pm

Re: Linear Amplifier ALC interfacing

Postby w-u-2-o » Fri Mar 15, 2019 5:47 pm

Wow, that is a lot of random factoids there, Ken!

Short answer: ALC is NOT required to use an amplifier with any radio that uses either PowerSDR mRX PS or Thetis as the client software and Apache Labs hardware. (Note: I cannot confidently make the same assertion about piHPSDR or SDR Console client software.)


1. With one exception (see below), the Apache hardware is, as far as I know, not susceptible to overshoots at the start or end of a transmission.

2. The ALC built into the software will absolutely, 100% prevent any excursions above the drive level setting. It is probably the most sophisticated ALC in any amateur radio available. It is a look-ahead compressor/limiter with a user adjustable compression factor (in Setup > DSP > AGC/ALC, the "Max Gain" and "Decay" in the ALC section). Because it is a limiter and NOT a clipper, and because the look-ahead function will always drop the gain as required to prevent exceeding the selected drive level, there is never any danger of the radio over-driving an external amplifier, and there is no danger of significant additional IMD (unlike the horrible clipper ALC used in the old legacy Flex PowerSDR software). Since there is never any over-drive condition there will never be any ALC output from the amp.

Now of course this entire system is predicated on the operator never doing something stupid like driving an external amplifier with 100W. If you have an amplifier with no protection against gross over-drive it would be possible, without ALC, to make a stupid and potentially costly error. However most modern amp's have protection circuits for such things. Hence once the drive is properly adjusted, again there will never be any ALC output. If the amp is generating ALC voltage you are driving it too hard. Turn the drive down until amp ALC output is nil. Use the two-tone feature in the software to assist in setting drive levels, as two-tone is generated at a perfect 0dB ALC level.

There are no controls that you can manipulate in PowerSDR mRX or Thetis that can defeat or work around the ALC. The CESSB overshoot control only deals with audio processing. CESSB, and every other audio processing function in the software, comes before the ALC and has no bearing on ALC operation other than if you've got everything turned up to the proverbial "11" the ALC will be in a limiting mode most of the time. By the way, with the advent of the CFC compressor, I strongly advocate never using CESSB or COMP (and COMP must be on for CESSB to work). COMP and CESSB are old and primitive compared to the CFC compressor.

Lastly, it does bear mentioning that there was, at one time, a project that added hardware and software to accept and process an ALC voltage for these radios. You can see vestiges of that in Setup > Transmit > Limit Drive on Ext. Amp Overload. However, as far as I know that project went nowhere and was never used by any except a very few people because it is simply unecessary. The only documentation I know of is in the Yahoo Group files section here.

Cautionary note: I would urge caution with the 8000. The amplifier topology of the 8000 is a bit strange and I have heard, anecdotally, of a few amps that died an early death at the hands of an 8000. The 8000 can and will generate some very high voltages when run into high impedances, like when you are tuning up. My 8000 certainly does, although it seems to get along with my KPA500 amp just fine.


Posts: 299
Joined: Fri Apr 14, 2017 3:47 pm

Re: Linear Amplifier ALC interfacing

Postby K9RX » Sun Mar 17, 2019 1:24 pm

Same response as before Scott.....
operator never doing something stupid like driving an external amplifier with 100W. If you have an amplifier with no protection against gross over-drive it would be possible, without ALC, to make a stupid and potentially costly error.

As I've explained ad nauseum the purpose of having ALC is NOT, I repeat not to act as some form of compression. It is SOLELY to protect the amplifier from - you said it - dumb mistakes. Since its trivial in the circuitry required and software to support I STILL content it SHOULD be in this radio.

Second "stupid/dumb mistake?" ... how about this - PSDR AS IS will not always remember the power setting it was last at on a band - it's not unusual for me to change bands and instead of being at the level needed for my amplifier, worst case of about 40W, a drive setting of typically 50 - 60, the drive is instead at 100%! I've intentionally reduced the max drive on all bands to about 75W (on my 8000). I run with and without the amp all the time. If I were to always use the amp then I'd agree with you - but that isn't practical in my estimation. What I do, use it/don't use it, is much more common I believe. That 75W on my Expert 2K is well over 2000W out in the MAX position.

So I don't think it is sufficient to pass blame on to the user - the software already has issues with wrong power settings - as it is if I'm in a hurry or tired I have (not so much any more as I try to be even more careful) hit FT8 with 1000W only to have to frantically turn the power down or hit HALT.

A product should ALWAYS be designed to be bulletproof - user protected. The user should NEVER be able to do something that causes damage. Period. And for the trivial cost of a few dollars this is one thing that, to me and I'm sure others that as usual remain quiet, is lacking.

If one wanted to use the full power range of the rig - 200W - or stated another way if I didn't reduce it to a max of say 75W I wonder if all amps protection circuits would protect them every time? MFJ?! I think it irresponsible to assume that will be the case. The ALC is on the amp - most all amps - it should be in the radio. It worked GREAT on my FTdx5000 to the Expert 2K... if I accidentally left the rig at 200W the system would protect itself and throttle back.

User avatar
Posts: 2332
Joined: Fri Mar 10, 2017 1:47 pm

Re: Linear Amplifier ALC interfacing

Postby w-u-2-o » Sun Mar 17, 2019 2:53 pm


I'm all for more features in PowerSDR, or, more probably Thetis, since that is the future. Sadly, as we all know, there are no developers who are very interested in working on the user interface portion of the code in any substantive way. We've tried to bribe KE9NS, but he didn't bite. Unless and until we can find a dev. who likes UI work, we are stuck. And those are the breaks in the world of open source code. And this is, as always, my not-so-veiled plea for any dev's that might be reading this. Nevertheless, I still think the pluses of what we have outweigh the minuses. The moment they don't I'll be moving on to the better solution, whatever that is.

That said, you could try to build and employ the ALC hardware that was referenced above, if you really wanted ALC. The code for that is in PowerSDR and Thetis.

I'm trusting my stalwart KPA500 to keep me stupid-proof. And, trust me, I think I've tested every alarm it has! :D



Return to “Amplifiers & Tuners”