On Board Processing

If you are able to use a good multiband processor, which also has proper peak control, it would benefit from bypassing the internal compressor/limiter of the AMT5000.

Leaving the internal comp/lim inline, in this case, does reduce audio quality and peak control, this is what i have done.

Paul.

Hello BOARDMAKER.

In both the AMT3000 and AMT5000 the compressor is no longer inline when the control is adjusted fully CCW.

What I wish I'd said in my opening post, and what is very important to point out, is that the modulation control (the limiter) is no longer inline when it is adjusted FULLY CLOCKWISE, which totally removes the limiting. It is not something anyone should do without having another limiter installed, but it is the way to completely turn control over to the outboard device.

You are right that if the limiter/modulation control were set CCW it would impose limiting at a very low audio level and not be a desirable situation.

Hope this information is useful.

Carl,

The audio is still routed through the processor ic, with any setting of the controls.

Your settings will stop the processor chip from actively compressing/limiting the audio through it.

Paul.

Yes, what you say is true, but since the processing itself is not taking place, the audio is passed as it is without being changed internally within the transmitter.

To correct our language we could say that the processing is not inline, only the line is inline.

With the on-board processing shut down there is no remaining problem that will impact your outbroad processor.

This is all a very interesting theoretical discussion.

But there's theory, and then there's reality.

If, and I repeat, if, there is degradation introduced to the signal in this manner, it would be very, very tiny.

And in Part 15 broadcasting (which is what we're discussing, isn't it), noise which is far greater than any such degradation by many orders of magnitude is introduced very quickly into your signal by enironmental and man-made factors (a legal Part 15 signal at best is probably around 0.1mw output due to antenna inefficiencies, and not the 50,000 or 10,000 or even 1,000 watts found in licensed stations).

When I had my Rangemaster, I was able to hear it (i.e., the programming could be heard above the noise) anywhere from 1-2 miles away, depending on direction, which is about the most you can expect from a legal installation.  But I could start to hear very low level static at the end of my street, if I listened carefully, and that in itself would dwarf any such signal degradation (if it had indeed existed).

If I was attempting to decide which Part 15 transmitter to purchase, this so-called issue wouldn't even enter into the discussion.  With Part 15, you do need good sound to start with (which virtually all but the cheapest transmitters provide), but range is everything to maintain that good sound above noise levels.

I suspect that the original poster over at that other website has ulterior motives for bringing up the AMT3000 & 5000 transmitters as examples - he appears to take delight in slamming SSTRAN at every opportunity, regardless of fact.

And while you can pick at every company and transmitter in this hobby to some degree, the facts show that the AMT5000 has the potential (if it is tuned correctly) to give you the most range, with great sound (and that has been shown both theoretically and in practice).

The AMT5000 is a very good design, and it is capable of good audio quality.

I am quite critical on audio quality, and i am fortunate to have a 9100b/2 optimod.

The main reason i bypassed the AMT5000 internal processor ic, was to sort out LF tilt issue, so the peak control was vastly improved, and also to improve audio quality, in this case.

I will soon be feeding a C-quam exciter to the AMT5000, so i will have to bypass the amt500's pll section to acheive this.

Paul.

LF (low frequency) tilt had not been previously entered as an element in this conversation which retrospectively puts me out of tempo with the entire discussion up to this point.

Giving a description of the, to most readers, unfamiliar concept of LF tilt, would be helpful to them.

Before concluding that the signal chip in either SSTran transmitter is the cause of signal distortion I would be suspicious about other stages of the audio chain both at the input and output of the 9100b/2 Optimod,  since we are in a mix and match or mis-match world between conumer and professional gear, given balanced to unbalanced circuits and the range of inter-stage levels involved, eg., -10dB, 0dB, +4 dB and so on.

My experienced impression is that the audio frequency response and the transient response of the SSTran transmitters is fabulous and vastly superior to many other transmitters.

Another surprise for me is to learn that you have by-passed the AMT5000 audio processing chip, not previously mentioned. On that point it would be very informative to part 15 readers to see a diagram of how best to make this modification.

Finally, the stereo mod being planned is something many here have dreamed about, so we hope that project gets posted.

Audio quality for the sake of itself (and your own listening) is all fine and good.  But other than that, you may get superb audio quality for several hundred feet or so until the ubiquitous AM static starts to creep into your pristine signal.

For the vast majority of Part 15 broadcasters, it's a non issue, and that's all I'm saying.  The lack of the ability to bypass on-board audio processing shouldn't be a reason for most potential broadcasters to pass on a particular transmitter.

One key for AM range is plenty of modulation. Although I haven't tested any of them, from what I've read over the past few years, different transmitters will only modulate to a certain percentage as received out of the box, and that's generally been less than 100%. 

My Procaster, using only the built in processing that comes with the transmitter gives me +125% and - 99% which is about the best you can do without introducing much very noticeable distortion. If I back off the modulation to 100%, then down to 80%, each lower level is clearly and demonstratibly giving me less range. 

The concept of firing up various transmitters to determine their potential range is somewhat misleading unless that test includes measuring with modulation.  This is one detail that irks me about certain AM transmitter tests.  For example, I do not believe that a Rangemaster would have comparable range to a Procaster if tested with program material, as they are received out of the box. The Procaster will simply modulate better due to onboard processing, giving it a range advantage. Use of the Inovonics 222 processor as recommended for the Rangemaster, I believe, would put the two neck and neck. I have NO numbers on this, just going on experience and experiments.

If onboard processing is up to snuff, the need to bypass it would be more of a preference than a necessity.  But if the onboard processing isn't going to give you 100% at a bare minimum, then you want to do something about it!  And some compression and limiting is going to help give you solid modulation with varying source material.

For all you who like to "EQ" your sound because you think it needs "more bass" for example.  I hope you're not setting up your EQ by listening to a radio!  EVERY radio will sound different and have dramatically different frequency response. Setting your EQ to the way you think it should sound while listening to a radio is wrong in so many ways.  Not only are you trying to decide the EQ curve, or "tone control" that you think you like best, you are trying to force your choices upon the listener. For the most part, especially on AM, these days audio going into a transmitter is pretty darn flat -- once you adjust for rolling off frequencies that are counting in the modulation mix but can't be heard on the air.  You'd be surprised how many stations pump up the bass, and all they wind up doing is overmodulating with a signal that no one can actually hear, that the transmitter can't reproduce accurately, that makes the sound muddy.  You have to know when enough is enough. this is why there are som many FM stations that sound like crap. Over processed to meet the managers, owner, or program directors concept of "we need to be LOUD and have a lot of BASS to compete".  Of course a high quality processor can help avoid a lot of those troubles, but few Part 15 stations are going to put in a $5000 processor (or more LOL). 

TIB

The basic built in processing in the AMT5000 is very useful to make consistent audio, but it has no tight peak control.
Also the nrsc pre-emp option is enabled after the processing, so this would further make peak control less defined.

Don't get me wrong, the built in basic processing is very useful, far better than none !

The Inovonics 222 has dynamic pre emphasis limiting, as well as final 10khz filtering, and tight peak control, so is in a different ball park to the basic AMT5000 processing.

The 222 is still a bit basic in the main limiter section, after all it was meant to be an add on nrsc compliance box, the main limiter is basic, but gets the job done.

I will get a pic done of the bypass mod on the AMT5000, as well as the info, if you want it.

Paul.

I am currently unable to upload my website of the picture of the mod on the amt500, getting error socket ............

The upload here does not have the option for direct upload of pics, only url option.

I will see if i can sort out my own webspace upload over the next day or two.

Paul.

Much appreciation Paul for taking on the task of getting that information to the thread... it's very interesting as a step to seriously modulate an AM signal.

Meanwhile I went looking for a reference to "LF tilt" in audio measurement, and so far have been reminded that the luminance bar on an NTSC test pattern shows "LF tilt"...

I also very vaguely remember the Aphex Aural Exciter with "Big Bottom", which may be a circuit for compensating LF tilt... I can't find my manual, but when I do I'll say something about it.

Boomer, a fellow broadcaster in carrier current, normal part 15 and abnormal higher power radio, has contacted Worldround Radio with this complete introduction to the advanced subject of LW Tilt:

LF Tilt usually relates to the frequency response of a transmitter's audio circuit in broadcasting, and the ability of the audio stage to pass a square wave.

To pass a square wave without tilting it, the audio system in the transmitter needs to be wide and flat, beyond the usual frequency range you'd use on the air.

Why worry about squared waves, when normal program audio is sines? Well, that's so you can use modulation techniques like clipping, which produces square waves. 
If you don't use clipping, you really don't need to worry about tilt all that much! If you do clip to max out your audio, then you should consider going tilt free.

If the squares go through a circuit without enough LF frequency response, they'll probably tilt, when they should be flat across the top of the square. The tilt effectively adds peaks that max out the modulation capability of the circuit before it should, meaning lower perceived volume level from the listeners' radios.

Adjustment is by scope, feeding in a low frequency square wave through the processor, for example 50-60 hz, then looking at it on the scope to see if it's a flat top, and if not, adjusting the tilt compensation control. In FM, you can use a deviation monitor or meter to adjust tilt, sending a square to the transmitter, and adjusting for lowest modulation level on it, which indicates you're removing the peak that's causing the false high modulation.

Here's a processor maker's page on it about AM, but it also applies to FM as well, any transmitter really. 

LINK TO BREAKAWAY BROADCAST PROCESSOR
Thank You Boomer

Here is the picture of the bypassed AMT5000.

Remove c39, and also remove c28.

Fit a new 100u/25v cap, with it's -ve lead to c28's -ve empty slot
directly, and the +ve lead of this new cap bent upwards and cut so this can be soldered to a short jumper wire.

Next look at s21 on the pcb, it is empty and has 2 holes, solder the juper wire to the s21 slot hole that is nearer the c24 capacitor position.

This mod connects the o/p of the differential amps output through a capacitor to the input of the modulator.

Also for good lf response you need to replace both c35 and c38 for 100u/25v
caps each (each +ve cap lead facing toward the back of the amt5000.

Finally replace c6 for a non polarised 10uf or greater capacitor, the reason for a non polarised here is the voltage across it can reverse depending on the setting of the r1 power setting.