Separate names with a comma.
Discussion in 'Amplifiers' started by Shockwave, Sep 5, 2019.
He has asked me about it recently so it is possible he has made statements regarding it.
so my assumption that this being related is confirmed.
he didn't let the cat out of the bag by any means
I can confirm that I briefly described what I was doing with him on two occasions (once almost 2 years ago and last month) as I've done here but not everything you've mentioned is correct or current.
And neither did I.
it's all good.
we await the unveiling and don't forget to share that video here first.
Ok enough with the CW and PEP mumbo jumbo. The proof is in the pudding. Where is the damn video already?
Video introduction showing the new linear amplifier design:
Second video showing the first two tone SSB test:
This is my first time making a video so I'm learning things about how the camera interacts with the scope. We were a bit rushed right down to writing the labels on in pencil. I should note that this two tone test was the hardest because it shows less output than normal speech PEP. The exciter was struggling to make 8 watts under this test. The next videos will be better.
The prototype shown here is only demonstrating the capabilities of a single output transistor for educational purposes. It by no means represents what can be done combining additional stages.
Yeah, and BBI won't show the inside of his 50-Volt stuff either.
How do you expect me to recoup two years of research and development if I give it all away with schematics and internal parts layout photos? All of the 50 volt stuff has been public knowledge for over a decade. Everything inside this amplifier has never been done before.
I wouldn't either. If the Chinese don't copy it the meth head at the cb shop will.
Inevitably someone will and will probably leave out everything they didn't understand like the electronic protection which makes the whole thing bullet proof. I forgot to demonstrate that circuit in action with this video but won't next time.
Got no answers for that. The "open-source" folks just tell you that if you build your gadget well and for a decent price they will sell. This does rather imply you'll have to have your widget built by a Chinese contractor.
My experience with bowl boxes of both empty and solid-state construction has been that the guy with the lowest price also offers the shortest life expectancy from an amplifier.
His price is low because he cut corners and left out all the stuff that makes it harder to blow up. Any part that doesn't contribute to a robust wattmeter reading is left out.
Adopted a rating standard for this kind of amplifier, the "KBK" rating.
As in "Keys Between Kabooms". Kinda like the old term "MTDT" in the bad old days of twitchy mainframe computers that would shut down if the machine-room temperature rose five degrees. Stood for "Mean Time To Downtime". Fifty years ago, this rating was a big deal for anyone buying a million-dollar mainframe computer.
But people buying amplifiers don't ask how long the thing will last. They want to know the price and how many Watts they'll see on the wattmeter.
My summation for anyone who asks me about having an amplifier built is to choose between the two main options:
1) Cheap to buy, expensive to keep and run.
2) Expensive to buy, cheap to keep and run.
Pretty well priced myself out of the business of building custom amplifiers. My estimate would always include the extended-lifespan features.
And a price higher than the other guys. Seems the other guy usually got the job.
Not a big deal. When the cheap-to-buy bowl box breaks down I get hired to revive it.
There is another angle to this. Is the effort to hide the details really because this is something new, wonderful and unique? But in the mind of a skeptical customer, the more effort you see being made to hide technical details, the more mundane they turn out to be.
The question becomes "Are they hiding it so I won't know how stupidly cheap it was to build and sell for this premium price?"
I think I know better. Built too many to think this.
But the prospective buyer probably won't know that.
And the problem of maintainability still means something. If you succeed in making it fail-proof, you won't be in that business for long. Once everybody who wants one has one, sales plummet.
If it's potted in resin, you won't be able to fix it any more than the ECU in your car when a 50-cent power transistor goes bad inside that block of resin. It gets replaced as a whole unit. Nobody's gonna chisel out that transistor and scab in a new one.
Is that the 'repair' rationale for this amplifier if it develops a fault? Sell them a new one and ditch the failed unit? Like an ECU at Auto Zone?
If you make the guts accessible for reasonable repairs, you also give away the technical details.
And if you pot the important components, when it breaks it's just a brick.
The elephant in the room is probably still China.
But that's the bad news. No easy answers.
So, if you're still awake, I do have one relevant question. Does your design include any protection from, or sensing of excessive drive power?
Modern insulated-gate FETs have this one weak link in the chain. The "indestructible" RF transistors that NXP and Motorola,er "ON Semi" show on YouTube are all being abused on the output side of the circuit. One schmuck with a radio that's too big can defeat all that ruggedness with one key of the mike and blow it up.
I've been contemplating a way to reliably sense excessive RF voltage on the transistor gates. Rectifying the RF-input voltage and feeding the DC into a zener. When the zener turns on, the relays drop out.
Probably wouldn't be fast enough, but I haven't let go of the idea.
Haven't heard any discussion of how to do this.
I've never been the type to squeeze every last drop out of any amplifier. If more power was needed, more stages were added. Enough so that no one device is run beyond 50% of its capacity. Just doing that with decent components usually yields an amplifier that could sustain a short on its output. As you already know this is not the weak link with modern devices. The gates need to be protected with a circuit that can handle the unexpected in a fault condition and it must do so within milliseconds at most.
The heart of my protection board is centered around a circuit I learned from Perkin Elmer. It uses an opto coupler that triggers an SCR at virtually the speed of light. You will see in the next video that this circuit has the ability to protect my PA stage from overdrive conditions that would kill the BLF288XR. Unfortunately I can't share the part of the circuit responsible for sensing but you'll see its ability to sense and protect against overdrive conditions.
These are the areas that will be potted. Areas that are key to this circuits ability to function but have decades of proven reliability to insure they won't need repairing. You won't find the output transistor potted or any other active part that could fail. Passive components that are potted could never fail due to overkill. Serviceability will remain long after I'm gone but hopefully the prying eyes of counterfeiters will be blurred for some time.
Okay, so my hands are not completely clean.
In the late 80s we adopted a setup for adding channels to the Cobra 148-type radios of the day. Needed only two 2-position switches to get channels 42,43 and 44. Those would get skipped unless you added a third switch if you used the published mods.
The difference was that we used a NPN transistor to invert the pin-11 wire coming from the channel selector. Tying pins 10 and 11 together, and feeding them the "upside down" logic state from the selector moved the "+32" channels to a stretch of the channel selector that would now only skip channel 41. The reason for coming up with this was to use only two original front-panel switches, the Tone and Bright/Dim to get those three "third-switch only" channels. Using only built-in switches was a lot more popular than the "porcupine" approach with toggle levers sticking out the top, bottom or sides of a radio.
A young guy who considered himself a Cobra 148 expert brought in a radio to have us install extra channels our way. We offered to do it while he waited, but they went out for lunch. Explained they only wanted to have our setup to copy. A point or two for honesty, I suppose.
Hokay.... Couldn't resist.
While they were gone I took a handful of 1N4148 tiny glass diodes and hooked them one by one to a 45-Amp power supply. There would be a faint, but visible flash from inside, and then it would be an open circuit.
Took several of these "dummy" non-diodes and wired them here and there around the PLL chip. Figured if the young guy copied it exactly he would have an enjoyable and satisfying puzzle to solve. Pretty sure he would have a one or two-channel radio by the time he had copied where all the diodes were.
Don't ever remember hearing back from the guy.
Probably a good thing.