Where else to leave the remaining 100nF RelCap ?
On the floor !
Look : http://img695.imageshack.us/img695/2617/parasoundjc1.jpg
Last edited:
It's not the feedback that makes the difference. It's the difference between BJT output stages and FET output stages. A simple BJT emitter follower is inherently unstable with a capacitive load -- period. The problem only becomes worse as the output stage becomes more complex (eg, complementary, or a double, or a triple, or error correction, or feedback).
It is trivial to make a power amp with a MOSFET output stage that is stable into any load without using inductors nor Zobel networks. On the other hand, it is extremely difficult to do the same with a BJT output stage. I worked on it for several months before I developed a technique that worked. None of our power amps (MOSFET or BJT) has ever used an output coil nor a Zobel network. They are unconditionally stable with any load.
This subject is only treated well in one book:
http://www.amazon.com/Handbook-Analog-Circuit-Design-Dennis/dp/0122542401
Dennis Feucht was a designer for Tektronix during their "golden age", working on their wide-bandwidth vertical deflection amplifiers. But even this book does not give enough information to solve the problem totally. It only offers a few clues. As you can see by the prices, the book is long out of print and extremely hard to come by. The high prices reflect the worth of the book.
Hi Walt, compromising the input common mode range (remember the 845?) further is not good for a general purpose part.
He also has several columns in analogZONE - the independent space for analog engineering that are often real gems.
I bought his initial book on CD and had it printed and bound. It sits just next to Walt's books
jan didden
Hi Scott, thnx, but that may be an apples vs. oranges compare.
For the ad825, take the drains away from the +Vs rail, tie into upper current mirror emitters. I can't see that this compromises CM range.
My general point was/is, why not use the drain current that is wasted?
Best regards,
Walt
Hi Walt,
If you read the fine print at any major corporation's web site you'll probably find wording to the effect that anything sent to them unsolicited becomes property of the company. Being a large corporation with money they view themselves (or the corporate lawyers do) as a target for law suits. If they were to respond it would open the door for outsiders to claim that some of their ideas were used in a design and so on. This is probably why they do not respond.
I wrote to several major semiconductor companies stating that I had a solution to a classical problem in digital design and that I was interested in selling the IP. I got no response from most of them, one sent my entire letter back a year later claiming that they had not read it, LOL!
I think you are agreeing with me, or am I misreading you, however the point is that some are hopefully not idiots. My patent attorney friends tell me that they do get a fair number of crazies claiming to have invented all sorts of things such as anti-gravity devices, etc.
There are many well known stories about this. Villchur had a working prototype of the acoustic suspension loudspeaker that he probably did not want to show without a non-disclosure, and he went to several major loudspeaker companies all of which told him his idea would not work.
Yes we agree, SOP (standard operating procedure) IP (intellectual property). I might exaggerate, we have been approached by folks that thought they had the best "sounding" amplifier as well as a guy who thought microwaving (oven) op-amps permanently removed Johnston noise. You can't proceed on totally subjective claims.
As for the Villchur story, at some level so what? His ideas made it out. Are there world changing ideas out there languishing because of the chicken and egg situation, I want an NDA/we don't sign NDA's? Maybe, it takes two to come to loggerheads.
Last edited:
So what? While the idea was simple no one else figured it out, and if you ask me it is one of the most significant breakthroughs in loudspeaker design - yet all those companies were probably too arrogant to listen to him. I don't know about the NDA for sure, it is my speculation. I think there is serious case of not invented here (NIH) going on with most people/companies. I don't buy into all the claims of the acoustic suspension system but it certainly did allow for a smaller box which led to the vented camp also figuring out how to make them smaller.
Actually, Harry Olson broke Villchur's patent, citing prior art. Ed V. did a service to the audio community, AND he made big money from it. However, he as was much 'sales' as anything else, and many people were diverted to his approach, (including me) when other approaches would have worked as well or better. What we REALLY needed in those early days, was a subwoofer to be added for the very bottom, and less excursion in the bass and midrange.
Many years ago I looked into having a standard chip custom packaged in a standard DIL case to access some nodes that normally were not accessible. With the '825, such an option would bring the input stage pos nodes out separately from the Vcc node. I seem to remember we could get a 1000 packaged for something like 5 grand. Plus the price of the naked dies of course. Sounds like a great group buy
Hopefully Scott could point us in the right direction.Any takers?
jan didden
Hi Walt,
Thanks for chiming in. I'm actually a bit surprized that no commercial power amplifiers (that I know of) have used this circuit; maybe it just has not been written up publicly in the past.
Cheers,
Bob
We haven't run out of complementary jfets as of yet! That is why we are not reaching for alternatives. We will never really run out of P channel jfets, just CHEAP, low noise, P channel jfets. Linear Systems is taking up the slack AND they make lots of jfets monthly, quality bipolars too.
It will just be more expensive, kind of like it is with tubes today.
It will just be more expensive, kind of like it is with tubes today.
- Status
- Not open for further replies.