bcarso
Well-known member
JD and Peter: if you feel inclined, an exectuive summary of Vittoz would be appreciated. :green:
Active Design Using Inverters?
- Thread starter skrasms
- Start date
Help Support GroupDIY Audio Forum:
clintrubber
Well-known member
[quote author="bcarso"]JD and Peter: if you feel inclined, an exectuive summary of Vittoz would be appreciated. :green:[/quote]
done :wink:
done :wink:
clintrubber
Well-known member
Sorry guys, you'll have to use something else... the inverter-amp happens to be patented by Raytheon in 1999 :shock: :roll: 
5963094
I don't see much new stuff in there...
(presently a 'broken' incomplete file, maybe they're withdrawing this patent at the moment :wink:
http://www.pat2pdf.org/patents/pat5963094.pdf )
5963094
I don't see much new stuff in there...
(presently a 'broken' incomplete file, maybe they're withdrawing this patent at the moment :wink:
http://www.pat2pdf.org/patents/pat5963094.pdf )
bcarso
Well-known member
That's....amazing. Must be that complex feedback network that is such a masterstroke of utility and novelty :razz:
> (presently a 'broken' incomplete file, maybe they're withdrawing this patent at the moment
http://www.pat2pdf.org/patents/pat5963094.pdf )
Pat2PDF is a handy hack, but can choke.
Try Google: http://www.google.com/patents?q=5963094
> Sorry guys, you'll have to use something else...
The Raytheon patent appears to I Claim many ordinary or specific things in combination. Claims 1-8 are all old-hat. Claim 9 says "wherein amplifier is a monolithic structure", so unless you have a silicon fab in the attic, you can't run afoul of it.
I think it is a cellphone/walkie-talkie building block for a commercial product.
If you copy it EXACTLY (clone the phone), Raytheon will try to demand payment.
Seems to me that just re-scaling to 75 ohms evades the specific Claim, and arguing infringement would only profit the lawyers.
Also note that it Claims class AB operation. Going to the start of this thread we had the CD4xxx CMOS misused as linear amplifiers; these are VERY Class A. I guess you could use a couple IRF MOSFETs with a 6V battery and make a class AB amplifier the same way.
http://www.pat2pdf.org/patents/pat5963094.pdf )
Pat2PDF is a handy hack, but can choke.
Try Google: http://www.google.com/patents?q=5963094
> Sorry guys, you'll have to use something else...
The Raytheon patent appears to I Claim many ordinary or specific things in combination. Claims 1-8 are all old-hat. Claim 9 says "wherein amplifier is a monolithic structure", so unless you have a silicon fab in the attic, you can't run afoul of it.
I think it is a cellphone/walkie-talkie building block for a commercial product.
If you copy it EXACTLY (clone the phone), Raytheon will try to demand payment.
Seems to me that just re-scaling to 75 ohms evades the specific Claim, and arguing infringement would only profit the lawyers.
Also note that it Claims class AB operation. Going to the start of this thread we had the CD4xxx CMOS misused as linear amplifiers; these are VERY Class A. I guess you could use a couple IRF MOSFETs with a 6V battery and make a class AB amplifier the same way.
bcarso
Well-known member
If you make the rail(s) low enough you can make 4000-series ~class AB. In fact outlying high-Vth-magnitude parts inside can make this a nuisance, since you will have lower gain and if used for a xtal "gate" oscillator it may not start.
OTOH worse is having leakage currents from dirty boards bias the input of such an oscillator away from the max gain region, close to bottoming or topping out at the output. This was happening on a Harman speaker with a TSSOP uC in the right satellite---it would sometimes take it a few minutes to start. It was a wakeup call for cleaning up the process, although some guarding techniques could have been pursued as well. We also made the feedback R a little smaller.
OTOH worse is having leakage currents from dirty boards bias the input of such an oscillator away from the max gain region, close to bottoming or topping out at the output. This was happening on a Harman speaker with a TSSOP uC in the right satellite---it would sometimes take it a few minutes to start. It was a wakeup call for cleaning up the process, although some guarding techniques could have been pursued as well. We also made the feedback R a little smaller.
There's something about large scale manufacturing that sharpens your thought process and points out any weak links in a design. :idea:
JR
JR
bcarso
Well-known member
[quote author="JohnRoberts"]There's something about large scale manufacturing that sharpens your thought process and points out any weak links in a design. :idea:
JR[/quote]
Oh yes. I've tangled from time to time with a few in here who feel that having one or a few pieces work fine is sufficient reason to bless production quantities, at least in principle. But then it is a DIY forum, so that attitude is understandable and even justifiable, as long as that production remains hypothetical.
JR[/quote]
Oh yes. I've tangled from time to time with a few in here who feel that having one or a few pieces work fine is sufficient reason to bless production quantities, at least in principle. But then it is a DIY forum, so that attitude is understandable and even justifiable, as long as that production remains hypothetical.
clintrubber
Well-known member
[quote author="JohnRoberts"]There's something about large scale manufacturing that sharpens your thought process and points out any weak links in a design. :idea:
JR[/quote]
So true... and then we're not there yet, designing stuff in newer IC-processes requires more and more software knowledge to a point that the word 'transistor' has become pretty rare at the coffee machine could you synchronize the database for me ? Please check in that version... I need a tiling script... Oh, we need to upgrade to a newer checkset-version & start filling in the checklists to get our mask-request granted...
That's the fun of DIY, a run of one, it skips all that madness... :thumb:
JR[/quote]
So true... and then we're not there yet, designing stuff in newer IC-processes requires more and more software knowledge to a point that the word 'transistor' has become pretty rare at the coffee machine
could you synchronize the database for me ? Please check in that version... I need a tiling script... Oh, we need to upgrade to a newer checkset-version & start filling in the checklists to get our mask-request granted...That's the fun of DIY, a run of one, it skips all that madness... :thumb:
bcarso
Well-known member
And also remember that in the corporate world, ten thousand atta-boys are wiped out by one Oh S**t!
and all those quality award plaques on the wall don't mean shine, when some vendor's SPC wasn't statistical enough, and you have lines down, with containers being recalled...
I will apologize in advance because this sounds arrogant but building one of something. or even what passes for manufacturing from many small companies, does not require a really comprehensive understanding of all the technology involved. And even with some understanding "stuff" still happens.
I think the high performance IC designers are the creme of the analog design crop these days. While it's fun to complain about the computer encroachments into chip design, that's better than eyeballing a rubylith (sp?) for accuracy. I never made an IC from scratch but had some involvement with semi custom ICs back when manual drafting was involved, don't want to go back to that.
I'd love to hear the real story behind some of the new super opamps. I bet there's some grey area between process variability and the more clear cut design decisions.
JR
I will apologize in advance because this sounds arrogant but building one of something. or even what passes for manufacturing from many small companies, does not require a really comprehensive understanding of all the technology involved. And even with some understanding "stuff" still happens.
I think the high performance IC designers are the creme of the analog design crop these days. While it's fun to complain about the computer encroachments into chip design, that's better than eyeballing a rubylith (sp?) for accuracy. I never made an IC from scratch but had some involvement with semi custom ICs back when manual drafting was involved, don't want to go back to that.
I'd love to hear the real story behind some of the new super opamps. I bet there's some grey area between process variability and the more clear cut design decisions.
JR
bcarso
Well-known member
[quote author="JohnRoberts"]and all those quality award plaques on the wall don't mean shine, when some vendor's SPC wasn't statistical enough, and you have lines down, with containers being recalled...
I will apologize in advance because this sounds arrogant but building one of something. or even what passes for manufacturing from many small companies, does not require a really comprehensive understanding of all the technology involved. And even with some understanding "stuff" still happens.
I think the high performance IC designers are the creme of the analog design crop these days. While it's fun to complain about the computer encroachments into chip design, that's better than eyeballing a rubylith (sp?) for accuracy. I never made an IC from scratch but had some involvement with semi custom ICs back when manual drafting was involved, don't want to go back to that.
I'd love to hear the real story behind some of the new super opamps. I bet there's some grey area between process variability and the more clear cut design decisions.
JR[/quote]
You know when you look at those new "super" National Semi parts that are rated for an abs max of +/- 17V, that some pretty serious agony was going on there. (I'm going to get some to play with as early as tomorrow btw)
IMO the only thing worse than extrapolating from the construction of a single sample to the presumption of no-problem production, is the uninformed use of simulators.
I will apologize in advance because this sounds arrogant but building one of something. or even what passes for manufacturing from many small companies, does not require a really comprehensive understanding of all the technology involved. And even with some understanding "stuff" still happens.
I think the high performance IC designers are the creme of the analog design crop these days. While it's fun to complain about the computer encroachments into chip design, that's better than eyeballing a rubylith (sp?) for accuracy. I never made an IC from scratch but had some involvement with semi custom ICs back when manual drafting was involved, don't want to go back to that.
I'd love to hear the real story behind some of the new super opamps. I bet there's some grey area between process variability and the more clear cut design decisions.
JR[/quote]
You know when you look at those new "super" National Semi parts that are rated for an abs max of +/- 17V, that some pretty serious agony was going on there. (I'm going to get some to play with as early as tomorrow btw)
IMO the only thing worse than extrapolating from the construction of a single sample to the presumption of no-problem production, is the uninformed use of simulators.
clintrubber
Well-known member
[quote author="JohnRoberts"]While it's fun to complain about the computer encroachments into chip design, that's better than eyeballing a rubylith (sp?) for accuracy.[/quote]
There are 'CAD-increases' on a few fronts: better able to simulate things, being able to tackle larger systems, better autorouting etc etc. That's good!
I previously hinted at yet another: the hassle for newer processes to get your layouts prepared for processing: increased attention required checks, tiling-hassle, non-tech checklists etc.
Regards,
Peter
There are 'CAD-increases' on a few fronts: better able to simulate things, being able to tackle larger systems, better autorouting etc etc. That's good!
I previously hinted at yet another: the hassle for newer processes to get your layouts prepared for processing: increased attention required checks, tiling-hassle, non-tech checklists etc.
Was before my time, but our dept still has a few of those sheet-sets for history's(sp?) sake.
Regards,
Peter
clintrubber
Well-known member
[quote author="bcarso"]I have a funny story about using "high-voltage" CMOS (like the 4XXX series) for ~linear amplifiers that I will tell someday if you are all good (notices people fleeing for the exits).[/quote]
Lest we forget.... let's bump :wink:
Lest we forget.... let's bump :wink:
bcarso
Well-known member
In a screaming CEDIA push at the moment.
clintrubber
Well-known member
Didn't want to rush you, you'll understand, just a friendly bump :thumb:
Bye,
Peter
Bye,
Peter
jdbakker
Well-known member
Bump for:
JDB.
['s not a midwinternight, but still...]
bcarso said:
JDB.
['s not a midwinternight, but still...]
bcarso
Well-known member
jdbakker said:
I guess this is not going to die.
OK. So many years ago I am designing a multimedia subwoofer amplifier and need a gate or two out of a hex or perhaps a quad package. I also need some gain for the auto-on circuit. So, having used them before as mediocre amplifiers I decide to do it again. This time I also devised a way to use one gate as the bias source for the inputs of a cascade of the others, having established that the threshold voltages within a package were fairly well-matched (I don't recommend this since the spec is not guaranteed, but I played closer to the edge in those days).
There was one character in the organization who was generally despised for his non-stop meddling. He had it in for me and many others. When I ran afoul of some waaay-substandard electrolytics in an admittedly somewhat risky design, he was really digging in to get me fired. And when he saw the design for the sub amp he racheted up the campaign.
Mike Watts, as mentioned earlier in this thread, was heading up our group at the time. He told me about sitting in a meeting with various high-level execs including Mr Character, who was bad-mouthing eloquently about what an idiot I was to attempt to use 4000 series CMOS as amplifiers. He cited his own experience as showing how they didn't work, and decried this stupid National Semiconductor series of articles that led him astray.
Mike said Do you recall who wrote those articles? Mr Character did not, although he was quick to point out that they ought to have been horsewhipped or worse.
Mike waited a moment, and then said: I wrote those.
Mike told me this story with great glee, as it was one of the rare occasions when Mr Character was completely at a loss for words.
