Nice work as usual, Pavel.
OT posts removed. Stay away from the personal remarks.
I honestly can't remember. I spoke with an apps engineer at Siemens (now Infineon) who knew exactly what the problem was, but I just don't remember now. That was about 18 years ago.
I tested many different brands and only IR had this problem and only on their P-channel parts. The people at IR were not helpful at all. First they said they would look into it. Then they said that the devices were not spec'ed for linear applications and that it met all of the published specs for switching and please go away. That was it. End of story.
The funny thing was that other companies had to get licenses to make the parts with the exact same part numbers. I assume that they also received some of the information on the fabrication process. But whatever it was that IR was doing, nobody else copied it, that's for sure.
When Nelson published his article in 2003, I got a second round of samples of a half dozen different part numbers for repeat testing. I wanted to see if they had fixed anything in the interval. Nope. They still had the exact same problem, so despite the problem in their process, they continued doing the wrong thing.
And how many times have we been told there is no problem? My guess would be the other folks found it, fixed it and were not required to tell under the terms of agreement!
My experience even with Bipolar Junction transistors is that the P parts never quite match the N parts. This I believe has to do with the dominant carrier mechanism.
But the other issue of interest that is interesting is the guys with the science backgrounds always ask why and the engineers ask how to I get around the problem!
My experience even with Bipolar Junction transistors is that the P parts never quite match the N parts. This I believe has to do with the dominant carrier mechanism.
But the other issue of interest that is interesting is the guys with the science backgrounds always ask why and the engineers ask how to I get around the problem!
I'm somewhere in the middle. While I was very curious as to what was going on, I was much more interested in building an amplifier that we could sell so that our company could stay in business. Once I found out the reason, I was satisfied. In the old days I never kept a notebook because I could remember everything.
Hah!
That's one reason I keep all of e-mail -- it often serves as a notebook. I often ask guys at work to e-mail me things so that I'll have a written record. But I'm still too lazy to keep a notebook. Sometimes I'll work out an idea with pencil and paper but then I'll just transfer it to the computer.
I'm sorry to have upset you Charles. This is old stuff, and we have discussed
it publicly before:
http://www.diyaudio.com/forums/solid-state/28853-sound-vmosfet.html
It also came up on other occasions, including postings from 6/3/2007 in
Bob Cordell's BJT vs Mosfet thread. Elesewhere I recall that Syn08 came
up with the "quantum mechanical" explanation.
I don't seem to recall claiming a discovery, but I have documented the
phenomenon from my own perspective for the benefit of the DIY
community. including a spiffy graph which was imbedded in my article
on Mosfet testing.
By the way, I still use IRFP9240's in some amplifiers.
it publicly before:
http://www.diyaudio.com/forums/solid-state/28853-sound-vmosfet.html
It also came up on other occasions, including postings from 6/3/2007 in
Bob Cordell's BJT vs Mosfet thread. Elesewhere I recall that Syn08 came
up with the "quantum mechanical" explanation.
I don't seem to recall claiming a discovery, but I have documented the
phenomenon from my own perspective for the benefit of the DIY
community. including a spiffy graph which was imbedded in my article
on Mosfet testing.
By the way, I still use IRFP9240's in some amplifiers.
Really nice stuff. Unfortunately, the bulk of it seems to have been
discontinued. Do you know which ones are still available?
Hi,
Do we now... Lets see.
I believe Charles, John and myself (BTW, I am unsure I belong into such august company, especially John had decades more experience and has been kind enough to share a lot...) have always insisted that the first rule is:
"If it makes you happy, it can't be bad." (Sherryl Crow)
What you do get from some of us who have been around a bit and have routinely experienced The Hour of the Bewilderbeast (Badly Drawn Boy) again and again when things do not quite work as we had every right to expect, is the occasional cryptic comment that shares something we have discovered, generally empirically.
Now I am quite busy and I am equally that others like John and Charles are also not blessed with a life of leisure or possession of that most desirable and rare of all "far out physics" artefacts, namely a genuine warranted time machine that adds another 24 Hours to each day.
The result is that such bits of experience are often passed in a quite direct, simple and short way. We have limited time, so it often takes the form of "try this", usually simple things things to try (like ordering a sample of a good performance J-Fet Op-Amp, prying a Bipolar input Op-Amp from it's socket or desoldering it's 8 legs and putting the J-Fet Op-Amp in it's place). No, you do not get twenty pages of explanations and another 20 pages of AP-2 Graphs. Not because we can't, but because it usually is not important enough.
And yes, by suggesting that such a simple thing may provide improvements, I guess it can come across as "my way or the high way" but it actually simply is "My Way" (Frank Sinatra). We are offering these experiences simple as our experiences and recommendations to TRY.
A wise man once said: "A good teacher offers practice, a bad one offers theories."
Now to those who read this advise all possible ways are open. They may choose any path. From simply accepting it as "the saying of Guru" (which I suspect neither John, Charles and others like them are comfortable with, certainly I am not) all the way to demanding the results of a DB Test or an appy polly logie for saying it instead.
The beauty of all this is that "in the end it doesn't even matter" (Linkin Park)
As the Arabs say, "The nature of rain is the same, but it makes thorns grow in the marshes and flowers in the gardens."
You describe the human condition. Anyone who is off the steep part of the learning curve has given up.
Has become a fish that swims with the stream.
Has joined the herd of sheep on their merry way to slaughter and is happy to be like that...
Ciao T
Do we now... Lets see.
I believe Charles, John and myself (BTW, I am unsure I belong into such august company, especially John had decades more experience and has been kind enough to share a lot...) have always insisted that the first rule is:
"If it makes you happy, it can't be bad." (Sherryl Crow)
What you do get from some of us who have been around a bit and have routinely experienced The Hour of the Bewilderbeast (Badly Drawn Boy) again and again when things do not quite work as we had every right to expect, is the occasional cryptic comment that shares something we have discovered, generally empirically.
Now I am quite busy and I am equally that others like John and Charles are also not blessed with a life of leisure or possession of that most desirable and rare of all "far out physics" artefacts, namely a genuine warranted time machine that adds another 24 Hours to each day.
The result is that such bits of experience are often passed in a quite direct, simple and short way. We have limited time, so it often takes the form of "try this", usually simple things things to try (like ordering a sample of a good performance J-Fet Op-Amp, prying a Bipolar input Op-Amp from it's socket or desoldering it's 8 legs and putting the J-Fet Op-Amp in it's place). No, you do not get twenty pages of explanations and another 20 pages of AP-2 Graphs. Not because we can't, but because it usually is not important enough.
And yes, by suggesting that such a simple thing may provide improvements, I guess it can come across as "my way or the high way" but it actually simply is "My Way" (Frank Sinatra). We are offering these experiences simple as our experiences and recommendations to TRY.
A wise man once said: "A good teacher offers practice, a bad one offers theories."
Now to those who read this advise all possible ways are open. They may choose any path. From simply accepting it as "the saying of Guru" (which I suspect neither John, Charles and others like them are comfortable with, certainly I am not) all the way to demanding the results of a DB Test or an appy polly logie for saying it instead.
The beauty of all this is that "in the end it doesn't even matter" (Linkin Park)
As the Arabs say, "The nature of rain is the same, but it makes thorns grow in the marshes and flowers in the gardens."
You describe the human condition. Anyone who is off the steep part of the learning curve has given up.
Has become a fish that swims with the stream.
An externally hosted image should be here but it was not working when we last tested it.
Has joined the herd of sheep on their merry way to slaughter and is happy to be like that...
Ciao T
Last edited:
Yep.
It is old. I was the one who told you about the effect, way back in October, 1994. But obviously the word hasn't gotten around yet, because there are plenty of people still using those parts.
Yes you did, although you did not describe it in a way that I would appreciate.
Interestingly, the techs at Adcom noticed it about the same time also, but
they expressed it as "this Harris part has lower distortion".
Nonetheless, I know people who subjectively prefer the IR part in
some designs. I feel certain you would appreciate that.
Interestingly, the techs at Adcom noticed it about the same time also, but
they expressed it as "this Harris part has lower distortion".
Nonetheless, I know people who subjectively prefer the IR part in
some designs. I feel certain you would appreciate that.
INCONTHTHTHIEVABLE!
An externally hosted image should be here but it was not working when we last tested it.
se
I personally want to thank Charles Hansen for telling me about the IR Pchannel problem. In fact, I would have missed it, and used the IR Pchannel devices in the CTC Blowtorch. It was only his insistence that this was a serious irregularity, and his giving me the Harris alternative, that kept me from using the IR's everywhere.
These are the sort of problems that Charles and I work hard to find. Why would we DELIBERATELY want to use a 'defective' device in one of our products, unless it was absolutely necessary? However, Charles did NOT owe me this help, in fact, he might have kept it a 'trade secret'. Many designers would have. Thanks again, Charles.
These are the sort of problems that Charles and I work hard to find. Why would we DELIBERATELY want to use a 'defective' device in one of our products, unless it was absolutely necessary? However, Charles did NOT owe me this help, in fact, he might have kept it a 'trade secret'. Many designers would have. Thanks again, Charles.
Maybe you could ''design in'' its gfs droop with frequency trait for OLG and NFB curve, thus tailoring the resulting THD spectrum, or be safer in AC gain or DC regulation loops for high frequency oscillations with no/less external compensation. There is a way we can build magic with what we miss from optimum even in life sometimes. Its even called subcultures, popular music genres, etc. Nothing interesting ever came out of perfectly linear suburbs.
- Status
- Not open for further replies.