Boron and phosphorous are still pretty common, there are some SiGe processes around. In digital there some things like even hafnium which I don't really know about.
For most analog things the aluminization is still aluminum slightly doped with silicon. There are a few secret sauces for making good contacts to thin film, etc., I think I posted a link to a patent with how complicated it can get. I haven't used copper or gold metalization but processes do exist, I'm not sure but I think the metals usually have something added. Again a process patent search will turn up more info than you can absorb.
At TI we were always shrinking die , or at least trying to for a greater yield. Sometimes a guess would make big difference. For example gold flash under slice was found to improve all parametrics.
I hope they have frozen the really nice analog and class-d parts.
The revised dies can meet specs but subtle undesirables can surface months after the new die is released. Its an easy trap to fall into.
So happy new year
Bruce
-
I hope they have frozen the really nice analog and class-d parts.
The revised dies can meet specs but subtle undesirables can surface months after the new die is released. Its an easy trap to fall into.
So happy new year
Bruce
-
Si-Au eutectic has been use to mount die (one way) for as long as I can remember.
Member
Joined 2009
Paid Member
Halfnium Oxide is also useful for optical properties when used on silicon image sensors.
the biggest dissapointment for me at our local jazz festival was discovering that they used a sound reinforcement system. It sounded bollocks - would have been so much better if they had just turned it off
Yup, cost is king, always. That's why the golden age for some types of components is now long gone.
How much more performance can we expect to obtain from Opamps over the next 5 years - ?
It was in vapor applied just prior to probe. Did improve all of the good things.
Cleaned off prior to dice and pack. I do not know how effect is picked up again.
Guess as you say on header.
Yes that was standard process from my day 1 and I'm sure before. Don't where your version of the folklore comes from. Other than some backside treatments to getter noise I don't see any basis for any other improvements, gold is usually a bad thing.
Last edited:
Thanks.
As I understand it, the starter substrate wafer is predoped and then the processing starts layer by layer.
Is there a standard polarity (p or n) for the substrate or is this variable according to manufacturer or analog device series ?.
Analog Integrated Circuit Design
Dan.
As I understand it, the starter substrate wafer is predoped and then the processing starts layer by layer.
Is there a standard polarity (p or n) for the substrate or is this variable according to manufacturer or analog device series ?.
Analog Integrated Circuit Design
That sounds just like you Scott.
Dan.
Last edited:
Here's a class that should not glaze anyone's eyes over, it has info on a wider variety of processes than I have used. There are many choices now especially with the oxide or junction isolated bi-polar ones included. There is a more comprehensive list here of the acceptor and donor atoms used.
EE 432 : Topics
I'm drilling down on processes and elements used in different manufacturers opamps as possible causes in subjective findings, aside from internal schematic considerations.
Is there a 'house sound' according to fab processes ?.
I have read in proaudio circles of 'opamp staples' subtly changing sound according to age of manufacture, IOW production changes implemented to 'industry standard opamps' in order to accommodate/comply with new processes and wafer sizes etc, and possibly including schematics changes.
Any info on this, and what is the usual substrate wafer pre-doping polarity ?.
Dan.
Is there a 'house sound' according to fab processes ?.
I have read in proaudio circles of 'opamp staples' subtly changing sound according to age of manufacture, IOW production changes implemented to 'industry standard opamps' in order to accommodate/comply with new processes and wafer sizes etc, and possibly including schematics changes.
Any info on this, and what is the usual substrate wafer pre-doping polarity ?.
Dan.
I heard a famous(?) story from The Analog Dinosaurs Dinner a decade ago.
Analog IC company XYZ sampled some preproduction parts to Nutty Audio Company ABC. Time passed. Then ABC called up and asked for a few dozen more.
"We've handed every one of them out as samples, and won't be getting any more back from fab for another five weeks."
"Can't you do anything? We need them desperately!"
XYZ reluctantly replied, "Well we do have a couple hundred units in low temp long term reliability testing that ought to shake loose in a week. BUT these units have been soaked in liquid nitrogen for 1000 hours so they aren't perfect representations of what our production output might be."
ABC said, "Didn't you hear me? We are desperate! We'll take anything! Send them right over!!"
So XYZ packs up the test victim devices and sends them to ABC.
ABC calls back in a week and says "These sound better! MUCH BETTER! We want you to only sell us chips that have been soaked in liquid nitrogen for 1000 hours!! But we can't pay a penny extra for that. In fact we think you should give us a discount, because we discovered this tremendous improvement and called it to your attention."
Analog IC company XYZ sampled some preproduction parts to Nutty Audio Company ABC. Time passed. Then ABC called up and asked for a few dozen more.
"We've handed every one of them out as samples, and won't be getting any more back from fab for another five weeks."
"Can't you do anything? We need them desperately!"
XYZ reluctantly replied, "Well we do have a couple hundred units in low temp long term reliability testing that ought to shake loose in a week. BUT these units have been soaked in liquid nitrogen for 1000 hours so they aren't perfect representations of what our production output might be."
ABC said, "Didn't you hear me? We are desperate! We'll take anything! Send them right over!!"
So XYZ packs up the test victim devices and sends them to ABC.
ABC calls back in a week and says "These sound better! MUCH BETTER! We want you to only sell us chips that have been soaked in liquid nitrogen for 1000 hours!! But we can't pay a penny extra for that. In fact we think you should give us a discount, because we discovered this tremendous improvement and called it to your attention."
Hi Bill, any chance to listen yet?
Also hoping to hear from bear, and anyone else who may wish to participate.
And, please don't feel bad if anyone gets a negative result. We still want to hear about any and all results, as it's good data useful for understanding what is reasonable for most people to hear.
Markw4,
I would do these types of test if I had my system up and running but to do that on my laptop with cheap headphones seems rather silly. I highly doubt that the audio section in my HP laptop is up to a critical test and the cheap Sony headphone I have aren't something I use to listen to music really.
I would do these types of test if I had my system up and running but to do that on my laptop with cheap headphones seems rather silly. I highly doubt that the audio section in my HP laptop is up to a critical test and the cheap Sony headphone I have aren't something I use to listen to music really.
Understood. Thanks for your interest though, and hope you get your system up and running so you can enjoy it sooner rather than later.
Well, 'fraid I may be a no-show on your tambourine/not cow-bell test.
I've been scared off plugging in this computer into the DAC board by tales of dire driver hell. The board is an "import" that is not likely to be "current production" and it is screwed down too, so to see it I may have to either find old jpegs (possible) or unscrew it and re-photograph... it's nothing special, but it sounds rather good (with the opamp that is now in the output DIP-8 spot), surprisingly so.
If I post the pix, you can all have a very very good laugh!
(oh! just got a brilliant idea... I could go over to a fellow audio type friend's place, he runs JRiver and does higher bitrate stuff routinely, bring my dongle, and have a go at it... yeah, that might work... ok.)
I've been scared off plugging in this computer into the DAC board by tales of dire driver hell. The board is an "import" that is not likely to be "current production" and it is screwed down too, so to see it I may have to either find old jpegs (possible) or unscrew it and re-photograph... it's nothing special, but it sounds rather good (with the opamp that is now in the output DIP-8 spot), surprisingly so.
If I post the pix, you can all have a very very good laugh!
(oh! just got a brilliant idea... I could go over to a fellow audio type friend's place, he runs JRiver and does higher bitrate stuff routinely, bring my dongle, and have a go at it... yeah, that might work... ok.)
1) Taken literally there is no such test in anyone's normal Rel/Qual plan that I have ever heard of, functionality is not possible at that temperature.
2) The correct answer would be we will sent you 10 pieces of each unbranded and if you can sort them blind we will do it, otherwise you agree to pay more in advance.
3) There's probably a cheat that can tell if the encapsulant has indeed been taken to -196C vs -55C
Last edited:
I doubt it, the claims usually don't pay attention to process, etc. I give them as much credence as Burson's claim that you can plug their op-amp into a socket intended for a VFA or CFA and see an "improvement".
That is true economics has forced transfer from 4" to 6" and to 8" wafers and all new equipment in most cases. This is not a trivial process and the public does not get to see the cases where the first try is a total failure. The schematic changes would be rare though. Occasionally a part on a new process might, for instance, oscillate at -40C rather than -60C due to some subtle change in characteristics so a cap needs an extra .1pF.
As for staples I can't imagine more variety out there than with the 5534, it must exist in literally dozens of forms maybe even 50 or more.
I'm still restless on some sort of definitive Golden ear hardware AB test for op-amps.
I see a couple of references here on hardware fixtures.
Looking at Ian Hegguln's : SUPPLEMENT TO THE
LINEAR AUDIO CUBE AMP VOL 8 ARTICLE
Have been googleng for two days and can't find any live links to the thesis or any other add on filters. Checked the Diyaudio and AP site too.
Avoiding software/firmware, as have written it for years and would prefer a hardware add on.
Two ports, passive BNC in -> BNC out device.
Any links or suggestions for the items in Ian's article?
Thanks
I see a couple of references here on hardware fixtures.
Looking at Ian Hegguln's : SUPPLEMENT TO THE
LINEAR AUDIO CUBE AMP VOL 8 ARTICLE
Have been googleng for two days and can't find any live links to the thesis or any other add on filters. Checked the Diyaudio and AP site too.
Avoiding software/firmware, as have written it for years and would prefer a hardware add on.
Two ports, passive BNC in -> BNC out device.
Any links or suggestions for the items in Ian's article?
Thanks
Attachments
On the semis, kinda off-topic- Not to start a new discussion , just my thoughts.
I've read the 1950 "Electrons and Holes" book over and over, just can't get past the theory that holes travel at a different speed vs electrons. Schokley had some difficulty too as seen in the preface.
-bruce
I've read the 1950 "Electrons and Holes" book over and over, just can't get past the theory that holes travel at a different speed vs electrons. Schokley had some difficulty too as seen in the preface.
-bruce
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.