You have it backwards, the failed parts which are all noisy and hence have no value were accepted for credit. What about that is unfair? It was stated that 20% failed so the typical part still passed.
As someone else said in the end it is probably better for everyone for him to sell them on the secondary market since we can not sell the returns as typical product.
Last edited:
Price doesn't matter. I worked for two years with single die which were 220 dollars per chip. The cost of the chip is irrelevant.
Spec, or typical as shown on a graph?
Again, there is a significant difference between a graph of a typical part, and a specification. Did they have a hard specification they guaranteed?
jn
Burning Amp
There will be schematics in my presentation going back 60yr. Several discrete op-amps at least 40yr. old if anyone here is interested. It is surprising how early JFET's were introduced as the input device of choice and how often 2N2222's, 2N3906's, 2N2907's and 2N3904's were the devices of choice. Almost universally 2N5912's were the FET's of choice.
There will be schematics in my presentation going back 60yr. Several discrete op-amps at least 40yr. old if anyone here is interested. It is surprising how early JFET's were introduced as the input device of choice and how often 2N2222's, 2N3906's, 2N2907's and 2N3904's were the devices of choice. Almost universally 2N5912's were the FET's of choice.
Last edited:
The choice of devices for discrete gain stages that you found is not surprising.
We used the TO-18 packaged 2222 and 2907, as well as their larger equivalents, the 2N2219 and the 2N2905 in 1967 at the Ampex Instrumention Division. We also used them at the Ampex audio department in 1968. These were pretty good complements, reasonably fast, and usable over quite a current range. Their fundamental weaknesses are relatively low beta, and relatively low first breakdown voltage. Even Ampex Audio had higher beta devices specified for input service. We used a version of the 2N5089 for this.
The problem with this is that these devices can be considered 'boilerplate devices' general purpose, but nothing exceptional. Often these devices had evolved EVEN TODAY in simulation programs, yet they are really 'obsolete' for SOTA designs.
We used the TO-18 packaged 2222 and 2907, as well as their larger equivalents, the 2N2219 and the 2N2905 in 1967 at the Ampex Instrumention Division. We also used them at the Ampex audio department in 1968. These were pretty good complements, reasonably fast, and usable over quite a current range. Their fundamental weaknesses are relatively low beta, and relatively low first breakdown voltage. Even Ampex Audio had higher beta devices specified for input service. We used a version of the 2N5089 for this.
The problem with this is that these devices can be considered 'boilerplate devices' general purpose, but nothing exceptional. Often these devices had evolved EVEN TODAY in simulation programs, yet they are really 'obsolete' for SOTA designs.
I've asked Mark to post it in the articles section even though it is just a lot of slides mostly. I am trying to annotate them enough so you don't have to be there. A large section of these schematics were a gift from a friend and represent years of reverse engineering, frequently having all component values. They are all technically obsolete and no longer have any IP to worry about.
I will probably convert it to a PDF and add some text sections in any case.
scott:
thanks. sounds interesting and i look forward to reading it.
now, if someone would do the same for classic test gear, i'd be a happier camper.
mlloyd1
thanks. sounds interesting and i look forward to reading it.
now, if someone would do the same for classic test gear, i'd be a happier camper.
mlloyd1
scott:
thanks. sounds interesting and i look forward to reading it.
now, if someone would do the same for classic test gear, i'd be a happier camper.
mlloyd1
I could pdf the test circuits from an IMPACT II semiconductor test platform from the early 80's...
I doubt anyone's interested in that however. It used a PDP-11 computer with this 12 inch diameter magnetic disc..
jn
12" HDD, stuff
I unfortunately had some experience with really large HDDs in the late 1980's. We (Concept Design) had developed the DAAD (Digital-Analog Analog-Duplicator), weighing in at around 600 lbs and incorporating 2 12" platter HDDs each storing 1GB. At the time this was a HUGE HDD (both size and storage capacity), but it was needed to store 90 minutes of cassette master sampled at 16-bit, 44.1kHz. When it was working it was infinitely better than a 1/2" tape loop bin in all respects. My role in that project was the analog output filter/line amp design.
However, when those HDDs crashed, the noise would make you jump out of your skin! We ended up rebuilding them ourselves in our clean room, which is yet another story.
The analog output to the duplicating slaves was from a 16-bit hybrid military D>As costing over $1500 each, because they were guaranteed to be monotonic at 4MHz, a speed required because we ran a 64:1 dupe ratio. This meant we would load the disc at 1-4x (source speed limited) and play back at 64x. Crazy stuff!
I was so glad when RAM cost came down to the point where we could get rid of those huge HDDs...what a PITA!!! The memory upgrade made the DAAD units pretty reliable.
OK, back to bickering.
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC Chapel Hill, NC
WXYC Chapel Hill, North Carolina - 89.3 FM
1st on the internet
I unfortunately had some experience with really large HDDs in the late 1980's. We (Concept Design) had developed the DAAD (Digital-Analog Analog-Duplicator), weighing in at around 600 lbs and incorporating 2 12" platter HDDs each storing 1GB. At the time this was a HUGE HDD (both size and storage capacity), but it was needed to store 90 minutes of cassette master sampled at 16-bit, 44.1kHz. When it was working it was infinitely better than a 1/2" tape loop bin in all respects. My role in that project was the analog output filter/line amp design.
However, when those HDDs crashed, the noise would make you jump out of your skin! We ended up rebuilding them ourselves in our clean room, which is yet another story.
The analog output to the duplicating slaves was from a 16-bit hybrid military D>As costing over $1500 each, because they were guaranteed to be monotonic at 4MHz, a speed required because we ran a 64:1 dupe ratio. This meant we would load the disc at 1-4x (source speed limited) and play back at 64x. Crazy stuff!
I was so glad when RAM cost came down to the point where we could get rid of those huge HDDs...what a PITA!!! The memory upgrade made the DAAD units pretty reliable.
OK, back to bickering.
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC Chapel Hill, NC
WXYC Chapel Hill, North Carolina - 89.3 FM
1st on the internet
Attachments
Attachments
![WinchesterDiskPack.big[1].jpg](/community/data/attachments/354/354581-0cbfa2549b5cee1bf6d8255df5a75f12.jpg)
Charles,
Apologies if this has already come up, but I guess you did check into the OPA860?
Jan
Big ...??? Anything less than the size of a washing machine, with single disk, isn't in the game ...
- Status
- Not open for further replies.