+- HP Forums (https://www.hpmuseum.org/forum)
+-- Forum: HP Calculators (and very old HP Computers) (/forum-3.html)
+--- Forum: General Forum (/forum-4.html)
+--- Thread: Saturn processor die-shot (/thread-18106.html)
Saturn processor die-shot - Gene Dorr - 03-07-2022 05:34 PM
A few months ago I finally got an HP-71B in terrible condition (I could never bring myself to do this to a working HP-71B), and took it apart to make photos of the guts similar to the ones in the July 1984 HP Journal. I've always been enamored of the Saturn processor, so I commissioned John McMaster to make a die shot of it. This is not the original 1LF2, but the second iteration 1LK7.
Here's the shot : https://siliconpr0n.org/archive/doku.php?id=mcmaster:hp:1lk7-0001
Enjoy!
Gene
RE: Saturn processor die-shot - Thomas Klemm - 03-07-2022 06:39 PM
This is so cool.
Thanks a lot!
PS: I like the umbrella.
RE: Saturn processor die-shot - Didier Lachieze - 03-07-2022 07:07 PM
Impressive picture !
RE: Saturn processor die-shot - TomC - 03-07-2022 07:51 PM
Gene:
Very Nice! Thank you for doing this.
I hope someone (with more patience than I) can label the pinouts. It would be interesting to know the function of each 'geographic' section.
(...on a somewhat related note, There are some interesting die photos/descriptions/animations of the 6502 online)
TomC
(03-07-2022 05:34 PM)Gene Dorr Wrote: A few months ago I finally got an HP-71B in terrible condition (I could never bring myself to do this to a working HP-71B), and took it apart to make photos of the guts similar to the ones in the July 1984 HP Journal. I've always been enamored of the Saturn processor, so I commissioned John McMaster to make a die shot of it. This is not the original 1LF2, but the second iteration 1LK7.
Here's the shot : https://siliconpr0n.org/archive/doku.php?id=mcmaster:hp:1lk7-0001
Enjoy!
Gene
RE: Saturn processor die-shot - Paul Berger (Canada) - 03-07-2022 10:00 PM
(03-07-2022 06:39 PM)Thomas Klemm Wrote: This is so cool.
Thanks a lot!
PS: I like the umbrella.
That would seem to be the logo for Corvallis a die I examined for an unknown chip has the umbrella with HP inside it.
Paul
RE: Saturn processor die-shot - Paul Berger (Canada) - 03-07-2022 10:07 PM
(03-07-2022 05:34 PM)Gene Dorr Wrote: A few months ago I finally got an HP-71B in terrible condition (I could never bring myself to do this to a working HP-71B), and took it apart to make photos of the guts similar to the ones in the July 1984 HP Journal. I've always been enamored of the Saturn processor, so I commissioned John McMaster to make a die shot of it. This is not the original 1LF2, but the second iteration 1LK7.
Here's the shot : https://siliconpr0n.org/archive/doku.php?id=mcmaster:hp:1lk7-0001
Enjoy!
Gene
What equipment was used to make the die shot? I have photographed chip dies using a camera with bellows and a microscope objective for a lens I have managed to achieve high magnification, filling a frame with a small area on a die, but lighting is tricky as is setting up you need to get the axis of the lens perpendicular to the die as depth of field is non-existent at that kind of magnification and if not perpendicular you cannot get the whole frame in focus.
Paul.
RE: Saturn processor die-shot - Gene Dorr - 03-07-2022 10:52 PM
I don't know the details, but I believe he shot multiple images, each of a small section of the die, and then stitched them together. If you download the full image and blow it up, there are a few areas where there seems to be slight discontinuities -- I think these are the "seams" where the separate images were put together.
RE: Saturn processor die-shot - KimH - 03-08-2022 08:04 AM
This is VERY cool, thanks for making it happen. And for the sacrifice of a chip!
I happen to have a Z9 printer and this will definitely be printed as a wall poster behind my TITAN showcase…..
(03-07-2022 05:34 PM)Gene Dorr Wrote: A few months ago I finally got an HP-71B in terrible condition (I could never bring myself to do this to a working HP-71B), and took it apart to make photos of the guts similar to the ones in the July 1984 HP Journal. I've always been enamored of the Saturn processor, so I commissioned John McMaster to make a die shot of it. This is not the original 1LF2, but the second iteration 1LK7.
Here's the shot : https://siliconpr0n.org/archive/doku.php?id=mcmaster:hp:1lk7-0001
Enjoy!
Gene
RE: Saturn processor die-shot - Gene Dorr - 07-03-2022 08:39 PM
(03-07-2022 10:07 PM)Paul Berger (Canada) Wrote: What equipment was used to make the die shot? I have photographed chip dies using a camera with bellows and a microscope objective for a lens I have managed to achieve high magnification, filling a frame with a small area on a die, but lighting is tricky as is setting up you need to get the axis of the lens perpendicular to the die as depth of field is non-existent at that kind of magnification and if not perpendicular you cannot get the whole frame in focus.
I forwarded your question to McMaster, and he (finally!) replied:
I use an Olympus BH2-UMA illuminator based system. Sounds like he is using a custom camera setup, so his mileage will vary depending on how well he built it. That said, it sounds like his problem is he's using darkfield illumination (ie oblique / from the side) whereas I'm using epi illumination (the light comes from the objective itself, hits the sample, bounces back into the objective). In short epi illumination uses a half silvered mirror between the objective and the camera to achieve this.
Gene
RE: Saturn processor die-shot - Wes Loewer - 07-04-2022 04:27 AM
(03-07-2022 06:39 PM)Thomas Klemm Wrote: PS: I like the umbrella.
So where is this umbrella? I can't seem to find it.
RE: Saturn processor die-shot - Didier Lachieze - 07-04-2022 04:39 AM
(07-04-2022 04:27 AM)Wes Loewer Wrote: So where is this umbrella? I can't seem to find it.
In the bottom left corner.
[attachment=10837]
RE: Saturn processor die-shot - Wes Loewer - 07-04-2022 05:10 AM
(07-04-2022 04:39 AM)Didier Lachieze Wrote:(07-04-2022 04:27 AM)Wes Loewer Wrote: So where is this umbrella? I can't seem to find it.In the bottom left corner.
Ah, thanks. I had looked at all the corners, but I was zoomed in too much.
RE: Saturn processor die-shot - Chr Yoko - 07-04-2022 06:35 AM
Super Nice picture !
This looks like a 47 Pins processor.
RE: Saturn processor die-shot - Gene Dorr - 07-04-2022 09:06 PM
Here's the IC sent to McMaster.
Gene
[attachment=10838]
RE: Saturn processor die-shot - pier4r - 07-07-2022 05:26 PM
Nice picture!
Can someone describe roughly the parts?
I would guess the very repetitive blobs (upper left, bottom left) is memory. Bottom right seem like registers (still very repetitive).
Upper right or central right should be were ALU and memory access functions are?
RE: Saturn processor die-shot - Thomas Klemm - 07-07-2022 07:10 PM
(07-07-2022 05:26 PM)pier4r Wrote: Can someone describe roughly the parts?
Something like this would be cool: Attempt to forcibly reverse a section of the HP35 ARC chip layout - full process record
RE: Saturn processor die-shot - Jonathan Busby - 11-17-2022 05:36 PM
I know as a *fact* that all the Yorke and Clarke SoCs ( and probably anything derived from them such as Bert ) use a PLA for at least most of the combinatorial logic in the embedded Saturn CPU, if not the rest of the chip -- I don't know if this is the case with the 1LK7. If it is the case then I'd expect to see a large regular grid for the AND plane and an abutting grid for the OR plane. In the die shot, the structure toward the upper left-hand corner looks like it may be a PLA.
On another note, if you're really into reverse engineering vintage ICs, I'd recommend a metallurgical microscope from AmScope as they're pretty affordable considering the alternatives and they are well suited to HP vintage ICs as even the Yorke SoC and Saturn were manufactured with a *HUGE* process technology ( we're talking something like 0.5 microns ).
Regards,
Jonathan
RE: Saturn processor die-shot - brouhaha - 11-17-2022 10:10 PM
(03-07-2022 05:34 PM)Gene Dorr Wrote: I commissioned John McMaster to make a die shot of it. This is not the original 1LF2, but the second iteration 1LK7.
Thanks very much for doing that, Gene!
I've had John decap and photo other chips for me, but hadn't yet had him do calculator chips. It's been on my to-do list.