Pictures of my case mods

464 views
Skip to first unread message

terry-...@glaver.org

unread,
May 1, 2024, 11:13:39 PM5/1/24
to PiDP-10
As I mentioned in an earlier post, the rear cover does not seat properly if you have soldered right-angle connectors in place for the MAX232, fans, power switch, etc.

I've had a steady stream of orders arriving from Digi-Key and Amazon as I experiment with a variety of modificiations.

At this point the case is physically complete and I'm happy with how it turned out. The Pi is running the emulator quite happily, but I still need to finish up the cabling for the serial ports and fans.

I'll provide a full write-up with parts lists, etc. once everything is together, fully cabled, and working.

Here's the system running blinky:

PXL_20240502_014657322-crop-s.jpg

And here's the back (yes, I know one of the fan grills is flipped, I'll fix it when I finish up the cabling). You can also see a pair of CH340 USB-to-serial adapters on the Pi. These will probably get black heat shrink tubing in the final assembly, rather than the clear that's on them now. They'll get cabled up to the MAX232 and then to the pair of DB25 connectors on the rear panel:

PXL_20240502_014317015-crop-s.jpg

Speaking of DB25 connectors, here they are. You'll note that they're male, as that's what DEC used.

PXL_20240502_014418178-crop-s.jpg

Lastly, here's a closeup of how I extended the rear panel. These are 6-32 x 1/4" hex spacers mounted in the holes that would normally use the self-tapping screws supplied with with kit. Since these are machine screw thread, I soaped up the threaded ends before screwing the spacers into the plastic posts in the case. Be careful to not overtighten. This allowed me to use metal hardware for mounting the fans, joining the various pieces of the back panel, and so on without needing to worry about clearance issues.

PXL_20240502_014352921-crop-s.jpg

oscarv

unread,
May 2, 2024, 10:30:09 AM5/2/24
to PiDP-10
Terry,

Brilliant!

>> These are 6-32 x 1/4" hex spacers mounted in the holes that would normally use the self-tapping screws supplied with with kit. Since these are machine screw thread, I soaped up the threaded ends before screwing the spacers into the plastic posts in the case. Be careful to not overtighten. This allowed me to use metal hardware for mounting the fans, joining the various pieces of the back panel, and so on without needing to worry about clearance issues.

Can you perhaps post some details on those spacers? They seem to be the solution for the (errr) design mistake where I did not leave enough clearance to solder in the fan power pin headers without hitting the back panel. Thank you!
If anyone wants to use the fan power headers, this looks like being the recommended approach.

Kind regards,

Oscar.

wjegr...@gmail.com

unread,
May 2, 2024, 12:16:32 PM5/2/24
to PiDP-10
Excellent. I have cooling fans, haven't installed them yet. My plan is to use the key switch to turn them on and off. I really don't want the noise all the time, since it's sitting on a bookcase right next to where I sit to watch TV and listen to music (streamed via another rPi, of course). But when people see it and ooh and aah, I want to turn on those puppies. Also, there is the issue with the decorative plate not really fitting if you have mounted the rPi cooler fan on the rPi card, which I did. I'll probably work up a 3D printed replacement with holes for the offending screw heads, but maybe a tiny standoff extension would be better, although it would be so small I'm not sure any exist. I'll post the stl file when I get around to it. Finally, I also thought about using the indirect USB power connection. I have the connectors. Has anyone bothered with this? I think it would look cleaner and also eliminate stress on the rPi connector.

oscarv

unread,
May 2, 2024, 2:00:19 PM5/2/24
to PiDP-10
The simpler option, that I've used almost all the time, is to buy 5V fans with a USB connector. Then you just plug them in to the Pi. See the web pages for some pointers, they're a few dollars on Aliexpress and the like. Not bad either.

For the cover panel on top of the Raspberry Pi - with the little cooler fan assembly clipped on, the solution is very simple. Drill two holes in the cover panel to give the mounts of the fan unit the space they need:

pi cover panel fix.jpg

(Picture has green panels - I'm still working with my last prototype... I too am waiting for a PiDP-10 kit as it happens!)

Kind regards,

Oscar.

DR

unread,
May 2, 2024, 7:49:14 PM5/2/24
to pid...@googlegroups.com
I've seen some general allusions to the price of a new PDP-10 from DEC
back in the day ranging from $100,000 to $2,000,000, which seems like
quite a range.

Some of the photos show quite the impressive systems, from a big rack
and the now familiar PiDP-10 console with a couple supporting racks
along side, to a longer one with a couple of tapes attached and many
cabinets.

There is much discussion about Oscar's dedication to reproducing the MIT
lab's machine so if one were to take that Institution's setup for an
example, where did that fall into the cost range, and what WERE to cost
ranges for the delivered machines?


I had a neighbor who worked for CRAY and he would follow a shipping
machine to the location and help (with many other technicians) with
install and initial certification (I guess it just doesn't work or not
work when the ON button is pushed), before the receiving University or
Weather modeling or air flow modeling group got the keys to it.


Labor alone for such would seem expensive, but he'd be gone for months,
it seemed.  Now incredible horsepower shows up in a small box from our
friendly UPS, DHL, FedEx drivers....


(Like I've heard,  wanting to go back to the 'good old days' doesn't
apply to medicine or dentistry.  You want the advancements. )

oscarv

unread,
May 2, 2024, 8:14:15 PM5/2/24
to PiDP-10
Dale,


On Friday, May 3, 2024 at 1:49:14 AM UTC+2 daleea...@gmail.com wrote:
 reproducing the MIT lab's machine so if one were to take that Institution's setup for an
example, where did that fall into the cost range, and what WERE to cost
ranges for the delivered machines?

At the far, far upper end I would imagine - given the huge Moby memories and all the peripherals.
Lars will know better, but the amount of disk storage that the stuffed-full ITS disk images offer, would have been unbelievable at the time.

We're all ultra-rich in Raspberry Pi land :-)

I sometimes look at photos of DEC factories. If you'd told DEC management in 1979 that one 'random dude' at his dinner table (as I was up till a year ago) would be able to outproduce DEC's 1970s factories, thanks to technological progress? 
That technology would progress so rapidly that that became not just possible but easy - would they have believed it? Because for the 8/I, 11/70 and soon KA10 it is literally true. One more expression of Moore's law I guess.
It is weird.

But wonderful. I'm having fun!

Kind regards,

Oscar.

Andy

unread,
May 2, 2024, 8:31:34 PM5/2/24
to PiDP-10
Do share DEC factory photos .. and I think we have just a few of your panama operation... some more wouldn't hurt😊

... the stuff you and your team are doing is very significant.  Just compare the activity on this forum compared to the 8 or 11....

oscarv

unread,
May 2, 2024, 9:27:04 PM5/2/24
to PiDP-10
>> Do share DEC factory photos .. and I think we have just a few of your panama operation... some more wouldn't hurt😊

Not the best pictures, but from the very nice 1968 Small Computer Handbook, where DEC talks a bit about their production (https://bitsavers.org/pdf/dec/pdp8/handbooks/SmallComputerHandbook_67-68.pdf)

Screenshot from 2024-05-03 03-05-59.pngScreenshot from 2024-05-03 03-07-27.png

Mind you, by 1969 DEC had multiple factories (but others here will know much more than I do).

But I do know this production line, AKA my dinner table. This table did over 10,000 PiDP-8 and PiDP-11s in 9 years. 
I do not maintain a clean desk policy in the living room, on work days :-)
 
Screenshot from 2024-05-03 03-10-42.png

But this is the new, er, Production Facility since last year:

...arrival of the PiDP-10 cases in February. That was a good day:

Screenshot from 2024-05-03 03-17-28.png

Jose will be angry, because he *does* maintain a clean desk policy. But this was my desk at the end of one day in Panama, not his fault:

 Screenshot from 2024-05-03 03-22-05.png

Hmm. Maybe more information than you asked for :-)

Kind regards,

Oscar.


oscarv

unread,
May 2, 2024, 9:41:23 PM5/2/24
to PiDP-10
One more photo won't hurt, this is the PiDP-10 with a PDP-10 at the RICM museum last month. Mind you, the original machine's switch colours have faded!

Screenshot from 2024-05-03 03-40-46.png

terry-...@glaver.org

unread,
May 2, 2024, 11:38:54 PM5/2/24
to PiDP-10
  Here's the parts list. You'll probably need to copy it to a mono-
spaced font to get the columns to line uo.

  The first is a spacer I use between the Pi and your back plate
(to avoid the fan pop-ins). You need to trim the nylon threaded
part (I put a metal nut on part-way, then snip it, then unscrew
the nut to re-form the threads if necessary.

  The second is the back panel spacer. It will probably only work
if the self-tapping screws have NEVER been installed. I recommend
putting soap on the threads and using a nut-driver to ensure it
goes in straight. Be careful to not over-tighten.

  The third is the matching screw for the back panel spacer, pro-
vided for completeness (some people may not have 1/4" long 6-32
hardware handy).

  The fourth is the DB-25 I used on the back panel.

  The fifth is the DB-9 version if people prefer those. Note that
both of these are male. Tweak the part number for female if you
want those connectors.

  The last item is the case fan I'm using, from Amazon. This is
a 2-pack.

Part #            Description              MFG                Digi-Key          Mouser

M2100-2545-SS        HEX STANDOFF M2.5X0.45 SS 5MM      RAF Electronic Hardware    1772-1543-ND          761-M2100-2545-SS
8717            HEX STANDOFF #6-32 ALUMINUM 1/4"  Keystone Electronics        36-8717-ND          534-8717
9306            MACH SCREW PAN HEAD SLOTTED 6-32  Keystone Electronics        36-9306-ND          534-9306
K85X-BD-25P-BRJ        CONN D-SUB PLUG 25POS VERT SLDR      Kycon, Inc.            2092-K85X-BD-25P-BRJ-ND      806-K85X-BD-25P-BRJ
K85X-ED-9P-BRJ        CONN D-SUB PLUG 9POS VERT SLDR      Kycon, Inc.            2092-K85X-ED-9P-BRJ-ND      806-K85X-ED-9P-BRJ


Fan:
GDSTIME 5Volt 120mm Fan, 1500RPM 120mm x 25mm DC Brushless Cooling Fan
https://www.amazon.com/dp/B07Y7XB6G2

On Thursday, May 2, 2024 at 10:30:09 AM UTC-4 oscarv wrote:

oscarv

unread,
May 3, 2024, 6:55:44 AM5/3/24
to PiDP-10
Terry,

Thank you! I will buy & try!


>>  The second is the back panel spacer. It will probably only work
>> if the self-tapping screws have NEVER been installed.

That worries me a bit... I will experiment.
Much appreciated!

Kind regards,

Oscar.

terry-...@glaver.org

unread,
May 3, 2024, 1:26:57 PM5/3/24
to PiDP-10
You might need to do something like glue the hex standoffs in. The issue is that the self-tapping screws provided in the kit enlarge the holes in the case as they cut their threads.

You can use a regular 6-32 screw and see if it can get a good "grip" in the first 1/4" or so. If that works, then the spaces should work. 8-32, which is the next SAE size up, is too big and might split the posts. Metric hardware uses a much finer thread pitch, so even if there's a metric size that fits, it probably won't get a good enough "bite" in the plastic to stay put.

Ideally, we'd use a hex standoff that was self-tapping on one end and 6-32 on the other. I looked on McMaster and Fastenal (the two big US vendors of oddball parts) and wasn't able to find that type of standoff.

They are available from specialty manufacturers, but without any prices listed (and are probably in the "if you have to ask, you can't afford it" range) and probably with very large minimum orders:
https://www.trfastenings.com/Products/Catalogue/Spacers-and-Pillars/Threaded-Pillars/Hexagon-Self-Tapping-Male-Female

Steven A. Falco

unread,
May 3, 2024, 1:42:43 PM5/3/24
to pid...@googlegroups.com
Might it be possible to drill and tap the case for 8-32 standoffs? I've done that with many 3-D printed projects, and PLA at least tolerates tapping.

Steve
> --
> You received this message because you are subscribed to the Google Groups "PiDP-10" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com <mailto:pidp-10+u...@googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-10/712ce361-c70c-4a97-8bea-b5fcee62801an%40googlegroups.com <https://groups.google.com/d/msgid/pidp-10/712ce361-c70c-4a97-8bea-b5fcee62801an%40googlegroups.com?utm_medium=email&utm_source=footer>.

John Peterson

unread,
May 3, 2024, 4:16:03 PM5/3/24
to DR, pid...@googlegroups.com
I worked at Apple in the late '80s, in the building that had their Cray X-MP. Once in the printer tray I saw the invoice for regular maintenance Cray charged Apple for the machine. It was just over $30,000 - each month.

--
You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-10/c4d4f505-8007-41f0-a6b2-f9c3da6ef5f7%40gmail.com.

oscarv

unread,
May 3, 2024, 7:14:47 PM5/3/24
to PiDP-10
These self-tapping screws are (apparently) quite special for the purpose of tapping into the types of plastics. So if the thread of your inserts is different, they'll eat another path into the plastic. So yes, then glue might be a good idea.

BTW - had to import these screws from a German company, the mold designer explicitly told me not to use cheap screws from 'anywhere' as they will dull the thread into the plastic much more quickly. 

Kind regards,

Oscar.

terry-...@glaver.org

unread,
May 3, 2024, 7:57:08 PM5/3/24
to PiDP-10
Yes, I can believe that you need special screws. Too small and they strip & spin, too large and they crack the molded post.

The ideal standoff would be one with the same thread as the existing case screws on the male end and a suitable metric or SAE thread on the female end. I think 3/16" would be perfect - I went with 1/4" just because of availability, but that leaves a slight gap between the back panel and the top of the case.

I mentioned this in a private email a while ago, but I'll repeat it here for the benefit of everyone else:

When re-installing a self-tapping screw, once you have the screw end in the plastic, do the counterintuitive thing of turning it counterclockwise (loosening) instead of tightening. When you feel the screw drop a little, that means it is sitting in the threads it cut previously. It will both tighten more easily and vastly reduce wear on the plastic from cutting new threads over and over each time you reinstall it. I learned this on an HP assembly line tour nearly 50 years ago (yikes!) but it doesn't seem to be common knowledge.

Glenn Babecki

unread,
May 3, 2024, 11:17:52 PM5/3/24
to PiDP-10
Terry,
Wow, someone else also knows this trick!  This works for most every screw, but extremely useful for self-tapping screws.

On Friday, May 3, 2024 at 7:57:08 PM UTC-4 terry-...@glaver.org wrote:
When re-installing a self-tapping screw, once you have the screw end in the plastic, do the counterintuitive thing of turning it counterclockwise (loosening) instead of tightening. When you feel the screw drop a little, that means it is sitting in the threads it cut previously. It will both tighten more easily and vastly reduce wear on the plastic from cutting new threads over and over each time you reinstall it. I learned this on an HP assembly line tour nearly 50 years ago (yikes!) but it doesn't seem to be common knowledge.


I don't recall how many decades ago I learned this trick from one of my uncles.  This particular uncle was the one that inspired me to go into electrical engineering.  He was employed at Hughes Aircraft and retired after his last role as system integrator for Hughes satellites that were launched by rocket and the shuttle.  Weird coincidence/fact was that he applied to be a mission specialist for shuttle satellite deployment.  Fortunately for him but unfortunately for Ronald McNair, McNair got the position as a mission specialist on the Space Shuttle Challenger mission STS-51-L; the last Challenger mission.

  Glenn

oscarv

unread,
May 4, 2024, 12:42:55 PM5/4/24
to PiDP-10
On Saturday, May 4, 2024 at 5:17:52 AM UTC+2 glenn....@gmail.com wrote:
Terry,
Wow, someone else also knows this trick!  This works for most every screw, but extremely useful for self-tapping screws.

It's one for the Building Instructions page.

Karl Schuneman

unread,
May 4, 2024, 2:38:43 PM5/4/24
to PiDP-10
Will heat sets work in these bosses?

Helium Phoenix

unread,
Sep 16, 2025, 4:55:42 PM (11 days ago) Sep 16
to PiDP-10
Terry & Oscar,

I picked up some of these spacers/screws from DigiKey to try, as I had a similar issue with mounting double-row right-angle pin connectors for the MAX232/onboard-serial connectors, along with other connectors.

I fortunately had not yet used the original included screws to mount the main panel to the housing.  The spacers worked like a charm.

I've almost got the unit complete.  Both fans, serial ports mostly wired, still waiting on a new pair of USB-TTL adapters to finish the rest.  Also have to decide about re-doing the resistors on the Pi Serial Console connector, as another thread on here mentioned a potential (not confirmed) issue with the voltage divider on it.  Very much looking forward to having it completed and running full time (once I find a place to enshrine it!)

-- Leonard

John Francini

unread,
Sep 25, 2025, 11:56:54 AM (2 days ago) Sep 25
to Helium Phoenix, PiDP-10
A question on the fans. I've been using Noctua fans on all my builds, and find them to be quite lovely and quiet. Are the fans necessary, or are they there just to provide some verisimilitude to a real KA-10? If so, maybe Noctua fans would be too quiet?

Thanks and regards,

John



--
John Francini <fran...@mac.com>
"It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts." -- Sherlock Holmes, A Scandal in Bohemia

--
You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.

Steven A. Falco

unread,
Sep 25, 2025, 1:32:44 PM (2 days ago) Sep 25
to pid...@googlegroups.com
Fans are not necessary.

But if you want it to sound like the real thing you'll need a massive number of very loud fans - at least that has been my experience working in machine rooms.

My desktop computer makes more than enough noise for me. :-)

Steve

On 9/25/25 11:56 AM, 'John Francini' via PiDP-10 wrote:
> A question on the fans. I've been using Noctua fans on all my builds, and find them to be quite lovely and quiet. Are the fans necessary, or are they there just to provide some verisimilitude to a real KA-10? If so, maybe Noctua fans would be /too /quiet?
>
> Thanks and regards,
>
> John
>
>
>
> --
> John Francini <fran...@mac.com>
> "It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts." -- Sherlock Holmes, /A Scandal in Bohemia/
>> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com <mailto:pidp-10+u...@googlegroups.com>.
>> To view this discussion visit https://groups.google.com/d/msgid/pidp-10/127dc017-6f58-474e-bd19-77cad681323en%40googlegroups.com <https://groups.google.com/d/msgid/pidp-10/127dc017-6f58-474e-bd19-77cad681323en%40googlegroups.com?utm_medium=email&utm_source=footer>.
>
> --
> You received this message because you are subscribed to the Google Groups "PiDP-10" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com <mailto:pidp-10+u...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pidp-10/CE322C5E-3B14-41FA-82F4-E054B0B803CD%40mac.com <https://groups.google.com/d/msgid/pidp-10/CE322C5E-3B14-41FA-82F4-E054B0B803CD%40mac.com?utm_medium=email&utm_source=footer>.

Adam Morris

unread,
Sep 25, 2025, 1:38:16 PM (2 days ago) Sep 25
to PiDP-10
The fans are there to provide soundscaping not cooling.  You do not need the fans if you want silence.

So, yes the noctua fans would probably not be something you would want to add.  Either no fans, or cheap fans if you want it to sound like a big noisy computer.

Adam

John Francini

unread,
Sep 25, 2025, 1:48:21 PM (2 days ago) Sep 25
to Adam Morris, PiDP-10
... and they'd need to be big, noisy AC fans - ISTR that there was a substantial 60-120-240 Hz component to the fan noise...  but I could be wrong, as I failed to check directly when I was in the LCM machine room in 2019, soaking up the sights and sounds while surrounded by KLs, KI, KA, and KS systems.  It just sounded lovely.

Thanks and regards,

John

--
John Francini <fran...@mac.com>
"It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts." -- Sherlock Holmes, A Scandal in Bohemia
--
You received this message because you are subscribed to the Google Groups "PiDP-10" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.

Richard Cini

unread,
Sep 25, 2025, 1:57:14 PM (2 days ago) Sep 25
to Steven A. Falco, pid...@googlegroups.com
I have a 6’ DEC rack with two 8” (I think) AC fans. Makes quite a racket but it’s a great sound. 

Rich


To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-10/c9860e82-1161-4afb-8545-c5c4963c9f34%40gmail.com.
Message has been deleted

Malcolm Ray

unread,
Sep 25, 2025, 2:48:16 PM (2 days ago) Sep 25
to pid...@googlegroups.com
I haven't bothered with fans in my PiDP-10. But if I wanted that
authentic data centre sound, I think it would be better to add an
amplifier and loudspeaker. Why should the PiDP-1 get all the audio fun?

Bradford Miller

unread,
Sep 25, 2025, 6:18:39 PM (2 days ago) Sep 25
to Malcolm Ray, PiDP-10
Agreed. One needs to hear the constant hum of power conditioners, Liebert AC units, line printers, ttys, tape drives, drum seeks, possibly paper tape punching and multiple mechanical keyboard noises in addition to the fans to feel “at home” :-). Surprisingly, some have recorded such sounds and posted them to youtube. Or maybe unsurprisingly: I recall “escaping” to the machine room on numerous occasions to avoid the chatter of grad students in the office next door to mine...
> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-10+u...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/pidp-10/92de52541fa01bddc7abfff687d157a0925b761e.camel%40apathetic.org.uk.

pbi...@gmail.com

unread,
Sep 26, 2025, 3:34:15 AM (yesterday) Sep 26
to Richard Cini, Steven A. Falco, pid...@googlegroups.com

If you meant the fans in the top of the rack they aren’t piddlin’ little 8” fans, each is 10” by 4” deep.  Serious stuff :-}.

 

Example: https://www.rfparts.com/cl2l2-comair-rotron-caravez-axial-fan-1650-rpm-550-cfm-1-5-amps-61-watts-115-vac-60-hz-used.html

terry-...@glaver.org

unread,
Sep 26, 2025, 5:17:00 AM (yesterday) Sep 26
to PiDP-10
On Friday, September 26, 2025 at 3:34:15 AM UTC-4 pbi...@gmail.com wrote:

If you meant the fans in the top of the rack they aren’t piddlin’ little 8” fans, each is 10” by 4” deep.  Serious stuff :-}.

DEC certainly did know how to handle cooling. Once they started outsourcing power supplies, things (both power and cooling) weren't as good.

H960 cabinets had big round fans at the top that moved a lot of air. Many BA11 chassis used a pair of big fans, side-by-side. Those didn't have round housings, but were built with narrowed sides so two could fit side-by-side in the width available.

Back in the day, I replaced all of the fans in several PDP-11/70s with new Torin TA450 fans with ball bearings. Powering up was a magnificent sound as all the fans spun up with identical speed / sound. I forget how many fans that was - 16, I think. It beat changing them one at a time as they failed.

On my PiDP-1, I have two fans controlled by the rear panel key switch, but that's nowhere realistic enough. So I bought this USB audio adapter: https://www.amazon.com/dp/B0CYTB9SWW and my plan is to play digitized "computer room" audio via the Pi 5 running the PiDP-10 into an amplifier / speakers. One thing that's helpful to know - if you plug that adapter into the Pi, no new hardware shows up. It only enumerates once a 3.5mm audio cable is plugged into its output. If anyone else wants to experiment with this, maybe we should start a collection of various computer room audio files. We'll definitely need one with the VT5x "feeper" sound, which the Hacker's Dictionary refers to as 'The "feeper" on a VT-52 has been compared to the sound of a '52 Chevy stripping its gears.'

Ric Werme

unread,
Sep 26, 2025, 10:38:08 AM (yesterday) Sep 26
to PiDP-10
The PiDP-10 with or without fans does a horrible job of emulating the sound of a KA10 power-on event.  The CPU bays each have a bank of squirrel cage blowers at the floor level air intakes.  When power was first applied I was always startled.  As the blowers spun up, they emitted a sound more like jet engines spinning up (but quicker).  Before my flinch ended, that sound faded and a pure white noise rush took over.  I've never quite understood the first sound, I suspect that as the blowers came up to speed the back pressure reduced the siren-like effect of blower blades cutting through the air.

My PiDP-10 currently has a Pi-4 and no fan.

Ric Werme

unread,
Sep 26, 2025, 10:41:52 AM (yesterday) Sep 26
to PiDP-10
"One needs to hear the constant hum of ... drum seeks, "

The Byrant drums I'm familiar with (DEC RM10-B and Bailey Meter 756) have fixed heads, one per track.  They had zero seek time (and zero sound).

Whit Turner

unread,
Sep 26, 2025, 11:23:31 AM (24 hours ago) Sep 26
to Ric Werme, PiDP-10
There was also the RD10 (Burroughs). The only noticeable sound (other than a low rumble and the air compressor) was when the heads crashed :-).

Whit

Ric Werme

unread,
Sep 26, 2025, 11:41:16 AM (24 hours ago) Sep 26
to PiDP-10
The RD10 is a fixed head disk with a very large platter.  At DEC's Mill they learned to position the drive with the axis perpendicular to the motion of the building so that when a delivery truck backed into the loading dock the impluse wouldn't trigger a head crash!

Adam Morris

unread,
Sep 26, 2025, 1:04:55 PM (22 hours ago) Sep 26
to pid...@googlegroups.com
IBM P series always seemed to do the same thing.  In their case the blowers start at full speed and as the hypervisor comes online it throttles them back to a much quieter speed.  I always assumed that they were engineered so that one blower could provide sufficient air flow for a fully laden chassis in the event of the other blowers failing.

DR

unread,
Sep 26, 2025, 1:38:53 PM (22 hours ago) Sep 26
to pid...@googlegroups.com
I never heard on of them crash a head, but I will bet it cannot come
close the sound that a Univac FastRand made when it crashed.

We had one hooked to the 1108 at the U of Wis computing site in
Madison.  I was the only operator on duty when there was a slight power
bump (at least the lights flickered) and I cannot recall how the power
to the computer was furnished.  Anyway the screeching was far worse than
a multi-car pile up and things came to a rapid pause.  After they
brought in another one, a second one was sort of in backup mode but I
was told they were very expensive.

sunnyboy010101

unread,
Sep 26, 2025, 1:44:09 PM (22 hours ago) Sep 26
to PiDP-10
Yea, little 5v fans just don't provide the "right" sound...

I have stored in a box in the attic a Sun V440 4U server that has fans that sound like a 747 just before takeoff when they run. At least under Solaris v8 they didn't throttle, so always ran flat-out when I was using the server. It sat in a basement surrounded by acoustic foam panels and still could be heard upstairs.

Even my 1U Intel Xeon servers have that "jet engine" sound when the power is first applied. Fortunately they quickly throttle down thanks to temp. sensors and an OS (linux) that understands both fans and temp. sensors. But occasionally something glitches and the fans revert to full power which only a full power cycle can fix.

As for my PiDP10, it sits on a bookshelf in my office, where I'd much prefer quiet to authentic.

Besides, for full authenticity you need the raised floor with bone chilling air conditioning plus the sound of many loud printers and several noisy disk packs and 9-track tape drives. No thanks. :-D
-R

sunnyboy010101

unread,
Sep 26, 2025, 1:45:46 PM (22 hours ago) Sep 26
to PiDP-10
I have a crashed multi-platter disk pack from a Harris computer circa 1981. Nice oxide layers with one circle of shiny aluminum on several platters. Big heavy thing.

terry-...@glaver.org

unread,
1:53 AM (9 hours ago) 1:53 AM
to PiDP-10
On Friday, September 26, 2025 at 1:45:46 PM UTC-4 sunnybo...@gmail.com wrote:
I have a crashed multi-platter disk pack from a Harris computer circa 1981. Nice oxide layers with one circle of shiny aluminum on several platters. Big heavy thing.

Back in the day, it was common to turn these into clocks, as (analog) clock movements of the correct size were available inexpensively.

[In]famously, a user at SPC asked "Where's my home directory?" after a disk crash, and I handed him one of the platters from the crashed drive (while the backup was being restored). 
Reply all
Reply to author
Forward
0 new messages