Looking for a DCJ11-AE CPU

[Paulo Rebordão]

I know this isn't the most appropriate forum, but I'm looking to buy this chip and can't seem to find one that: A) reasonably priced, B) not coming from China. 

Any help appreciated

Paulo Rebordão

[Henry Bent]

On Sun, 8 Jun 2025 at 04:49, Paulo Rebordão <pjreb...@gmail.com> wrote:

I know this isn't the most appropriate forum, but I'm looking to buy this chip and can't seem to find one that: A) reasonably priced, B) not coming from China. 

What, if I might ask, is the problem with buying a chip from China?

-Henry 

[Henry Bent]

Also there's one here in the US for what I would consider a reasonable price. https://www.ebay.com/itm/115303919787?_skw=DCJ11-AE

They're collector's items at this point, and it's a seller's market when you have one new in a box.

-Henry 

[Bob Darlington]

I have a couple new in box, with absolutely no sign that they were ever pushed into a socket.   I bought them from China for roughly $50 each (shipped) to the USA.  They come and go when it comes to availability.   What I have matches the ad you included in your email.  18MHz units.   Whatcha got planned?  I am planning on building some computers around them.

Have fun,
Bob

--
You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/CA%2BgrvcrUaOU5RLSDidGJhXQ%2Bx15%2BeZ40_0nZuStjYj_ys7snZA%40mail.gmail.com.

[John H. Reinhardt]

Most you find on Ebay will be the -09 variant (see chart below), but be aware when you purchase to make sure.  There is one place selling the -04.  Not bad in itself but it may not be the most recent version.  Keyways (the link provided by Harry Bent) has NOS. sealed in OEM packaging like the picture from Bob Darlington.  Their stuff is good and I've bought a number of CPUs from them to plug into M8190 boards with defective or missing CPUs.  They are among the latest date codes (98xx) being chips manufactured by Harris Semi for DEC.  I , along with a number of others on the VCFED.ORG DEC forum have bought DCJ11's from zoe_0112, a seller from China.  They are the same OEM packaged and the couple I bought have worked and tested good in M8190 boards.  Currently the store is empty so I don't know if they ran out of chips or are taking a break in the current uncertain tariff environment.  Previously they had sold over 300 chips.  In searching I see a number of new China sellers.  I have no experience with them, nor do I know of anybody that does.  They may be just as good as zoe_0112, but I can't say.

If you know you are going to use more than one, this Keyways listing is a slightly better deal - 2 for $189
https://www.ebay.com/itm/145462544440

Of course, if you expect to use A LOT of them, then you can't beat this - 100 for $7999.99  :D
https://www.ebay.com/itm/111476137790

DCJ11 Variations
Model No.	Part No.	Speed	Notes
57-19400-04	DCJ11-AC	15 MHz
57-19400-07	DCJ11-AA	20 MHz	DECserver 550 CPU (clocked at 18 MHz)
57-19400-08	DCJ11-AC	15 MHz
57-19400-09	DCJ11-AE	18 MHz	M8190-AD/AE, M8981

--

John H. Reinhardt

[John Hudak]

+1 on the Keyways recommendation.  I've known Mitch for a long time and the few things I've gotten from them have been good. I've found their prices to be relatively in line with the market but sometimes on the high side. YMMV

--
You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pidp-11/51367839-9044-4fde-8758-d0723f54a201%40thereinhardts.org.

[David Bakin]

I'm extremely interested in this as I have two DCJ11-AE , -09 variant (similar original boxes to Bob's above, except labelled "DA80100000" instead of "DA80300000").  Also an FPJ11 except that isn't in a box, it's just stuck into some insulating foam.

I would love to build a PDP-11 to go with my PiDP-11 console BUT I'm scared to touch it: I have little electronics circuit design experience (I'm a kit builder) and these chips use some MOS variant that is pre-CMOS technology, I believe, and I think has different "rules", including rules about power voltage supply sequencing.  So I'm afraid of burning the things up trying to bring them up by trial and error.

So, very interested, especially for

* tips on how to proceed to experiment with this chip without (much) risk of frying it

* actual circuits!!

I HAVE A DREAM, and that dream is a GEN-U-INE PDP-11 maxed out with RAM with an FPGA acting as QBus  <-> peripheral bus adapter, with standard DEC QBUS interface programming for some disk adapter with an SD card for media or maybe just a flash chip, also for networking (e.g., DEUNA or something to standard ethernet), IOW, emulate actual adapters so PDP-11 operating systems boot up, also the floating point chip.  

-- David Bakin

P.S. My NEXT GLORIOUS DREAM after getting my DCJ11/FP11 real PDP-11 going is to do something similar with my PiDP-10 - an FPGA implementation of a KA-10 (a -20 with its memory architecture is out of the question, I'm sure ...)

To view this discussion visit https://groups.google.com/d/msgid/pidp-11/CAE%3DviVos9CQGo6nPybUtgjwPzOiEAzBcFdbTGKn57wwHEKBi%2Bg%40mail.gmail.com.

[Ed Porter]

On 08/06/2025 20:40, David Bakin wrote:

> I would love to build a PDP-11 to go with my PiDP-11 console BUT I'm
> scared to touch it: I have little electronics circuit design experience
> (I'm a kit builder) and these chips use some MOS variant that is pre-
> CMOS technology, I believe, and I think has different "rules",
> including rules about power voltage supply sequencing.  So I'm afraid of
> burning the things up trying to bring them up by trial and error.
>
> So, very interested, especially for
>
> * tips on how to proceed to experiment with this chip without (much)
> risk of frying it
> * actual circuits!!
>

You could start here https://www.chronworks.com/J11/ or here
https://www.5volts.ch/pages/pdp11hack/ . I suspect you'll have a hard
time making them work with the PiDP-11 lights, though.

I too have a couple of DCJ-11s (from Keyways, ebay link in earlier
reply) and I really should get round to unwrapping them and doing
something similar.

-ed

[Jon Crowell]

It's a major design project to take a J11 chip and design a working computer system.   The last design in DEC was the KDJ11-D/S (M7554) for the PDP 11/53 system.  The CPU was q-bus based but takes a few custom gate arrays and other logic to create a working memory interface, required UARTS , TOY clock, Interrupt handler and a QBUS interface.   The chip was J11 was a very old design and fairly primitive by todays standards.  It has a list of quirky things that aren't documented.    Your better off buying some old Qbus parts and building a real working DEC bootable system.  

https://bitsavers.trailing-edge.com/pdf/dec/pdp11/1173/EK-KDJ1D-UG_KDJ11-D_May87.pdf   

The above has a good overview of the 11/53 processor and shows the hardware block diagrams.   It had everything you needed to boot except a disk controller.  We shipped this with the RQDX3 (MSCP) and RD31/32, RD53/54 disks.     The CPU has the console and memory on the full sized qbus module.

Jon

[Ed Porter]

On 08/06/2025 21:52, 'Ed Porter' via [PiDP-11] wrote:
> On 08/06/2025 20:40, David Bakin wrote:
>
>> So, very interested, especially for
>>
>> * tips on how to proceed to experiment with this chip without (much)
>> risk of frying it
>> * actual circuits!!
>>
> You could start here https://www.chronworks.com/J11/ or here https://
> www.5volts.ch/pages/pdp11hack/ . I suspect you'll have a hard time
> making them work with the PiDP-11 lights, though.

And the link I *meant* to post...

https://web.archive.org/web/20160908152649/http://www.cs.ubc.ca/~hilpert/e/pdp11hack/index.html

-ed

[Johnny Billquist]

I have the DCJ11 Microprocessor Users Guide, which have the information
about all the pins, the power requirements, the power up sequencing,
clocks, and all other stuff.
I can't remember where I got it from, but it should be possible to
locate online.

EK-DJC11-UG-PRE. October 1983. It's a preliminary version, but I think
it should be good enough to design hardware around.

And as people mentioned, it's not CMOS. Not sure I'd call it pre-CMOS.
It's NMOS. CMOS did also exist back then, but wasn't the technology
chosen. But basically, it's just fed +5V on two pins, and then you have
clock, address/data and various control pins. I haven't designed
anything with it, but it don't look too bad. But yes, you should
obviously always be careful about static electricity on any pins.

Johnny

--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol

[John Hudak]

With all due respect, hardware design from scratch with a J-11 chip is  more than a bit challenging.  There are a few major hurdles: 1) finding and using the IC's of the day which makes interfacing the J11 to the the rest of the world.  Requires a knowledge Standard TTL, Fast TTL, Schottky TTL, High power TTL, Low power TTL & Advanced Schottky TTL( AS-TTL) and knowing what can be used and what can be substituted. In addition, you need to have a good working knowledge of positive and negative logic, gate minimization, and tradeoffs with timing among signals.   2) Assuming you can determine the proper signals and timing to other major building block chips, there is the issue of using PALs, or CPLD's to efficiently provide the required signals with out using a slew of TTL gates/iC. 3) interfacing 5VDC logic devices to 3.3 VDC logic devices.  Assuming the necessary TTL devices are NLA, and you manage to develop a subset of logic implemented in CPLDs, there may be the case that the only devices around to implement the rest of the design are 3.3VDC devices.  If you never had done any major hardware design using 68K family chips or i486 family of chips, you are in for a very steep learning curve.  It is just not about gluing the large functional block chips with wires.  I am an experienced cpu board level designer and implementer starting with 8-bit 6502's through i486, amd2900 bit slice devices into ARM 64 devices. I don't want to rain on your parade but you may be better off looking at implementing a PDP11 system at the board level.  Given DECs 'quarks' in various qbus backplanes, and board level compatibility, it will be challenging enough to  get all of that right.  

IIRC, acquiring the cpu board from a DEC server that contains the J11 and is essentially the basis for 11/53, can be converted (via a different set of ROM images)  into a 11/53 workalike when the correct complement of Qbus boards are attached to it., e.g disk controller, memory expansion, SLUs, etc.  

Somewhere in my files I have the details but haven't looked at it in over 10+ years.

good luck!

[Anton Lavrentiev]

> hardware design from scratch with a J-11 chip is  more than a bit challenging

FWIF, there's "PDP-11/Hack de Luxe" using this CPU, which can be used as a guide:

https://forum.vcfed.org/index.php?threads/pdp-11-hack-de-luxe.1252532

Based on (an earlier version of) the PDP-11/Hack project

To view this discussion visit https://groups.google.com/d/msgid/pidp-11/CAE%3DviVpUBz0Y7sWTaqP0j-Ax2BnVg0LNzzCy0%3D0cxvkEqGu2MQ%40mail.gmail.com.

[Sytse van Slooten]

That manual is also on Bitsavers, same as all the other docs you'd need to build a system based on J11.

Let's not make this into some kind of rocket science or mythology, there's nothing there that is out of reach if you're willing to spend a couple of years of your life.

But also, be clear that it'll take way, way more than just a couple of weeks or a few hours every other Thursday, and most of those hours will be spent on issues caused by not properly reading the documentation that is freely available. And what you'll end up with is the same as someone else has done before. Several times over.

Feel free not to read this :-)

> --
> You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.

> To view this discussion visit https://groups.google.com/d/msgid/pidp-11/81430f41-2b00-4314-b511-3178fb3f4d2e%40softjar.se.

[Johnny Billquist]

I was searching for it on bitsavers, but couldn't find it. Where is it?

Also, for sure, you don't cobble together a board in a few hours, but if
you have some experience in designing boards, it's not too bad. One
thing though is that you really also need to implement a DL11, since the
MicroODT of the CPU will expect to interact with it.

Johnny

[Damian Wildie]

Worth a look:  http://s100computers.com/My%20System%20Pages/PDP11%20V2%20Board/PDP11%20Board%20V2.htm

[Jon Crowell]

The J11 was a messy project at DEC...  It was driven by Bob Supnik and later Dan Casaletto as DEC was building it's internal semiconductor capibility.   Here are the internal specs for the Data Chip and Control Chip.  The chips were designed and fabricated by Harris later became intersil.  There was a skunkworks project in DEC to lift the J11 design clean it up and drop it in a new simple package in the late 80's but it never got finished.  

http://www.bitsavers.org/pdf/dec/pdp11/j11/J-11_Data_Chip_Specification_Jul82.pdf

http://www.bitsavers.org/pdf/dec/pdp11/j11/J-11_Control_Chip_Specification_Jun82.pdf

Jon

[Paulo Rebordão]

Thanks for the tips. I'll definitely buy one from Keyways.

It will go into my M8190 board.

[terry-...@glaver.org]

On Sunday, June 8, 2025 at 8:19:21 PM UTC-4 Jon Crowell wrote:

There was a skunkworks project in DEC to lift the J11 design clean it up and drop it in a new simple package in the late 80's but it never got finished.  

I'd be interested in hearing more about this. You're either using the existing J-11 microcode (hopefully with optional bugfixes that can be enabled/disabled), which means you're pretty much constrained to the control and data chip design, which is where a lot of the mess is, or you're throwing out the microcode which means you're essentially designing another PDP-11 processor. The latter isn't impossible - QED did several, Mentec eventually did several, and so on. But they all ran on (possibly customized) merchant silicon, not from-scratch designs.

Bob Supnik states that that the J-11 was intended to be a capstone for the PDP-11 processor family. It nearly became a tombstone instead. At one point parts of DEC were using me as a back-channel to Harris because they couldn't get answers from the inverted "U" of messaging and management - information was being omitted / changed from Harris engineering -> Harris management -> DEC management -> DEC engineering.

It also didn't help that DEC couldn't decide if the J-11 was for their own systems only, their own systems plus certain vendors that didn't compete with them directly (Atex, for example), or something they wanted to sell to anyone. This led to the J-11 marketing manager (at the time a very nice lady named - IIRC - Cathy Berida) having to call everyone who placed an order for a DCJ11 - it didn't matter if it was 10,000 pieces or just one - to ask them what they were going to do with it.

The Harris-marked 98xx datecode CPUs were "contractual obligation" CPUs for Mentec and other customers with long-term supply commitments, because DEC wanted to get rid of that before the sale to Compaq.

[Lawrence Fisher (RealTimeCat)]

I was working for Plessey Peripheral Systems in Irvine, CA, in the mid 80s. We had started collaborating with Mentec on the J11-based CPU/Memory/Serial Q-Bus cards. Shortly after that DEC pulled the J11's from the OEMmarket. I heard a rumor that an RIP J11 note card was sent to one of our corporate-level managers from someone inside Digital at the time.

[frankg]

Maybe there is hope for the J-11 in a PIDP-11. FPGA brings UNIX V1 to the DEC J-11

[timr...@gmail.com]

That would be cool if I could upgrade my J-11 in my Pro-350.  I wonder if that would work?

[terri-...@glaver.org]

I'm pretty sure the 325 and 350 both used F-11 (11/23) CPUs, while only the 380 used the J-11. I'm not sure if an adapter board could be created or not (complicated by the fact that the MMU is an add-on chip in the F-11 and integrated in the J-11). It would certainly be an interesting exercise. 

[Tim Radde]

Yeah, my bad.  Was getting over excited so read it wrong.  Dang.

--
You received this message because you are subscribed to a topic in the Google Groups "[PiDP-11]" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pidp-11/VAZoNbEPJcU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/fa55afca-b177-42ed-beb2-c90327318345n%40googlegroups.com.

[Johnny Billquist]

Considering that the Pro-350 have an F11, that would be quite a feat...

Johnny

On 2025-10-06 01:49, timr...@gmail.com wrote:
> That would be cool if I could upgrade my J-11 in my Pro-350.  I wonder
> if that would work?
>
> On Sunday, October 5, 2025 at 12:28:53 PM UTC-5 frankg wrote:
>
> Maybe there is hope for the J-11 in a PIDP-11. FPGA brings UNIX V1

> to the DEC J-11 <https://hackaday.com/2025/08/30/fpga-brings-unix-
> v1-to-the-dec-j-11/>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com

> <mailto:pidp-11+u...@googlegroups.com>.

> To view this discussion visit https://groups.google.com/d/msgid/pidp-11/

> c78bf90b-aef0-4ebc-8eb6-ea6a2cfc1029n%40googlegroups.com <https://
> groups.google.com/d/msgid/pidp-11/c78bf90b-aef0-4ebc-8eb6-
> ea6a2cfc1029n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Digby R.S. Tarvin]

Interesting discussion. Getting one of those LSI PDP11 implementations up and running does sound like a very interesting project - but I have always assumed that, like most microprocessors, they don't bring out all of the signals required to implement the iconic lights and switches PDP front panels, It would obviously be different with the earlier discrete machines, but pins are usually at a premium with microprocessor implementations. If that is the case, then the FPGA option would seem a lot better for a PiDP targeted processor, as I assume you could bring out any signals you wanted if you have enough pins.

Does anyone know how close the LSI implementations can get to having a front panel? 

Regards,

DigbyT

To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/471411a3-c655-499b-9d2c-f0540b507454%40softjar.se.

[terri-...@glaver.org]

On Monday, October 6, 2025 at 4:58:47 AM UTC-4 digb...@gmail.com wrote:

Interesting discussion. Getting one of those LSI PDP11 implementations up and running does sound like a very interesting project -

People have implemented them on everything from perfboard to full S-100 coprocessor cards. All you need for simple playing around is power, clock, and something that emulates a DEC UART. You don't need any RAM to talk to console ODT, but that gets boring quickly. Memory is a little more complex, but with the availability of large-capacity new old stock static RAM chips you can just connect them and not have to worry about DRAM refresh, etc. Peripherals are more complicated, but with something like PDP11GUI or TU58FS you could load paper tape BASIC and have a full computer on something the size of a postcard.

| Does anyone know how close the LSI implementations can get to having a front panel? 

No LSI had anything more complex than controls for On/Off, Halt, Restart, and (rarely) remote diagnostics enable and displays for power on and run as shipped from DEC. The various BA23/123 "cabinet kits" for the CPU could have a selector switch (for baud rate) and one or more 7-segment displays (for diagnostic status).

You have access to the address and data, the type of bus cycle being implemented, etc. You can implement the display register for 16 output LEDs and the switch register for 16 input switches, along with halt and restart. What you don't get is state maintained internally by the CPU: I/D, 16/18/22 addressing mode, K/S/U, and other miscellany.

There's a user project to create one, which you can read about here: https://www.retrocmp.com/tools/qprobe The basic board just has LEDs on the "user end" of the board, but there's an optional extender that brings the lights out to a BA23/123-sized insert.

[Digby R.S. Tarvin]

Thanks - that is more or less what I expected. So you could create a PDP11 computer with an IMSAI style front panel,  but if the objective is to drive your PiDP11 hardware as an accurate simulation of an original 11/70, a software or FPGA emulation would probably be a better approach. With the bonus of being able to add instrumentation that the original never had. 

 The DCJ11 would still be nice as an 'authentic' CPU for verifying the accuracy of instruction set emulation, and just knowing it was a step closer to running on 'real' hardware. Without the noise and power requirements of a full 11/70.

Regards,

Digby

--

You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pidp-11/1b50cedc-76dd-40c6-851b-ad3c1ff9d067n%40googlegroups.com.

[Paulo Rebordão]

I have a couple of DCJs lying around unused. Is there a "ready made" project where they can be plugged and do some "useful" work? 

Paulo Rebordão

You received this message because you are subscribed to a topic in the Google Groups "[PiDP-11]" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pidp-11/VAZoNbEPJcU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pidp-11+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pidp-11/CACo5X5hYoYYobV%3D7P2_E_rqwOfMU5NhCqAH%2BTUQfv0H_KQ9Zrw%40mail.gmail.com.

[frankg]

As an alternative to a full panel it is possible to have blinken lights. The retrocmp site sells two different panels that can work with ODT PDP-11s. They also sell an interface board if your 11 is a Q-Bus system. QProbe2023