0011 A standalone Motorola 68000 CPU tester
![Adrian's Digital Basement ][](http://yt3.ggpht.com/ytc/AIdro_k7CrQRe71qN6c0Ua7A1P_I8KkxjqLa70u0zYeCvdvnKw=s48-c-k-c0x00ffffff-no-rj)
HTML-код
- Опубликовано: 8 ноя 2021
- Welcome to SMMC 0011! This donation allow you to quickly test 68000 CPUs to see if they are working and not fakes.
-- Video Links
68000 Freerun Tester (doesn't appear your can buy them)
mercurycoding.com/blog.html
Support the channel on Patreon:
/ adriansdigitalbasement
Adrian's Digital Basement (Main Channel)
/ @adriansdigitalbasement
-- Tools
Deoxit D5:
amzn.to/2VvOKy1
store.caig.com/s.nl/it.A/id.16...
O-Ring Pick Set: (I use these to lift chips off boards)
amzn.to/3a9x54J
Elenco Electronics LP-560 Logic Probe:
amzn.to/2VrT5lW
Hakko FR301 Desoldering Iron:
amzn.to/2ye6xC0
Rigol DS1054Z Four Channel Oscilloscope:
www.rigolna.com/products/digi...
Head Worn Magnifying Goggles / Dual Lens Flip-In Head Magnifier:
amzn.to/3adRbuy
TL866II Plus Chip Tester and EPROM programmer: (The MiniPro)
amzn.to/2wG4tlP
www.aliexpress.com/item/33000...
TS100 Soldering Iron:
amzn.to/2K36dJ5
www.ebay.com/itm/TS100-65W-MI...
EEVBlog 121GW Multimeter:
www.eevblog.com/product/121gw/
DSLogic Basic Logic Analyzer:
amzn.to/2RDSDQw
www.ebay.com/itm/USB-Logic-DS...
Magnetic Screw Holder:
amzn.to/3b8LOhG
www.harborfreight.com/4-inch-...
Universal ZIP sockets: (clones, used on my ZIF-64 test machine)
www.ebay.com/itm/14-16-18-20-...
RetroTink 2X Upconverter: (to hook up something like a C64 to HDMI)
www.retrotink.com/
Plato (Clone) Side Cutters: (order five)
www.ebay.com/itm/1-2-5-10PCS-...
Heat Sinks:
www.aliexpress.com/item/32537...
Little squeezy bottles: (available elsewhere too)
amzn.to/3b8LOOI
--- Links
My GitHub repository:
github.com/misterblack1?tab=r...
Commodore Computer Club / Vancouver, WA - Portland, OR - PDX Commodore Users Group
www.commodorecomputerclub.com/
--- Instructional videos
My video on damage-free chip removal:
• How to remove chips wi...
--- Music
Intro music and other tracks by:
Nathan Divino
@itsnathandivino 
Наука
I've been a viewer for quite awhile now, and always enjoy your retro electronics adventures. Thanks for taking the time to showcase my device! It seems its usefulness is limited, given the issues it currently has, but I'm about to dust off the schematics and make Revision B in the near future. :)
Hey Mercury 13 - really interested in this - I have a bunch of 6800 and 6808 processors to test. have you got the schematics for your board available please. I would like to buy one too please. cannot seem to add to cart in your webstore.
@@alanjones3617 I haven't yet got the schematic recreated, but I do have some boards available.still.
@@mercury0x000d cool may i buy one please? is it blank pcb or made up?
@@alanjones3617 It's a blank PCB. How can I contact you?
Will you try a bad cpu from your dead parts bin to test the tester? It would be interesting to see what the indicator shows for a known dead cpu.
Okay, but then you cut the shot where you're covering the label AFTER you took your hand away! 😂
I maxed the video resolution and went frame by frame but the address is not readable, not even the huge number of the tracking.
I mean, this tester is kinda "okay"... I guess? No limiting resistors on the LEDs is a really weird choice (no need to load the IO drivers that heavily so they go into current limit mode; if they have a failure, it'll show up anyway if you have a current limiting resistor), but in any case, this only runs NOPs on the chip at different frequencies. This is such a basic test you can run on it and that would be an understatement. I would add a ROM that excercises all the opcodes, registers, maybe interrupts and bus transactions handling. I certainly won't take the outcome of this tester as a definite "this 68k works" or not.
Fully agree. The test is so minimal as to not be worth spinning up a board. Adding just an atmega one could emulate rom and ram. Non-trivial software effort though.
@@AppliedCryogenics Yeah, like these other IC testers which do test a whole bunch of logic ICs, and I saw one that even exercised a z80 or a 6502.
Despite the fact of the missing current limiting resistors the signetics part is most likely an nmos version of the cpu with lower high levels (ttl logic vs. cmos) whereas the others are maybe cmos parts.
Yes Signetics is an original NMOS type, with weak pull ups, while the others are very likely all CMOS, so have faster pull ups. The NMOS one might function at 16MHz, but very likely the bus drivers will not be able to drive any more than a single LSTTL buffer 2cm away from the chip, as the pull up will not reliably pass a logic 1 threshold in the time before the clock edge, combined with propagation delays through the decode logic, latches the data on the data line in or out.
He mentions adding resisters as a to-do in his post about the tester. Though he missed the case of a faulty led destroying an address line.
A lot of current is floating through the 68000 VCC pin when all LEDs are on. The bond wire is glowing inside. I am wondering why no drop resistors were installed.
4:31 Revision B will have current limiting resistors to protect the LEDs.
@@Okurka. I guess the 68000 also needs to be protected. :-)
Glowing bond wire? No way. Those led's are drawing no more that 15mA. Even with all the LED's the current draw won't exceed 400 mA. Unless of course if your comment was exaggerated sarcasm.
@@alexsinclair2012 The leds are pretty bright in the video so I guess it could easily be 30mA per LED. Bond wires are very thin, but the glowing was a joke indeed. Perhaps Adrian could check the total current when all leds are on.
@@alexsinclair2012 When designed properly, signal LEDs should draw about 15mA. But with no current limiting resistor it can be much more. Modern IOs has current limiting inside, maybe here as well. But with no current limiting, LEDs will burn in moment :)
To quote from their blog: "There are a few tweaks which need to be made for the future Revision B board: [...] Install current-limiting resistors on all the LEDs. In cases of catastrophic CPU failure such as a severe short, full current could be passed directly through the processor to the LEDs, damaging them." So yeah, definitely a known issue.
Thanks for pointing that out! :)
You must of heard me screaming as I was about to ask about the LED's being so bright, I thought maybe it’s just your Cameras exposure. Always make me wince when I seem them being driven so hard, so totally agree dropper resisters would be a must for the next revision
*have
Even if it is still in spec for those CPUs to be able to handle the datalines being shorted to ground, in the case of catastrophic CPU failure or a fake part, it could get pretty exciting...
Gotta keep in mind -- not all instructions are equal -- just because a processor can NOP at 16 MHz, doesn't mean all the functions, APUs, etc. will work -- and even if it does all "work" -- it doesn't mean the math results, branch instructions, etc. will always be handled correctly at those speeds. -- It also doesn't mean the the electrical characteristics are appropriate / in spec. -- At a 200% overclock, I would definitely use a heatsink though.
I ignored your content for almost a year now. But as the days get shorter I will be binge watching your videos again.
1:40 Not sure about Motorola MC68000 0.07" raster variant, but Hitachi made HD63000PS-8 , -10 and -12 versions in Shrinkdip package, actually both of Motorola's licensees Hitachi and Toshiba seemed to LOVE to offer the goods in Shrinkdip versions, and japanese OEMs liked to use those saving 30% in width and height too.
Adrian, your prolific output continues to astound me! Keep up the good work!
There is a project from the editor-in-chief of Canard PC in France called the Universal Chip Analyzer which is designed to comprehensibly test many different types of CPUs. Right now only Intel architectures are supported from 8080/Z80 up to 286 but a version for MOS/Motorola architectures is in the works.
Interesting!!! I'm currently working on getting a SWTP 6800 (not 68000!) system to work. I built it in the 1970's. Even have the CT64 terminal to go with it.
Chip rated max freq is based on worst case voltage and temperature, both of which affect performance. The manufacturer would have some margin on the spec as well. So when you are testing at room temperature, with a 5.0 volt supply, you would expect it to run faster than at the rated speed. Also, as you mentioned, if the yields are good, there may not be enough slow parts, so faster parts would be downgraded and sold for less.
Those CPUs are rated so that 99% of production passes. So if only 98% can run stable at 16mhz, the whole lot may be rated at 12mhz. Silicon lottery.
Simplest way to drop LED brightness is to cut the common ground lead and put in a pair of series connected 1A diodes. Will drop around 1V2 extra, and then the LED current will be much lower, and saves placing the 16 220R resistors in each line. Still will have LED current varying according to number on, but the drivers will not be so stressed. Did that on a filament display conversion to LED, as the new batch of LED displays were very much higher efficiency over the prototype batch, and thus the display came out too bright. The 2 diode drop with the existing resistors used, 82R 0.125W, made it nicely match the other displays and not flood them out.
My experience when it comes to overclocking is that the old CPU:s could be overclocked by a ~lot~. If I remember correctly (at the very least until the 486) they didn't REALLY make different CPUs at all, they just had their die and made a shitload of CPU:s and it it was during the quality testing that told them whether f.ex. a specific 286 CPU was at 6 MHz, 8, 10, or 12.. Because of how stable it was and how much heat they generated (There is always slight differences in production).. and this was when people ASSUMED you wouldn't use a heat-sink.. I had a 486sx20 which I got a sink and a fan to and ran at 50MHz and it was stable and lasted..
That said; You shouldn't assume that just because this processor CAN run at a specific speed mean that it TRULY can run at that specific speed. These types of testers usually just run a "NOP" command. I don't know the 68000 that well, but I think it didn't exactly had a nop command, but whatever command one chose to run the CPU with it will be a fairly simple statement.. probably one that executes the fastest of all commands. I don't know how the timing on the 68k work, but I would suspect that SOME of the instructions it handles take more than just the one clock-cycle (It was a rather complicated and expensive device, so I assume it have multiplication and division as op-codes).. so.. addition is probably slower than the actual command that this processor can run at this speed.. and.. mul/div should take longer than one clock-cycle.. but.. are the number of clock-cycles per instruction hard-coded OR can the processor DETECT if the last operation isn't done and it need to add a wait-state?.. I wouldn't know; The second alternative does sound a BIT complicated for the era.. either way... Each instruction is probably more or less "close" to the time-limit it has, so.. just because the nop work fine doesn't mean ALL instructions would.
Running NOPs probably isn't a good overclocking test. The real problems would be more likely to pop up in the ALU, conditional jumps, or even memory access.
To improve this board, they ought to build a ROM that runs test instructions meant to make sure that all registers work, status flags, conditional jumps, etc.
There is a 68000/Z80 test program for Sega Genesis that runs a suite of tests like that. I don't know where I got it, but it's very useful!
I think it would need at least a small amount of RAM too, in order to be able to test instructions that affect the stack.
That's a good idea. Maybe I'll work some kind of ROM socket into the next revision. :)
you can cut the ground and solder a resistor for the leds
Can help but rewind the intro. So perfect.
*Moral of the story* : Doesn't matter how fast you can count, there will always be LED's in front of you, flashing at a rate you can comprehend.
EDIT: Get one of the slow motion camera channels to record D1 - D12 activity!
Where's the Slow-Mo Guys when you need them?? :)
To check if a processor is ok with a specific clock frequency, I would run the processor at a lower voltage, like 4,5V or even lower. If it runs there, there will be enough margin when running at 5 volts.
I use to have a ceramic 68000 with the gold lid. Sadly I have no clue where it went to. Still have a few of the plastic packages floating around as well as QFP packages.
In the 1980s both Dr Dobb's Journal and Byte magazines had a 68k cpu tester, like the 6502 tester you shown before, using No Op bytes to run the entire address space.
That article was also reprinted in the Dr Dobb's Toolbook of 68000 Programming. I think there is a PDF of that book online.
Ahhh the Keystone state. You can ride the historic steam trains by day, and go to Hershey at night.
I want one of these things - the amount of counterfeit, remarked 68000's out there is astounding.
I do have some unpopulated boards left over, if you're interested.
@@mercury0x000d Ooh, yes please!
@@ukmk3supra Cool, how can I contact you?
There are indeed 16mhz 68000's like the one in the Atari Jaguar which is a 16 mhz CPU running as 13.295 mhz which is half of the RISC processors in the Jaguar. The Hitachi CMOS versions can run up to 20mhz.
I would not use that without resistors, it can kill the cpu without a LED dying. Way too much current as the voltage drop has to appear within the cpu, bondwires etc.
68030 was the first processor i used after a childhood 6502. I'm old.
I wonder if those LEDs being brighter is an indicator that the CPU is drawing more power at a given clockspeed?
2:43 - there's A1-A23, along with "UDS" and "LDS". UDS = Upper data strobe and LDS = Lower data strobe. These enable one or both of the bytes of the bus depending on what access is being made.
This is what I needed, I had to build one and it wasn't that fun. Also I don't think the overdriven leds are really an issue for short tests because they are white and they can be driven this way. Still, at least 100ohm resistors would be nice to have
next project? cutting traces and installing resistors in series to the LED's???
Funny coincidence: my current project - featuring an 8085 - also has a "free run" test mode integrated. These might have been a common way of testing circuits. If the actual circuit has a provision for it, you can also test all the address decoder logic chips. The concept is a nice part of history though; I'm pretty sure flashing LEDs won't help with GHz CPUs 😜
2:43 Well, yes, and no... the design implements a 32-bit instruction set with 32-bit registers and a 24-bit address bus. The internal data bus, however, is indeed 16-bit. [Edit: additional technical pedantry]
Pittsburgh represent!
Woohoo! :)
What does this test say about the condition of the processor. Address lines work, but what about the rest of its registers. I think a better test would be a program that tests all functionality of a processor and display's al the results.
Just testing it is not totally dead, and that at least the clock, reset, bus arbitration, and instruction decoder more or less are functional. Better would have been to simply add on a single fast EEPROM that is mapped into the top half of the address space, and have a simple program in it that checks out each on chip instruction in turn, with output simply being a latch on the data bus selected when the ROM is not accessed. Simple, and the on chip registers can store enough that you simply have a output that runs a series of leds to show all is working, using a big timing loop to slow the tests down, so it takes a second or two per full test loop. Only needs one fast 1M 8 bit flash device, and a tiny bit of glue logic, plus a bus latch for the LED drive.
Hi from Pennsylvania
I believe the original Motorola and Signetics 68Ks used an NMOS process. Your relabeled CPUs are probably the later HCMOS process chips with better high output capability, hence brighter LEDs.
Hello Adrian... I realize this is somewhat a dated video... but I wanted to ask if there was some way you can get in touch with the person/persons at Mercury Coding to see if they plan a B version of this board.... I am happy to reverse engineer this one (using images from their site) to make one of my own, but i'd rather give credit where credit is due and see if they are willing to do a B board or if they were going to post the PCB files on their site.... Thank you in advance! great content on the channel! Cheers.
I would love to see a PiStorm stuck in to this thing
How much current can those lines source or sink ? ........cheers.
Where could I find the schematics, to build something like this? I'm just thinking about a new hobby. Digital Electronic design is an interest.
5:45 - only the top "68010" is definitely real. The other two are possibly rebadges. I ordered a bunch of 68010's from china and only one of the vendors I bought from was selling real ones. There were a bunch of 4 MHz 68000s rebadged as 68010s, which sucks, because the 68010 is the one which supported an MMU and allowed the vector table to be relocated. The original couldn't properly support an MMU because the bus error exception (page fault, etc.) frame didn't include enough information to restart the instruction the exception occurred for. There were a couple of systems with an MMU and the original 68000, but they did exotic things. One workstation actually ran two 68000 chips, with one CPU a single instruction behind the other. If the "leader" encountered a page fault, it would trigger an interrupt on the "main" CPU before the instruction which had the error. It would then handle the page fault and restart the "leader" in the correct state. Obviously not ideal, as the 68000 was an expensive CPU in the day and two CPUs meant twice the power consumption.
There are LEDs with built in dropping resistors… might want to look around online for some. My work had a customer years ago that had quite a few boards using these LEDs. Unfortunately that customer is long defunct and all the documentation is long gone so I wouldn’t even know what brand they were.
There is a black dot on the anode for those, and they are designed for specific voltages. I have only seen 12v ones, but I am sure they make 5v ones.
at least one resistance for all leds needed.
FYI if you're looking for a good source for 68000 series chips try and find some AMF bowling accuscore curtain wall chassis they call I'm 7100 all the way up to 7500 series and they come with all the support ships.
You can contact places like us bowling I know there's one in California or Tucker Bowling Supply in Texas
Those LED's would drive me nuts. Really dumb they made it like that.
I'd like to know how 68k detects bus contention (berr).
There likely is built-in protection for data and address pins but I don't know the mechanics behind it.
It cant detect bus errors per se. Something external has to assert the BERR signal to tell the CPU there is an issue with the current bus cycle.
If there is contention that causes data corruption you might end up with an invalid instruction exception, or bad operands that cause other issues like reads/writes to bad memory locations (that might result in a bus error) or erroneous arithmetic results.
hello, I should test a Motorola 68060, is there anything to do this?
thank you
Probably a bit exxaggerated but in my opinion running NOP Just indicates the PC is running normal and nothing else. I would recommend at least an ADD instruction to sum the address output with a constant number and check result with an XOR logic or so. That will bring in ALU alive, heat the CPU and is a good indicative of stability
You're right, NOP free-running is only the most basic of tests. However, that's all I needed, so that's all I designed it to do. :)
Never seen yellow bubble wrap before
Well A0 isn't needed for instruction fetching, but would be needed for byte data access. IIRC there isn't an actual A0, but a HBE and LBE signal, the HBE would be the equal to A0.
Good point.
Yep. A0 is split between the upper and lower data bus strobe pins.
But when it fetches a 16 bit word it doesn't do it one byte at a time so no, HBE isn't really much like A0 at all.
Can that cpu source that much current to directly drive (overdrive) the leds? *shudders* I would put transistors for the leds. Great content as usual.
I would price buffer chips first, but same concept.
The NMOS chips with totem-pole style drivers can drive a tremendous amount of current for a "logic" IC. It would have been completely fine if they'd put 1k resistors or something before the LEDs.
There are LEDs with internal current limiting resistors, they may be those type, just guessing. Whoever designed that board clearly knows what he is doing so I wouldn't imagine him forgetting a detail like that and stress the CPU with high current drawn from the IO pins.
The 68k does not use A0 since it is it accesses data in 16 bits so the first 2 byes for instance are covered even without an A0 line. The 386 did something similar
I think that's the CPU my Atari ST had.
this test can tell you if for sure a chip is bad but it can't tell you if a chip is actually good. Testing nops really only useful to see if the chip is a 68k or not and if it works at the most basic level. Can't really test if its a good chip or not without more detailed test of the instruction set really you need to test alu and memory access too.
this is ok if the leds have intenal resistors
Not to nit pic but please board devs, white and blue leds are for illumination not indication. Please use red, green, amber, etc.. Can't see / read anything if you have a led flash light pointing at your face.
Also wonder if you could step through the free run process to read each driver line at a time instead of it just counting away at whatever MHz you have it set to, but I know almost nothing about this CPU..
I'll take that under advisement! I may have some green ones laying around somewhere... :) As for single-stepping the CPU, yes that is possible through use of the Halt line which could be implemented with a little extra logic added to the board.
Didn't someone send a whole bunch of 68000s, or was that a different CPU.
covering the package with a hand… and yet leaving 1 or 2 frames for people to decode stuff from 😀
Wonder if a PiStorm would work in there.
I would have put a 220 - 330 ohm resistor connected to each led.
Does anyone know what clock divider chip it uses? I'd like to make one of these.
I used a 74HC4520 dual 4-bit counter, but you could use any basic counter IC. For more output frequencies, you could even use a 74VHC4040 counter, which has twelve stages instead of the eight the 4520 offers.
👍
I've got some 8MHz rated 68k's that run just fine at 16MHz.
Someone I know has a 12MHz part that runs fine at 32MHz.
Other members of the 68K family were similar; I've heard of 75 MHz 68060s running at over 100 MHz with proper cooling!
Oh yeah, after watching this video my eyes feel like Paul Atreides' in Dune Messiah, I have to type this comment entirely by feel! I'd definitely be bodging in some resistors if I were going to use it on any chips I paid any decent money for. But aside from that I like the idea.
I don't know about what they did in the 68000 world, but it is not uncommon in the Intel-compatible world to sell CPUs marked (and often locked) down to a lower speed than they're capable of under testing.
It'd be so much nicer without those pesky white/blueish LEDs.
The 68000s are said to all run with way higher clocks than announced. Probably you could run even all of these with 25MHz, and even 33 or more with proper cooling.
Yep, the whole 68K family was quite overclocking tolerant, which could make for great fun with modern cooling solutions!
Running those LEDs without any current limiting is silly, and potentially dangerous for the chip. Better still would be to buffer the LEDs through bus buffers, this then protects the chip further... plus there are times in the cycle when the data bus shouldnt be loaded at all. Some facility to check BR/BG(ACK) would be important too
I would probably have designed this to take a 9v barrel jack type supply, and have an onboard 5V regulator. Two in fact - one to supply the chip, another to supply supervisory circuitry... i would also then build in an overcurrent detect to shut off the CPU supply and warn if it draws too much current. Especially important with the ATX supply - imagine if the CPU was a short inside, with the typical 5V current available from an ATX supply, you would have smoke and burnt traces very quickly!
This will kill the cpus by overloading it.
I wonder if there is an adapter for xLCC MC68k's?
Not yet! If I can find / design some kind of ZIF-style socket for those, I will definitely create one.
so these are 3.3V LEDs on 5V without resistors? lol
I want to FreeRun the Digital/Harris J11 Processor. I wonder how hard it would be? What could go wrong? 😂
*stares at tester*
*Immediately blinded*
I'd recommend frosted low intensity LEDs for indicators like this rather than clear bright LEDs. Kids these days. 😆
Tell the manufacturers of consumer goods that while you are at it also. Watching this on a monitor with a blindingly bright white LED just under the screen in the bezel for a power indicator and have several others that use obnoxiously bright blue LEDs for the same purpose, which is just insanity to me. You want a nice dim red LED like on the early C64's for this sort of power indicator.
The LEDs could have integrated resistors?
It doesn't. It says on their blog that the lack of resistors is a known issue.
Ketchup.
By the end the LEDs got really-really bright, that doesn't look good. And I don't think it's because of the 3rd CPU you used, if you were to go back to the first ones I think they would still be as bright if not brighter :/
The 3rd one could be a (more modern) CMOS version, which actively drives the high outputs and can source a lot more current an the old HMOS (a little updated NMOS) versions.
It'd be pointless to run it for hours anyway. All you'd be doing it checking it can execute NOP instructions. Almost worthless as an extra indication that it'll run actual software properly for hours.
Unless a 16-bit processor is designed to do byte accesses, it will not have A0.
68k can do byte access but it uses a different pin other than an address pin to select either byte one or byte two.
@@GORF_EMPIRE Ah yes, it uses /UDS and /LDS pins.
@@gusbert Yes sir... somewhat similar to how the 386 does byte access.
2:40 The Motorola 68000 is a 32-bit processor.
16/32
It has 32-bit wide registers, but the databus is only 16 bits wide.
@@BertGrink Correct. It's really a 16 bit processor with 32 bit "enhancements". Even the ALU isn't fully 32 bit, i.e., the multiply instructions can only do 16x16 bit with a 32 bit product. If you want to multiply two 32 bit numbers, you'll need to write a software routine.
why don't you just cut the LED ground trace in the corner and install a resistor? haha
Running NOPs with DTACK grounded
Yep, it's technically "ORI.B #0, D0"... which is basically a NOP when nothing has been initiated lol By the way, nice work on your 68060 build!
Are you sure the LEDs aren't the type with an integrated current limiting resistor? Also, they are pulsed, so that can increase their peak current handling before thermal runaway. Any idea what his EPIC project was?
It’s not the leds to worry about but the cpu under test. It will sink so much current for the 23 leds that it can be destroyed running for too long. The fact that the most leds are pwmed effectively doesn’t help much because I’ve not limited inrush current only puts further stress on other parts of the cpu than the address line drivers.
4:31 Revision B will have current limiting resistors to protect the LEDs.
@@Okurka. Yes, I’ve already read this and other commenters mentioned this, I only wanted to to comment on the op‘s comment - no worries 😉
@@0toleranz In his video, he's concerned about an LED failing in a short which would then damage the CPU. Also, on the designer's page he's looking at on his phone, the designer was planning to use current limiting resistors in a (never made) future version to avoid damaging the LEDs due to a CPU fault, not to prevent damaging the CPU. In the MC68000 datasheet, I don't see a max. current output per pin specified. I would have installed current limiting resistor from the start to protect the CPU, not the LEDs, but perhaps the designer knows something about the per pin output current capability of chips from that era that I don't.
pea
It looks like this is just a 'simple' NOP tester, that let the CPU run thru it's adress space. You can simply do this on bread board too. Put EA71 (0100 1110 0111 0001) at the data bus, the cpu it self will then run from 0 to ffff ffff (more precise from what ever is stored at $4, it reads EA71 EA71, so it starts at EA71 and runs to $FFFF FFFF for the first run) and wraps around to 0 again, VCC, GND and 3-4 more lines to VCC or ground to set the bus correctly thats it.
You can't execute instructions from an odd address, so the opcode you wire in has to have the LSb clear - otherwise you get an address error exception.
It's fine to use an opcode of 0000 anyway, this translates to "ori.b #0, d0" which might as well be a NOP since it has no effect, except setting the Z flag in the status register.
Philadelphia is a port city. On the East Coast.
what a useless thing, the 68000 is very stable.
to bad... the 68k freerun tester ist no longer available :(
68000 can easily overclock to 28MHz, as you can buy 28MHz OC Amiga 500 Turbo cards, like the old Supra Turbo 28, and the new Amiga 500 HC508CR accelerator cards.
Yep, most of entire the 68K line was very tolerant of overclocking. :)