• Please review our updated Terms and Rules here

PDP 11/45, Part 5

Hi All;

Dave, Thank You for Your input..

Remember, Or maybe You didn't see, but anyway Posting #192, and #194, States that I show a "1" in the Most left Position..
Which would make it an '8'..
On '0275, I will check, BUT Most likely I read a '0' for a 'D' in my Listing..

On '027, that was corrected back before posting #192 or so, I thought, and that we agreed must be that we changed it for the Good..

"" Are your Exx.bin files listings from the actual ROMs "" NO !!
These are from Your Hex files after I copied them to my NoteBook.. Making sure I copied Correctly, I also have to check that my main Binary Listing from my NoteBook is the same, I have done one check after I copied Your Hex files into the NoteBook.. But, I need to do another.. For clerical reasons, my writing and my typing..
As Stated in the above Posting # 199, I may have a problem in at least one of the proms, First I need to verify my Theory about what I found and secondly I then need to build a circuit for advancing the address, by another method, than by an eight position switch, which I know how to do with a single button and a Counter IC such a 74192 or 74193 or something like that..
And then I will go through all of the Proms in Pairs, also Remember that I might have at least one other prom that may be Bad, and possibly more..
Later, this week, I will as I can Actually Check out what Proms that are Readable and verify them against what we have found..
I Just want to make sure what we have typed in is correct, Before I have any New proms made..

THANK YOU Marty
 
Last edited:
Thanks for the confirmation on what I was looking at from your listings.

So we still need to 100% compare my field listings with the DEC documentation to identify more of my transcription errors.

Back to filling the Baptistry with water...

Dave
 
Hi Marty, Dave:

Marty, thanks for sending along the bin files. I probably won't have time to look at them until later in the week, though.

Re. post #200, taking a quick look at the ROM listings I have in the 11/45 engineering drawings, I agree with Marty's value for words 122 and 123, and Dave's value for word 275.

Marty, if you have your ROMs socketed, you might consider sending them at some point to either myself or Don -- I think either of us could quickly read them out on our burners, and also generate checksums. This would be another good check on the transcriptions that are being made here by yourself and Dave. I'd do this with my own, but neither of my RACs is socketed and I don't want to go in and desolder all the ROMs since everything is working at the moment.

cheers,
--FritzM.
 
Hi All;

Dave, Thank for the confirmation..

From Post # 196, ""Dave, on the Addresses that I have checked, Except for '0122 and '0123, the rest of them are fine up to Address '0175, I haven't checked any further, yet..
Addresses '0176 thru '0303 are all good..
I have checked from Addresses '0304 thru '0377 and they are all Good, Mine had a few mistakes that I corrected.. ""

Everything from '000 to '377 Has been check from my copy of Your Hex file, so once we get these three or four discrepancies Ironed out, then I would say that Your Hex files are OK..

Fritz, Thank You for Your confirmation..

"" Marty, thanks for sending along the bin files. I probably won't have time to look at them until later in the week, though. ""
No Problem, If we need it, You are another source of confirmation..

"" Re. post #200, taking a quick look at the ROM listings I have in the 11/45 engineering drawings, I agree with Marty's value for words 122 and 123, and Dave's value for word 275. ""

Good to Know, THANK YOU..

"" Marty, if you have your ROMs socketed, you might consider sending them at some point to either myself or Don -- I think either of us could quickly read them out on our burners, and also generate checksums. ""
That Basically is the Plan, But, as much as is possible, I want to check the files that we have Made, before having any New Proms made..

"" This would be another good check on the transcriptions that are being made here by yourself and Dave. ""
Yes, which once we get these last three or four possible problems solved, then I can send You the needed updated files, and then If from the files You can do a checkSum check for final Verification..

"" I'd do this with my own, but neither of my RACs is socketed and I don't want to go in and desolder all the ROMs since everything is working at the moment. ""

I agree with You, that, I wouldn't want You to do that..
I most likely Once we get everything worked out, I hope to Have Don Program a set and what extra's we need to make my set complete, mainly because He has the Needed Proms already.. But, If You can do a checksum of the Files themselves, that would be a big help in Verifying what we have, before any Proms are Programed..

THANK YOU Marty
 
Hi All;

I found at least some of the source of my problem..
I did not have the Higher voltage buss tied to the lower voltage buss on the Breadboard..
So I am in the process of wiring things back up, after pulling some wires trying to see what the problem was..
Once I get it back to working order I will let You know what I find on the proms..

Yes, it now works like it's supposed to work..
Now, that doesn't mean that all of the Addresses have been checked, that is next..
And I need to make the one button address Increment circuit..

I thought I knew How to build an incrementer circuit, but, I am missing something,
I have tried using 74LS283's, 74LS193's, 74LS381's and 74LS181's, and I am missing something..
Of course, it's probably, something so simple, and I am overlooking it..
I am going to try a 74LS374 and an 74LS181 together..

THANK YOU Marty
 
Last edited:
A 74LS193 should work on its own.

Assuming you want a RESET and COUNT UP function wire as follows:

CLR (pin 14) normally '0'. Pulse to '1' to clear the counter.

/LOAD (pin 11) to '1' permanently.

I would wire the A, B, C and D inputs to '0' or '1' permanently (we are not using the parallel LOAD function - so I don't permit floating inputs).

The DOWN pin (4) needs wiring to '1' (we don't use DOWN so it needs to be disabled).

The UP pin (5) should normally be a '1'. Pulse it to '0' and then '1' to count UP. The count will take place on the '0' to '1' transition of UP.

/CO (pin 12) is the carry output to the next stage UP pin if you need a larger counter than 4 bits.

Dave
 
Hi All;

Thank You, Dave for the Information.. I knew that the 74193 should work by it self..

There are some things going on Now, that I can't share, and so my mind is Mush right now..

One Question, On /CO (pin 12), where so I wire it to for the Next stage ??

THANK YOU Marty
 
Wire the /CO pin of the earlier stage to the UP pin of the next stage.

Put your feet up and do something to relax if your brain is much - other than wire-wrapping of course!

Dave
 
Hi All;

Dave, Thank You for Your answer, I had found it before Your answer,
and I had actually stated so, in the Last posting, but I forgot to save my answer..
So of course it didn't show up as such..
My latest problem that I am trying to working out is it, goes to many cycles in the push of a button, I had tried 7474, 74175, but the closest to working is Our circuit from the 8i were we combined a 74L00 and a 7474 to get a narrow enough of a pulse, I am also looking into some of my other circuitry to see if that might be causing a problem..
No it still does the same thing..


THANK YOU Marty
 
The 'classic' way to debounce a pushbutton is with a change-over push-button and a couple of cross-connected gates (as in http://electronics.stackexchange.com/questions/69952/improving-a-push-button-debouncing-circuit).

If you only have a single push-button (push to make - release to break) either use a CR network (C to 0V, R to +5V, Pushbutton across C and 0V - you will have to experiment with the CR values) and a schmitt gate to 'square up the edges' or use something like a 555 timer with the push-button to trigger the pulse.

Lots of options here...

Dave
 
Hi All;

Dave, Thank You for the Suggestion,,
I had tried the old Cross-coupled Nand Gate circuit..
I had not seen the circuit that You Linked to above..
BUT, at this point, I am not sure if it is Debouncing or the 74193's that I have themselves..

"" and a schmitt gate to 'square up the edges' or use something like a 555 timer with the push-button to trigger the pulse. ""

I will need to find it again, I think it might be Ben Eater who had a 555 timer One-shot pushbutton circuit, that I will look into..

https://www.youtube.com/watch?v=81BgFhm2vz8

OK, after messing around for awhile, and adjusting the timing on the 555, I got it to Count correctly..

THANK YOU Marty
 
Last edited:
Hi All;

Thank You DDS for the Link, This looks like one of the first ones that I originally ran across..
I have looked for people making CPU's/ALU's from TTL's for a long time.. And every once in awhile I find someone new to me..
I have looked at quite a few of His Video's, just not lately, so I have a few to catch up on, that I haven't seen..
Especially his newest Ram video's..
I don't have enough Breadboards to do what He has done, I would need to more than Double what I have at the present..
But, Thank You for the Link, Hopefully others might take an interest in what He is doing..
He presents thing very well..
If You like the other one DDS, This one is even better

https://www.youtube.com/watch?v=35zLnS3fXeA

Or this one

https://www.youtube.com/watch?v=a73ZXDJtU48

I have corrected the bin file for E99, which I will be sending to Dave and Fritz..

Well, Looking at both E90 and E89, it has multiple Errors, and I have only looked at the first 16 addresses..
Some places look OK, but for the most part, it is not right.. I don't know if my circuit is Loading too much for the Prom, or if the circuit is still not correct and so reading things wrong, or it's just this Bad..

I am going to see If I can Build Fritz's FPGA Circuit, now that I have the two Low Voltage IC's that I would need, We'LL see if I can Read a Prom at 3.3 Volts as Power instead of 5.0 volts for power..

THANK YOU Marty
 
Last edited:
Marty,

You do/did remember that these ROMs are open-collector and require pull-up resistors on the data lines?

All you should need to do to these ROMS to read them is feed a power supply, 8 address lines, 4 pull-up resistors on the data lines and ground the two not chip select signals.

Dave
 
Hi All;

Dave, Thank You for the reminder, and Yes, I do have pull-ups on the Data Lines and I have it on the Address lines as well..
I have tried the address lines both with and without pull-ups and it made no difference in the errors..
Here is what it shows --

Address, Board, Notebook
'000, E5, E5
'001, D7, 4D
'002, 4D, 4D
'003, A4, 91
'004, 4D, 52
'005, 6E, A9
'006, 91, A9
'007, EC, A9
'010, 52, D7
'011, EC, A4
'012, 52, 6E
'013, AD, EC
'014, A9, EC
'015, 6E, AD
'016, A9, 6E
'017, E5, E5...

I have Put the Proms back in the 11/45 and the 11/45 is Now Dead in the Water..

So, as soon as we can Verify that what we have in the bin files that I have created, I will then Send what I have to Don and Hopefully they would be correct, and once I would get them back from Him.. I can see if that Helps any..

I put in the M8123 as a sanity test, just to make sure that something else was not Broken, and it came up and it will Deposit and Examine..
So, that tells me either my M8103 is No Good or the Proms in my M8103 have gotten worse..

So, back to You Dave, Once You say that all of the Files look OK, I will send them on to Don..

THANK YOU Marty
 
Last edited:
It never rains it pours :-(!

I have checked the Texas Instruments data sheet with the 11/45 schematics and the address and data lines are used 'consistently' between the two (i.e. DEC haven't pulled any 'nasties' like swapping address or data bits around).

Looking at your columns of hex digits - I can see the data there - but they are not at the same addresses??? I was just wondering if your address wiring is correct (i.e. have you actually wired ROM A0 (pin 5) to the least significant bit of your counter etc.).

Can you just explain a bit why you have two hex digits for each address (8 bits) rather than one (4 bits)? When I took your BIN files I guessed that you had 'packed' two sets of ROM fields in a single byte to save on space. Is this the case - or am I missing something somewhere?

YES: There's a pattern!!!

--
D7 - D7
--
A4 - A4
--
6E - 6E
--
EC - EC
--
EC - EC
--
AD - AD
--
6E - 6E
--
E5 - E5

The ODD numbered addresses from the BOARD data column match your NOTEBOOK data column sequentially from address '010.

The EVEN numbered addresses from the BOARD data column match your NOTEBOOK data column sequentially from address '000 (with the exception of '012 = 52 matching with '005 = A9 - which I suspect may be a transcription error as everything else is OK).

I need a little bit of time now to work out what your wiring error is...

Dave
 
Last edited:
Hi All;

Dave, Thank You for Your Help..

"" Looking at your columns of hex digits - I can see the data there - but they are not at the same addresses??? I was just wondering if your address wiring is correct (i.e. have you actually wired ROM A0 (pin 5) to the least significant bit of your counter etc.). ""

I think so, but I can recheck it, and Draw it out on Paper and submit it to You for You to Look at, As You have said in the Past, another set of eyes..

"" Can you just explain a bit why you have two hex digits for each address (8 bits) rather than one (4 bits)? ""

That was the Way My Hex Editor was set up, and I was/am unsure of How to change it to a 4 bit word, Having said that I did go Vertically down the Listing, it just stored the Hex numbers as 8 bit Pairs, I didn't think it would make any difference as long as I had the numbers in the correct order..
Are You saying I don't have the Hex numbers in the correct order, is Your Question or of more of why I had the longer Word Length..

"" When I took your BIN files I guessed that you had 'packed' two sets of ROM fields in a single byte to save on space. Is this the case - or am I missing something somewhere? ""
No Your not missing something, I just don't know alot about my Hex editor and it's operation.. It's the Binary that I am concerned with, not the Hex..

"" I need a little bit of time now to work out what your wiring error is... ""

I will copy down what I have wired up and get it to You..
No need to copy it down, see next posting..

THANK YOU Marty
 
Last edited:
Hi All;

I Found the Problems, I had the lower counter wired wrong, and that was based on my Address Leds which was wired wrong only I thought it was correct.. And I had a funny miswired 74193, which was not the case..
So, once I wired the first 74LS193 up correctly and stepped through it I was getting the Correct Code at the different Addresses..
So, then it was a matter of wiring of the Address Led's to their correct place..
And I NOW get the Correct code at the Correct address !!!!
The first 16 are Now a perfect match..

THANK YOU Dave for pointing me in the right Direction, and then it was a matter of correcting the Problem once I looked at it on Paper..
Like You have said before, another set of eyes, even at such a long distance, are a help..

And Now I can Look at the proms in Pairs, that way I only had to advance the addresses eight times instead of with only one Prom at a time, where I would be advancing my addresses sixteen times..
I can Hopefully locate which Proms are Bad and which ones are OK..

Dave, Thanks for Your Insight..

THANK YOU Marty
 
Last edited:
Hi All;

Good News and Bad News..

First the Bad News, I found One bad Location in Prom E89, Address '056, it had an 'A' and it should have been a 'C'..

Now the Good News, all of the other Locations were fine and they Matched what Dave Had in His Hex File, for proms E89 and E90..

I did check the Bad Location from the Listing and it is supposed to be a 'C'..

Now to the Next two proms to be checked.. OK, I have checked the next two Proms.. And, You wouldn't believe it..

But, at Address '056 Prom E91 has '4' and it should be '0' !!

Now to the Next two proms to be checked.. OK, I have checked the next two Proms..

E97 and E98 Both check out OK.. No Errors..

Now to the Next two proms to be checked..

THANK YOU Marty
 
Last edited:
Somewhere between work and home on the motorway I see you sorted out your wiring error(s)...

It is interesting the issue is at address '056...

I have just checked and the SN74187 is a mask programmed ROM and not a PROM. The titanium-tungsten (Ti-W) PROMs had a failure mode where the 'blown' fuses (logic '0') could re-grow - thus making the PROM bit location change from a '0' to a '1'. This failure mode could make the memory location 'glitch' whilst the fuse was regrowing and you could get an unreliable/intermittent PROM value before it went permanently wrong.

I don't - however - think this is the problem in your case.

The SN74187 is mask programmed and not a 'fuse-blow' device. To change a value from 'C' (1100) to 'A' (1010) would require two simultaneous faults. One from '0' to '1' and the other from '1' to '0'. That makes the event extremely unlikely and the '1' to '0' flip doesn't follow the Ti-W failure mode.

It is interesting that the next pair of ROMs were also different... Hmmmm.... I think you may have come across a microcode 'fix' here!

I entered the microcode from the engineering drawings on bitsavers - which reports to be revision 'H'. I have just looked at the documents that came with my 11/45 and the RAC engineering drawings are revision 'F' (much earlier - but my 11/45 is quite a low serial number).

Do you have any idea what your RAC card revision is Marty (or I suppose the next best thing is the part number on the E89 ROM). E89 (Rev 'H') is supposed to be '23-C02A2' (as is Rev. 'F').

This may indicate your 11/45 ROMs may not be faulty - but have a 'fix'. We can compare the values when you have dumped them.

Carry on with the good work Marty!

Dave
 
You must log in or register to reply here.
Back
Top