Enterprise Forever
:UK => Hardware => Topic started by: gflorez on 2014.February.06. 02:10:48
-
Can this chip work in the EP?
I Remember I adapted it to my Amiga 500 copying the schematics from a Commodore 501 memory expansion.
It connects directly to the processors data bus, not by serial bus.
I will look for the Amiga schematics.
-
Here is my clumsy but functional project:
-----------------
Aquí esta mi torpe pero funcional proyecto:
-
Can this chip work in the EP?
It is possible, but exist a Clock Board for Enterprise. (http://gafz.enterpriseforever.com/Galery/Bovitmenyek/Picture/Clock_and_Calendar~1.jpg)
It is use the 146818 (or compatibe) RTC chip, which are used millions of PC motherboards.
Schematic (http://gafz.enterpriseforever.com/Hardware/PCB/Picture/Clock&Calendar-1.jpg)
Component placement. (http://gafz.enterpriseforever.com/Hardware/PCB/Picture/Clock&Calendar-2.jpg)
In my ZozoTools included handler program for this card.
-
Gyula Mészáros seems like God, he is in almost all interesting projects of that old time...
On the othe side today you are like JC..........
--------------
Gyula Mészáros parece Dios, está en casi todos los proyectos interesantes de aquellos viejos tiempos...
Pero hoy en día tú eres como JC..........
-
However, note that the official Amiga clock has fewer components.
Only three resistors, three capacitors, two diodes, a variable capacitor, a crystal, the battery and finally the chip.
It also uses only four lines of the address bus and four data bus, RD and WR, 0v and 5V.
Even I was surprised when I saw my "bad copy" working....
-----------------------
De todas formas, observa que el reloj oficial de Amiga tiene muchos menos componentes.
Solo tres resistencias, tres condensadores, dos diodos, un condensador variable, un cristal, la batería y por último el chip.
También utiliza solo cuatro lineas del bus de direcciones y otras cuatro del bus de datos, RD y WR, 0v y 5vcc.
Hasta yo me sorprendí cuando vi que mi "mala copia"me funcionaba....
-
That little expansion port at the side of the Microteam EXDOS card, can it be for a RTC?
-------------------
Ese pequeño puerto de expansión a un lado de la tarjeta EXDOS de Microteam, ¿puede ser para un Reloj en Tiempo Real?
-
However, note that the official Amiga clock has fewer components.
Yes, but I'm sure it is connecting for any predecoded port in the Amiga. If you want to build this to Enterprise also need to add address decoding circuits (U3,4,5).
If the expansion connector on your card have a +5V (at pin A4/B4) then the 7805 circuit not needed.
-
That little expansion port at the side of the Microteam EXDOS card, can it be for a RTC?
Do you think at the other side as the machine connection?
-
Do you think at the other side as the machine connection?
If yes then possible but currently nobody try it :-)
It is have a predecoded port signal and also +5V. You can omit many components of the original card.
It is at different I/O address but I can modify the clock program for you.
-
Yes, that 13+13 pins.
Okay, but first I have to finish the current project (EPROM / SRAM), only was just curious. Later I'll remind you.
------
Si ese puerto de 13+13 pines.
Bueno, pero antes tengo que terminar el proyecto actual(EPROM/SRAM), solo era curiosidad. Mas adelante ya te lo recordaré.
-
That Microteam expansion port is ideal for the RTC, as I can easily weld pins and do a insertable little card, horizontal over the EXDOS card, occupying very little space....
It appears to have all the necessary signals, except +9v but it can be taken from the EP edge connector.
-----------------------------
Ese puerto de expansión de la Tarjeta Microteam es ideal para el Reloj en Tiempo Real, ya que puedo soldar fácilmente pines y crear una pequeña placa insertable horizontalmente encima de la tarjeta EXDOS, ocupando muy poco espacio....
Parece que están presentes todas las señales necesarias, menos +9v, pero puede traerse del conector de expansión del EP.
-
Please change topic to "Real Time Clock on a Microteam EXDOS", please.
-
Done.
-
One more favour please. Can you make a scheme with all components and wires, something easy that an "Ic breaker" could follow?(while I wait...)
The diagram you brought is full of corrections, and is confusing.
------------------------
Te agradecería un favor mas. ¿Puedes hacer un esquema con todos los componentes y conexiones, algo fácil que un "rompe Integrados" pueda seguir?(mientras espero...)
Es que el diagrama que mostraste está lleno de correcciones y es confuso.
-
And, how about this?
isn`t it simpler?
-
And powering with only 5vcc. With accumulator or battery.
Sorry, the picture was taken with the Android snapshot.(pressing volume down + power)
-
Dallas DS1287 (http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=7&ved=0CF8QFjAG&url=http%3A%2F%2Fdev-docs.atariforge.org%2Ffiles%2FDS1287.pdf&ei=8v8AU8TCFqrX7Aaip4G4Cw&usg=AFQjCNH1BsYv_VYG7tgfHyciwmUTTxOs9g&sig2=Xofr3onPZLv2wUdrrVWrCg) compatible with the originally used MC146818, but including all subcircuits!
I think it is need just wiring for the MICROTEAM card expansion port!
It looks the current part number is a DS12887/DS12C887.
-
OH! thanks. Then I will equivocate less ...
Only the decoder part is now necessary. So, I can follow the original design. The resonator¿?(CD4069U) of the crystal was the part I found more difficult to understand. There are easier approaches on the Web.
Thanks again, Zozo.
-
I have a doubt about -CS as, in the original scheme is only active(-) if the two voltages (+9 v and +5 v) are present. But I have read in the data sheet of the MC146818 it must(can) be set permanently to 0v if not used.
-------------------------
Tengo alguna duda con -CS, que en el esquema original solo esta activo(-) si están presentes los dos voltajes(+9v y +5v). Pero he visto en el informe del MC146818 que se debe(puede) dejar permanentemente a 0v si no se usa.
-
On the other side, -CS is ignored with voltage lower than 4,25 .
-
It seems to me that the little 13+13 connector on the Microteam EXDOS card is for another WD177X and so to manage two more floppies.
It lacks the necessary A3 to A7(*) address lines, or a -WR signal, but there are nine pins that can be modified to suit.
B6, B7, B8, B9, B10, A6, A7, A10 and A11
--------------------
Me parece que el pequeño conector de 13+13 de la tarjeta EXDOS Microteam es para otro WD177X y por lo tanto para gobernar dos floppies mas.
Carece de las lineas de direccion necesarias A3 a A7(*), o de la señal -WR, pero hay nueve pines que pueden ser modificados para esas funciones.
B6, B7, B8, B9, B10, A6, A7, A10, y A11.
* Edit: addresses A2 to A7
-
It seems to me that the little 13+13 connector on the Microteam EXDOS card is for another WD177X
No, it is used for EPROM programmer, SIO or PIO interface, etc
and so to manage two more floppies.
The EXDOS (with original Shugart pinout) can handle four drives. Only the PC pinout limit to two drives, but you can add one Shugart standard drive as third to two PC drives.
It lacks the necessary A3 to A7 address lines
Not too important for simple devices as the RTC chip.
, or a -WR signal
It is as combined predecoded signals.
-
No, I'm not interested in more floppies, only want to follow strictly the original Mészáros design so not force you to modify your ZTools.
But....... If that means easy work for you.... can you guide me on how I can select, write and read the chip so that your program can interact with it?
The WD177X is accessed with only two address lines, we can do the same. And the -WR can be substituted with a spare negative output from the IC10 or IC11 of the Microteam EXDOS card, the "unwanted" pin signals that I mentioned earlier. But then is necessary you can access the signal via software to interact with the RTC.
I don't understand fully electronics(nor deep programming), only a little electricity and logic, so I would need you to specify which signals are more appropriate to access the clock.
Specifying, in Mészáros design U1 and U5B, you can put the address lines not present (A2-A7) to 5v and "nothing" will change?
-
Before I send you the Microteam card, I would like also send you the prototype of RTC to try if it works and then you can adapt your Zozotools.
As we only have A0 and A1 on the Microteam expansion connector, and to follow close the original Mészáros design, what do you think of changing the 74ls30(U4) with a 74LS10(three input nand gate)?
And, what if we substitute -WR by -P38 or -P30 (m or l) coming from IC10B (Microteam)?
But I still don't have A7.
I don't know if the clock can be accessed then.
-
Now I find some time for study how work the clock board:
U5B and A4 do full decode for I/O address access at ports 7Eh,7Fh
A1-A6 needed to be 1, A7 0, IORQ 0, and M1 also 1. M1 is important, because the IORQ active at Interupt accept cycle with M1. Non active M1 can identify the valid I/O operations. (Not using the M1 are the one of the bugs in the Spectrum Emulator card original version.)
U5A and U5C&D decode the port signal to separate AS and DS signals for the RTC chip.
AS will be 1 when port 7Eh accessed, DS will be 1 when 7Fh accessed.
On the MICROTEAM EXDOS card we have a predecoded signals:
Valid I/O siganal decoded with M1 and IORQ:
NW1A: 1Ah write
NO1B: 1Bh write
X19: 19h read
Y1A: 1Ah read
NI1B: 1Bh read
Just address decoded, these need additional decoding with IORQ and M1
P30: 30-37h
P38: 38-3Fh
All are active low.
For the RTC I think the best way use a XOR gates (74LS86 or compatibles) for combine the 1Ah read/write as AS, and 1Bh read/write as DS
One gate for NW1A and Y1A, if one (and only one) of them 0 then the output will be 1 as the RTC need it.
Another gate for NO1B and NI1B.
Third gate used for invert the RD signal for the RTC:
One pin the RD, another to +5V then the output are inverted go to the RTC RW pin.
When using the Dallas chip, then connect CS just to GND. It is internally disabled when no VCC.
-
Thanks for the information. That simplifies the project some more...
Then the U3 74LS245 is no more needed, is it?
-
Yes it is not needed, same present on the MICROTEAM card (near to the edge connector).
-
Then -M1 and A0-A7 doesn't needed?
Do you say something like this?:
Edit: the pins of the third gate now are correctly connected...
-
Then -M1 doesn't needed?
We use signals which are predecoded with M1. (The others (P30/P38) need a M1 and IORQ decoding)
Do you say something like this?:
Yes.
-
Sorry, I edited after you reply.
And A0-A7 is no more needed?
-
I'm still waiting for the Dallas chip, but I can build the prototype with that valuable information....
-
And A0-A7 is no more needed?
Yes.
-
The only aspect I don't like of that Dallas chip is the ten years limit of the internal lithium cell...but always there is someone that has overcome the obsolescence of a particular item:
http://www.mcamafia.de/mcapage0/dsrework.htm
Here is the entire page just in case it would disappear:
Always obsolescence built out....
-
I have a least 20 years old Dallas chip, will try to rework it.
-
At last I've ended the RTC. Here are some pictures:
https://www.dropbox.com/s/o0hr6ai9q31fz4g/1.jpg
https://www.dropbox.com/s/xoaak4q6hbsl93k/2.jpg
https://www.dropbox.com/s/8efwxlo5t1g4y9d/3.jpg
https://www.dropbox.com/s/ojrrg3w2n34hiky/4.jpg
https://www.dropbox.com/s/q29ac94wdzxr2gd/5.jpg
Or inside a "RTC on a Microteam" directory inside the FTP ftp://ftp.enterpriseforever.com (http://ftp://ftp.enterpriseforever.com)
-
Very nice! Do you tested it?
-
No, I can`t. My Microteam EXDOS hangs the EP every few seconds. On the other hand I don`t know how to access it.
Have you read your private messages?
-
No, I can`t. My Microteam EXDOS hangs the EP every few seconds.
Ok, I will look at what happened.
On the other hand I don`t know how to access it.
It is possible from BASIC with IN and OUT commands.
But a working machine is needed :oops:
Have you read your private messages?
Yes, and I'm wrote my work address to you.
-
And the clock, does it react?
Will look later, now I'm away in the weekend :oops: but I am going to modify the ZT Clock for handling the different port addresses, and see what happening on Monday :-)
-
Now tried your clock circuit, the system randomly freeze if it is attached :-( and also can't be detected.
With original clock card (form Gyula Mészáros) working fine.
Now thinking what is the problem...
-
Have you set up the clock? Seems that first you must put the oscilator to work, writting a 010 in bits 4,5 and 6(DV0,DV1,and DV2) of register A. Also you have to set bynary or BCD mode in all ten bytes of the calendar registers.
-
Oh sorry, if you can`t acccess it you can`t write at all.....
-
Good news!
Your clock card started working! Mostly :oops:
After everything virified at least 5 times, I changed the Pin 1 MOT from GND to VCC, from Intel mode to Motorola mode.
Then it is working! So the fault when I guess the Z80 (which are unofficial upgraded version of Intel 8080 and developed by ex-Intel engineers) need the Intel mode instead the Motorola :oops:
Anyway the original clock card don't use Pin 1.
Some problem left: the year randomly changed to E5h :-) I think this is software problem, something different between the original Motorola and the new Dallas RTC chips.
-
Motorola then....like the Amiga...
You are a genius...
-
I downloaded tons of datasheets for look what happened in the RTC market in the last 20 years...
Not exist any good comparison table about the MC146818 compatibles :evil: Hard work to dig out all informations.
Probably now I can understood the Dallas part numbers :oops:
-
About the Dallas family (not the Ewings :-) ) :
The origin are the Motorola MC146818 which will be very popular when started using in the IBM/AT.
It is have a total 64 bytes, 10 are the time/date, 4 control registers, 50 free for the user (PC store the Setup variables here).
Then many company produce compatible chips, Dallas Semiconductor company called their type as DS1285.
Then come the idea: why needed lot of additional components for the RTC chip? Then bend up some pins of the DS1285, and put to the IC top the crystal oscillator and lithium battery. The complete module called as DS1287. For use this in a system, only address decoding logic needed.
For the future types, also the ...85 ending the normal IC, and ...87 the complete module.
Another letter used, A at ending, for example DS1287A, this means it is have a RCLR input pin, which is clear the user memory bytes, but don't the time/date. This is the CMOS Clear jumper in PC motherboards.
At the next type, the memory size doubled, 128 totaly bytes, 114 free for user. These called as DS12885/DS12887/DS12885A/DS12887A.
Them coming the Y2K panic and Dallas make a hw hack for bad sw :-D this serie got a C letter...
The PC BIOS store the century byte at 32h address. DS12C885/DS12C887/DS12C885A/DS12C887A also use this byte as century, and load 20 to this location after 99/12/31 23:59:59... probably this function only interested at this moment :-) the datasheet don't talk about it what will happen after 2099/12/31...
The next secret in the naming the + signal, for example DS12C887A+
This is not a any interesting function, just mean ROHS compatible.
Latest serie the R series.
DS12R885 can charge rechargeable battery. But the DS12R887 only exist in a BGA version :-( In the DIP version use same battery as the older series, the charge circuit disabled, then this type called as DS12CR887... ...also a C letter, but different than the previous C, crazy naming :evil:
This serie have a RCLR pin at default, then no more ...A types.
-
Some problem left: the year randomly changed to E5h
It is only at the writing. If once set correctly, then works without problem, also on 10Mhz.
Try to find the timing difference or something :oops: in the tons of datasheet...
-
What timing difference?
-
What timing difference?
When writing RTC registers.