Looking for Programming Specs for Hitachi 64180
Greg LaPolla
anyone knows where I can find one, I would very much appreciate it.
Thanks
David Emrich
Try www.zilog.com
David.,
Greg LaPolla <gr...@qualityteleservices.com> wrote in message
news:tfk97sc13rf0nkgv1...@4ax.com...
Richard Erlacher
<gr...@qualityteleservices.com> wrote:
>I am looking for a programmers guide for a Hitachi 64180 chip. If
>anyone knows where I can find one, I would very much appreciate it.
>Thanks
===================================================
I don't know about a programmer's guide, but I have, right here in my
lap, the HITACHI "HD64180 8-bit High-Integration CMOS Microprocessor
Data Book.
What questions do you want to answer?
I personally never had one of these, and, now that ZILOG offers a
really FAST Z-80, I'm not interested in the SLLOOOOWWWW HItachi part
as a CPU kernel for a CPU board. I'm planning to struggle through
bringing up a BIG BOARD, probably this weekend, running at 10 MHz with
1 MB of RAM on board in the form of 80ns 256K DRAMs. That 64180 won't
run that fast. I like hot-rodding the old hardware.
Dick
David Emrich
news:3876185c....@mindmeld.idcomm.com...
Funny, I've got handfuls of HD64180R10's at home. They clock at 10MHz quite
happily! In fact they'll go higher than that even though they're not
supposed to. The R6's were 6MHz, but almost everyone that used them fed
them with a 12.288MHz xtal to get nice baud rate multiples. I suspect that
there'd be no problem running the R10's at 24.576MHz for all the same
reason. After all it's only 2.5% over.
Rumour has it that there was an R16 too.
David.
Ag@Whd
<dem...@ihgtech.com.au> scribbled these words:
>It's almost exactly the same as the Zilog Z180.
>Try www.zilog.com
So how does it differ???
It's certainly similar to the Z80 (opcode fetch is half a clock cycle
earlier).
Adrian
WWW WWW Adrian Gothard
WWW ww WWW White Horse Design
WWWWWWWWWW
WWWW WWWW rot13 to reply: j...@mrgarg.pb.hx
---
Designers of GPS-based satellite tracking systems for vehicles
Richard Erlacher
ARROW's web site and found the cost of the ZILOG 20 MHz part to be
under $10, for those who aren't looking for a use for their 64180's.
Now, what do you mean by 2.5% ove?r Over what? Don't you mean
2500% ?
My first 4 MHz NMOS Z80A happily ran at 12 MHz. I know what's
possible if you operate outside the spec's. However, you get what you
deserve if you do that. (Not that it's always bad)
My comment simply points out that one can speed up his system without
adding any strange hardware. You just speed up the clock, and plug in
the appropriately specified parts. Most Z80 systems used those
rather awkward Z80 peripherals which essentially required you to
upgrade your peripherals as well as the CPU if you hot-rodded your
CPU. I've found it's not necessary. Just the memory has to be speedy
enough to keep up.
Dick
David Emrich
Ag@Whd <w...@catalpa.zetnet.zet> wrote in message
news:8unb7ssuiu0je0pfl...@4ax.com...
> On Fri, 7 Jan 2000 11:19:52 +0800, "David Emrich"
> <dem...@ihgtech.com.au> scribbled these words:
>
> >It's almost exactly the same as the Zilog Z180.
> >Try www.zilog.com
>
> So how does it differ???
Aparently, the biggest difference is what almost everyone calls a "bug" with
the 64180's interrupt handling in (I think) mode 2.
It drops a cycle (one of the acknowledge cycles from memory, this is about 6
years ago, and I haven't touched them since), I guess to improve latency,
but it meant that it didn't work at all nicely with standard Z80 peripherals
(It did of course work with all the on-chip peripherals, so the point was
moot if you did everything in the 64180 and had no external peripherals).
>
> It's certainly similar to the Z80 (opcode fetch is half a clock cycle
> earlier).
In that respect, it's identical to the Z180 (not just similar). The Z180 is
actually quite a way from the Z80, it has on-board crystal oscillator
dividers, timing control, and some neat peripherals, eg, two serial ports,
some DMA, DRAM refresh control, etc.
David
David Emrich
> For sure, HAVING the parts on hand makes a difference. I checked
> ARROW's web site and found the cost of the ZILOG 20 MHz part to be
> under $10, for those who aren't looking for a use for their 64180's.
Yeah, sad isn't it. I remember saving up for nearly a whole year to buy a
set of Z80 CPU, PIO and SIO once.
>
> Now, what do you mean by 2.5% ove?r Over what? Don't you mean
> 2500% ?
Yeah, you're right, partly, I was assuming an R12 (which AFAIK didn't
exist). For an R10, it clocked internally at 10MHz, but took a 20MHz
crystal (divided by two before clocking the CPU). So instead of 20 MHz, I
gave it 24.576MHz. That meant the core was singing at 12.288MHz instead of
10 (which is where my math went wrong, I assumed 12.288 instead of 12, which
gives 2.5% over speed). That means that it's 2.288 MHz over the limit.
Which, over 10, gives 0.2288, or 23%. In other words, it's going 23% faster
than it should. (and it survived that too! Sounds more remarkable than I
first thought).
>
> My first 4 MHz NMOS Z80A happily ran at 12 MHz. I know what's
> possible if you operate outside the spec's. However, you get what you
> deserve if you do that. (Not that it's always bad)
Grin!
> My comment simply points out that one can speed up his system without
> adding any strange hardware. You just speed up the clock, and plug in
> the appropriately specified parts. Most Z80 systems used those
> rather awkward Z80 peripherals which essentially required you to
> upgrade your peripherals as well as the CPU if you hot-rodded your
> CPU. I've found it's not necessary. Just the memory has to be speedy
> enough to keep up.
Yup. Although skewing in your address decode can cause non-linear
nightmares, since it doesn't change with increased bus speed.
David.
Jeff Jonas
>For sure, HAVING the parts on hand makes a difference. I checked
>ARROW's web site and found the cost of the ZILOG 20 MHz part to be
>under $10, for those who aren't looking for a use for their 64180's.
*whimper* When I was designing and breadboarding a Z80-A in college
in 1984, I paid maybe $10 for that, holding the more expensive
6MHz Z80-B parts for the next version (which I never got to build).
But considering the complexity of the '486, Pentium and such CPUs,
why isn't there a gigaHertz Z80 by now? :-)
>My first 4 MHz NMOS Z80A happily ran at 12 MHz.
I'm curious if there's a clear test of when to back off for unreliable
operation? Unless you really know the chip mask,
how do you know what's first to break at higher speeds?
Perhaps it's some particular instruction that stresses the CPU
more than the others, or is more time critical.
Many PC users are keen on "overclocking" but I'm yet to see a more
scientific approach on how to determine reliable operation and how
to back off the speed. I'd expect some kind of burn-in program that you
run for 10+ hours to have confidence that the system will run ALL
operations reliably.
>My comment simply points out that one can speed up his system without
>adding any strange hardware. You just speed up the clock, and plug in
>the appropriately specified parts. Most Z80 systems used those
>rather awkward Z80 peripherals which essentially required you to
>upgrade your peripherals as well as the CPU if you hot-rodded your
>CPU.
While I recall being disappointed that the Z80B at 6MHz never had all
the peripheral support of the Z80A (such as DMA),
I wonder why your experience makes you call the Z80 peripherals "awkward"?
I'll defer to your real world experience but all my Z80 architecture
designs & planning taught me that the Z80 CPU to peripheral interface
was TERIFFIC when used in mode-3 vectored interrupt!
In later years, I've salvaged a LOT of equipment that used the Z80 and SIOs.
In fact, ther dual sync/async Serial I/O chips seemed to rule
the serial port world! A friend claims the Zilog SIO chip was used
by Apple at 1MHz data rate for Apple-talk.
It was kinda the tail wagging the dog: the Z80 peripherals were
so desirable that the Z80 CPU was dragged along
and was probably the least expensive part!
--
Jeffrey Jonas
jeffj@panix(dot)com
The original Dr. JCL and Mr .hide
Douglas Beattie Jr.
:why isn't there a gigaHertz Z80 by now? :-)
There is -- it's just not a production part.. If you need a really
really fast CPU based on a standard core, you can have it for a price.
I talked to the company who made a 250_MHz 65C02 last year too.. they
told me how they did it, but wouldn't tell me who the client was...
(probably uncle Sam.)
--
Douglas Beattie Jr. http://www2.whidbey.net/~beattidp/
Shawn Sijnstra
"Douglas Beattie Jr." wrote:
>
> :But considering the complexity of the '486, Pentium and such CPUs,
> :why isn't there a gigaHertz Z80 by now? :-)
>
> There is -- it's just not a production part.. If you need a really
> really fast CPU based on a standard core, you can have it for a price.
>
> I talked to the company who made a 250_MHz 65C02 last year too.. they
> told me how they did it, but wouldn't tell me who the client was...
> (probably uncle Sam.)
Either that or someone might be planning an anniversary edition
Apple IIe or Vic 20 that runs really fast :)
Shawn
Steven N. Hirsch
<beat...@shell.whidbey.net> wrote:
>:But considering the complexity of the '486, Pentium and such CPUs,
>:why isn't there a gigaHertz Z80 by now? :-)
>
>There is -- it's just not a production part.. If you need a really
>really fast CPU based on a standard core, you can have it for a price.
>
>I talked to the company who made a 250_MHz 65C02 last year too.. they
>told me how they did it, but wouldn't tell me who the client was...
>(probably uncle Sam.)
With the proper licensing, and mucho bucks worth of synthesis and
simulation tools, you could put a 65x02 in a programmable logic
chip. Heck, maybe you could even do it at home (let me know if you do)
<g>.
Steve
Douglas Beattie Jr.
>
> Douglas Beattie Jr. <beat...@whidbey.net> wrote:
> >
> >I talked to the company who made a 250_MHz 65C02 last year too.. they
> >told me how they did it, but wouldn't tell me who the client was...
> >(probably uncle Sam.)
>
> With the proper licensing, and mucho bucks worth of synthesis and
> simulation tools, you could put a 65x02 in a programmable logic
> chip. Heck, maybe you could even do it at home (let me know if you do)
>
** ** ** ** ** ** ** ** ** ** ** ** ** **
Okay Steve... here's the URL:
http://www.free-ip.com/6502/index.htm
And I'll betcha the 64180 will be next!
Ag@Whd
<dem...@ihgtech.com.au> scribbled these words:
>> >It's almost exactly the same as the Zilog Z180.
>> >Try www.zilog.com
>>
>> So how does it differ???
>
>Aparently, the biggest difference is what almost everyone calls a "bug" with
>the 64180's interrupt handling in (I think) mode 2.
>
>It drops a cycle (one of the acknowledge cycles from memory, this is about 6
>years ago, and I haven't touched them since), I guess to improve latency,
>but it meant that it didn't work at all nicely with standard Z80 peripherals
>(It did of course work with all the on-chip peripherals, so the point was
>moot if you did everything in the 64180 and had no external peripherals).
>
>> It's certainly similar to the Z80 (opcode fetch is half a clock cycle
>> earlier).
>
>In that respect, it's identical to the Z180 (not just similar).
I said Z80 not Z180.
>The Z180 is actually quite a way from the Z80, it has on-board crystal oscillator
>dividers, timing control, and some neat peripherals, eg, two serial ports,
>some DMA, DRAM refresh control, etc.
I meant in programming terms, as that is the subject of this thread!
It has a around a dozen extra instructions, repeated string moves to I/O
ports, etc.
Richard Erlacher
<www.scenix.com>
I seriously doubt that anyone's built a 250 MHz 6502 yet. The 6502
core is a good candidate for embedded us in an FPGA/CPLD, but the
technology to do it has been available only a few months.
The 650x core is available at various sites in VHDL or Verilog format
but though the circuit operates at a high speed, it doesn't operate at
anything approaching an exact multiple of the clock rate on the old
NMOS version. The limiting factor is the speed with which one can
perform the register-to register operations THROUGH the ALU, in this
case an adder/subtractor, as the ALU in the 650x core was used not
only for arithmetic on the operand data, but for arithmetic on the
addresses as well.
The Z80 is also "out there" in VHDL and VERILOG format, but the Z80
takes about twice as much silicon resources as the 650x versions that
are out there. Moreover, the 650x cores that are available so far,
are not a reproduction of the extremely silicon-thrifty core used in
the "real-McCoy" but rather just reimplementations of the instruction
set. If one took the time to recreate the actual core design used by
MOS Technology, it would probably take well under 1k gates and
probably could be pushed to about 50 MHz in current 5-volt technology,
and perhaps to somewhere within the 100-200 MHz range in the fastest
CPLD/FPGA technology.
What you have to keep in mind is that the ALU has to consist of an
adder/subtactor, a multiplexer to operate in parallel with that but
from the same data sources and with the same destination set, and the
instructions are nearly all register-to-register moves through the ALU
with some transfers in parallel with some of the registers in the
internal register set but routed to the address bus and to/from the
data bus. I've even considered the design from the standpoint of an
implementation using a 4-bit ALU, and using it twice per instruction,
with the entire core operated from a 4x clock. That imposes its own
restrictions on speed, though.
If you visit www.6502.org you can get some pointers to various sites
which are still involved in 6502-oriented stuff. You can still buy
6502's, by the way, at speed ratings up to 14 MHz. The Z80 can still
be bought at speeds up to 20 MHz. I'd say that's going to be enough
for most folks, since the peripherals are all slower than that.
Keep in mind that a 1 MHz 6502 can probably outrun a 4 MHz Z-80 by
about 15-20%. Now multiply by the difference in clock rates available
to day and I think you'll have your hands full!
On Sun, 30 Jan 2000 17:43:54 GMT, shi...@adelphia.net (Steven N.
Hirsch) wrote:
>On Wed, 12 Jan 2000 22:18:03 -0800, Douglas Beattie Jr. \
> <beat...@shell.whidbey.net> wrote:
>
>>:But considering the complexity of the '486, Pentium and such CPUs,
>>:why isn't there a gigaHertz Z80 by now? :-)
>>
>>There is -- it's just not a production part.. If you need a really
>>really fast CPU based on a standard core, you can have it for a price.
>>
>>I talked to the company who made a 250_MHz 65C02 last year too.. they
>>told me how they did it, but wouldn't tell me who the client was...
>>(probably uncle Sam.)
>
>With the proper licensing, and mucho bucks worth of synthesis and
>simulation tools, you could put a 65x02 in a programmable logic
>chip. Heck, maybe you could even do it at home (let me know if you do)
><g>.
>
>Steve
>