Q: F1 fuse replacement / SID silence
Andreas of Shape
I recently found out that a broken fuse (f1 250v 312 1-1/2a) was the
reason why the SID didn't work anymore. I need a new one, but the
question is: Does this fuse need to be exactly the same? I bought one
with the following specs:250v 8/10 a... It seemed to work, except that
one of the sound channels was silent... Any tips? I also removed cr7 and
swiched to 12v so that I could use an old 6581 SID. I have a new c64 II.
Hope someone can help me out. Thanks!
Andreas
John Iannetta
Andreas of Shape said,
"I recently found out that a broken fuse (f1 250v 312 1-1/2a) was the
reason why the SID didn't work anymore. I need a new one, but the
question is: Does this fuse need to be exactly the same? I bought one
with the following specs:250v 8/10 a... It seemed to work, except that
one of the sound channels was silent... Any tips? I also removed cr7 and
swiched to 12v so that I could use an old 6581 SID. I have a new c64 II.".
Run this program to check the three voices:
10 s=54272:fori=stoi+24:ready:pokei,y
20 next:m$=chr$(13):printchr$(147)
30 s$=chr$(19):print"press <return>."
40 geta$:ifa$<>m$then40
50 prints$:print:print
60 print"voice";a/7+1
70 pokes+a+4,17:fori=1to800:next
80 pokes+a+4,16:a=a+7:ifa=21thena=0
90 goto40
100 data0,48,0,0,0,0,240
110 data0,48,0,0,0,0,240
120 data0,48,0,0,0,0,240
130 data0,0,0,15
All three voices should sound about the same. The 0.8 A fuse should work just
fine, so long as you don't use the Datassette, or have anything connected to
the user port. Do you have a voltmeter?
--
When backing up your hard drive, shift into reverse gear S M O O T H L Y.
John
Nate Dannenberg
> question is: Does this fuse need to be exactly the same? I bought one
> with the following specs:250v 8/10 a... It seemed to work, except that
Yes. Remove this fuse *IMMEDIATELY*.
You need one that reads "250V 1A". In other words, a one-amp fuse. The above
fuse is rated at 8 amps evidently. My guess is that the 10 means a maximum of
a 10-amp surge (such as during power-up).
> one of the sound channels was silent... Any tips? I also removed cr7 and
> swiched to 12v so that I could use an old 6581 SID. I have a new c64 II.
Did you replace the capacitors attached to pins 1-4 of the SID? I forget the
reference numbers on the board, but if you look at them, they are marked "223"
and usually followed by a letter. These are 22000 pF caps, which are standard
for the 8580.
To use a 6581 they must be replaced with poly capacitors marked "471". These
are 470 pF capacitors, which are the standard for use with the 6581.
Without the proper capacitors, your filters will not work properly, which can
make one or more voices seem silent.
If you've done everything ecxactly as outlined in Commodore World (this appears
to be where you got your upgrade (downgrade?) info from), then your problem
lies in the SID chip, which must be replaced to correct the problem.
_/ _/ _/
_/_/ _/ _/_/ _/_/_/ _/_/
_/ _/ _/ _/ _/ _/ _/
_/ _/ _/ _/_/_/_/ _/ _/_/_/_/
_/ _/_/ _/ _/ _/ _/
_/_/_/_/_/ _/ _/ _/_/_/ _/ _/_/ _/_/_/ _/_/_/_/_/
natedac(at)southwind(dot)net \_/Ż\ http://www2.southwind.net/~natedac
_____________________________________________________________
| GCS d- s++:++ a-- C++ UB>++ L>+ E- W++ N++ K- w--- M- V- PS |
\ PE Y+ PGP- t+ 5 X+ R tv+(++) b+ DI(+) D+ G e+ h+ r- y- /
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
Bil Herd
John Iannetta (76703...@CompuServe.COM) wrote:
Hi John!
How are things, still doing the expert tech support after all of these years huh?
It's amazing how much I am sure I have forgotten about the tech details though I
still remember a few things like the expression on a programmer's face when you drop
a PET 2000 on his foot. Still amazes me that people continue to use the C64/C128
stuff after all of these years. Keep up the good work, catchya around.
Bil Herd
Stephen Judd
In article <6n8ebt$d1n$1...@news.jersey.net>,
As in, "Senior Hardware Design Engineer at Commodore and C128 project
leader" Bil Herd?
-S
John Iannetta
Bil Herd said,
"It's amazing how much I am sure I have forgotten about the tech details
though I still remember a few things like the expression on a programmer's face
when you drop a PET 2000 on his foot. Still amazes me that people continue to
use the C64/C128 stuff after all of these years."
Great to hear from you again!!! To readers of this Newsgroup who don't
recognize Bil's name, he was one of the designers of the C-128. And the
reference to dropping the PET on a programmer's foot no doubt referred to the
friendly rivalry between the hardware engineers (animals) and the software
crew. If I can get Bil's permission, I'll upload to comp.binaries.cbm a text
file in the libraries of the CompuServe CBMAPP forum. It is Bil's description
(edited by Marte Brengle) of those hectic weeks before the C-128 was announced.
We miss you at CBMAPP, Bil; stop around sometime.
Bil Herd
>ju...@merle.acns.nwu.edu (Stephen Judd)
>How are things, still doing the expert tech support after all of these years huh?
>still remember a few things like the expression on a programmer's face when you drop
>a PET 2000 on his foot. Still amazes me that people continue to use the C64/C128
>
>Bil Herd
>As in, "Senior Hardware Design Engineer at Commodore and C128 project
leader" Bil Herd?
-S
I've been called worse. :)
<guilty as charged>
Bil
GaryNine
In article <6nc7ga$qqs$1...@news.jersey.net>, bh...@zeus.jersey.net (Bil Herd)
writes:
> I've been called worse. :)
Certainly not by anyone in this newsgroup.
> Still amazes me that people continue to use the C64/C128
> stuff after all of these years.
It still amazes me, and from what I can tell, most other users in this
group, that you and your fellow engineers managed to create a system so
reliable and so open-ended that many of us will _never_ have to switch
computers. And those of us that do, usually come back for the sheer joy. Thank
you!
GaryNine
Cameron Kaiser
gary...@aol.com (GaryNine) writes:
> It still amazes me, and from what I can tell, most other users in this
>group, that you and your fellow engineers managed to create a system so
>reliable and so open-ended that many of us will _never_ have to switch
>computers. And those of us that do, usually come back for the sheer joy. Thank
>you!
I also saw Dennis Jarvis, part of the C65 development project, here off and on.
I have some old postings also from Fred Bowen. What happened to him?
--
-------------- The Commodore 64 lives: http://computerworkshops.home.ml.org/ --
Cameron Kaiser (posting with a Commodore 128) | "When in doubt, take a pawn."
cdkaiser@concentricMUNGEnet | -- Mission: Impossible
-- personal page: http://calvin.ptloma.edu/~spectre/ --------------------------
Stephen Judd
In article <6nc7ga$qqs$1...@news.jersey.net>,
Bil Herd <bh...@zeus.jersey.net> wrote:
>>ju...@merle.acns.nwu.edu (Stephen Judd)
>
>>As in, "Senior Hardware Design Engineer at Commodore and C128 project
>>leader" Bil Herd?
>
Well I think that's awesome. I'm very happy to metaphorically shake
your hand, and would like you to know that I use that wonderful computer
of yours all the time, and if I don't get a degree I'll hold you
directly responsible :).
Now, I have just two questions:
1. Who the hell thought up that 2MHz Z80 kludge?
2. Why the hell is the VDC interface just two bytes wide?
:) Speaking for the 128 programming community :)
evetS-
Nate Dannenberg
> Now, I have just two questions:
>
> 1. Who the hell thought up that 2MHz Z80 kludge?
>
> 2. Why the hell is the VDC interface just two bytes wide?
And I have two more (somewhat more wordy) questions:
3. Why did Commodore, for whatever reason, use one of the buggy
VIC-II chips as the model for the two versions found in the C128
and C128-DCR?
4. About the SID chips used in the C64 and C128: Why was the
decision made to change the filters and the $D418 "digi"
capability (as it is called today)? Could these not have been
left in the new design?
Yours is an excellent machine, despite the few design flaws as noted above.
With it's simple architecture that is nearly 100% compatible with the C64,
it is amazing what has been added to these machines over time. As an
example, I have taken a C128-DCR, added some enhancements to it, and a
number of peripherals, and built it all into a PC tower case. If you would
like to learn more about this tower, browse to my URL at the bottom of this
message, and take the link to "C128 Tower Info".
Other examples can be found in back issues of CMD's magazine, Commodore
World. This is an 8-issue a year magazine, dedicated to the use and
useability of the C64 and 128, and includes a few programs, hardware tips
and tricks, new, and more. Browse to http://www.cmdweb.com for more info.
Brett Tabke
Nate Dannenberg <nat...@onyx.southwind.net> wrote:
>> Now, I have just two questions:
>>
>> 1. Who the hell thought up that 2MHz Z80 kludge?
$10 that was Bowen's doing...
>> 2. Why the hell is the VDC interface just two bytes wide?
I bet we could guess a possible answer to that one Nate.
A:Cheaper,
B:easier to do because we have to wait around for the vdc side ram
refresh/update/pick nose.
C:Off the shelf - it was there: no custom ic design to be done -
drop it and go. (ever notice how a CGA card and the VDC registers
are almost identcale?)
D:The amiga is racing towards release - would you want to put out a
good gfx 8bit machine when in a short while your gonna want to push
amigas to the 64 community? It was an interum machine.
>And I have two more (somewhat more wordy) questions:
> 3. Why did Commodore, for whatever reason, use one of the buggy
> VIC-II chips as the model for the two versions found in the C128
> and C128-DCR?
> 4. About the SID chips used in the C64 and C128: Why was the
> decision made to change the filters and the $D418 "digi"
> capability (as it is called today)? Could these not have been
> left in the new design?
5: Why did the prototype 128's have 64k 80 col vid ram?
6: What was the extra rom slot "really" designed for?
7: Who wrote "The Commodore 128 Graphics Expander" that was
release by HoraSoft with a CBM copyright on it, within months
of the 128 release? (ie: was this what was to go into rom?)
8: What are you doing these days?
...ah to heck with pestering you about all of the above, how about a
couple good shop "war stories" instead? :) You know the standard
fare, "what was Fred Bowen REALLY like...lol (smile Fred).
Good to see you Bill.
Brett Tabke
--
-- Brett Tabke (phdss at writeme.com)
-- cbm hp: http://www.netins.net/showcase/phdss/
Todd S. Elliott
In article <+TAn1IIel...@netins.net>, ph...@writeme.comm wrote:
>....ah to heck with pestering you about all of the above, how about a
> couple good shop "war stories" instead? :) You know the standard
> fare, "what was Fred Bowen REALLY like...lol (smile Fred).
>
regarding the days when the c128 was being designed. Those text files are
still available at CompuServe, AFAIK.
-Todd Elliott
Bil Herd
Stephen Judd (ju...@merle.acns.nwu.edu) wrote:
: Well I think that's awesome. I'm very happy to metaphorically shake
: your hand, and would like you to know that I use that wonderful computer
: of yours all the time, and if I don't get a degree I'll hold you
: directly responsible :).
: Now, I have just two questions:
: 1. Who the hell thought up that 2MHz Z80 kludge?
: 2. Why the hell is the VDC interface just two bytes wide?
: :) Speaking for the 128 programming community :)
Hi Stephan,
Well that would fit, I never got a degree myself... high school or otherwise (damm Indiana
school system). The up side was I was an engineer by the age of 20 so I had a few years
more experience back when that mattered. I recommend that you have a standby abacus at
least if all's you have is a '128. :) What programs run on the old stuff???
Well the answer to the first question is contained below... It's something I wrote to Jim
Brain awhile back when he asked.
Now the answer to the second question is rather long and drawn out... yup, almost a story
in itself. Yup.. a seperate post ya might say...
At any rate, here is part of the story of the Z80...
From bh...@zeus.jersey.net Thu Jul 2 22:37:19 1998
Date: Mon, 20 Oct 1997 00:32:41 -0400 (EDT)
From: Bil Herd <bh...@zeus.jersey.net>
To: Jim Brain <br...@mail.jbrain.com>
Subject: Re: Z-80 in the 128 -- what for? (was: What made you want a C64/128 in the first place?) (fwd)
<Jim had asked somthing to the effect of what court ordered us to install the Z80>
It wasn't mandated by court order, it was mandated by a 23 year old
engineer that realized that marketing had gone and said that we were 100%
compatible. This turned out to be a hard nut to crack as no-one knew
what C64 compatibility meant. Companies who designed cartridges for the
C64 used glitches to clock their circutry not realizing that the glitches
were not to be depended on, etc.
The Z80/CPM cartridge didn't work on all C64's, no-one had really taken
the time to figure out why. Someone noticed that a certain brand of the
address buffer used in the CPM cart worked better than others so someone
concluded that it must be the timing parameters that made a difference.
This wasn't true, it was a very subtle problem that dealt with the way the
6502, the Z80 and the DRAM had been interlaced together. So here we had
a CPM cart that didn't work with all C64's and it worked even
less reliably with the C128 even though the timing parameters in the C128
were far better. In my opinion you couldn't call the C128 compatible
with the CPM cart as it only ran 20% of the time when tested overnight.
ALSO, I worked hard to make sure the C128 had a reliable power supply. I
was told "no fuse'..... opps one got in there by accident... in fact it
was easily accessible... darn it anyway. However, with the wide
variations in minimum and maximum power supply requirements we couldn't
handle the CPM cart, it needed an additional .5 amp because of some
wasteful power techniques that were used in it. I couldn't foot the
bill for an additional .5 amp that might only occasionally be used.
SO, with that said, I accidentally designed the Z80 into the next rev of
the board. We designed the C128 in 6 months from start to finish
INCLUDING custom silicon, these were records back then, the Z80 was added
around the second month.
NOW, we were weeding through just what C64 compatibility meant. Twice a
day someone would find some silly stupid thing that we had never even
heard of before... (color character mode with IO swapped out in 2K ultra
max mode???). Three cartridges that were developed by Commodore fell into
the batch of things that didn't work, the first was "Wizard of Wor" and
"Solar System"... after looking at it for 4 hours I was starting to get
concerned, I usually could fix anything in four hours or particularly
hard problems overnight. (More problems got fixed between 1:00am after
the bars closed and 6:00am when management wasnt around.) Finally I
tried them on C64s which were hard to find as everyone stole them and
took them home. Whew it crashed there too, within seconds. This was
enough to prompt the head of the IC design group who suddenly remembered
that there had been a design flaw in the cartridge ROMs the good news was
we shipped the same ROMs to Atari for their Christmas rush.
NEXT was the Commodore Speech Cartridge. It just plain crapped when plugged
in. Oh oh, this was designed by "engineers", "texan engineers" no less.
Turns out the cartridge would monitor the bus and when it saw the C64 try and
load its reset vectors it would suddenly SLAM the game cartridge line down
which, with luck, would swap out the C64s onboard Roms and replace key values
with their own. Now these signals were never meant to be toggled in real
time, they were sensed when the system was reset and that was it usually.
This slowed us down by about an hour in the development of the C128. I
decided to go ahead with the Z80 mod. The only Z80s we had were in the
Sinclair computers that we used to use for doorstops. My prototype that sits
in my basement still says Timex on the Z80. Now I knew the Z80 booted from a
different spot in memory than the 6502 did (C64 processor). We wrote the
code for the Z80 by hand punching it in to an EPROM burner at 3:00 in the
morning. It worked by starting up and looking at the cartridge port with out
tripping the hated "grab the bus" mechanism that crashed the C128 before it
could initialize the MMU. When it saw the cartridge sitting out there like
an ogre on the doorstep the Z80 would initialize the MMU into C64 mode so
that the cartridges instructions would make sense. This mode is also forced
whenever the Commodore key is held down during boot. So it turns out that
the Z80 was vital in making the C64 compatible.
1)So now the Z80 made it so that we didnt have to work with the flawed
CPM cart and keep our lofty goal of 100% capatibility.
2)The Z80 also made it so we could boot any C64 cartridge cleanly into
C64 no matter what sillyness lay beyond the cartridge interface.
3)Having the Z80 on board allowed us to keep a good quality switching
power supply as it didnt have to be bigger than we needed.
4) Commodore Australia called us to tell us that they would personally
rip every Z80 off of each mother board if we insisted in shipping with
the Z80. This was enough assurance for me ! ;)
You have to watch what you ask me... there is a story for EVERYTHING.
Bil
-------------------------------------------------------------
|Bil Herd, President |Personal and Commercial Accounts|
|----------------------------| Leased Lines - Dedicated ISDN |
|InterActive Network Systems | Web Hosting - Web Authoring |
|----------------------------| Network Consulting |
|(609) 227-4428 | |
|(609) 227-3943 fax | http://www.jersey.net |
|--------------------------------------------------------------
|Serving Burlington,Camden,Gloucester,Mercer&Atlantic Counties|
-------------------------------------------------------------
Bil Herd
Todd S. Elliott (ey...@erols.com) wrote:
: -Todd Elliott
Fred was, well the word is, short. He is still short to this day I believe. He used to
wear these silver balled "deely boppers" (like the killer bees wore on SNL) only his were
all glittery and such. Managment hated them. We liked them because since Fred was shorter
than the Herman Miller cubicle furniture, we could tell where he was by the deely boppers
wondering around, kinda like those flags on dune buggies. He would also make beep noises
sometimes.
As Terry Ryan once said, (about 1 month ago actually), all I learned about parenthood I
learned from Fred Bowen... all I learned about Fred Bowen I learned from Mad Magazine.
Bil
Bil Herd
Nate Dannenberg (nat...@onyx.southwind.net) wrote:
: 3. Why did Commodore, for whatever reason, use one of the buggy
: VIC-II chips as the model for the two versions found in the C128
: and C128-DCR?
We had about 6 months to do the design and 4 custom IC's AND if we did anything too
different we would loose compatibility.... Ask me about the time we moved the dot on the
letter I in the font rom and crashed a program... go ahead, ask.
: 4. About the SID chips used in the C64 and C128: Why was the
: decision made to change the filters and the $D418 "digi"
: capability (as it is called today)? Could these not have been
: left in the new design?
This would have had to been after my time, I took the SID chip in my wirewrap prototype out
of the guy next to my office's C64. I warned him not to leave that damm Jumpman music
playing all night.
: Yours is an excellent machine, despite the few design flaws as noted above.
: With it's simple architecture that is nearly 100% compatible with the C64,
: it is amazing what has been added to these machines over time. As an
: example, I have taken a C128-DCR, added some enhancements to it, and a
: number of peripherals, and built it all into a PC tower case. If you would
: like to learn more about this tower, browse to my URL at the bottom of this
: message, and take the link to "C128 Tower Info".
The only compatibility problem I knew of was my fault for not being paranoid enough. I
left the 2mhz bit in the Vic chip visable in C64 mode. I could have made it "go away" but
didn't cause we I was feeling good about things that day instead of paranoid. That'll
teach me.
: Other examples can be found in back issues of CMD's magazine, Commodore
: World. This is an 8-issue a year magazine, dedicated to the use and
: useability of the C64 and 128, and includes a few programs, hardware tips
: and tricks, new, and more. Browse to http://www.cmdweb.com for more info.
We had a better MMU ready but was kept from putting it in... would have been 1mbyte support
native. We did have one of the earlier DMA engines though... from Basic no less.
Bil
Martijn van Buul
It occurred to me that Bil Herd wrote:
> Ask me about the time we moved the dot on the letter I in the font rom and
> crashed a program... go ahead, ask.
>
causes some programs to crash)(ah well... Some geeks might have been using
the character ROM as program code....)
> The only compatibility problem I knew of was my fault for not being paranoid
> enough. I left the 2mhz bit in the Vic chip visable in C64 mode. I could
> have made it "go away" but didn't cause we I was feeling good about things
> that day instead of paranoid. That'll teach me.
That's not a bug, it's a feature which enables me to run Elite a bit more
smoothly :)
--
Martijn van Buul, mart...@mud.stack.nl
Tijntje@OuterSpace - 131.155.141.166 3333
Nicolas Welte
Bil Herd wrote:
>
> The only compatibility problem I knew of was my fault for not being paranoid enough. I
> left the 2mhz bit in the Vic chip visable in C64 mode. I could have made it "go away" but
> didn't cause we I was feeling good about things that day instead of paranoid. That'll
> teach me.
>
What about the CAPS/LOCK bit in the processor data register? Some C64
programs do only run with CAPS/LOCK pressed. And I guess some software
also doesn't like the missing SID chip mirrors in the $D500-$D7FF area.
But it still is a great machine, I love mine :-) And there are still so
many unknown features in this machine, at least for me.
Nicolas
Nicolas Welte
Brett Tabke wrote:
> >> 2. Why the hell is the VDC interface just two bytes wide?
>
> A:Cheaper,
> B:easier to do because we have to wait around for the vdc side ram
> refresh/update/pick nose.
Actually I think it would've been possible to interleave CPU and VDC
accesses to the video RAM, if the VIC chip wouldn't have been in the
system. I think all 6845 based video systems (see also next question)
allow direct access to video RAM from the CPU. The problem is, the VIC
chip already uses the unused clock cycle of the CPU to refresh RAM and
get it's graphics data from there, so there wasn't any time left to map
in VDC RAM accesses. Perhaps there were also problems with synchronizing
the switchable 1/2 MHz CPU clock and the also switchable 1/2 MHz VDC
character clock. Who knows? Perhaps Bil Herd ;-)
> C:Off the shelf - it was there: no custom ic design to be done -
> drop it and go. (ever notice how a CGA card and the VDC registers
> are almost identcale?)
No surprise. CGA (and Hercules) cards use the 6845 chip, which looks
like a predecessor of the MOS 6545 chip, which has even more
similarities to the VDC chip. But while both of these need to have a lot
of external logic, the VDC is a fully integrated color graphics chip
with text and graphics modes in both monochrome and color.
The 6545 and 6845 data sheets are at funet and example schematics for a
6845 implementation can be found at Andre Fachat's page.
Nicolas
Nate Dannenberg
> We had about 6 months to do the design and 4 custom IC's AND if we did
> time we moved the dot on the letter I in the font rom and crashed a
> program... go ahead, ask.
This makes sense. Actually the "buggy" refers to the VIC's habit of
producing dots now and then when you access a register. Example... Execute
this two-liner on a C128 and watch the screen fill with a pattern of dots,
present until you stop the loop:
10 POKE 53265,32 (to blank the screen)
20 POKE 53280,0: GOTO 20
The screen is supposed to just turn black, but the VIC likes to insert a
little grey dot right on the edge of the final write cycle that stores the 0
to 53280 ($D020). You can imagine what the screen does when it's very
active.
> This would have had to been after my time, I took the SID chip in my
> wirewrap prototype out of the guy next to my office's C64. I warned him
> not to leave that damm Jumpman music playing all night.
Actually this was Mr. Yannes' field, however I figured you might know why
the choce was made to move to the 8580 even though it lacked the
undocumented features the 6581 had, which includes the two mentioned above.
Btw, thanks.... Now I've got that music stuck in my head! :-)
> The only compatibility problem I knew of was my fault for not being
> paranoid enough. I left the 2mhz bit in the Vic chip visable in C64 mode.
> I could have made it "go away" but didn't cause we I was feeling good
> about things that day instead of paranoid. That'll teach me.
Actually this feature has found many uses; C64 programs that can benefit
from a little added speed when used on a C128, take advantage of this.
Novaterm 64 v9.6, for example, can use the C128's VDC chip from C64 mode,
and when doing so, it also uses the C128's 2MHz mode.
> We had a better MMU ready but was kept from putting it in... would have
> been 1mbyte support native. We did have one of the earlier DMA engines
> though... from Basic no less.
Why would they not allow this 1MB MMU to be used? In retrospect, a hack has
since been developed by Marko Mäkelä (on this newsgroup), to provide this
capability, up to 1MB of internal RAM, much of which is usable even from
BASIC.
_/ _/ _/
_/_/ _/ _/_/ _/_/_/ _/_/
_/ _/ _/ _/ _/ _/ _/
_/ _/ _/ _/_/_/_/ _/ _/_/_/_/
_/ _/_/ _/ _/ _/ _/
_/_/_/_/_/ _/ _/ _/_/_/ _/ _/_/ _/_/_/ _/_/_/_/_/
natedac(at)southwind(dot)net \_/¯\ http://www2.southwind.net/~natedac
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Nate Dannenberg
> Nate Dannenberg <nat...@onyx.southwind.net> wrote:
> >> Now, I have just two questions:
> >>
> >> 1. Who the hell thought up that 2MHz Z80 kludge?
>
>
>
> I bet we could guess a possible answer to that one Nate.
Actually, you would be answering these for Steve Judd, as Bil already did.
The two following these were my questions.
I think Steve wanted to know why they chose a two-register format, as
opposed to directly mapping the 37 or so registers the VDC actyually had.
Hence, why use an I/O port style of access (this is what it reminds me of,
having coded a few VDC things in the past), rather than mapping the VDC's
registers (I think it has 37 of them) right into the C128 memory map.
One normally does:
LDA #$register
STA $D600
loop BIT $D600
BPL loop
LDA #$data
STA $D601
It would have saved considerable time and lots of program code to simply be
able to LDA #$data: STA ($D600+#$register), like you do with the SID chip
and other I/O devices.
_/ _/ _/
_/_/ _/ _/_/ _/_/_/ _/_/
_/ _/ _/ _/ _/ _/ _/
_/ _/ _/ _/_/_/_/ _/ _/_/_/_/
_/ _/_/ _/ _/ _/ _/
_/_/_/_/_/ _/ _/ _/_/_/ _/ _/_/ _/_/_/ _/_/_/_/_/
natedac(at)southwind(dot)net \_/Ż\ http://www2.southwind.net/~natedac
_____________________________________________________________
| GCS d- s++:++ a-- C++ UB>++ L>+ E- W++ N++ K- w--- M- V- PS |
\ PE Y+ PGP- t+ 5 X+ R tv+(++) b+ DI(+) D+ G e+ h+ r- y- /
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ
Local H/BCI/PSW/SYS
>This makes sense. Actually the "buggy" refers to the VIC's habit of
>producing dots now and then when you access a register. Example...
Execute
>this two-liner on a C128 and watch the screen fill with a pattern of dots,
>present until you stop the loop:
>
>10 POKE 53265,32 (to blank the screen)
>20 POKE 53280,0: GOTO 20
>
>The screen is supposed to just turn black, but the VIC likes to insert a
>little grey dot right on the edge of the final write cycle that stores the
0
>to 53280 ($D020). You can imagine what the screen does when it's very
>active.
>
$C000 and make necessary changes):
LDA #$20
STA $D011
LDA #$00
STA $D020
JMP $1307
then do 'FAST:SYS4864' from BASIC. Looks like a nice pattern of dots. The
only thing that bugs me about the so-called 'raster-sparkle' is that if I
run any routine that is not tied to a stable raster IRQ and it happens to
change a register in the visible screen area, then it looks horrible. Of
course, I try to time my unstable routines to change in the off-screen area
:)
Bil Herd
On Fri, 3 Jul 1998, Nate Dannenberg wrote:
> > We had about 6 months to do the design and 4 custom IC's AND if we did
> > anything too different we would loose compatibility.... Ask me about the
> > time we moved the dot on the letter I in the font rom and crashed a
> > program... go ahead, ask.
>
> This makes sense. Actually the "buggy" refers to the VIC's habit of
> producing dots now and then when you access a register. Example... Execute
> this two-liner on a C128 and watch the screen fill with a pattern of dots,
> present until you stop the loop:
>
> 10 POKE 53265,32 (to blank the screen)
> 20 POKE 53280,0: GOTO 20
There were several worse problemes at the time, we had to work carefully
to make the CES show that January. We had sparkles in Multi-color
character mode (or some mode like that, it was 14 year ago) and "tear" in
80 column mode. We found that if you adjusted the power supply down to
4.75 volts the MCC sparkles went away and if we turned it to 5.25V we
stood a chance of getting the 80 column chip to work. Good thing that we
were using adjustable supplies. ;)
>
> The screen is supposed to just turn black, but the VIC likes to insert a
> little grey dot right on the edge of the final write cycle that stores the 0
> to 53280 ($D020). You can imagine what the screen does when it's very
> active.
>
In the real old days I used to call this "thump" after a bad trait that
electronic organs used to have (When you pressed a key, part of the charge
used to turn on the audio gate ended up getting into the audio gate
itself, hence a thump could be heard at the beginning of each note.)
The amazing thing about the VIC chip is that it works. These were the
days when there were NO electronic checking programs to ensure that the
chip matched the schematic and vice versa. They would literally spend
months crawling around tables pushed together with huge pen plots,
measuring the lines that represented the different layers of the chip with
a little hand ruler. You could be a brilliant designer but your chip
could suck or not work of you didn't have the patience to carefully check
every transistor on the chip. A mind numbing job. Surprisingly, the
really good IC designers were not only very good at design but also
checking. My brain would have locked up after checking 40-50 thousand
transistors, I don't know how they did it.
> > This would have had to been after my time, I took the SID chip in my
> > wirewrap prototype out of the guy next to my office's C64. I warned him
> > not to leave that damm Jumpman music playing all night.
>
> Actually this was Mr. Yannes' field, however I figured you might know why
> the choce was made to move to the 8580 even though it lacked the
> undocumented features the 6581 had, which includes the two mentioned above.
>
I remember a little more now, I believe they replace the single pole
simple filter with a digital filter, probably a switched cap filter.
There are several reasons why hanging a big ole cap off of a chip isn't as
desirable as later solutions. (For example: when you turn off the power,
where does the charge on the cap go... not to mention the voltage on the
cap now excedes the VCC voltage on the SID chip)
Also, the 8580 was newer tech, more producable, more testable. Testable
is very important.
> Btw, thanks.... Now I've got that music stuck in my head! :-)
>
> > The only compatibility problem I knew of was my fault for not being
> > paranoid enough. I left the 2mhz bit in the Vic chip visable in C64 mode.
> > I could have made it "go away" but didn't cause we I was feeling good
> > about things that day instead of paranoid. That'll teach me.
>
> Actually this feature has found many uses; C64 programs that can benefit
> from a little added speed when used on a C128, take advantage of this.
> Novaterm 64 v9.6, for example, can use the C128's VDC chip from C64 mode,
> and when doing so, it also uses the C128's 2MHz mode.
>
> > We had a better MMU ready but was kept from putting it in... would have
> > been 1mbyte support native. We did have one of the earlier DMA engines
> > though... from Basic no less.
>
> Why would they not allow this 1MB MMU to be used? In retrospect, a hack has
> since been developed by Marko Mäkelä (on this newsgroup), to provide this
> capability, up to 1MB of internal RAM, much of which is usable even from
> BASIC.
>
Politics. It was designed, in fact we advertised the expandability...
hence a dma engine was involked.
Bil Herd
Nate Dannenberg
> stood a chance of getting the 80 column chip to work. Good thing that we
> were using adjustable supplies. ;)
heh.. indeed
> checking. My brain would have locked up after checking 40-50 thousand
> transistors, I don't know how they did it.
I'm barely able to draw small 5 or 6-chip schematics, let alone read LSI
diagrams.
> desirable as later solutions. (For example: when you turn off the power,
> where does the charge on the cap go... not to mention the voltage on the
> cap now excedes the VCC voltage on the SID chip)
Hrm.. I've never considered this.. Perhaps the SID has one or more bleeders
in it for this.
> Politics. It was designed, in fact we advertised the expandability...
> hence a dma engine was involked.
Figures as much. By DMA engine, does this imply the presently used DMA
features of the 17xx series REU's? I think I'm missing something here.
Nate Dannenberg
> True, but on some machines this is VERY important. Example: my C128-DCR
> (my spare, not my tower) has two 6581's in it. It produces very low output,
Ack! I meant to say my C128-DCR (spare) has *8580's* in it, NOT 6581's!
Andy Finkel
>Nate Dannenberg (nat...@onyx.southwind.net) wrote:
>: 4. About the SID chips used in the C64 and C128: Why was the
>: decision made to change the filters and the $D418 "digi"
>: capability (as it is called today)? Could these not have been
>: left in the new design?
>
>This would have had to been after my time, I took the SID chip in my wirewrap prototype out
>of the guy next to my office's C64. I warned him not to leave that damm Jumpman music
>playing all night.
If I recall correctly, the filter changes were an attempt to make the filter
characteristics predictable and uniform; the original SID filters were
implemented with straight resistors, rather than a ladder resistive network,
so they tended to vary from lot to lot. (I remember Bob Yannis telling me
that it was one of his few regrets in the original SID chip design, in
response to my bitching about doing some speech SW). Of course, its been a
few years, memories fade, and I was just doing the software :-)
andy
Microsoft is a "scrappy" company in exactly the same sense that Godzilla is a 'scrappy' monster.
Andy Finkel
>It occurred to me that Bil Herd wrote:
>> Ask me about the time we moved the dot on the letter I in the font rom and
>> crashed a program... go ahead, ask.
>>
>Euh.... What about that story!? (I can hardly believe that changing a _font_
>causes some programs to crash)(ah well... Some geeks might have been using
>the character ROM as program code....)
It's 100% true. A major paint package was reading the font out of rom to make
the characters in its title screen. When we moved the dot in the A, the fill
routine in the paint program went insane and crashed.