Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

TransWarpGS Clone

691 views
Skip to first unread message

Conrad

unread,
Nov 18, 2012, 5:47:32 PM11/18/12
to
I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively unchanged. I am in the process of cloning the TransWarpGS. I should have a Gerber file ready by February. The first step will be a direct clone to get it to market as soon as possible. The FPGA used on the original card is possible to get right now, but the supply will dry up before long, so the next step will be to reengineer it to use a more modern FPGA, to make it a bit more futureproof. Keep in mind the process has already begun and there are no barriers to it continuing unhindered as of yet, so expect this to be ready before long. The IIGS has always been close to my heart, and I'm tired of seeing so many people be barred for using this machine to its full potential by the insane cost of TWGS or ZipGSX. I'd like to hear from anyone who would be interested in ordering a TransWarpGS when they are ready for sale, so I can get an idea of interest. Looks like initial target price might be around CAD$150-CAD$175. Before anyone starts with skepticism or hostility let me just say now that I absolutely understand your reactions. An announcement like this probably seems too good to be true, but it isn't! After my troubles with my recently purchased TransWarpGS (my ROM eraser hasn't arrived yet so I haven't started the repair process yet) I realized that there are so many people out there who would love to have an accelerated GS but don't want to pay the going rate for an accelerator. I have found a way to make this project actually possible, and the process has already begun. I don't want to say a lot now until I have some things nailed down, but expect a full explaination shortly. So guys, let's hear some questions and comments!

Conrad

D Finnigan

unread,
Nov 18, 2012, 6:21:02 PM11/18/12
to
I'm definitely interested.

I think I speak for everyone here when I say that I hope you succeed! :-)


Christopher G. Mason

unread,
Nov 18, 2012, 7:14:17 PM11/18/12
to
On 11/18/2012 5:47 PM, Conrad wrote:
> I have some news for the Apple IIGS community.

Hopefully this goes further then that "GSBarndoor" project. Those guys
were all talk, no action (or just plain trolls). That price would be a
bargain if it is cheaper then the going ebay rate ($350-400 nowadays!).
Assuming someone managed to crack the FPGA on the ZipGSX, it would be
easier to clone in theory, but I'm a TWGS fan myself (gotta love that
built-in CDA). The discrete logic gives it an edge in overclocking. With
modern parts, the heat and power use should go down.

gid...@sasktel.net

unread,
Nov 18, 2012, 7:36:58 PM11/18/12
to
On Sunday, November 18, 2012 4:47:32 PM UTC-6, Conrad wrote:
> I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively unchanged. I am in the process of cloning the TransWarpGS. I should have a Gerber file ready by February. The first step will be a direct clone to get it to market as soon as possible. The FPGA used on the original card is possible to get right now, but the supply will dry up before long, so the next step will be to reengineer it to use a more modern FPGA, to make it a bit more futureproof. Keep in mind the process has already begun and there are no barriers to it continuing unhindered as of yet, so expect this to be ready before long. The IIGS has always been close to my heart, and I'm tired of seeing so many people be barred for using this machine to its full potential by the insane cost of TWGS or ZipGSX. I'd like to hear from anyone who would be interested in ordering a TransWarpGS when they are ready for sale, so I can get an idea of interest. Looks like initial target price might be around CAD$150-CAD$175. Before anyone starts with skepticism or hostility let me just say now that I absolutely understand your reactions. An announcement like this probably seems too good to be true, but it isn't! After my troubles with my recently purchased TransWarpGS (my ROM eraser hasn't arrived yet so I haven't started the repair process yet) I realized that there are so many people out there who would love to have an accelerated GS but don't want to pay the going rate for an accelerator. I have found a way to make this project actually possible, and the process has already begun. I don't want to say a lot now until I have some things nailed down, but expect a full explaination shortly. So guys, let's hear some questions and comments!
>
>
>
> Conrad


What kind of speeds do you hope to achieve? Is 40 Mhz illogical?

Rob

Sam Latella

unread,
Nov 18, 2012, 7:59:08 PM11/18/12
to
Count me in for 1 possibly 2!!!

Excellent News!

Sam


Bill Garber

unread,
Nov 18, 2012, 8:01:38 PM11/18/12
to

<gid...@sasktel.net> wrote in message news:bb9a6b75-6d40-44dc...@googlegroups.com...
On Sunday, November 18, 2012 4:47:32 PM UTC-6, Conrad wrote:
>> I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement
>> and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively
>> unchanged. I am in the process of cloning the TransWarpGS. I should have a Gerber file ready by
>> February. The first step will be a direct clone to get it to market as soon as possible. The FPGA
>> used on the original card is possible to get right now, but the supply will dry up before long, so the
>> next step will be to reengineer it to use a more modern FPGA, to make it a bit more futureproof.
>> Keep in mind the process has already begun and there are no barriers to it continuing unhindered
>> as of yet, so expect this to be ready before long. The IIGS has always been close to my heart, and
>> I'm tired of seeing so many people be barred for using this machine to its full potential by the insane
>> cost of TWGS or ZipGSX. I'd like to hear from anyone who would be interested in ordering a Trans-
>> WarpGS when they are ready for sale, so I can get an idea of interest. Looks like initial target price
>> might be around CAD$150-CAD$175. Before anyone starts with skepticism or hostility let me just
>> say now that I absolutely understand your reactions. An announcement like this probably seems
>> too good to be true, but it isn't! After my troubles with my recently purchased TransWarpGS (my
>> ROM eraser hasn't arrived yet so I haven't started the repair process yet) I realized that there are
>> so many people out there who would love to have an accelerated GS but don't want to pay the
>> going rate for an accelerator. I have found a way to make this project actually possible, and the
>> process has already begun. I don't want to say a lot now until I have some things nailed down,
>> but expect a full explaination shortly. So guys, let's hear some questions and comments!
>>
>> Conrad
>
> What kind of speeds do you hope to achieve? Is 40 Mhz illogical?
>
> Rob

For the sake of being reasonable, let's see it getting done before we ask for astronomical feats. 8o)

Bill Garber
http://www.sepa-electronics.com/



Conrad

unread,
Nov 18, 2012, 8:36:09 PM11/18/12
to
@Bill: My thoughts exactly. I don't want people to start thinking this will be a miracle device. Thank you.

@Rob: The first edition will be a straight up clone of the TWGS, so the current ceiling on speed will apply. Don't forget that the '816 as currently available could never come close to reaching 40MHz. Its possible that when I take the next steps in doing some reengineering, if and only if we can develop an '816 softcore for an FPGA (no opensource ones are currently available, although WDC does offer an RTL '816) that the current ceiling on speed could be surpassed. That is way in the realm of uncertainty though. I'm not even going to be thinking of things like that until the first version is out the door.

@Christopher: Don't worry, I am definitely not all talk. Like I said, the process has already begun. The problem with the ZipGSX is that it doesn't use an FPGA but rather an ASIC (for those unfamiliar with the term, Application Specific Integrated Circuit, in laymans terms a custom chip). The steps involved in reverse engineering just the ASIC are far more complicated then the whole TWGS cloning project. Maybe when the whole TWGS project is concluded we can think about going down that road, but honestly I don't see a point if the TWGS is available again. Maybe a project to just reverse engineer the ASIC so people can ressurrect cards that have died from fried ASICs, but again that's very far in the future if it even happens at all so I'd rather not even start going down the road of discussing it right now, I just want to focus on the current project.

@D Finnigan: Thanks for the support! I'll keep you updated and let you know as soon as preordering starts.

@Sam: Excellent! I'll make sure you are notified as things progress and as soon as you can place a preorder!

magnusfalkirk

unread,
Nov 18, 2012, 8:57:35 PM11/18/12
to
Count me in for at least one as well! A TranswarpGS is all I need to make my GS a perfect machine, at least in my humble opinion.

Steven Hirsch

unread,
Nov 18, 2012, 8:57:52 PM11/18/12
to
On 11/18/2012 05:47 PM, Conrad wrote:

> The FPGA used on the original card is possible to get right
> now, but the supply will dry up before long, so the next step will be to
> reengineer it to use a more modern FPGA, to make it a bit more futureproof.

Unless the bitstream for that late-80s Xilinx part is directly applicable to
more recent FPGAs, you have your work cut out for you. There have been some
steps taken towards reverse engineering raw bitstreams into logic
architecture, but it's anything but trivial.

All that said, I wish you the best of luck. It would be great to see this
come about!

In theory, I would think that a modern GS accelerator could be built with a
WDC 65816, a recent generation Xilinx or Altera part and a couple of glue
chips. If I had more free time and about half the knowledge of the Mike
Mahons and Alex Freeds of the world, I'd be glad to help :-).

Steve

Egan Ford

unread,
Nov 18, 2012, 9:37:54 PM11/18/12
to
On 11/18/12 5:47 PM, Conrad wrote:
> so I can get an idea of interest

Interested.

Bill Garber

unread,
Nov 18, 2012, 10:31:25 PM11/18/12
to

"magnusfalkirk" <dean....@gmail.com> wrote in message news:374a3673-bf64-4579...@googlegroups.com...
Me too, although I prefer the ZipGSX, if you make a good solid TWGS, I'd be more than happy to front you $350
for two of them, shipped. I've had both TWGS and ZipGSX, and the Zip was the more stable card by far, unless
it (the TWGS) was just messed with too much. Well, I was able to get all of my accelerators to run with stability,
and sell them off. Now I wish I had kept one of each, among other things that I have sold off, including a bunch
of 'Apple 3.5" drive interface' cards (Superdrive controllers). BTW, 8MHz is plenty for me, not a speed junkie. 8o)

Bill Garber
http://www.sepa-electronics.com/



Conrad

unread,
Nov 19, 2012, 12:09:59 AM11/19/12
to
@Steve: You're absolutely right. I very much doubt there is a new device that is compatible with the raw bitstream, although its possible there is a more available, slightly less old device that is, which could be an intermediate step. When I was talking about reengineering, I was thinking more along the lines of using a logic analyzer to study the interaction between the FPGA and the rest of the device. The FPGA used on the TWGS is VERY basic, in fact I believe it was Xilinx's first commercial FPGA. I think it only has 64 logic elements. This should make it much easier to deal with from a reverse engineering perspective, since there can't be that much functionality in there. If all else fails there are services out there that you could send a IIGS with a functioning TWGS installed to and who would send you back VHDL/Verilog code, for a price of course. This is always an option if nothing else works. Don't worry, I'm not under the impression that having a bitstream is equivalent to having the VHDL/Verilog code. I think you are right that it wouldn't be that hard to develop a new accelerator, but this is even easier and will get a product out the door even faster. :P

@Bill: From my understanding the ZipGSX will always be more stable if you run it with an older '816, since the ZipGSX ASIC does a better job of correcting the deficiencies of the older '816 die then the TWGS GALs. From what I understand, if you use one of the Sanyo-designed '816s with updated GALs and 32k of fast (<20ns) cache, the performance should be equal up to a certain MHz, but the Zip will probably top out 1MHz-2MHz faster than the TWGS when overclocking. When you say you will front me 350 dollars for 2 new TWGS cards shipped, do you mean you are willing to prepay? I am not ready to accept prepayments yet, but when I am that would definitely be awesome, since it would certainly help pay for the materials and processes and get the cards out the door faster. I will let you know when I am set up for things like that. That goes for anyone else as well, if people are wanting to prepay to get a card from the first batch that would help get everything done faster, but again I'm not ready to accept payments yet however I will let people know as soon as I am. Also, if anyone wants to lend a TransWarp GS to me, that will help get things done faster, since my card is currently malfunctioning and that could cause a problem with the reverse engineering process. Any card that someone lends will not be destroyed, the reverse engineering process is non-destructive, and I would return it when finished along with a brand new TWGS for the trouble, since its cheaper for me to borrow one and give a new card for the trouble than it is for me to buy a new working TWGS for the purpose.

@magnus and @egan: Excellent! I will add you both to the list of interested parties!

Conrad

unread,
Nov 19, 2012, 12:09:59 AM11/19/12
to

Antoine Vignau

unread,
Nov 19, 2012, 1:22:05 AM11/19/12
to
Dear Conrad,
Please count me in for two.

I've sold all my TWGS, I would have been delighted to send one to you.

Good luck with the project,
Antoine

Hugh Hood

unread,
Nov 19, 2012, 1:26:18 AM11/19/12
to
in article 1ab297d4-0278-4a7f...@googlegroups.com, Conrad at
conrad...@gmail.com wrote on 11/18/12 7:36 PM:

> @Rob: The first edition will be a straight up clone of the TWGS ...


Conrad:

I realize that your initial goal (and a laudable one) is to create a
"straight up clone", but if at all possible for either the initial 'clone'
or for a future revision, the TransWarpGS, regardless of running speed,
_really_ could use significant improvements to its power consumption and
heat generating specs.

Those (2) things (which the ZipGSX has in its favor) strongly impact the
reliability, not only of the accelerator itself, but also of the entire
machine when you consider the effects of power/heat on the other cards. Yes,
fans help, but ...

I run my IIGS's (and have been for almost 20 years now ) in our office 5
days a week from 9-5 and the Zip's have never gone south on me. I wish I
could say the same for my one and only TransWarpGS. I use it intermittently
and keep a special System Saver GS with a big (super-loud) fan in it just to
keep it usable. It's stable at 10 MHz, but man it can cook!

Figure out a way to beat the heat (and power), and I'll beat a path to your
door for several.

That's because I've got back-up parts for just about all my stuff _except_
for the accelerators. I've bought a few 'used' Zips over the years for $$,
but unfortunately, I wouldn't count any of them as being reliable.

It seems almost every one of them has experienced some sort of failed
'upgrade' attempt on the part of the previous owner(s), with the result
being damage to the RAM sockets and/or oscillator socket. I've even got one
where the owner cut the trace supplying 5 Volts to the processor and instead
is running the 12 Volt line through some sort of resistance to drop it down
to about 6 Volts for the processor. This one also has reliability issues.
The irritating thing is that even these untrustworthy ZipGSX's were very
pricey. I was going send them all off to Henry for repair, but that seems
unlikely now.

Good luck with the project. Keep us posted as you proceed.





Hugh Hood

P.S. Has anyone ever tried to track down any of the old Zip folks to see if
the ASIC spec is laying around somewhere collecting dust awaiting
resurrection? I know it was patented, but how specific did the patent
application get?


Conrad

unread,
Nov 19, 2012, 2:56:26 AM11/19/12
to
@Hugh: First off, I can do repairs on those Zip cards for you, its absolutely no problem at all. I know how to repair things like pulled traces and obliterated through holes. I have things like circuit writer and silver epoxy, so I can fix them up no problem. I just wish the TWGS I bought had such an obvious problem. Hopefully when I upgrade all the outdated stuff on it then it will come back to life. Anyway, in regards to your comments, the first thing I am going to do will literally be a perfect copy, and the reason for that is I know how easy it is for a few little changes to snowball into a stalled project. I absolutely will do exactly what you suggested, but I will first make a perfect clone available, just for all the people who could do with just a plain old TWGS. I don't mean to sound rude and I want you to know I absolutely value your input, I just want to get a perfect clone out before worrying about anything else, just because I don't want this project to stall before something meaningful is produced. In regards to contacting the ZipGSX people, I don't know if its been done but I'll certainly try it. I think its a real possibility that if they gave us anything, it would be a set of metal masks for a no longer produced gate array, but there is also a possibility it could be something much more valuable. I'll let you know if anything comes of it.

@Antoine: It's too bad I didn't announce this sooner then:(. I thank you so much for your offer even though I am too late. It was very kind of you. This was the reason I was trying to win your auction a while back, I just didn't want to announce it prematurely and get peoples hopes up before I was sure it was possible to do what I wanted to do. I will absolutely put you down for two! Thanks again Antoine!

J. Vernet

unread,
Nov 19, 2012, 4:37:26 AM11/19/12
to
Le 18/11/2012 23:47, Conrad a écrit :
> I have some news for the Apple IIGS community.
>
> Conrad
>
Count 1 for me, and even 2 if the price is below $150 each !

I do not need a 40Mhz TWGS ;). Even a 8/10/12 Mhz one will make me happy !

My two TWGS are close to dead now.

Is it so difficult
JV (remove _NOSPAM)

Conrad

unread,
Nov 19, 2012, 6:03:17 AM11/19/12
to
Excellent! I will add you to the list! My TWGS may end up dying for the cause as I may be able to cut development costs by using a destructive reverse engineering process. Its a tough decision to make. I know it will die to bring life to many more but its still a tough choice. What do you guys think?

Chuck Morris

unread,
Nov 19, 2012, 8:28:38 AM11/19/12
to
I would likely be interested in buying one. I already have 2 TWGS in my ROM1 and ROM3 machines, but it would probably be a good idea to have at least 1 backup.

sdspi...@hotmail.com

unread,
Nov 19, 2012, 9:14:03 AM11/19/12
to
Count me in for a couple. 8 to 12 mHz would be all I need.

Ken C

gid...@sasktel.net

unread,
Nov 19, 2012, 12:31:49 PM11/19/12
to

> @Rob: The first edition will be a straight up clone of the TWGS, so the current ceiling on speed will apply. Don't forget that the '816 as currently available could never come close to reaching 40MHz. Its possible that when I take the next steps in doing some reengineering, if and only if we can develop an '816 softcore for an FPGA (no opensource ones are currently available, although WDC does offer an RTL '816) that the current ceiling on speed could be surpassed. That is way in the realm of uncertainty though. I'm not even going to be thinking of things like that until the first version is out the door.


Although, my comment may have seemed more like a joke. I would just like to add, that I do need one that is quite fast. Greater than 16 Mhz. 8 - 12 Mhz is ok for speeding up the finder and is 2 to 4 times faster than the IIGS fast speed of 2.8 Mhz. Games, obviously do not need an accelerator, nor word processing. But the one aspect that requires the high acceleration even on modern computers, is, obviously graphics. I have reverse engineered Dream Graphics, lsw compression, to compress the IIGS SHR screen. And I have written a driver to let Prodos 8 use an entire 8 GB flash drive. A fairly fast accelerator would definitely be in the works for me.

Now if I could only convince one of you hardware geniuses to build me a graphics card that can display at least 640x400 and support thousands of colors. And plug into a modern monitor. Then I will be all set.



gid...@sasktel.net

unread,
Nov 19, 2012, 12:34:04 PM11/19/12
to
On Monday, November 19, 2012 11:31:49 AM UTC-6, gid...@sasktel.net wrote:
> > @Rob: The first edition will be a straight up clone of the TWGS, so the current ceiling on speed will apply. Don't forget that the '816 as currently available could never come close to reaching 40MHz. Its possible that when I take the next steps in doing some reengineering, if and only if we can develop an '816 softcore for an FPGA (no opensource ones are currently available, although WDC does offer an RTL '816) that the current ceiling on speed could be surpassed. That is way in the realm of uncertainty though. I'm not even going to be thinking of things like that until the first version is out the door.
>
>
>
>
>
> Although, my comment may have seemed more like a joke. I would just like to add, that I do need one that is quite fast. Greater than 16 Mhz. 8 - 12 Mhz is ok for speeding up the finder and is 2 to 4 times faster than the IIGS fast speed of 2.8 Mhz. Games, obviously do not need an accelerator, nor word processing. But the one aspect that requires the high acceleration even on modern computers, is, obviously graphics. I have reverse engineered Dream Graphics, lsw compression, to compress the IIGS SHR screen. And I have written a driver to let Prodos 8 use an entire 8 GB flash drive. A fairly fast accelerator would definitely be in the works for me.
>
>


Mistype. Should be lzw compression

Conrad

unread,
Nov 19, 2012, 2:21:39 PM11/19/12
to
@Rob: No I don't take your comment as a joke at all. Not one bit. You have stated your need. I absolutely agree with you that a super fast IIGS would be amazing for certain applications. I take it from your comment regarding hardware geniuses that you're more of a software guy? My explaination regarding why the TWGS cannot reach those speeds without a major redesign might have been a bit heavy on hardware lingo, so I will explain it a bit better. Currently, the 65816 processor that is the center of the IIGS and the TransWarpGS has an official top speed of 14MHz according to its designers (Western Design Center or WDC for short). Its possible to push it a tiny bit past that, so let's say the real top speed is 15-16MHz. In order for this barrier to ever be broken, one of two things will have to happen: WDC would have to release a new version of the 65816 with a faster top speed, or someone would have to design a 65816 in some hardware description language such as VHDL or Verilog. The first option is pretty much out of the question unless something motivates WDC to do so, which probably will never happen. That leaves us with the second option. This is not trivial either, as it requires the CPU to be reverse engineered. WDC does offer what is called an RTL (Register Transfer Level) description of the 65816. This means they have described the internal interactions of the 65816, thus allowing someone who knows what they're doing to model the CPU in a hardware description language. Its still not simple though, it would still require a significant amount of work, and then the TWGS itself would have to be redesigned to use an FPGA (field programmable gate array, a reconfigurable logic device) instead of a physical CPU. Like I said, once the clone is finished I will definitely look at trying to do this, but for now I only want to focus on the clone, just to make sure it gets done.

@Ken and @Chuck: Excellent! I will add you guys to the list! Maybe just so I can keep track of everyone who is interested you guys can email me with your declared interest, how many you want, and your shipping address, just so I can file it all away for later on. My email is conrad(nospam).(beatitspambot)russo@(blahgoawayspammers)gmail.com. I think you guys can figure out how to redact it so I don't get spammed by usenet trolling spambots.

Sean Fahey

unread,
Nov 19, 2012, 3:13:21 PM11/19/12
to
On Sunday, November 18, 2012 6:14:43 PM UTC-6, Christopher G. Mason wrote:

> Hopefully this goes further then that "GSBarndoor" project. Those guys
> were all talk, no action (or just plain trolls). That price would be a

Meh, I wouldn't be too hard on them. I believe they ran into unforeseen technical problems and then there was some entropy after that. I don't think they were trying to make an exact clone, but a better, stronger, faster accelerator. <queue Six Million Dollar Man Theme>

STYNX

unread,
Nov 19, 2012, 3:27:13 PM11/19/12
to
Im interested as well!
If you are in need a TWGS, i may have a spare one.
-Jonas

Conrad

unread,
Nov 19, 2012, 5:22:50 PM11/19/12
to
@Sean: You're absolutely right. There is so much stacked against people succeeding at classic computing projects, you have to at least give them credit for trying. The difference between me and most other people involved in these kinds of projects is that right now this has my undivided attention. This is my day job right now, I have nothing else besides my family life to interfere. I am going to be doing projects for several other classic platforms as well as offering many open source projects for on demand production, but my first project is the TWGS. By the way, I like the Six Million Dollar Man reference. :)

@Jonas: Excellent! I'll keep you updated as the project progresses!

Steven Hirsch

unread,
Nov 19, 2012, 8:18:03 PM11/19/12
to
On 11/19/2012 02:21 PM, Conrad wrote:

> That leaves us with the second option. This is not trivial
> either, as it requires the CPU to be reverse engineered. WDC does offer
> what is called an RTL (Register Transfer Level) description of the 65816.
> This means they have described the internal interactions of the 65816, thus
> allowing someone who knows what they're doing to model the CPU in a
> hardware description language. Its still not simple though, it would still
> require a significant amount of work, and then the TWGS itself would have
> to be redesigned to use an FPGA (field programmable gate array, a
> reconfigurable logic device) instead of a physical CPU.

Somehow I think that WDC gets paid some healthy bucks for commercial use of
that design. And, as you point out, RTL intended for ASIC synthesis is going
to look one heck of a lot different than one targeted at FPGA fitting.

Steve

Conrad

unread,
Nov 20, 2012, 2:18:37 AM11/20/12
to
> Somehow I think that WDC gets paid some healthy bucks for commercial use of
>
> that design. And, as you point out, RTL intended for ASIC synthesis is going
>
> to look one heck of a lot different than one targeted at FPGA fitting.
>
>
>
> Steve

Forgive me if I'm incorrect here as I have never designed an ASIC, only done FPGA work, but I was under the impression that RTL coding is very similar between ASIC and FPGA design. There are some different methodologies used, some things need to be considered for one but not the other, and sometimes certain aspects of the design need to be changed to fit a particular FPGA's structure, but there should not be a significant difference between RTL designed for one and RTL designed for another. There are development environments available that will accept RTL and facilitate it's conversion to either an ASIC or VHDL, or even a bitstream for a particular FPGA, but they are beyond the reach of hobbyists, at least if you are buying them legitimately. As far as licencing costs of the RTL '816 core from WDC, I'll bet they want you to buy 10000+ licences at a similar price per license to what you'd pay per unit buying that many physical chips. There is an opensource implementation of the '816 instruction set but it's incomplete and I don't know if it implements the 16-bit side of things or just the 8-bit part. Once the clone is finished I am going to try and find out what is involved in licensing the RTL core from WDC and how it all works, but that's a while off yet.

Steven Hirsch

unread,
Nov 20, 2012, 8:26:56 AM11/20/12
to
On 11/20/2012 02:18 AM, Conrad wrote:

> Forgive me if I'm incorrect here as I have never designed an ASIC, only
> done FPGA work, but I was under the impression that RTL coding is very
> similar between ASIC and FPGA design. There are some different
> methodologies used, some things need to be considered for one but not the
> other, and sometimes certain aspects of the design need to be changed to
> fit a particular FPGA's structure, but there should not be a significant
> difference between RTL designed for one and RTL designed for another.

I don't design ASICs either, but I've spent the last 16 years writing
verification software for folks who do and am going by what I've heard from
them.

> There are development environments available that will accept RTL and
> facilitate it's conversion to either an ASIC or VHDL, or even a bitstream
> for a particular FPGA, but they are beyond the reach of hobbyists, at least
> if you are buying them legitimately.

Yes, the synthesis, placement and optimisation tools come from folks like
Cadence, Mentor and Synopsys and cost upwards of $20-30k per seat. They are
all heavily protected with software licensing schemes and pirating them would
do you no good at all.

> As far as licencing costs of the RTL '816 core from WDC, I'll bet they
> want you to buy 10000+ licences at a similar price per license to what
> you'd pay per unit buying that many physical chips. There is an opensource
> implementation of the '816 instruction set but it's incomplete and I
> don't know if it implements the 16-bit side of things or just the 8-bit
> part. Once the clone is finished I am going to try and find out what is
> involved in licensing the RTL core from WDC and how it all works, but
> that's a while off yet.

It would be interesting to know but, yes, one thing at a time :-).

J. Vernet

unread,
Nov 20, 2012, 9:28:52 AM11/20/12
to
When you speak about the WDC 65816, does this chip:
http://www.westerndesigncenter.com/wdc/w65c816s-chip.cfm

wil work with your project (and in a stock IIgs, à 2Mhz) ?

9$ :]

I do not even know 'til now that the chip (and the 65C02)are still in
production.

y'a plus qu'à....

JV

Erik Holter

unread,
Nov 20, 2012, 3:51:04 PM11/20/12
to
I will buy one.

Erik Holter

Conrad

unread,
Nov 20, 2012, 5:59:11 PM11/20/12
to
@Steven: You have much more experience with ASICs then I do and I appologize for my misunderstanding. I read over the WDC site again and saw that they do in fact offer an '816 FPGA core. So that could work for us, but again its a long way off. You are right that pirating such tools would be completely useless as any products produced with such pirated tools would be forbidden fruit. Either way its all just conjecture right now.

@JV: Yes that is the CPU that will be used on this card (the PLCC version) and could also be used to replace the cpu of a stock IIGS (the DIP version) although it would offer no speed increase on its own. I am aware that its still available and I will be acquiring the CPUs I need from WDC or a distributor. The problem with these CPUs is that they officially top out at 14MHz, although you can push them a bit faster then that. If we wanted to make for example a 40MHz TWGS we would need an alternative. That is why we were discussing an RTL implementation of the CPU.

@Erik: Excellent! Please email me with your shipping address, how many cards you want, how you will pay, and what oscillator you want installed. You can reach me at conrad.russo<nospamnospam>gmail.com. Replace <nospamnospam> with @.

Steven Hirsch

unread,
Nov 20, 2012, 9:33:42 PM11/20/12
to
On 11/20/2012 05:59 PM, Conrad wrote:

> @Steven: You have much more experience with ASICs then I do and I
> appologize for my misunderstanding. I read over the WDC site again and saw
> that they do in fact offer an '816 FPGA core. So that could work for us,
> but again its a long way off.

Heh, always bear in mind that I'm a software guy who knows enough about
hardware to be dangerous :-). It does make sense that they would have
soft-cores for both FPGA and ASIC synthesis - likely at prices intended to
keep WDC in business.

> You are right that pirating such tools would
> be completely useless as any products produced with such pirated tools
> would be forbidden fruit. Either way its all just conjecture right now.

Beyond the obvious illegalities involved, my point was that you wouldn't be
able to run them even if you did get your hands on a distribution CD. Every
industrial-grade EDA tool I'm aware of uses FlexLM license management. It can
be difficult even for those of us on the inside to get the proper incantations
and permissions to run the blasted things.

Steve

J. Vernet

unread,
Nov 21, 2012, 5:17:27 AM11/21/12
to
Le 20/11/2012 23:59, Conrad a écrit :

> @JV: Yes that is the CPU that will be used on this card (the PLCC version) and could also
>be used to replace the cpu of a stock IIGS (the DIP version) although it would offer no speed
>increase on its own. I am aware that its still available and I will be
acquiring the CPUs I
>need from WDC or a distributor. The problem with these CPUs is that
they officially top out at 14MHz,
>although you can push them a bit faster then that. If we wanted to make for example a 40MHz TWGS we
>would need an alternative. That is why we were discussing an RTL
implementation of the CPU.

If fact, you'd better try to rebuild a complete "IIgs in a chip" like
the Amiga minimig project (http://www.minimig.net/), if you want to
build a ultra faster IIgs. They rebuilt an Amiga using an FPGA (and a
68000...For your project, you will need, also, to build an 65816 in FPGA).

For my part, it's not what I want/ I only want a working TWGS, at 14Mhz,
with the maximum amount of cache if possible, but even a stable/working
7Mhz/8k will make me happy !


JV

Conrad

unread,
Nov 21, 2012, 8:01:07 PM11/21/12
to
@Steven: Yes I would imagine the cost of their RTL core might be prohibitive, but there's always the chance that someone there has fond memories of the IIGS and will make something available that we can use. That must be annoying as hell to deal with DRM that doesn't let even authorized users use their software. I would want to pull my hair out.

@JV: You're right that logically that makes sense. However, I think you would have more happy people with a high speed accelerator that they can use in their existing IIGS then to have a minimig style IIGS workalike. For me anyway it just wouldn't feel the same. The accelerator card I will be making available will have 32k cache and can opperate at 14MHz so that will be fine!

Polymorph

unread,
Nov 22, 2012, 4:01:47 AM11/22/12
to
Hi Conrad,

Although I already own a Transwarp GS, I only have an 8Kb cache board. Will you be making cache boards that are compatible with the original Transwarp GS's?

If so, I believe there are a number of people like me that would be interested.

Cheers,
Mike

http://apple2.sytes.net/

kamel....@free.fr

unread,
Nov 28, 2012, 4:09:02 PM11/28/12
to
@Conrad,

Don't forget that we have the chance to share the IIGS CPU with the SNES/SuperFamicom.
This mean that other people can be working in a 65816 Softcore so you better look at them :)
Example : ece545.com/F12/reports/F10_SNES.pdf
in this PDF you can read "CPU
The CPU in the SNES is a modified Ricoh 5A22. We were fortunate to get a synthesizable Verilog description of the 65C816 from Western Design Center (WDC), which was used as the core in the SNES CPU. We added several peripherals to the CPU."
Maybe you can contact them and search for other people working on a Snes in a FPGA.

Regards
Kamelito

Kamelito

unread,
Nov 28, 2012, 4:23:52 PM11/28/12
to

stephane briancourt

unread,
Dec 1, 2012, 6:24:10 AM12/1/12
to
hello,

This is a great news.
I am very interested by this project and want to buy 1 or 2 TW clone asap.
thanks
SB


Conrad

unread,
Dec 6, 2012, 5:17:20 PM12/6/12
to
This is excellent! I totally forgot about the SNES community. I have also contacted WDC, and Bill Mensch replied to me. I think we might be able to sort something out for a future expanded accelerator. Like I have said before though, this is a ways off yet, and such a project will only take place if I recover my costs on the clone, so please don't hold off on the clone hoping for a better version some time soon. The clone is concrete reality, an enhanced accelerator is still just speculation. I just wanted to be clear on that.

In related news I have opened a blog for my new venture called Retrology Computing. I will be posting a new thread about Retrology here shortly, but to make a long story short, I am trying to pick up where Reactive Micro has left off, but I will be supporting a variety of classic platforms in addition to the Apple II. I have some other minor projects for other platforms on the go, but the TWGS clone is my first big one which will get my attention before any other big projects, so don't worry about it being lost among other projects :). You can read the blog here: http://retrologycomputing.blogspot.ca and post any comments or questions you feel like. I am going to post a topic about this now, so look there for more details.

Conrad

sland...@gmail.com

unread,
Dec 8, 2012, 3:31:00 AM12/8/12
to
Im down for 1 as well. I really cant afford a real TWGS, price was astronomical. My GS is already tricked out just need the transwarp and its done

Conrad

unread,
Dec 8, 2012, 6:12:47 AM12/8/12
to
The cards that will be made available are real TransWarpGS cards in every sense of the word, they look identical and function identically. The price is just cheaper because there is no longer a finite supply. I know what you mean though. Send me an email with your details and I will keep you updated as things progress.

Erik Holter

unread,
Dec 9, 2012, 4:23:52 PM12/9/12
to
My order for your clone card stays firm!! And I will for sure order any new future versions as well :-)

Ritcho

unread,
Dec 11, 2012, 8:53:16 AM12/11/12
to
Em domingo, 18 de novembro de 2012 20h47min32s UTC-2, Conrad escreveu:
> I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively unchanged. I am in the process of cloning the TransWarpGS. I should have a Gerber file ready by February. The first step will be a direct clone to get it to market as soon as possible. The FPGA used on the original card is possible to get right now, but the supply will dry up before long, so the next step will be to reengineer it to use a more modern FPGA, to make it a bit more futureproof. Keep in mind the process has already begun and there are no barriers to it continuing unhindered as of yet, so expect this to be ready before long. The IIGS has always been close to my heart, and I'm tired of seeing so many people be barred for using this machine to its full potential by the insane cost of TWGS or ZipGSX. I'd like to hear from anyone who would be interested in ordering a TransWarpGS when they are ready for sale, so I can get an idea of interest. Looks like initial target price might be around CAD$150-CAD$175. Before anyone starts with skepticism or hostility let me just say now that I absolutely understand your reactions. An announcement like this probably seems too good to be true, but it isn't! After my troubles with my recently purchased TransWarpGS (my ROM eraser hasn't arrived yet so I haven't started the repair process yet) I realized that there are so many people out there who would love to have an accelerated GS but don't want to pay the going rate for an accelerator. I have found a way to make this project actually possible, and the process has already begun. I don't want to say a lot now until I have some things nailed down, but expect a full explaination shortly. So guys, let's hear some questions and comments!
>
>
>
> Conrad

Hi Conrad!

Congratulations by the project!

I'm very interested!

Big Hug,
Ritcho.

chris....@gmail.com

unread,
Dec 15, 2012, 12:32:41 AM12/15/12
to
On Sunday, November 18, 2012 2:47:32 PM UTC-8, Conrad wrote:
> I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively unchanged. I am in the process of cloning the TransWarpGS.

I'm late to the party here, but I'm interested in a TWGS clone as well.

Also, best of luck with your endeavor - as a fan of a variety of systems, it's nice to see somebody step in with the intent of providing needed hardware across a lot of different platforms.

Cheers,
Chris Hafner

fx

unread,
Dec 18, 2012, 2:23:51 PM12/18/12
to
I would also like to register my interest in purchasing one!

rwal...@gmail.com

unread,
Dec 30, 2012, 7:33:38 AM12/30/12
to
On Sunday, November 18, 2012 4:47:32 PM UTC-6, Conrad wrote:
> I have some news for the Apple IIGS community. Keep in mind this is a preliminary announcement and I'm not sure of the timeframe yet, but if nothing changes the timeframe should stay relatively unchanged. I am in the process of cloning the TransWarpGS. I should have a Gerber file ready by February. The first step will be a direct clone to get it to market as soon as possible. The FPGA used on the original card is possible to get right now, but the supply will dry up before long, so the next step will be to reengineer it to use a more modern FPGA, to make it a bit more futureproof. Keep in mind the process has already begun and there are no barriers to it continuing unhindered as of yet, so expect this to be ready before long. The IIGS has always been close to my heart, and I'm tired of seeing so many people be barred for using this machine to its full potential by the insane cost of TWGS or ZipGSX. I'd like to hear from anyone who would be interested in ordering a TransWarpGS when they are ready for sale, so I can get an idea of interest. Looks like initial target price might be around CAD$150-CAD$175. Before anyone starts with skepticism or hostility let me just say now that I absolutely understand your reactions. An announcement like this probably seems too good to be true, but it isn't! After my troubles with my recently purchased TransWarpGS (my ROM eraser hasn't arrived yet so I haven't started the repair process yet) I realized that there are so many people out there who would love to have an accelerated GS but don't want to pay the going rate for an accelerator. I have found a way to make this project actually possible, and the process has already begun. I don't want to say a lot now until I have some things nailed down, but expect a full explaination shortly. So guys, let's hear some questions and comments!
>
>
>
> Conrad

Count me in for a clone, and some day a faster one - if it happens. An accelerator is all my IIgs is missing.
-Ryan

Antoine Vignau

unread,
Dec 30, 2012, 7:57:33 AM12/30/12
to
Too bad, I've sold 5 accelerator for the IIgs over the past few months!
av

Tempest

unread,
Dec 30, 2012, 12:53:56 PM12/30/12
to
Will the design be improved any? I already have a nice Transwarp card, so I'm wondering if there would be any reason to buy this one and sell my old one.

Conrad

unread,
Dec 30, 2012, 7:31:26 PM12/30/12
to
On Sunday, December 30, 2012 9:53:56 AM UTC-8, Tempest wrote:
> Will the design be improved any? I already have a nice Transwarp card, so I'm wondering if there would be any reason to buy this one and sell my old one.

Hi everyone! I have a big announcement! I will be posting a new thread announcing this, but I just wanted to make the announcement here too! I have been dealing with a hard disk crash on one of my main machines and have not had time to post updates on this project, but I have still been working on it! The day has finally arrived!

I just wanted to let you know that preordering is open now for the TWGS clone. The preorder price is $180 Canadian Dollars. This price includes a 14MHz card with 32k cache. You can specify a different speed at no cost, or you can request multiple oscillators to be included at a cost of $3.00 Canadian dollars per extra oscillator. This price does not include shipping, but shipping is not payable until the card is ready for delivery. Delivery right now is set for the end of February, but it could be 2-3 weeks sooner or later then that depending on the postal service. I will be taking preorders for the next two weeks, after which the preordering will be closed. The card will go on regular sale 1 week after preorder shipping starts and the regular sale price will be $200 Canadian. The card is quite popular, so preordering ensures you will get one of the first batch. Payments can be sent via paypal to conrad...@gmail.com or via MoneyGram for those who don't have PayPal. Either method is fine, neither is preferred. Please let me know if you are interested in preordering.

To answer your question Tempest, the first edition will just be a clone. No improvements. It is possible that a later version will be produced with enhancements, but for now the IIGS community just needs a new source of accelerators!

Kevin

unread,
Dec 30, 2012, 8:59:25 PM12/30/12
to
dang, I need to get my butt in gear on my sd project which is coming up 3 months late
0 new messages