6502.org Forum  Projects  Code  Documents  Tools  Forum
It is currently Mon Jul 08, 2024 6:01 am

All times are UTC




Post new topic Reply to topic  [ 61 posts ]  Go to page 1, 2, 3, 4, 5  Next
Author Message
 Post subject: [Contest] TIM history
PostPosted: Thu Jan 16, 2020 1:10 am 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
(I’m travelling at the moment and am writing this on an iPad all from memory, so it’ll have to be quite brief, hopefully I’ve got everything right)

The 6530-004 is a 6532 part with the TIM monitor ROM preprogrammed. This was sold as a kitset with a sample schematic and was one of the first 6502 development systems (and probably the simplest ever made). The TIM monitor was written at MOS technology (on contract) and is one of the first pieces of 6502 software.

I’m going to base my design off the schematic that came with the TIM kit, and build up the example system. I will omit the current loop interface as that is not very useful now, but will include what they called the EIA interface, which is a very simple but standard serial port. I’ll try and get period accurate chips for everything, but can’t source a 6530-004 so will have to use the 6532 and EPROM. If I could figure out a way to hide a tiny ROM under the 6532 I’d like to do something like that, but it doesn’t look like there is anything that can fit. I even thought of getting one of those chip fakers that cause us so much grief to make me a 6530 out of a 6532 (in looks, not functionality). I plan on adding some 2114 RAM to bring the RAM up to 1k. The TIM schematic has a simple mechanism to allow remapping of the 6532 RAM into the vector areas so that the vectors are in RAM after startup. The reset is a manual button that you need to press after powering up.

I’ll add an expansion edge connector to the PCB to allow for more useful functionality. I’m thinking of implementing a EEPROM programmer expansion board as that is one of the things that everyone needs starting out. So in theory your could buy your preprogrammed TIM ROM and then you’d be able to make your own ROM’s from then on.

An important part of this project is documenting the TIM and it’s history (or pointing to existing information on the web). I’d also like to see if I could get Tiny BASIC running too. It’ll require more ROM and RAM but will easily fit in the 8k address space.


Top
 Profile  
Reply with quote  
PostPosted: Thu Jan 16, 2020 9:03 am 
Offline
User avatar

Joined: Thu Dec 11, 2008 1:28 pm
Posts: 10838
Location: England
Sounds great!


Top
 Profile  
Reply with quote  
PostPosted: Thu Jan 16, 2020 9:14 am 
Offline

Joined: Thu Jul 20, 2017 9:58 am
Posts: 91
I still have a few 6530-004 left:

Attachment:
6530_TIM.jpg
6530_TIM.jpg [ 4.16 MiB | Viewed 2243 times ]


And I guess you already know this site with Information about it:

http://retro.hansotten.nl/6502-sbc/tim-6530-004/


Top
 Profile  
Reply with quote  
PostPosted: Fri Jan 17, 2020 5:55 am 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
fhw72 wrote:
I still have a few 6530-004 left:


That’s probably most of the world’s remaining supply, and ceramic too. I’ve been trying to find some Rockwell branded ones but they are impossible to get.

fhw72 wrote:
And I guess you already know this site with Information about it:

http://retro.hansotten.nl/6502-sbc/tim-6530-004/


Yes, some really good information there. The Rockwell manual is a great resource. My schematic is based off of that.


Top
 Profile  
Reply with quote  
PostPosted: Fri Jan 17, 2020 9:33 pm 
Offline

Joined: Thu Jul 20, 2017 9:58 am
Posts: 91
Did Rockwell also produce 6530-004? I've got a few (ceramic) 6532 from Rockwell.... but no 6530


Top
 Profile  
Reply with quote  
PostPosted: Fri Jan 17, 2020 9:55 pm 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
fhw72 wrote:
Did Rockwell also produce 6530-004? I've got a few (ceramic) 6532 from Rockwell.... but no 6530


They did. Their TIM manual appears to be the best reference for TIM boards, but I think they moved on soon after to put a lot of effort into the AIM-65, which appears to be based on the KIM-1 type boards but went way further. There was quite a lot of software for the AIM-65 and you could get video and disk interfaces.


Top
 Profile  
Reply with quote  
PostPosted: Fri Jan 17, 2020 10:02 pm 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
My first bit of research has been into power supplies, based on the plan of building a period accurate power supply for the TIM board. I’m starting to realise how easy we have it now. My first path of investigation was to look at the Apple 1 power supply as it’s of the same period. The first thing that struck me was the giant capacitors. These seem really hard to get now and are very expensive. The next thing is the voltages, +12, -12, +5, and -5. There’s not a lot of need for all of these today, and our biggest issue is the transition to 3.3v. If I build the serial port as per the schematic I’ll need the +12 and -12. Mostly I think the -5 was used for older RAM chips. I won’t be going back that far so can eliminate that one, but it’s probably one of the easier ones to generate. Then I had a look for transformers, and they are getting hard to find and are expensive. They are heavy things so shipping is an issue (but not as bad as a CRT). So building a period accurate power supply is very simple from a technical sense, but the parts are getting hard to get.


Top
 Profile  
Reply with quote  
PostPosted: Fri Jan 17, 2020 11:06 pm 
Offline

Joined: Mon May 21, 2018 8:09 pm
Posts: 1462
Could you use a PC-type PSU for this? You'd need to find one that doesn't have a no-load shutdown, but it'll provide +12, +5 and -12 safely in a nice big metal box.


Top
 Profile  
Reply with quote  
PostPosted: Sat Jan 18, 2020 9:48 pm 
Offline

Joined: Mon Mar 02, 2009 7:27 pm
Posts: 3258
Location: NC, USA
Chromatix wrote:
... in a nice big metal box.

Not so big. Even the oldest PC SMPS's.

A PC type power supply is a great suggestion. You just need to put a jumper between the green and black wires to bypass the no-load shutdown. Then it'll run as soon as you throw the switch.


Attachments:
PC.PS.jumper.jpg
PC.PS.jumper.jpg [ 112.58 KiB | Viewed 2109 times ]

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502
Top
 Profile  
Reply with quote  
PostPosted: Mon Jan 20, 2020 3:40 am 
Offline

Joined: Mon May 21, 2018 8:09 pm
Posts: 1462
Well, it's a big metal box when you compare it to the average wall-wart, which is all you'd need to supply a single voltage these days.


Top
 Profile  
Reply with quote  
PostPosted: Mon Jan 20, 2020 11:07 pm 
Offline

Joined: Mon Mar 02, 2009 7:27 pm
Posts: 3258
Location: NC, USA
So check out a mini ITX power supply!

_________________
65Org16:https://github.com/ElEctric-EyE/verilog-6502


Top
 Profile  
Reply with quote  
PostPosted: Mon Jan 20, 2020 11:29 pm 
Offline
User avatar

Joined: Fri Aug 30, 2002 1:09 am
Posts: 8464
Location: Southern California
I haven't looked, but I think you could find an inexpensive wall wart or table-top triple power supply with +5V and ±12V. The ±12V (very non-critical, and doesn't even need to be regulated) is useful for a lot of things that might go on a computer board.

Every electronics hobbyist should have a triple bench power supply. I used to recommend the Global 1300 and 1310 which were inexpensive, but they've quit making it. Now they only have much bigger, more expensive ones with far more power than we need for these projects, so I need to find another one to recommend.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?


Top
 Profile  
Reply with quote  
PostPosted: Tue Jan 21, 2020 12:44 am 
Offline
User avatar

Joined: Sat Dec 01, 2018 1:53 pm
Posts: 727
Location: Tokyo, Japan
jds wrote:
My first bit of research has been into power supplies, based on the plan of building a period accurate power supply for the TIM board. I’m starting to realise how easy we have it now.... So building a period accurate power supply is very simple from a technical sense, but the parts are getting hard to get.

Yup. I had a similar idea for a PSU for my Commodore 64, having vague memories of my simple high-school power supplies built with just a transformer, diode bridge and perhaps a cap or two, and I too found out that transformers are pretty expensive and not entirely trivial to source these days. (Were they ever cheap? Or was I just used to stuff being expensive?)

So far nobody's given any suggestions that don't involve giving up period accuracy. If you are going to abandon that, certainly a modern PC power supply is workable, but you also need to remember that it's not unusual for these supplies to require a minimum load, as well as needing to have the internal switch shorted. And the smaller they are, the noisier the fan. Personally, I'd go with Garth's idea of just sourcing a smaller and quieter switching PSU. There are plenty out there that provide a half amp of -5V and -12V, 1-2 A of +12V, and 4 A or more of +5V. Examples from AliExpress include this, this and this.
Attachment:
sm-quad-psu.jpg
sm-quad-psu.jpg [ 51.38 KiB | Viewed 2027 times ]

Wall warts can also work: for my C64 I ended up going with two wall warts, one for the +5 VDC, (which is dead easy to find), and one for the 9 VAC (which also wasn't so hard, and the reason I couldn't just use my bench supply).

And yes, though not applicable to my particular C64 problem, now that I've had some experience I wish I'd spent a bit more and bought at least a two-channel bench supply. I live with an adjustable wall-wartish thing when I need a second voltage and I can't take +5 from my USB power supply, but I wouldn't use that into something without its own regulator because it's pretty tricky to adjust accurately.

_________________
Curt J. Sampson - github.com/0cjs


Top
 Profile  
Reply with quote  
PostPosted: Tue Jan 21, 2020 8:39 pm 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
cjs wrote:
So far nobody's given any suggestions that don't involve giving up period accuracy. If you are going to abandon that, certainly a modern PC power supply is workable, but you also need to remember that it's not unusual for these supplies to require a minimum load, as well as needing to have the internal switch shorted. And the smaller they are, the noisier the fan. Personally, I'd go with Garth's idea of just sourcing a smaller and quieter switching PSU. There are plenty out there that provide a half amp of -5V and -12V, 1-2 A of +12V, and 4 A or more of +5V. Examples from AliExpress include this, this and this.


I think that is the conclusion I am coming to, that one of these switching power supplies is a good realistic choice, but is also a bit like the old metal box power supplies that the AIM-65 used, or even at a stretch it's similar to the Apple ][. I found a Mean Well unit that had +12 and -12 and I could regulate from there. I don't need -5 and the +-12 is only for the serial port. In a modern design you could do away with that too by going direct from 5V to USB. At the time it was up to the user to provide a power supply in most cases, the power supply not being a part of the system.

Thanks for the suggested units, I hadn't found most of those.

Feature creep is setting in again, but I am trying to resist. I've got this idea of adding a card edge connector for expansion, with a similar layout to the KIM-1, or AIM-65. There isn't a footprint for this in KiCAD so it's a bit of work, and it's very limited with the 8k address space, but it could be a useful addition. I don't think there is any need to be compatible with an existing bus, but just similar in spirit.


Top
 Profile  
Reply with quote  
PostPosted: Tue Jan 21, 2020 9:00 pm 
Offline

Joined: Thu Mar 10, 2016 4:33 am
Posts: 170
I had a look at the address decoding last night. The issue here is that the TIM schematic uses a full 6502 with 64k address range. I wasn't sure if the ROM would work in a compressed 8k range.

The basics of this is that instead of having 64k to work with, your 8k range is repeated 8 times through the address range. An address, like FFFE, would be repeated 8 times at 1FFE, 3FFE, 5FFE, 7FFE, 9FFE, BFFE, DFFE, FFFE. So the requirement is that the original 64k addresses don't overlap when mapped modulo 2000. If they do then the TIM ROM can't be used with a 6504/8.

So we have the following memory map for a full TIM system:

    0000-0800 RAM (2k here)
    6E00-6E0F 6532 I/0
    7000-73FF TIM ROM
    FFC0-FFFF TIM RAM (mapped in after startup to allow altering vectors)

Does this map onto 8k? This is an address range of 0000-1FFF. It turns out that it does without any address clashes, and the address map would look like this on the outside of the CPU:

    0000-0800 RAM
    0E00-0E0F 6532 I/0
    1000-13FF TIM ROM
    1FC0-1FFF TIM RAM (mapped in after startup to allow altering vectors)

Internally this is repeated 8 times and all the addresses also appear at their required locations.

The TIM schematic uses 2111 RAM chips which are 512 x 4. I'm going to use 2114 chips which are 1024 x 4, as used in the AIM-65. The board should have two sockets for the minimal RAM and another 2 for another 1k expansion RAM. I have to keep in mind that this is a minimal system, it would be better to put effort into a 65C02 based board rather than trying to cram too much into this board. I still think I can add an EEPROM programmer expansion board and make this useful, but it's not really a practical design for today.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 61 posts ]  Go to page 1, 2, 3, 4, 5  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 21 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to: