While planning my new cpu boards, i remembered that quirk that i have with the r/w line in my current system, and how i had to add a resistor to it to prevent signal bouncing, i even recorded a video where when i touched the line with a screwdriver, the program would magically work fine...
Now when using ribbon cables or backplanes, what kind of resistors on the end would fit? 4.7k?
And what ic are good for using them as bus amplifiers, i have the 74ls645, but i guess there are better and faster ic...

Sounds like you had another problem altogether...

There's nothing specific to terminating a 6502 R/W line. You probably had a power distribution problem. I've seen weird things like these happen with poor power distribution in some of my earlier larger Wire-Wrap projects using the typical .030" wire, sometimes 3" in length...

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

I had that isolated case back few months ago, and it has been fixed...
Anyway the current sbc board is a huge mess of wires, and when i have to modify something it takes me at least 10 minutes to figure out what i am looking for .
But since i am building the new cpu board, i would like to put this topic at rest once for all times.
I know that the 65c02 has more fanout than the nmos, but since i am going for the nmos again(if i get a cmos, ill try it too), and that 6844 dma ic, i could use a bus amplifier to power the backplane properly. I know that there is no need for this in sbc computers, since they are small, but in my case i have tons of peripherals...

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

The 74ABT245 is much faster, and more powerful.
On ebay the cheapest offer is for 15 lots, lol.
As for the backplane that i am using, it is about 15cm long...

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

That is why i started the topic in the first place, but i am not sure how exactly should i rig the terminators...

I started this several posts ago and didn't get it posted yet because I've been trying to find the rule of thumb again--and I've been unable to--for how many inches per ns of rise time you can go before worrying about terminations. If you're at 4MHz and under, you can stay with HC speeds which I think will be fine for your 6" lines without terminations. In the automated test equipment shown on my project pages here, the actual STD bus backplane was about 5" long, and the boards gave a few extra inches to the signal path, and it used LS and HC logic with no terminations and no problems.

If you do terminate though, what resistance to use to properly terminate a line depends on the transmission-line characteristic impedance, which depends on the width of the line and its distance from the ground plane (.060"? .020"? Something else?), and also the dielectric constant (which for standard FR4 PCBs is 4.2). You'll want the lines narrow enough to get the characteristic impedance high enough that the logic family being used can reliably drive valid logic levels to the load. If the load is a little off, it's better, according to Dr. Howard Johnson, and it makes sense, to have them a little high than a little low. https://web.archive.org/web/20120302190 ... ws/5_1.htm
For CMOS logic where the middle of the "no man's land" for logic levels is 2.5V and it pulls up and down with the same strength (unlike TTL), it would be appropriate to have matched resistors to Vcc and ground, ie, to take the termination to Vcc/2. https://web.archive.org/web/20120302190 ... itTerm.htm and https://web.archive.org/web/20120302190 ... dition.htm

If you have loads along the length of the line though instead of having a "non-stop flight" so to speak, you have a more complex situation; also, when signals intentionally go in both directions, like for memory where you are reading and not just writing. You pretty much need terminations at both ends. https://web.archive.org/web/20120302190 ... dsterm.htm .

Actual ground bounce is a problem with ground connections though, not terminations. It is widely misunderstood. See my page on construction for good AC performance at http://wilsonminesco.com/6502primer/construction.html , particularly the links at the bottom, one being to a Fairchild ap. note on groundbounce.

_________________
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?

Well then, in my case i have the ideal nightmare .
I have the 15cm backplane and sockets along the way...
So my idea was to have the cpu board, and then the bus amp would drive the backplane, so one terminator should be on the cpu board, and the other at the end of the backplane. The backplane is a regular stripboard with tin on the strips, the cpu board is on one end, and the last peripheral module, is on the other end, where the distance between the module socket is almost constant(i left a bit more space for the crtc board). I have 4.7k resistor arrays, so i thought about using them... Unfortunately the professors kinda liked to avoid getting into deep conversations about terminators, because it is tricky... I would like to test the backplane, but my scope probe would interfere too much, so my findings would not be consistent...

Btw. How do you call that perforated boards like stripboards, but where every hole is a stand alone copper spot, prefboard protoboard(i kinda get confused every time)???

I haven't seen any defining standard on what exactly to call them-- perfboard, protoboard, solder-type breadboard, etc., and in your case, pad-per-hole.

_________________
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?

Yeah, Vectorbord comes to mind, but only for socketted applications and usually for WireWrap.

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

I forgot to mention, that my backplane nightmare doesn't have all of the address lines, so for peripheral that need more address space, i have to plug a separate ribbon cable...

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

The lower drive current will be ok for CMOS-only loads where the only current is what's needed to charge the load capacitance, and the idea was partly to do it more slowly to make it friendlier for design and construction that is not suitable for high speeds. ABT is great, but in this case, coupled with the long lines he's talking about it might be kind of like starting a child on on a very high-powered motorcycle for his first one-- not recommended. In looking up the numbers in my National Semiconductor books though, I found it interesting that the 74AC245 had ten times the drive current of 74HC245 (75mA sink and source, compared to 7.5mA).

_________________
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?

The ABT is almost 3 times faster than the AC. What if add resistors of 4.7k at the backplane end leading to gnd?