6502.org

I guess he's trying to simplify, but it's not just the impedance mismatch between the receiver and the transmitter in the general case, but between transmitter and the transmission line at the origin .. then at the other end the transmission line to the receiver.

A good callout. I was conflating point 1 and point 2 on that page of the primer, as I've been thinking about busses a lot lately. I'm still very much learning about these finer points of ... I'd say AC characteristics but that may not even be entirely the appropriate term. I watched that clip from 2:20-3:00 and was in fact reminded of Ben's video discussing "Why Build A Computer on a Breadboard" which really was my first introduction to some of this physics. I'll have to find some time when I'm more focused to watch that Jack Ganssle video in full.

My point was simply that, for many of the Eater builders coming, they've already decided they're going to use breadboards, and it's a build well known to work on breadboards. Then they're met with this: I'm obviously newish here, so I don't know the stats, but almost all of the questions that say they're building the Eater kit I have read on this board were not ultimately resolved by switching off of breadboards. I fully expect that advice becomes more important the more complicated the build gets. But with the simple "Garth-1" running at 1MHz with only one IC for all the glue logic, the problems I have noticed are more related to wiring mistakes, or problems with interrupts and reset. My sample size is extremely limited, so I defer to the many years of experience above me! Is pushing the hobbyist with an Eater kit to buy wirewrap materials the best direction to push them? What's the top priority to make an Eater kit build more reliable? I predict it's the Reset circuit...
The kits are fulfilled directly by Jameco, even if you order from the Eater website. It comes with Jameco packaging and a Jameco invoice with Jameco part numbers. I'm sure there's an economy of scale to the kits they sell for him, but I doubt he's buying up chips on Mouser and then shipping them to Jameco to package up.
And I just want to continue to underline that these intro level materials, where all you are doing is blinking LEDs, is very useful for folks who don't have the knowledge level needed to tackle Garth's primer. That was me, six months ago. There are still pages in the Primer that I read over several times when I need that knowledge (and can rarely pull it up from memory, so indeed, it's great that its there and static). When you're at that level, it's very exciting to have the computer on and blinking LEDs within the first 30m. Blinking 8 LEDs in sequence may be a successful and complete build for some folks. Yes, there's a whole lot more to go from there, but the enthusiasm and good will generated is something that helps people get into the hobby and learn the basic terms so they can learn the next steps. The folks who come from Ben's videos are "totally new". Garth's primer recommends folks to two out of print books if they're "totally new".
Hear hear! I certainly am enjoying sharing my perspective and constantly adding more notes and phrases to my journal to look up!

This is for sure!

I watch Ben's video's too, but it is plain to me with my few years of experience in digital electronics (merely 40+ of them) as I watch where those newbies enthusiastically starting out might run into problems. I still run into problems all the time. This is not simplistic stuff, but it can be portrayed that way.

It would be even less after Paypal dings him their commission of a few percentage points. But yeah, buying in quantity sure helps. It's good to see he has some kind of deal going where Jameco can make up the kit and drop-ship it. I get the SRAMs for the modules I sell for about half of what individuals would have to pay for just a few, and I buy from legitimate US distributors, not UTSource or eBay or China. Still, when I consider the time I put into each one, and into communicating with each customer, and into trips to the post office, etc., it is definitely not a money-maker. I just do it to promote our interest.
I am glad for the excitement he has stirred up and the new members he has (probably unintentionally) brought us.

(BTW, thanks, BDD, for the website compliments.)
Anyplace there's an impedance discontinuity, there'll be a reflection. And on a solderless breadboard, you don't really have transmission lines anyway, so what happens is uncontrolled and rather unpredictable. A transmission line is a specialized cable or other structure (like a trace against a ground plane on a PCB) designed to conduct electromagnetic waves in a contained manner.
You posted while I was writing, so I have not checked again yet to see if that's the video I'm thinking of; but I know he had one "proving" the good performance with a clock oscillator, but he did it right, with the ground return running right next to the signal line, whereas most builders will probably run the ground return over to the rail, down the rail to the end of the breadboard and up the next breadboard's rail, far from the signal line, making essentially a big loop, a big inductor, with taps along the way for other things. Surprising things can be accomplished if you do them right, but the successes will come from things the newbie won't catch. It is my understanding that the Cray-1 supercomputer used miles of wire-wrap and an 80MHz clock; but every signal line had its ground return twisted with it, in loads and loads of twisted pairs. Special insulation was also used to increase the propagation velocity to well above what it would be with more-common wire.

Some of the best learning does seem to happen when we take a look at a malfunctioning (or non-perfectly functioning) design and see what fixes it. I think that is indeed the video you're thinking of Garth, and pointing out the ground return makes me want to pull out my scope and try it both ways. I've learned a bunch (and I'm sure others have too) from folks on here pointing out what's wrong.

You don't have transmission lines in the engineering sense (realizable ideals, well defined within parameters, meeting specifications, etc..), but any length of wire/trace/cable is a 'transmission line' for sake of the science and the governing theory. So, yes I agree with you, any place there is an impedance discontinuity, or even a significantly abrupt change, there will be reflections or worse (situations where impedance changes with frequency, etc..). So, with solderless breadboards the impedances are all over the map. And that is only half of it. Bad connections due to corrosion of the bread board contacts, poor plating, incorrect wire size, low spring tension, corroded wire, etc.. causing supply spikes and ground bounce all add their magic spells to the situation. Even keeping wires as short as possible, the situation is worse than had been depicted in the video. Still, sometimes it works despite the odds against it.

This one?
Techniques for reliable high-speed digital circuits
On the last page (so far), enso gave a link to a very long but excellent video. One of the points made is that the signal energy does not travel in the copper, but in the dielectric between the signal trace and the return path. The copper only guides it.

Also wanted to confirm if that was the topic BigDino was referring to?
I actually noticed while doing the math in this thread that Jameco appears to be the cheapest place to get small volumes of W65C02s in the US at the moment. The ones I've gotten from Jameco all appear to be brand new / not pulls. At Mouser, the 1 unit price is $10.75. At Jameco, the WDC65C02 is $6.95! We may all be valuing from a big order Jameco made to fulfill the kits. Good news for the BOM of my own SBC which I had calculated only at Mouser so far.

Yes.

I came across this video while rounding up a few others I want to link to on my site.
"You must Unlearn what You have Learned"
https://www.youtube.com/watch?v=y4REmZlE7Jg
From about 11:48 to 25:54, he shows a little circuit he built up on a solderless breadboard that pulses the current, and shows on the 'scope how the voltage from the power bus to the ground bus is heavily affected because of the inductance in the solderless breadboard. He even used the power and ground that run close to each other, rather than one down one edge of the board and the other down the other edge, where his way is a lot better than what hobbyists usually do when the ground pin of an IC is on one side and the power pin is on the other. He's using a lot more current than we're usually talking about, but OTOH we're normally talking about much faster rise times which cause more problems. He puts the capacitor different distances from the load to show how much better it is to have it really close. He also shows that a bigger capacitor is not an improvement over a smaller one. In other parts of the video, he skewers other sacred cows.

$6.95 or so is what I remember paying from Mouser before the pandemic and chip shortage that drove up prices. Could it be that Jameco hasn't updated their prices to reflect the new scarcity?

I must admit, when I first joined the forum here (not that long ago), being strongly urged to wire wrap keep the tracks as short as possible, avoid breadboards at all costs, etc… I was a little disheartened. And thought, geez, if I don’t do all this to the letter, things just will not work.

Then you get reports that feel contradictory, like ‘at 1MHz you can get away with murder’.

So in a way, I chose to take some information with a grain of salt and just push on and try things. Which, to me, is one of the best ways to learn. “Try and see what happens”.

In doing so, I was able to overclock the W65C02 all the way up to just short of 40MHz at 5v, before things got really unstable. This was using 20cm of un-optimised stripboard, housing the CPU and a couple of other bits, running from a clock which was constructed on, you guessed it, ‘a breadboard’.

The fact that all of this worked on such a noisy setup had given me a huge amount of confidence in being able to construct what I want to. And I only want the 6502 to run at 12.5 MHz for my main project.

So while wire wrap might be the best case scenario for the guru’s here, it’s not the end of the world if you chose stray from that, for whatever reason that may be.

Excellent to hear that you got the result you needed, despite the initial discouragement.