W6502SBC, a simple 6502 sbc

[willie68]

a short update, as the first breadboard build didn't work, i'll give it another try.

Removing all and starting again, with an Arduino Mega as a clock source and bus monitor.
Aun i think, i found the problem with the first build. A13 had no connection from the CPU to the breadboard bar. So I opened up the breadboard and used the pliers to squeeze the brackets together a bit. Now it's working. Next I need to put the RAM back in and repeat my initial tests with the 6522.

[willie68]

Today I wanted to breathe a little more life into the SBC. Actually, I wanted to install the RAM, but you wouldn't have seen anything there, except in the monitor. So I decided without further ado to connect the LC display. I didn't find the direct method so convincing, so I connected the display to the 6522. (I might do that later with 4 bits or I2C) I connected it to port B with full 8 bits and 3 lines to port A like Ben Eater did in his video. And indeed it is alive. A few small changes to his source because of the different addresses and the RetroAssembler and I could already see the Hello World on the display. Here's the proof:

[BillO]

Nice progress.

The direct attachment does work quite well though.

http://www.6502.org/mini-projects/optrexlcd/lcd.htm

[willie68]

the first Proto PCBS have arrived...
found 2 bugs already...

[willie68]

because I constantly have contact problems with the breadboard, I assembled the PCB and eliminated the 2 problems. Lo and behold, plugged in my monitor and the LCD and it's up and running.
Next point: ZIF socket with connector for the EEPROM.

Now the whole thing looks more like an SBC.

[Dr Jefyll]

Hello, SBC!

[floobydust]

Congrats on getting your PCB up and running! Always a nice feeling when your system works after spending all the time.

Now it's time to start writing code! Get a 6551 and USB-to-UART and direct attach to your PC using a terminal program.

Have fun!!

[BigDumbDinosaur]

willie68 wrote:

because I constantly have contact problems with the breadboard, I assembled the PCB and eliminated the 2 problems. Lo and behold, plugged in my monitor and the LCD and it's up and running.

Excellent. It is amazing how much trouble a breadboard can cause.

[willie68]

Here is the current progress, not much but for me ok:
- working LC-Display with 4-Bit interface
- Jiffy Clock (20ms)

[willie68]

Just a short comment the R65C51 (from china) are driving me crazy. I have 4 of them, bought together, but every chip from the outside is different. 3 PlasticDip one ceramic, 3 with the Rockwell logo one without…
And in my PCB every chip behaves different. The ceramic chip doesn’t the oscillator, one give always an interrupt, bit 7 even if thee s no other bit set. One uses always 5 Bit serial, no matter what I put in the control register, one alway output %00000000, the last has a different baud rate.
I’ll give up to use the 6551 for my SBC, have ordered a pair of max3100…

[BigDumbDinosaur]

willie68 wrote:

Just a short comment the R65C51 (from china) are driving me crazy. I have 4 of them, bought together, but every chip from the outside is different. 3 PlasticDip one ceramic, 3 with the Rockwell logo one without…
And in my PCB every chip behaves different. The ceramic chip doesn’t the oscillator, one give always an interrupt, bit 7 even if thee s no other bit set. One uses always 5 Bit serial, no matter what I put in the control register, one alway output %00000000, the last has a different baud rate.
I’ll give up to use the 6551 for my SBC, have ordered a pair of max3100…

Welcome to the wonderful world of counterfeits and DOA pulls.

[BigEd]

We've discussed before why this happens and what one can do to mitigate the risk. Usually the seller will refund on request, because of the way reputation systems work. It's still worth trying to pick a seller which others recommend and which has a good reputation score, and it's worth buying more parts than you need because of the likelihood of getting a mixed bag.

[speculatrix]

willie68 wrote:

Just a short comment the R65C51 (from china) are driving me crazy.

When I was trying to lay my hands on various 6551s I reckon my success rate was about 33%. The rest were either duff, fakes or not the 6551s they claimed to be. Not all the bad chips were from China - but all the good ones weren't.

I did eventually end up with a bunch of good chips that, among other things, allowed me to experience first-hand their variety of bugs (which was part of the exercise). And I built a perfectly working serial port. But I soon moved on to the NXP SC28L92 and didn't look back.

[plasmo]

I started out my retro journey with 68K, Z80, and Z280. The used parts, mostly from China, were cheap and functional. Some, maybe 10%, suffered from mechanical issues of bent/weak/broken legs but vast majority are electrically functional, most can be recklessly overclocked 25%, 50% above its rated frequency. 65xx used parts broke the mold--used 6502 from same sellers are mostly non-functional, certainly none is functional above 4MHx. 65xx peripherals are similarly mostly non-functional and those that worked are buggy in many different ways.

I'm curious why 68K/Z80 used parts are mostly good, but 65xx are mostly bad?
Bill

[floobydust]

willie68 wrote:

Just a short comment the R65C51 (from china) are driving me crazy. I have 4 of them, bought together, but every chip from the outside is different. 3 PlasticDip one ceramic, 3 with the Rockwell logo one without…
And in my PCB every chip behaves different. The ceramic chip doesn’t the oscillator, one give always an interrupt, bit 7 even if thee s no other bit set. One uses always 5 Bit serial, no matter what I put in the control register, one alway output %00000000, the last has a different baud rate.
I’ll give up to use the 6551 for my SBC, have ordered a pair of max3100…

Well, just to add a bit more information....

- The current WDC W65C51 has a defect, where the transmit interrupt flag is stuck on, so that's a problem.
- Some 6551 versions require a 1 megohm resistor in parallel with the crystal to oscillate, some don't (the new WDC part does).
- Rarely will you get a 65(C)51 to run much higher than the rated speed (note: I only run mine with interrupt-driven receive and transmit).

I have quite a collection of them... early Rockwell 6551, Synertek 6551, Rockwell 65C51, about 4 different versions of W65C51 from WDC directly, some of the latest WDC released chips and likely one or two others. In general, I've managed to get pretty much all of them working, but there are a few problem ones in the mix.

Some years ago, I purchased some small quantity of Rockwell CMOS chips from utsource.net. This included 10- R65C51P4, 10-R65C22P2 and 10-R65C02P4. All CMOS 4MHz rated parts. So far so good... they work, albeit the R65C22P4 parts have very fragile leads... one broke off on a chip. In any case, as you already have the PCB setup for the 6551, You'll probably have good luck purchasing a few R65C51P4 from utsource.net, as they still show stock. My Micromon code (github) will drive a 6551 as a console with interrupt-driven and buffered receive and transmit, which will allow you to sort out which chips work and which ones don't.

https://www.utsource.net/itm/p/1524790.html