Yarrb - redesigned

[DaveLecky]

Hi Roland,

That would be the best next step/solution, if we lived in the same town it would be much easier.

I’m happy to pay the postage.

Dave

[roland]

That would be the best next step/solution, if we lived in the same town it would be much easier.

That can be arranged. The house next to my neighbours is for sale

I’m happy to pay the postage.

We'll settle that later...

[DaveLecky]

Hi Roland,

I look forward to receiving the parts for testing.

I could move to your neighbours house ok , but you will have to have a local , “ Church of Scotland “ Church nearby as my wife is a Minister.

Dave

[cjm]

Your 8255 is set to clear 3, we have seen this in the past. Unfortunately it's not just a setting. It is related to the clock signal. Some people solved it by replacing the 8255. Also adding a 100 ohm resistor in the clock wire that you soldered did in some cases solve the issue.

The best way might be cjm's patch to drive the i/o at 1MHz. I should have used that code for the Cpld. It can be changed but you need a special programmer for it.

I think that the ATOM's clock transition time is too slow for the CPLD, this results in occasional glitches in the clock which the 8255 cant quite handle.

Adding a 4 MHz clock module to the YARRB and updating the CPLD code is a great solution.

HOWEVER there is potential solution for the clock problem that doesn't involve updating the code for the CPLD or swapping chips around! The following mod buffers the clock signal though a spare logic gate and allowed my previously unstable atom/yarbb to work reliably.

viewtopic.php?f=44&t=19947&p=278886#p278886

[DaveLecky]

Hi cjm,

It’s certainly worth a try, I’ll have a go tomorrow.

Thanks

Dave

[roland]

I have already send a know good 8255 to David and I hope that it solves his problem. If so, he can return his "bad" 8255 to me and I can do some experiments with it. I quite like both your solutions but I prefer the extra clock on the Yarrb2 board. First of all, the socket is already on board and secondly, it takes less modifications to the Atom which makes it easier for less experienced people to install the board.

[DaveLecky]

Hi Roland,

I will hold back for now, and wait till your chips arrive and try that first.

Dave

[roland]

O no, please do the test with the extra inverter port if you like. That will keep you busy and save me some time

[DaveLecky]

Will do Roland.

Dave

[DaveLecky]

Hi Folks,

I have applied the changes suggested by cjm, the Atom showed the same results as in the following link.

It did however appear to go faster than before.

viewtopic.php?f=44&t=18020&start=210#p294864

The changes....

Dave

[chrisn]

I've just installed a Yarrb2 board into my Atom, and I was experiencing the exact same issues as DaveLecky. The mod suggested by cjm seems to have solved it for me

[Edit] The board is working fine at 1MHz and 2MHz (using ?#BFFE=70), but if I select 4MHz (?#BFFE=38) the Atom hangs.

[DaveLecky]

Hi chrisn,

Glad to hear you got yours working, where did you get the info on the commands...
Quote.
[Edit] The board is working fine at 1MHz and 2MHz (using ?#BFFE=70), but if I select 4MHz (?#BFFE=38) the Atom hangs.
I don’t see the in the Manual.

Dave

[chrisn]

Hi Dave, I hope you can get your Yarrb2 board working.

I worked out the values from the Yarrb manual, section 3.1 and 4.1.2:

3.1 Switching CPU clock frequency

You can switch between 1, 2 and 4 MHz in your software. Bits 6 and 5 of #BFFE let change you the frequency:

bit 5: set -> 4 MHz clear -> 1 or 2 MHz (depending on bit 6)
bit 6: set -> 2 MHz clear -> 1 MHz

Please not that when bit 5 is set then bit 6 is ignored.

Bit 5 is 32 and bit 6 is 64. I add 6 to these to enable AtomMMC, so the values I tried are 38 (for 4MHz) and 70 (for 2MHz).

[DaveLecky]

Hi chrisn,

Thanks for the info.

Dave

[roland]

The board is working fine at 1MHz and 2MHz (using ?#BFFE=70), but if I select 4MHz (?#BFFE=38) the Atom hangs.

That is expected .... your processor and 8255 must be able to handle this speed. I have installed a R65C02P4 and a NEC D8255AC-2 which work perfect at 4 MHz.

[DaveLecky]

Hi Roland,

Your testing IC’s arrived today so here are my results.

Test 1 - My original IC’s with no resistor fitted.
Header displayed ok, on shift break I get the menu as before which distorts as before.....no change there.

Test 2 - Your IC’s no resistor
Header displayed ok, on shift break I get the menu which is stable.
On selecting first game, Asteroids, it plays ok for a couple of minutes then goes into a endless scrolling up when killed.
On selecting other random games the either give errors, Name error 118 or Syn error 135, or just hangs with a square prompt top left.

Test 3 - My original IC’s with 47 ohm resistor fitted.
Header displayed ok, on shift break I get the menu which is stable.
Results are as Test 2 above.

Test 4 - Your IC’S with 47 ohm resistor.
Results are as Test 2 above.

Summary
Using both your IC’s improved things only in respect of not requiring a 47 ohm resistor to get a stable Menu.
Everything else operated in the same way, initially stable then deteriorating after a couple of minutes.
I’ve been using the asteroids game as a standard to keep tests the same, I take it, it doesn’t crash on everyone’s Atom......
If I leave the menu on screen for a few minutes then Asteroids doesn’t load at all, just the square prompt as described above.
At one point it appeared that the items on the menu were selecting themselves while I was scrolling through them,without my input, or I may just be going mad.......

Not very encouraging I’m afraid.

Dave

[cjm]

Hi Folks,

I have applied the changes suggested by cjm, the Atom showed the same results as in the following link.

It did however appear to go faster than before.

viewtopic.php?f=44&t=18020&start=210#p294864

The changes....
55B058C3-B73E-4217-8CD0-85F795A56847.jpeg


Dave

Hi Dave,

I cant tell from your photo - is pin 10 of IC44 bent out as per roland's installation guide?

Chris

[roland]

Is the power supply stable? After a few minutes the Atom starts to behave faulty. This often indicates a temperature related problem. What is the voltage when the problems are starting?

[DaveLecky]

Hi Chris,

Yes, the pin was out ok at installation, I did reinsert it during your tests, it is back in now.EDIT..IT IS BACK OUT NOW

Dave

[DaveLecky]

Hi Roland,

The voltage is 5.18 at switch on, and stays stable when the problems start.

Here is my power supply.

The only thing I can see which is not quite right, is that the chips on the AtoMMC board are touching the solder contacts on the bottom of the motherboard.

Either my header pins are too low or the female socket on the AtoMMC is too short, I realise it is height critical to get the lid on successfully.

Dave

[roland]

I don't think that it matters that the IC's touch the solder contacts. The IC's are not conductive.
The voltage is fine. Nice power supply btw.

Does the Atom run stable with a simple program like this:

Code: Select all

10 DO
20   PRINT "*   "
30 UNTIL 0

Does it run for 10 minutes or so? And is there any difference with or without MMC?

(Actually I don't have any clue what is going on in your Atom, so let's just try some simple things....)

[DaveLecky]

Hi Roland,
Voltage 5.19

It’s starting to deteriorate after 6 minutes.

After 10 minutes, getting worse, some + and : showing now, some inverted.

After 15 Minutes.....

Hit shift break after 18 minutes and entered A.

Voltage 5.19

Will leave it for a few minutes and remove the MMC.

Dave

[DaveLecky]

Hi Roland,
5.2 volts with MMC removed.
I’ve let it cool a bit, removed the MMC and ran your prog. again.

After a few minutes........

It starts again.

I have checked the temperature of the 6502, it’s 28 C between pins 1 to 8, and 24 C over the rest of the chip.....

Dave

[hoglet]

Dave,

Is this behaviour with the Noise Killer board fitted?

If so, it looks like there might be suprious writes happening to the video RAM.

Can you take a close-up photo of the Noise Killer board, so I can read the IC labels and see the jumper settings?

Can you do the same for the YARRB board?

Dave

[roland]

I was also thinking about the noise killer board after seeing this screen corruption .... is it possible to remove the noise killer board and re-install two 2114's (just for the text mode) and how does the test program behave without noise killer?

[DaveLecky]

Hi Hoglet,

Are these ok.

Dave

[DaveLecky]

Hi Roland and Hoglet,

I’ve removed the noise killer, replaced IC 29,30 and inserted the two 2114’s, Ic 41 and 42.

The MMC is has also been removed.

The header appeared ok, the prompt took a couple of presses, cntrl break.
I have input Rolands wee program and ran it for 10 minutes..... with no screen corruption.

Dave

[roland]

Do you have enough 2114's to fill the complete video memory? If so, how does arcadians run with the 2114's?

[DaveLecky]

Hi Roland,

Yes I have a full compliment of 2114’s.
Replaced the 2114’s, replaced the MMC.

I don’t have arcadians, I presume you mean asteroids on the MMC.

Started ok, but when killed, the screen started to scroll up endlessly as before,until reset.
I have played this game and passed this point a few times, before the error, a couple of days ago.

I am having better results with the modern creations......item B


Dave

[DaveLecky]

Hi Folks,

I am scrolling through the menu trying random programs, it certainly seems a bit more stable, it would normally have crashed starting programs by now....15 minutes up.

There is more noise on screen as expected, without the noise killer.

Dave