Yarrb - redesigned

emulators, hardware and classic software for atom + system machines
User avatar
oss003
Posts: 3278
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yarrb - redesigned

Post by oss003 » Thu Jun 18, 2020 8:33 pm

WOW Dave ...... :shock:
This is some serious high school CSI stuff you are doing ....... nice job .....

Greetings
Kees

User avatar
roland
Posts: 3862
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yarrb - redesigned

Post by roland » Thu Jun 18, 2020 10:40 pm

hoglet wrote:
Thu Jun 18, 2020 6:49 pm
I think it's likely one or more of these connections is broken, possibly due to damage to the PCB where the sockets for IC10/11 have been removed. All of these signals pass through IC11.
That location also explains why the Atom even doesn't boot when the VIA is fitted. I also once had troubles with these resistors:
1. In the Atom 2k14 where they were tied to +5V [-X
2. In an Atom where those resistors were bend and made a connection between their wires.

I suspected this earlier but I couldn't see any clear damage and John confirmed there were no short circuits in the IC 10/11 area.

Anyway, a very nice analysis done here Dave =D> IMHO this is a very educational post for troubleshooting without advanced equipment.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Fri Jun 19, 2020 1:37 pm

Dave

Your exhaustive investigation is much appreciated.

Here are the results of the tests you asked for.

I think I have muddled the waters as I had fitted a 6502 i obtained on line and not the original one in the Atom. The original one is now installed and the readings are enclosed as well as a photo.

The connections between the 6502 and the 6522 and the resistors R39, 40, and 41 are all working.

With the power off and the meter set to K ohms there was no reading between pins 8 and 34 of the 6502.

John
6502 readings 3.xls
(9.5 KiB) Downloaded 6 times
Current 6502.jpeg

User avatar
hoglet
Posts: 9241
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yarrb - redesigned

Post by hoglet » Fri Jun 19, 2020 3:28 pm

Gosh, this is turning into rather a marathon.
John Ferguson wrote:
Fri Jun 19, 2020 1:37 pm
I think I have muddled the waters as I had fitted a 6502 i obtained on line and not the original one in the Atom. The original one is now installed and the readings are enclosed as well as a photo.
Can I just confirm that the three full sets of measurements you previously posted were with the "fake" 6502?

That would explain two things:
- it would explain the high output voltages when driving a logic '1'
- it would explain the why you got better results with JP2 set to 2-3

I suspect the "fake" 6502 is a remarked WD W65C02.

To confirm this, with the "fake" 6502 out of the socket, measure the resistance between pins 1 and 21.

Could you also post a similar close-up photo of the "fake" 6502?
John Ferguson wrote:
Fri Jun 19, 2020 1:37 pm
The connections between the 6502 and the 6522 and the resistors R39, 40, and 41 are all working.
That's a rather a shame - I wonder if this connection is intermittent - or whether there is damage shorting two signals together.

The measurement set you re-posted yesterday (6502 readings 1.xls) very clearly pointed to an issue in this area.
John Ferguson wrote:
Fri Jun 19, 2020 1:37 pm
6502 readings 3.xls
Looking at those results ("Set 3)", there is something slightly inconsistent:

Code: Select all

                        (65C02) (65C02) (6502)
Pin     Name    Set 0   Set 1   Set 2   Set 3
7       SYNC    3.41    1.529   1.546   0.031
9       A0      0.046   2.83    2.83    0
10      A1      0       1.496   1.507   3.529
11      A2      0       1.509   1.495   3.541
12      A3      0       0.03    0.031   0
13      A4      0       4.503   4.481   0
14      A5      0       0.019   0.02    0
15      A6      0.03    0.019   0.02    0.032
16      A7      0.03    0.019   3.589   0.032
When you measured pin 7 (SYNC), the value was 0.031 which means the 6502 is not executing instructions. And the lower address pins also look static (3.5-3.6V is a solid "1" for an NMOS device)

However later on, there is evidence it is running again.

Code: Select all

                        (65C02) (65C02) (6502)
Pin     Name    Set 0   Set 1   Set 2   Set 3
17      A8      0.03    3.09    3.004   2.88
18      A9      0.048   4.5     0.925   2.727
19      A10     0.057   4.501   3.616   2.919
20      A11     0.056   5.116   3.621   3.51
22      A12     0.045   5.114   3.6     3.458
23      A13     0.049   5.114   0.028   3.516
24      A14     0.051   4.342   0.029   2.885
25      A15     0.078   5.114   0.464   3.498
Voltages of ~2.7-2.8V indicate a signal that is mostly high, but goes low approx 1 cycle in 6. And these address bits are consistent with the 6502 looping at #FF10 again.

I think what's happened here is that during the measurements the Atom rocked a bit, which can cause the BREAK key to be pressed and the 6502 to restart.

It might help to try to follow this recipe when taking a set of measurements:
1. Power up the Atom
2. Press and release BREAK (to make sure the 6502 sees a distinct reset)
3. Turn the Atom upside down, and carefully wedge it so that it can't rock and cause BREAK to be pressed again
4. Measure SYNC (pin 7)
5. Measure pins 1 - 40
6. Measure SYNC (pin 7)

By measuring SYNC before, during and after the other pins, you get an indication of whether anything has changed during the measurement that might affect it's consistency

It can also speed up measuring if you glue a pinout to the 6502 with pritt:
6502_pinout.pdf
(87.92 KiB) Downloaded 9 times
As you now have an NMOS 6502 fitted, you need to return JP2 back to pins 1-2. It's possible this is causing the 6502 to crash.

With the original 6502, have you ever seen the ACORN ATOM banner?

Dave
Last edited by hoglet on Fri Jun 19, 2020 4:42 pm, edited 1 time in total.

User avatar
hoglet
Posts: 9241
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yarrb - redesigned

Post by hoglet » Fri Jun 19, 2020 4:39 pm

John,

Here's an additional set of measurements you should make regarding R39/40/41, because I'm convined there is something amiss there.

You need to unplug the following parts (basically everything that connects to the data bus):
- the AtoMMC
- the YARRB board
- the 8K Video RAM and Noise Killer board
- the 8255

(This list would normally be longer, but you have already removed the other chips that connect to the data bus)

Set your multimeter to a range suitable for measuring 4.7KOhms.

Clip the black probe of the meter to a good ground point (I use the negative end of the electrolytic capacitor C11 in the bottom left corner of the PCB).

On the 6502 socket, use the red probe of the meter to measure the resistance (to ground) of:
- D0 (Pin 33)
- D1 (Pin 32)
- D2 (Pin 31)
- D3 (Pin 30)
- D4 (Pin 29)
- D5 (Pin 28)
- D6 (Pin 27)
- D7 (Pin 26)

Depending on the type of IC socket fitted, you might need to make this measurement on the solder side of the board to make a connection.

D1, D2 and D3 should read 4.7K; D0 and D4-D7 should be high impedence (i.e. no reading).

Then repeat the measurement on the 6522 socket.

Dave

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sat Jun 20, 2020 4:14 pm

hoglet wrote:
Fri Jun 19, 2020 4:39 pm
John,

Here's an additional set of measurements you should make regarding R39/40/41, because I'm convined there is something amiss there.

You need to unplug the following parts (basically everything that connects to the data bus):
- the AtoMMC
- the YARRB board
- the 8K Video RAM and Noise Killer board
- the 8255

(This list would normally be longer, but you have already removed the other chips that connect to the data bus)

Set your multimeter to a range suitable for measuring 4.7KOhms.

Clip the black probe of the meter to a good ground point (I use the negative end of the electrolytic capacitor C11 in the bottom left corner of the PCB).

On the 6502 socket, use the red probe of the meter to measure the resistance (to ground) of:
- D0 (Pin 33)
- D1 (Pin 32)
- D2 (Pin 31)
- D3 (Pin 30)
- D4 (Pin 29)
- D5 (Pin 28)
- D6 (Pin 27)
- D7 (Pin 26)

Depending on the type of IC socket fitted, you might need to make this measurement on the solder side of the board to make a connection.

D1, D2 and D3 should read 4.7K; D0 and D4-D7 should be high impedence (i.e. no reading).

Then repeat the measurement on the 6522 socket.

Dave
Readings on the 6502 and 6522 were correct with everything removed.

Measured the "fake" 6502 resistance between pins 1 and 21 was 4.3ohms.

I can not confirm that the 3 sets of measurements were with the same 6502.

More readings enclosed with normal 6502. Checked no change before after readings pin 7 read 0.99

JP2 now set to pins 1 and 2.

Acorn Atom banner was achieved once after screen rubbish random display and pressing break again. Takes some time after switch on before this is possible.

John
Recent buy of a 6502.jpeg
6502 readings 4.xls
(9 KiB) Downloaded 4 times

User avatar
hoglet
Posts: 9241
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yarrb - redesigned

Post by hoglet » Sat Jun 20, 2020 4:54 pm

Hi John,

Thanks for confirming that R39/40/41 are fine.

And thanks for another set of readings ("Set 4")

For the convenience of anyone else mad enough to be following this, here are the results todate:

Code: Select all

                        (65C02) (65C02) (6502)  (6502)
Pin     Name    Set 0   Set 1   Set 2   Set 3   Set 4
1       VSS     0.024   0.017   0.021   0.021   0.024
2       RDY     5.04    5.128   5.111   5.102   5.119
3       Phi1Out 5.06    2.416   2.38    2.089   2.13
4       IRQ     5.06    5.128   5.125   5.11    5.12
5       NC      0       0.004   0.25    0       0
6       NMI     5.01    5.128   5.123   5.1122  5.117
7       SYNC    3.41    1.529   1.546   0.031   1
8       VCC     5.08    5.133   5.128   5.112   5.123
9       A0      0.046   2.83    2.83    0       0.003
10      A1      0       1.496   1.507   3.529   0
11      A2      0       1.509   1.495   3.541   3.56
12      A3      0       0.03    0.031   0       3.507
13      A4      0       4.503   4.481   0       3.465
14      A5      0       0.019   0.02    0       3.45
15      A6      0.03    0.019   0.02    0.032   3.444
16      A7      0.03    0.019   3.589   0.032   3.443
17      A8      0.03    3.09    3.004   2.88    3.426
18      A9      0.048   4.5     0.925   2.727   3.189
19      A10     0.057   4.501   3.616   2.919   3.423
20      A11     0.056   5.116   3.621   3.51    3.421
21      VSS     0.02    0.016   0.016   0.02    3.43
22      A12     0.045   5.114   3.6     3.458   3.436
23      A13     0.049   5.114   0.028   3.516   3.43
24      A14     0.051   4.342   0.029   2.885   3.432
25      A15     0.078   5.114   0.464   3.498   3.432
26      D7      3.3     2.737   5.028   2.799   4.966
27      D6      0.024   0.707   5.025   0.696   4.949
28      D5      0.024   3.467   2.465   3.482   4.94
29      D4      0.024   2.73    1.597   2.736   4.905
30      D3      0.021   2.851   1.754   2.583   4.253
31      D2      0.021   1.249   1.682   1.254   4.212
32      D1      0.021   0.618   1.196   4.365   4.235
33      D0      0.024   2.194   1.786   4.856   4.822
34      R/W     0.07    5.109   4.97    3.578   3.588
35      NC      0       0       0.002   0       0
36      NC      0       0.3     0.003   0       0
37      Phi0In  0.026   1.748   1.795   1.821   1.906
38      SO      5.117   5.144   5.146   5.125   5.134
39      Phi2Out 0.01    2.59    2.66    2.414   2.162
40      RES     3.925   4.016   4.042   3.982   4.005
There are a couple of unusual things in the "Set 4" measurements.

1. I think there must be a mistake on pin 21, as this signal is ground but you have measured it as 3.43V.

2. These two results are inconsistent with each other:
- A0 (pin 9) is very close to 0V
- SYNC (pin 7) is at 1V

When fetching an instruction at address A, the 6502 bus will show the following activity:
- Cycle 0 - read address A
- Cycle 1 - read address A + 1

So I would expect to see A0 toggling pretty much every cycle on a running 6502. In your case, A0 is close to 0V, indicating the 6502 is stopped. Yet a SYNC value of 1V indicates the 6502 is running and fetching an instruction every 3-4 cycles. So we have a contradiction.

I can only think of two explanations for this:
- the 6502 is faulty
- the 6502 is being clocked at 2MHz or 4MHz (which is possible if something has written to the YARRB register at 0xBFFE)

At this point, I don't really know what to suggest next. I think we have exhausted what is possible with just a multimeter, and a logic probe isn't much better given where we are now.

If I had the Atom here it would be very easy to diagnose:
- I would swap the 6502 for a known good one
- I would connect an oscilloscope to the clock and check the frequency is 1MHz and the waveform is clean
- I would capture a trace of the system booting on a logic analyzer and feed the results through my "6502 decoder"

You don't have the ability to do any of these things.

I don't really have any further suggestions :(

Maybe Roland has some ideas?

Dave

User avatar
1024MAK
Posts: 9990
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Yarrb - redesigned

Post by 1024MAK » Sat Jun 20, 2020 5:16 pm

hoglet wrote:
Sat Jun 20, 2020 4:54 pm
...For the convenience of anyone else mad enough to be following this...
Don’t know about mad, thought I was already classed as that! Certainly a lot of head scratching going on... :shock: :x 8-[ :-#

Anyway, what I was going to ask, is this:

John, do you have any other machines that use 6502 processors?
Master 128s, Compacts, Commodore 64 and Atari’s don’t count (as they all use variations).

Are you able to program (burn) EPROM or EEPROM chips?

Are you any good at building projects on strip board?

If the answers to the above questions are negative, do you mind telling us which part of the world you live in, and if in the U.K. which part. A county is enough information.

Mark

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sat Jun 20, 2020 5:29 pm

hoglet wrote:
Sat Jun 20, 2020 4:54 pm
Hi John,

Thanks for confirming that R39/40/41 are fine.

And thanks for another set of readings ("Set 4")

For the convenience of anyone else mad enough to be following this, here are the results todate:

Code: Select all

                        (65C02) (65C02) (6502)  (6502)
Pin     Name    Set 0   Set 1   Set 2   Set 3   Set 4
1       VSS     0.024   0.017   0.021   0.021   0.024
2       RDY     5.04    5.128   5.111   5.102   5.119
3       Phi1Out 5.06    2.416   2.38    2.089   2.13
4       IRQ     5.06    5.128   5.125   5.11    5.12
5       NC      0       0.004   0.25    0       0
6       NMI     5.01    5.128   5.123   5.1122  5.117
7       SYNC    3.41    1.529   1.546   0.031   1
8       VCC     5.08    5.133   5.128   5.112   5.123
9       A0      0.046   2.83    2.83    0       0.003
10      A1      0       1.496   1.507   3.529   0
11      A2      0       1.509   1.495   3.541   3.56
12      A3      0       0.03    0.031   0       3.507
13      A4      0       4.503   4.481   0       3.465
14      A5      0       0.019   0.02    0       3.45
15      A6      0.03    0.019   0.02    0.032   3.444
16      A7      0.03    0.019   3.589   0.032   3.443
17      A8      0.03    3.09    3.004   2.88    3.426
18      A9      0.048   4.5     0.925   2.727   3.189
19      A10     0.057   4.501   3.616   2.919   3.423
20      A11     0.056   5.116   3.621   3.51    3.421
21      VSS     0.02    0.016   0.016   0.02    3.43
22      A12     0.045   5.114   3.6     3.458   3.436
23      A13     0.049   5.114   0.028   3.516   3.43
24      A14     0.051   4.342   0.029   2.885   3.432
25      A15     0.078   5.114   0.464   3.498   3.432
26      D7      3.3     2.737   5.028   2.799   4.966
27      D6      0.024   0.707   5.025   0.696   4.949
28      D5      0.024   3.467   2.465   3.482   4.94
29      D4      0.024   2.73    1.597   2.736   4.905
30      D3      0.021   2.851   1.754   2.583   4.253
31      D2      0.021   1.249   1.682   1.254   4.212
32      D1      0.021   0.618   1.196   4.365   4.235
33      D0      0.024   2.194   1.786   4.856   4.822
34      R/W     0.07    5.109   4.97    3.578   3.588
35      NC      0       0       0.002   0       0
36      NC      0       0.3     0.003   0       0
37      Phi0In  0.026   1.748   1.795   1.821   1.906
38      SO      5.117   5.144   5.146   5.125   5.134
39      Phi2Out 0.01    2.59    2.66    2.414   2.162
40      RES     3.925   4.016   4.042   3.982   4.005
There are a couple of unusual things in the "Set 4" measurements.

1. I think there must be a mistake on pin 21, as this signal is ground but you have measured it as 3.43V.

2. These two results are inconsistent with each other:
- A0 (pin 9) is very close to 0V
- SYNC (pin 7) is at 1V

When fetching an instruction at address A, the 6502 bus will show the following activity:
- Cycle 0 - read address A
- Cycle 1 - read address A + 1

So I would expect to see A0 toggling pretty much every cycle on a running 6502. In your case, A0 is close to 0V, indicating the 6502 is stopped. Yet a SYNC value of 1V indicates the 6502 is running and fetching an instruction every 3-4 cycles. So we have a contradiction.

I can only think of two explanations for this:
- the 6502 is faulty
- the 6502 is being clocked at 2MHz or 4MHz (which is possible if something has written to the YARRB register at 0xBFFE)

At this point, I don't really know what to suggest next. I think we have exhausted what is possible with just a multimeter, and a logic probe isn't much better given where we are now.

If I had the Atom here it would be very easy to diagnose:
- I would swap the 6502 for a known good one
- I would connect an oscilloscope to the clock and check the frequency is 1MHz and the waveform is clean
- I would capture a trace of the system booting on a logic analyzer and feed the results through my "6502 decoder"

You don't have the ability to do any of these things.

I don't really have any further suggestions :(

Maybe Roland has some ideas?

Dave
Your mention of #BFFE reminded me that when I fitted the Yarrb board I thought you had to set this location for ram or rom.

So I have probably done some harm. This might be the problem. It does not explain however that with the Atom striped back to the minimum it still
does not work properly. Maybe running the 6502 at a higher speed causes some damage.

I have now measured pin 21 as 0.02

John

User avatar
hoglet
Posts: 9241
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yarrb - redesigned

Post by hoglet » Sat Jun 20, 2020 5:34 pm

John Ferguson wrote:
Sat Jun 20, 2020 5:29 pm
Your mention of #BFFE reminded me that when I fitted the Yarrb board I thought you had to set this location for ram or rom.

So I have probably done some harm. This might be the problem.
I don't understand this comment.

Can you any explain a bit more exactly what you think you did wrong?

I didn't think you ever had the YARRB booting to the > prompt and be able to type anything in.

Dave

PS And make sure you don't miss Mark's post above
Last edited by hoglet on Sat Jun 20, 2020 5:56 pm, edited 1 time in total.

User avatar
1024MAK
Posts: 9990
Joined: Mon Apr 18, 2011 5:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Yarrb - redesigned

Post by 1024MAK » Sat Jun 20, 2020 5:38 pm

Trying to run this type of equipment at too high a speed for a short time is unlikely to kill anything, it would just cause the program to crash.

Mark

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sat Jun 20, 2020 7:19 pm

hoglet wrote:
Sat Jun 20, 2020 5:34 pm
John Ferguson wrote:
Sat Jun 20, 2020 5:29 pm
Your mention of #BFFE reminded me that when I fitted the Yarrb board I thought you had to set this location for ram or rom.

So I have probably done some harm. This might be the problem.
I don't understand this comment.

Can you any explain a bit more exactly what you think you did wrong?

I didn't think you ever had the YARRB booting to the > prompt and be able to type anything in.

Dave

PS And make sure you don't miss Mark's post above
Dave

When I received the Yarrb board there was no guide with it just a piece of paper showing the EEPROM contents and values to put in #BFFE. So I am wondering whether it was working and I typed into #BFFE and set the clock speed by mistake.

I can't remember whether this is was what happened. I am saying could I have accidently changed the clock speed.

If I had what are the implications?

What I do know is that before fitting the Yarrb board the Atom was working normally with the noise killer board.

So I am thinking what was it that changed the situation.

By the way the logic probe has just arrived.

John

User avatar
hoglet
Posts: 9241
Joined: Sat Oct 13, 2012 7:21 pm
Location: Bristol
Contact:

Re: Yarrb - redesigned

Post by hoglet » Sat Jun 20, 2020 8:01 pm

John Ferguson wrote:
Sat Jun 20, 2020 7:19 pm
When I received the Yarrb board there was no guide with it just a piece of paper showing the EEPROM contents and values to put in #BFFE. So I am wondering whether it was working and I typed into #BFFE and set the clock speed by mistake.

I can't remember whether this is was what happened. I am saying could I have accidently changed the clock speed.

If I had what are the implications?
I really don't think any damage would have resulted.

All that would have happened would the clock would have changed to 4MHz and the Atom would have crashed.

As soon as you hit BREAK, or power-cycled the machine, it would be back to running at 1MHz again.
John Ferguson wrote:
Sat Jun 20, 2020 7:19 pm
What I do know is that before fitting the Yarrb board the Atom was working normally with the noise killer board.

So I am thinking what was it that changed the situation.
But you have said that even before you fitted the YARRB there were some reliability issues, with the machine occasionally locking up:
It has always had a tendency to lock up from time to time.
So it already had an intermittent fault, which has just got a lot worse with all the prodding and poking.

My best guess at the moment (from what we saw in "Set 4") is that the 6502 is flaky. But is could also be a mechanical issue (a hairline crack in a PCB trace) which would be far harder to track down.

You'll have another 6502 to try soon right? Is this the Atom that's being returned?
https://www.ebay.co.uk/itm/Acorn-Atom-V ... 3613664395

That looks like a genuine 2MHz SY6502A to me.
John Ferguson wrote:
Sat Jun 20, 2020 7:19 pm
By the way the logic probe has just arrived.
Do have a play with it - you could be able to tell the difference between:
- a signal that is undriven
- a signal that is driven low
- a signal that is driven high
- a signal that is pulsing

Try it on the clock signal into the 6502; that should definitely be pulsing.

Dave

User avatar
roland
Posts: 3862
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yarrb - redesigned

Post by roland » Sat Jun 20, 2020 8:07 pm

John Ferguson wrote:
Sat Jun 20, 2020 7:19 pm
When I received the Yarrb board there was no guide with it just a piece of paper showing the EEPROM contents and values to put in #BFFE. So I am wondering whether it was working and I typed into #BFFE and set the clock speed by mistake.
All the registers of the CPLD (i.e. #BFFE and #BFFF) are reset to their default values after a power on. So whatever you typed it was undone at the next power cycle. It can't kill your Atom. I often try to switch to 4 MHZ just to see if the processor can handle it. Most don't but they keep working after a reset or power cycle.

I also don't know what might be wrong. I suggest either to learn to work with your logic analyser (connecting the wires and using the software) or send / bring your Atom to somebody with a scope, spare parts and knowledge of Atoms.
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sun Jun 21, 2020 8:16 am

Dave and Roland

I think that you have spent a lot of time helping me for which I am very grateful. The knowledge you both have makes membership of this forum so useful. I suggest we call this a day for the time being. I am reassured that the Yarrb board will continue to operate normally.

The present Atom was my first one for my collection which at one time stood at 3 machines. My experience in getting them to work was that the connections were never a problem for me in getting them to work as they usually worked when they were originally owned. Replacing the support chips usually corrected the fault and more rarely a major chip like the one for the operating system.

Standing back to review the situation I have in mind that the Atom prompt can only appear after many break key presses and only after a substantial amount of time. This suggests to me that what has happened is a coincidence with fitting the YARRB board and that a support chip is failing but starts to work a little better after being warmed up.

My plan now is to replace all the support chips. When I receive the Atom that is being returned to me I will be able to replace other major chips as well.

Now I have the logic probe it gives me an opportunity to understand more about the signals I should expect. I will try it on the clock signal as you mention.

Regards

John

Kazzie
Posts: 1600
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Yarrb - redesigned

Post by Kazzie » Mon Jun 22, 2020 11:19 am

1024MAK wrote:
Sat Jun 20, 2020 5:16 pm
hoglet wrote:
Sat Jun 20, 2020 4:54 pm
...For the convenience of anyone else mad enough to be following this...
Don’t know about mad, thought I was already classed as that! Certainly a lot of head scratching going on... :shock: :x 8-[ :-#
Label me as another one of those mad hatters following along. :lol:
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

BillG
Posts: 13
Joined: Wed Dec 17, 2014 6:56 pm
Contact:

Re: Yarrb - redesigned

Post by BillG » Tue Jun 23, 2020 4:32 pm

Roland kindly sold me a yarrb2 board which I have now built and successfully installed - works perfectly and is really impressive - I just need to work out how to use all its features. Thanks!!

Previously my Atom was set up with ATOMM2 to auto load on power up via IRQ to #A000 version - which was just very convenient. WIth YARRB is it possible to somehow default #A000 as ATOMM2 and use the IRQ to PL8 again - I realised after installing it got really upset with the link place because I think ATOMM2 is now at #E000 instead. It's just less convenient to have to poke into #BFFE and then LINK #E000 every time I switch on - I guess I must be lazy....?
Atom June 2020  Pre YARRB.jpg
Atom June 2020  YARRB not working if PL8 link on.jpg
Atom June 2020  YARRB works 1.jpg
Atom June 2020  YARRB works 2.jpg

User avatar
roland
Posts: 3862
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yarrb - redesigned

Post by roland » Tue Jun 23, 2020 4:37 pm

If you remove that IRQ link then the Yarrb should start with AtoMMC active while the rom is at #E000. Works perfectly in about 30 Atoms or so :)

I'll publish an user guide soon. It's almost ready.....
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

BillG
Posts: 13
Joined: Wed Dec 17, 2014 6:56 pm
Contact:

Re: Yarrb - redesigned

Post by BillG » Tue Jun 23, 2020 4:51 pm

That's interesting - one thing I didn't mention was that I modified my ATOMM2 board with a little hardware reset delay line into the PIC because the IRQ/#A000 version was a bit hit & miss on power up/break in my Atom (mine seems a bit slow to get going, so often seemed to miss the IRQ signal being sent from ATOMM2 - my only solution was to slow the ATOMM2 reset down to match - I note there's a new version that might have addressed this). I'll take the extra delay out and see if that works for how it is now). Cheers Roland - looking forward to the guide!

User avatar
oss003
Posts: 3278
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yarrb - redesigned

Post by oss003 » Tue Jun 23, 2020 5:30 pm

The #Exxx version of the AtoMMC rom uses a patch at #Fxxx so doesn't need an IRQ.
Please disable the IRQ with *CFG C0 and remove LK3.

Greetings
Kees

User avatar
roland
Posts: 3862
Joined: Thu Aug 29, 2013 9:29 pm
Location: Born (NL)
Contact:

Re: Yarrb - redesigned

Post by roland » Sat Jun 27, 2020 11:38 am

I have uploaded the second draft version of the Yarrb User Manual to the first post of this topic.

Or download it directly from here: download/file.php?id=57812

Any comments are welcome!
FPGAtom: 512 KB RAM, Real Time Clock and 64 colours
MAN WOMAN :shock:

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sat Jun 27, 2020 1:24 pm

roland wrote:
Sat Jun 27, 2020 11:38 am
I have uploaded the second draft version of the Yarrb User Manual to the first post of this topic.

Or download it directly from here: download/file.php?id=57812

Any comments are welcome!
Thank you Roland

I will study it with great interest.

Regards

John

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Mon Jun 29, 2020 5:34 pm

John Ferguson wrote:
Sat Jun 27, 2020 1:24 pm
roland wrote:
Sat Jun 27, 2020 11:38 am
I have uploaded the second draft version of the Yarrb User Manual to the first post of this topic.

Or download it directly from here: download/file.php?id=57812

Any comments are welcome!
Thank you Roland

I will study it with great interest.

Regards

John
Roland and Dave

I have made some progress in getting the Atom to boot with the YARRB board and noise killer board fitted.

If I put my finger on the 4mhz clock crystal the Atom will boot complete with the prompt sufficient to accept a key press. However it is not stable and removing the finger the screen crashes to a rubbish display. I am not pressing hard with my finger just touching.

The crystal is a new one and the connections on the board are sound according to my multimeter.

Can you comment on this phenomenon please as to further action or investigations I need to take.

Regards

John

User avatar
oss003
Posts: 3278
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yarrb - redesigned

Post by oss003 » Mon Jun 29, 2020 6:30 pm

Maybe you can demonstrate it at the virtual ABUG tomorrow .....

Greetings
Kees

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Mon Jun 29, 2020 7:06 pm

oss003 wrote:
Mon Jun 29, 2020 6:30 pm
Maybe you can demonstrate it at the virtual ABUG tomorrow .....

Greetings
Kees
I am not aware of the ABUG meeting our how I would do this though I have used Zoom before. I suppose I could prepare a video.

John

Kazzie
Posts: 1600
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Yarrb - redesigned

Post by Kazzie » Mon Jun 29, 2020 7:08 pm

John Ferguson wrote:
Mon Jun 29, 2020 7:06 pm
oss003 wrote:
Mon Jun 29, 2020 6:30 pm
Maybe you can demonstrate it at the virtual ABUG tomorrow .....

Greetings
Kees
I am not aware of the ABUG meeting our how I would do this though I have used Zoom before. I suppose I could prepare a video.

John
See here. :)
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

User avatar
oss003
Posts: 3278
Joined: Tue Jul 14, 2009 12:57 pm
Location: Netherlands
Contact:

Re: Yarrb - redesigned

Post by oss003 » Mon Jun 29, 2020 7:15 pm

Just ask on the ABUG thread to add you to the list and you will get an ID and Password to login for Zoom.

You could use a webcam or just locate your laptop so that we can see your Atom.

Greetings
Kees

cjm
Posts: 20
Joined: Fri Sep 01, 2017 3:19 pm
Location: Woking, Surrey
Contact:

Re: Yarrb - redesigned

Post by cjm » Tue Jun 30, 2020 5:57 pm

roland wrote:
Sun Oct 06, 2019 12:06 pm
Edit: I attached the user manual to this post to make it easier to find.
Thanks - that's very useful!
Atom | BBC Model B | RPi | Software professional | Electronics hobbyist

John Ferguson
Posts: 209
Joined: Sat Oct 07, 2017 12:35 pm
Contact:

Re: Yarrb - redesigned

Post by John Ferguson » Sun Jul 05, 2020 10:28 am

Since I had all that tremendous help from Roland and Dave as a result of fitting the Yarrb board i have carried out the following changes which has resulted in the Atom and Yarrb board working normally for 99% of the time.

The 4mhz crystal only working when you touched it was cured by fitting a replacement. A very strange fault.

All the supporting IC's and crystals have been replaced with the exception of the 81LS95's as I am waiting replacements.

J1 is left open and J2 closed with pins 2 and 3.

A great sigh of relief.

Now to get on with finishing the HDMI interface.

John

User avatar
anightin
Posts: 494
Joined: Thu Aug 23, 2018 2:03 pm
Location: Cambridge UK
Contact:

Re: Yarrb - redesigned

Post by anightin » Mon Jul 06, 2020 6:28 pm

Just a quick thank-you to @Roland so say what a brilliant product and manual. =D>

I must admit this was the 3rd YARRB I've installed now, with my most recent ATOM getting the upgrade treatment and it booted first time 8). This was a YARRB 2 board that has been waiting too long to be fitted.

I finally got round to installing it after a recap, composite mod, regulator bypass and new crystals. I tested the machine after each mod just to be sure and fitted the YARRB 2 last.

I'm just waiting for an SRAM for a Noise killer board from @hoglet (another item waiting patiently in the parts box) which I'll fit tomorrow after Farnell have delivered, and for @SirMorris to see if he has any ATOMMC left.
IMG_7320.jpeg
IMG_7321.jpeg
IMG_7318.jpeg

Post Reply

Return to “acorn atom and acorn system series”