Need plustest.dsk testbench 9 output on original CPC 6128

[freemac]

Hi,

I need WinAPE'plustest.dsk testbench 9 result on an original Amstrad CPC  6128.
www.winape.net/downloads.jsp > Plus Test

It's an ask from my MiST-board FPGAmstrad project cat co-worker, he seems motivated to reach a better WAIT_n generator than my... welcome him !

Thanks for help

[Gryzor]

Darn, my 6128 just got decommissioned...


Anyone?

[GUNHED]

Doesn't it bores one out if nobody helps? 2000 members...

Ok, I may got what you need....

System 1: CPC6128, CRTC2:
Test 9 works normal until the &EC codes, there are two errors marked with an "X".

&ED, &46: 2
&ED, &4E: 2

After &ED, &5D it suddenly stops working!


System 2: 6128 Plus:

&ED, &46: 2
&ED, &48  5
&ED, &49: 5
&ED, &4E: 2

&ED, &50: 5
&ED, &51: 5
&ED, &58: 5
&ED, &59: 5


After &ED, &5D it suddenly stops working! Probably a crash, since a spot appears on screen.

Any other help needed?

[Kris]

Here are my results (teste performed on CPC 6128 CRTC 1)


Pictures of each screen attached in the .rar file.

[GUNHED]

That's exactly what I got. However on an emulator the test doesn't crash, it continues.

[Kris]

Emulators are not perfect 


Anyway, if it can help to improve the CPC core for the MisTer ...

[remax]

On my CRTC4, it doesn't crash at ED Codes, the test goes to the end.


It seems to be the same 4 pages as Kris but it doesn't stop. Do you need the remaining pages ?

[freemac]

On my CRTC4, it doesn't crash at ED Codes, the test goes to the end.

It seems to be the same 4 pages as Kris but it doesn't stop. Do you need the remaining pages ?

Yes sure, it can be interesting for future, when I'll understand this testbench. In WinAPE it does pass, all testbench are welcome in CPC, I love them !


You've got also some crosses (failing tests) ?

[Kris]

I have CRTC 0 & 4 as well, I will do the tests this evening to be sure we have the same behaviour and results.

[arnoldemu]

I have CRTC 0 & 4 as well, I will do the tests this evening to be sure we have the same behaviour and results.

The ones that are failing and crashing are interrupt mode 0 and interrupt mode 2 related so rely on the data on the bus to give the appropiate interrupt vector or opcode at interrupt time.
I have seen it run to completion on a few CPCs.

Is everyone who is posting results running it on a bare CPC without any other hardware attached? (except Hxc/Gotek).

[Kris]

I did tests on a bare CPC, nothing connected...

[arnoldemu]

Here are my results (teste performed on CPC 6128 CRTC 1)


Pictures of each screen attached in the .rar file.

On my type 1 it doesn't crash:
6128
has shielding  and centronics connectors.
UM6845R (Type 1 CRTC)
AY-3-8912
Zilog Z084004PSC
40010 36AA
SED9420C
UM8272A
z70290
40015
40025
TOSHIBA TMP8255AP-5
EME-156
@Kris: What border colour(s) does the attached program show on your type 1?
The attached program attempts to make the same configuration as the plus test.CRTC Register 0F is selected, I register set to &BE, testing IM0/IM2 instruction, code at &8000.

[arnoldemu]

System 1: CPC6128, CRTC2:
Test 9 works normal until the &EC codes, there are two errors marked with an "X".

&ED, &46: 2
&ED, &4E: 2

Both of these are "IM 0".

After &ED, &5D it suddenly stops working!

This is IM 2.

[Kris]

Hi Arnoldemu,


I'm currently on holidays, far away from home, I will be able to test your .DSK next week.
Regards

[arnoldemu]

@GUNHED Does your CPC with type 2 have a 40010 GA or a 40007 GA?

[GUNHED]

It's a Schneider CPC6128, I don't know the number, have to look the next time I open the case.

[arnoldemu]

I have tested all my CPCs and all pass plustest without problems. These CPCs have different CRTC.

I now being to think this bus data is happens only with older Gate-Array. All my CPCs have 40010 gate-array except my type 4. All pass plustest.

My oldest computer is a 464 has Z70200 MC0002B with a 40010.
Did anyone test on a CPC with a 40007 GA?

[arnoldemu]

It's a Schneider CPC6128, I don't know the number, have to look the next time I open the case.

Please note down the information on all the ICs including the 244 and 373 near the GA.

[Kris]

Result is a white border on the Plus; will test on type 1 soon.

[GUNHED]

IIRC these commands are connected to IM 2? In this case there may be a floating bus problem.

[arnoldemu]

IIRC these commands are connected to IM 2? In this case there may be a floating bus problem.

Yes and IM0.

I am trying to find:* What is the value of the data on the bus?* Did an IC accidentally put it on there?* Is the data fixed?* Is it caused by a specific CPC mainboard and ICs? (This is why I wanted to know the numbers on the GA, 6845, 8255, 74LS373 and 74LS244 near the GA).

To make this problem happen then:* An IC needs to be responding (accidentally) to interrupt acknowledge and put data on the bus. 
* Something is responding to memory refresh from Z80 (where IR is the address on the bus) (The values and conditions are there on the bus but the CPC doesn't use it).
* The data latch controlled by the GA is holding data for longer than it should.
My original thoughts was that it was from the status register of CRTC type 1. My program tests this.
My CPCs don't fail the test so there must be other conditions that cause this.

[GUNHED]

&ED,&5D is RETN, in both cases (CPC6128 and 6128 Plus) I had the crash after this command.

Look at page 13 

[arnoldemu]

&ED,&5D is RETN, in both cases (CPC6128 and 6128 Plus) I had the crash after this command.

Look at page 13 

Thank you for the correction.
I now need a page of these 
I was running the instruction *timing* test not the interrupt wait state test.Doh.
I tested on my type 1 and yes it crashes on ed 5e, and yes 46 and 4e show the same as you describe.I can investigate

[arnoldemu]

Thank you for the correction.
I now need a page of these 
I was running the instruction *timing* test not the interrupt wait state test.Doh.
I tested on my type 1 and yes it crashes on ed 5e, and yes 46 and 4e show the same as you describe.I can investigate

Both my CPC type 0 and type 1 crash on test 9.
Arnold emulator, when in CPC emulation, does the same as plustest does on my CPC but shows the same result as TFM's type 2.
Test 9 is obviously not compatible with CPC.
I now need to work out if I can get the same results as Kris. He sees C instead of 2 for 46 and 4e

[arnoldemu]

I tried test 9 on my type 1 and it reports the same as my type 0 and type 2. So I have a different result to Kris.
Plus test will calculate a time of '2' for IM0 on CPC where bus value is 0x0ff.

Kris' type 1, seems to be putting 0x040 on the bus.

@Kris: please run the test in the disc attached to this post.

What colour bar do you see?
(for type 0 and type 2 I would expect to see orange).