... /> Flavio Matsumoto performed this test again, this time using the ... /> Machine type: unknown
Contention offset: 0x00

... (0x7f)
High port contention 2... failed(0x7f)
contention" and "iocontention" from fusetest ...

... and perhaps the contention pattern (although contention detection didn' ... br />
[*] The remaining test programs from [URL="http://wizard ... positions. In this test, you should determine ... contention, but everything else should work just fine. This test ...

... out the contention.[/QUOTE]

Thank you!

Test findcontention ...

... more money to spend on test-gear although I did manage ... for messing about with an IR helicopter I have, only to ... I can afford a proper test tool. I got back into ...

... LIST]
[*] Memory contention is identical to Spectrum 128K ... contention to keep everything synchronized. Although this is not a test ... specifically about contention, in practice ... the same contention pattern, since a different contention pattern ...

... /> [/LIST]
The performed test checks the timings of an ... label".

Executing this test on a Spectrum 128K produces ... main screen access regardless of contention.

Moreover, the ... sure).

The test program used here is called ...

... 'll also pass that last test :)[/QUOTE]
I just executed ... . Indeed it passed the last test, but not the first one ... is correctly emulating that memory contention and main video updates in ...

... GAL unrainer that alters memory contention and my DivIDe is a ... the case) and try the test again. I've noticed that ... is does the (incorrect) memory contention exist on all grey+2 ...

... signal which goes away during contention periods - and therefore avoid the ... make a test that simulates ZX contention, then go and test the clock ...

... signal which goes away during contention periods - and therefore avoid the ... make a test that simulates ZX contention, then go and test the clock ...

... !) it was time to test it on the real hardware ... make a good amount of contention, I just entered the ... two bouts of 6 cycle contention. The Spectrum clock, after ... too. A more complete test will be to count the ... what's expected. This test is a little while off ...

... mandatory. The STIME test shows the red segment ... /> FUSETEST reports a contention offset of 9 instead of ... br /> This is another test I cannot remember the name ... 0x3800. Spectaculator shows me contention at one cycle before ... 5V while my cartridge test was in use ( ...

... not know how to measure contention time. Is there already any ... can I find info about contention patterns? [/QUOTE]

You ... length as well as the contention patterns of various things using ... /timingtest-0.2.zip"]timing test[/URL] program.

...

... length as well as the contention patterns of various things using ... /timingtest-0.2.zip"]timing test[/URL] program.

Patrik ...

... and RUN for each next test.

The infinite loop ... there to check for possible contention fluctuations, but it hardly applies ...

... and RUN for each next test.

The infinite loop ... there to check for possible contention fluctuations, but it hardly applies ... .com.br/2013/04/timing-test.html]blog[/url].

... is important to confirm that contention doesn't start earlier ... should cover the first entire contention cycle.[/QUOTE]

... 8690 and 8850, the awaited contention patterns became apparent (see ... blogspot.com/2013/04/timing-test.html]same posting[/url] ...

... something?

Contention on Spectrum 48K with "normal ... [/IMG]

Contention on Spectrum 128K with "normal ... Contention on Spectrum +3 should start at cycle 14361 but this test ... is correctly emulating memory contention and port *FF just ...

... ]Ops! It seems this test is reporting contentions with a ... m missing something?

Contention on Spectrum 48K with "normal ... at cycle 14335 but this test indicates 14336:
[/QUOTE ... ]

Yes, this test counts frame cycles from the ...

... . It doesn't emulate memory contention at all, even for the ... ), or specially produced test snapshots. An obvious test would be to dump ...

OK, let's test this. Suppose we are ... />
tacts, lost due to contention in sp-1 -> tacts ... , lost due to contention in sp-2 -> overall ... 8=7.5 tacts of contention delays. Would you agree ... = 2.6 tacts of contention delays in average (when using ...

... ;707424]OK, let's test this. Suppose we are executing ... the first PUSH executed during contention. After this, the exact ... same 5T delay until the contention period finishes, or until ... rough estimate, because the actual contention delay is deterministic and it ...

... Fuse's test suite would catch any situations where contention should be ...

[QUOTE=Philip Kendall;707717]I'd once again hope that Fuse's test suite would catch any situations where contention should be applied. It might not though...[/QUOTE]

He he thanks, I'll go back and have a look at that :)

[QUOTE=introspec;708067]Do I understand correctly that this is only relevant when IR points to the contended memory?[/QUOTE]

Absolutely. The ULA isn't going to apply contention when the address on the bus isn't one it thinks is contended!

... reference from Arjun and my contention is now sorted thanks ... all![/QUOTE]

All contention tests passed OK. Please note ... this is reason why this test detects spectrum model incorrectly and ... attempt to test ULA late model as ...

... to fill time without incurring contention delays while the ULA is ... write to the screen while contention is going on, a long ... 128K timings in my initial test, as that's my 'default ...

... owing to the I/O contention.[/QUOTE]

The number ... the extra delays due to contention will make the same routine ... will need some time to test everything properly...

... />
True. The I/O contention would only be an issue ... the extra delays due to contention will make the same routine ... time to test everything properly...[/QUOTE]

ULA contention will iron ...

I uploaded the test program [URL="http://www.mediafire. ... routine could work on all contention models including Spectrum +3.