ZX Spectrum +2 Repair. (No. 7). Part 5.

[@beingatliberty]

The depth precision and clarity of your videos is a breath of fresh air.

[@FirstWizardZorander]

Wow, this is the most thorough repair to date, by far! Always a joy to watch your videos.

[@DavidPagett]

Loving this series, it has become a CSI style forensic examination :-) More please.

[@GroovyDrifter]

Hey! That's exactly the same that one of my grey +2 does... ramdom failures on most of the even pages. I used a divide with testrom3 loaded.

[@MrGot2k]

Great video! I'am waiting for the next part.

[@GadgetUK164]

Another brilliant video! I have the same logic sniffer there - from Dangerous Prototypes. I bought mine from Saleae. Look forward to the next part.

[@paulb4uk]

These videos are superb love the detail and explanations of what your doing very helpful.

[@0M9H4X_Neckbeard]

Wow really loving this channel, I wish I had your computer knowledge. I could fix anything! But then again all my stuff still works for now. Really want to pick up a Spectrum by the way, seems to be such an amazing system! Almost as rad as the Vectrex.

[@JoulesperCoulomb]

0M9H4X I almost always have rubber keyed 48K Spectrums available for sale and they can be quite cheap if you don't need it to be in the very best cosmetic condition. Send me a message via zx.zigg.net if you're interested.

[@croysk]

Great stuff!

[@pcuser80]

Is the memory refresh ok? Contended ram is mostly refreshed by the Ula, non contended by the cpu. The Trigger can be made very simple: if (address bus = 7EA2 and m1 = low) then start trigger.

[@JoulesperCoulomb]

pcuser80Your trigger suggestion might have worked, though it may be one of the combinations I tried. It is surprising how often unwanted triggering happens as the buses change state, passing through the configured trigger condition for just an instant. I found it best to be as specific as possible when configuring the trigger.

[@alexanderseverniy4054]

Damn! That escalated fast ))) From home repair to geek magic in 10 minutes........

[@obiwanjacobi]

Awesome videos, I've watched them all. Is that a Seeed Logic Sniffer you're using? And what software is that? Perhaps do a video on that? Tell us how you use it and what are its strengths and weaknesses? Thanx!

[@davidesparzaguerrero4545]

Can the faulty component (whatever is causing these memory corruptions) have been damaged by the previous repair? The problem is that I cannot remember anymore if the hangs and crashes used to happen also before then. I am pretty sure I remember myself playing plenty of games like Carrier Command all the way to the end for several days, and then it cannot be that whatever is wrong was so from the start. It's a long time ago, probably 1989-1990, when I was last able to load and use commercial programs in this computer, and the repair job was about there, I think precisely because the computer wasn't loading programs properly. I was 11 back then, when I took the computer to the repair workshop by myself, so probably they didn't detail to me what they had exactly done for that reason. In any case, using this logic analyzer has crossed the border of what I would consider "Trying to fix it as much as reasonably possible". Now you are, for me, in the land of legends...

[@ratos74]

when I select 48K mode on my +2 there will be no bankswitching, writing to port 32765 will have no effect.

[@JoulesperCoulomb]

That's how it's meant to be. However, as I demonstrated in part 4, it's possible to disconnect the signal that inhibits memory map bank switching.

[@ratos74]

@@JoulesperCoulomb Oh, I omitted the inhibit disable part of the video. However You can do an USR0 command in 128 basic and the machine resets into 48 basic but with memory paging enabled. Please have my excuse and respect!