atari 520ste problem with acsi2stm

[thejolz]

Thanks for the advice.
Yes, I have a Mitsumi power supply.
I bought a 68k HC, but it still isn't working. As you say, must be the power supply, so the next thing I'm going to do is set the values ​​recommended by Exxos on the power supply and change all the pullups. I'm using the ACSI2STM externally.

About the power supply, could I put a KBU1010 instead of a KBU1005?

Let's see if I have the same luck as Atari030

[Dlfrsilver]

Thanks for the advice.
Yes, I have a Mitsumi power supply.
I bought a 68k HC, but it still isn't working. As you say, must be the power supply, so the next thing I'm going to do is set the values ​​recommended by Exxos on the power supply and change all the pullups. I'm using the ACSI2STM externally.

About the power supply, could I put a KBU1010 instead of a KBU1005?

Let's see if I have the same luck as Atari030

Having the PSU recapped is the basic. you must be fine powering wise, and also, the current spikes from the PSU must be eliminated and lowered to the maximum (as seen on Exxos pages, in the pictures of oscilloscope measurements), by replacing the capacitors. The original ones were really bad to regulate in the PSU, so installing better ones is the right move to make.

Next, change the 2 big ones on the motherboard (they are important for stability and filtering as well on the MB).

Then, one important information : on my 2 STEs early revision, i have started by replacing the DMA chip FIRST, and it did not solved the corruption problem. Things ended when i bought from Exxos 2x 68000 "low current consumption". After replacing the original 68000 CPUs, the hard drive problem disappeared completely one end for all. The very possible problem about the HDD corruption is an "all-together" piece by piece problem. The DMA chip and the CPU are talking, bilateraly, but the stream/flux of data is running through traces that are 1) too thin, 2) too near, allowing some parasite noise to add up to the "data bits". The 68000 used in the STE is not per se the culprit, it's the motherboard fabrication as a whole. Outside the STE, the 68000 would do its job perfectly fine, but in the STE, it won't work 100% as intended. the 3) fact is that Atari for cut cost choosed to build them as 2 layers motherboard, when "normal good ones" are 4 layers.

So my personal choice would be :

1) PSU

2) 68000 Low Current consumption (less noise on the bus)

3) DMA chip replacement (if you do first the DMA chip, and it doesn't solve the problem, you will have to buy the 68HC000....).

Cheers !

[atari030]

One other important thing, one step at a time. Do one thing, test, then the next if it doesn't work. If you do multiple fixes it'll be a crap shoot trying to figure out what it was that actually solved the problem.

FYI, here is the chap from Atari Forum that had PSU problems.

https://www.atari-forum.com/viewtopic.php?t=44695

[russellnash]

In 1992 I bought a Protar hard drive and when I connected it to my STe it started corrupting the files after a short time. When I phoned Atari they said that the DMA chip needed replacing and offered to do so if I sent it to them. I didn't fancy that, so they sent me a new DMA chip which I paid £10 for and replaced it myself. It worked fine after that. Just a short anecdote.

[exxos]

A important step is to test the hard drive itself

viewtopic.php?t=43

The "bax DMA chip" thing doesn't tend to show in that test. But it proves if there is a problem elsewhere or not first.

With so many hard drives clones about these days, who knows how much testing they have all had in various machine configurations. Open source stuff gets ripped and changed all the time, probably by people who just want to make a quick buck rather than doing deep dive testing on multiple machines. So I would always question how much testing they have had. Even iffy SD cards or iffy sockets can be a factor often overlooked.

Take the BBAN hard drive I tested, it worked for months on @dad664npc STE but was completely unstable on mine. Those issues were fixed eventually but I think people are skipping over the basic steps of if the drive even works properly or not in the first place. Different tolerances on chips or batches of drives can make or break stuff. The pins used on some drives are to thick for the DMA port and likely causing damages as well. The list of problems is almost endless.

A lot of drives people power with phone chargers and the regulation is often bad. Not good enough for powering sensitive flash cards etc. They can corrupt cards and will show bad sectors if tested on a PC. Formatting them often fixes that issue. There are just so many things that can go wrong.. It doesn't help that the internet is full of the bad DMA crap and people jump to conclusions and don't consider much else. The DMA port connectors are 30+ years old now and some are bad and xorroded and need cleaning etc.

While recapping the PSU and HC CPU solves a lot of problems, if it's not improving things then your likely looking at the wrong stuff. If that hard drive test program doesn't pass like 99% of the time then the problem would tend to me, to indicate a problem related to the drive itself.

Again I can't emphasize enough that just because a drive works stable on one machine doesn't mean it's stable on another. There's a huge thread on the BBAN drive about all those type of issues.

[Dlfrsilver]

One other important thing, one step at a time. Do one thing, test, then the next if it doesn't work. If you do multiple fixes it'll be a crap shoot trying to figure out what it was that actually solved the problem.

FYI, here is the chap from Atari Forum that had PSU problems.

https://www.atari-forum.com/viewtopic.php?t=44695

It depends on too many factors. Some people reported that changing only the DMA chip fixed the problem, some others reported that (like me) changing the CPU & DMA was needed to solve the corruption, some others say that only changing the CPU did the trick for them.

It's a multipath solving problem. It's not a single path for everyone. there are 4 or 5 ways to get to the fixed result.