Weird Gotek problem in an STE?

[Badwolf]

Yes, I'm on 3.41 too, I think.

Good call to update the firmware before diving too deeply into the weeds!

BW

[Hellraiseruk]

Always best not to overlook the simple things :)

JD

[Badwolf]

Always best not to overlook the simple things :)

Absolutely!

Unfortunately updating to 3.42 didn't fix the problem, but I note 3.41 was when that 'don't wait' came in. I'll try downgrading to 3.40.

What PSU are you using there with the soft start?
I had assumed that the drive wouldnt work when i saw the screen scrambled, interesting you say it works and its only the OLED thats affected.

I'm actually using Exxos' PSU and this is in an Amiga 500, so it's not directly analagous. I turned my attention to the start-up delay as I know this has one and actually it's a bit too slow for the Amiga!

The Amiga's power on reset signal seems to assert for about 250ms so the CPU tries to start under a slight brown-out. Sometimes it'll reset a few times and then work and sometimes it'll just get stuck in a loop. I 'fixed' that by actually modifying the Amiga reset circuit to assert for a bit longer (changed C712 to 47k, if anyone's interested).

I put two and two together and perhaps made five, as initially every time I powered up the PSU before plugging in both the Amiga and the OLED would work. I've since noticed that doesn't appear to be true 100% of the time.

So now I'm a little more inclined to think it's a screen/firmware issue. Perhaps compounded by the brownout a bit?

Resetting the Gotek itself whilst powered doesn't help, BTW.

BW

[Maximilian]

Do you have a picture of the setup?
(Wiring etc)
And from the insides of the Gotek?

[RB80atari]

Hi,

Yes could be that the latest FF firmware updates are speeding up the power on time for the oled screen, pressumably earlier versions- prior to 3.4.1 took a bit little longer to initialize, i do not know...
Ive not revisited this yet as have others things i needed to look at, the STE has been put to the side for now.
But if you do go back to 3.4.0, let us know how you get on hopefully this will fix it.
Its strange how it didnt affect my other STE and it works fine in my Mega too.
I guess the other thing you could try would be to slew the power rail going into the oled itself?

Cheers

[Badwolf]

Do you have a picture of the setup?
(Wiring etc)
And from the insides of the Gotek?

I don't have one to hand, but mine's a normal Gotek-in-an-A500-with-OLED-instead-of-eight-segment-display setup.

What do you have in mind?

BW

[Badwolf]

But if you do go back to 3.4.0, let us know how you get on hopefully this will fix it.
Its strange how it didnt affect my other STE and it works fine in my Mega too.
I guess the other thing you could try would be to slew the power rail going into the oled itself?

Wilco. Yes, I don't see it on the Gotek in my Falcon which can run from the same supply too, so I suspect firmware or a variation screen to screen.

I also haven't noticed it happen when I run the A500 from my other external supply.

An interposer circuit that delays the 5V line would certainly be an option, but I'll try downgrading again when I get a chance.

BW

[Maximilian]

What do you have in mind?

It should be working..., so we must be missing something.
And a picture says more then a thousand words.

I was thinking of the chinesium wires that are going around, iron core and copper outside.
(you can use a magnet to check)
Those wires can have a couple of ohms over a short distance of wire...or maybe a bad connection.

I've been staring at a Gotek programming issue for an hour last week realizing that it was the wire itself, I even removed the connectors before I found the issue...

[Badwolf]

It should be working..., so we must be missing something.
And a picture says more then a thousand words.

I was thinking of the chinesium wires that are going around, iron core and copper outside.
(you can use a magnet to check)
Those wires can have a couple of ohms over a short distance of wire...or maybe a bad connection.

I've been staring at a Gotek programming issue for an hour last week realizing that it was the wire itself, I even removed the connectors before I found the issue...

Aha, gotcha. In this case I'm using the original A500 floppy power connector and the wires to the OLED itself are the original Gotek ones that went to the seven (plus one) segment displays, but no harm in running an independent power line to the OLED as a test, I suppose.

It's all buttoned up again now as I'm working on my STFM, but will try to get pictures of it in the act when I next have it running.

Cheers,

BW

[Maximilian]

I've also heard of defective screens that need an extra reset before working.
https://hxc2001.com/docs/gotek-floppy-e ... -oled.html
Note
If the OLED screen doesn’t appear to work and if you are sure about the OLED wire connections, try to reset the Gotek : Shutdown the machine, press the Gotek’s right button, switch on the machine and keep the button pressed during ~5 seconds. This will clear the internals settings and force the long screen detection process. Some OLED screens with factory defects need this to be detected properly the first time.