STE no reset on power up issue fix

[exxos]

This seems to happen very often now, but not just on STE machines. Where the machine doesn't reset "from cold" or doesn't reset on power on at all. Normally the work-around is to press the reset button and machine boots as normal.

There have been various reports of this when adding boosters, or changing the PSU etc. I had this issue yesterday while using a stock machine to test simms.

The reset circuit is pretty rubbish to start with, but like always, bad caps are mostly to blame, but not only that, the caps seem to be wrong values as well.

sters.JPG (34.88 KiB) Viewed 11326 times

The quick fix I did (as others have done and confirmed) is to solder a 22uF across C100 (for approx 33uF total) . Upping the value of C102 does nothing. In fact it seems C100 *should* be a higher value than C102, but the circuit shows C102 as larger. I haven't yet had time to look into this more.. but this is what I did..

IMG_4553.JPG (250 KiB) Viewed 11326 times

I would suggest changing C102 for a new capacitor, keeping 22uF as the value. C100 replace with a new 47uF capacitor. Voltages between 10V - 63V are typical. This will likely solve a lot of reset issues on the STE.

As reset circuits are similar with STFM, Falcon, MEGA etc, this fix will likely solve reset issues on those machines as well.

[exxos]

I decided to run a simulation on this circuit. As upping the value of C100 doesn't make much sense, even though it has been proven to work by multiple people.

sch555.PNG (16.34 KiB) Viewed 6830 times

First off the standard values.

BLUE = SYSTEM RESET
RED = C1 ( C100)
GREEN = C2 (C102)

1022.png (23.44 KiB) Viewed 6830 times

Then upping C100 to 50uF. Basically does nothing in terms of reset times.

5022.PNG (22.22 KiB) Viewed 6830 times

Then upping C102 to 50uF (C100 10uF). Which sd expected, increases the reset time.

1050.PNG (23.99 KiB) Viewed 6830 times

It is slightly puzzling why the value of C100 solves reset problems.

So I decided to remove C100 Completely simulate a total cap failure.

no10.PNG (19.3 KiB) Viewed 6830 times

So reset does not happen at all in that test. Would be odd to think the capacitor would fail completely though

[exxos]

In this test I simulated a slower ramp up of the 5V rail.

RED = 5V
GREEN = C100
BLUE = C102
DARK GREEN = RESET

slow5v.PNG (24.81 KiB) Viewed 6827 times

Even slower 5V ramp up.. and BOOM!

slow2.PNG (29.7 KiB) Viewed 6824 times

It is known my own PSU has a slow ramp up (soft start current limited) so I wonder if with the many Atari variations of the PSU, that there are likely various ramp up speeds on the 5V rail. Plus with various failing and replaced capacitors will also affect the ramp up speed. Even so a totally "stock" machine would still fail to start. So we cannot simply rule it out to be PSU issues. Part of the blame is on C100 failing. I think another part of the problem is that "reset high" isn't actually high for very long. It's mostly depending on the state of the 5V rail itself. Even so, upping the value of C100 does nothing even with a slow ramp up on the 5V rail. I also guess shorting C100 out across the reset switch for 30 years probably didn't do it much good

I guess is someone has a 4 channel scope, measure a real world reset circuit on a STE and see what is going on with the stock capacitors in the reset circuit, then a extra 22uF across C100.

[Higgy]

@exxos do you Have any info on the Reset fix for a C070789 Rev D? I am working my way through the ST FM fixes.
I can't see where are the corresponding C100 and C102.
I guess the caps will be physically located near the Reset Switch.
Thanks

[exxos]

@exxos do you Have any info on the Reset fix for a C070789 Rev D? I am working my way through the ST FM fixes.

Looks like C2 and C7.

STFM.JPG (48.54 KiB) Viewed 6328 times

[Pacman]

What's the final verdict on this matter?
My STe often suffers from this problem).

Will a better power supply likely improve matters?

[exxos]

Do what it says in the first post..

[Pacman]

Do what it says in the first post..

The following is what's confusing me (from the end of this posting) -emphasis in red by me:

Even so, upping the value of C100 does nothing even with a slow ramp up on the 5V rail.

[exxos]

Do what it says in the first post..

The following is what's confusing me (from the end of this posting) -emphasis in red by me:

Even so, upping the value of C100 does nothing even with a slow ramp up on the 5V rail.

Nobody helped and I took it no further. Anyone can measure if they want to study it. But the fix is still replace the caps.

[Pacman]

OK. Got it
Thanks