Oric Atmos resets after a few seconds ... when not grounded

[TomPhobos]

Hi,

as promised in my welcome post here is the problem that i have with my Oric Atmos.

It's a german Atmos that was sold to me as new old stock. So i bought it directly. When i got it first i measured the power adapter to check if the output voltage is too high. But it wasn't too high, nothing came out of it. OK, i checked out the parameters of the power supply and used a 9V universal one.
The Atmos started, but only a cursor appeared in the upper left corner. So i opened the computer and noticed that i wasn't the first one who did it. The warranty sticker was lift up before. Ok, so i finished opening it and here's the mainboard:



As you can see it's an issue 4 mainboard. It has 2 ROMs:



I pressed them i their sockets and luckily the computer started ... but only for a few seconds. Then it resets and hangs.
When i checked the mainboard i noticed that there was some work on it before:




There were also several cables on the back side of the mainboard (i don't have photos yet, i'll make them at the weekend).

OK, i read a bit here in the forums and first i replaced the 3 capacitors ... without success.
But when i checked if one of the components are getting hot suddenly the keyboard clicked. I was touching the modulator, so i grounded the computer. And this was it. When i connect the modulator or another part with ground connection to a "real" ground (radiator, ground of a power socket) the computer works. So it seems to be a problem with the ground line.
I checked many things, measured the voltages, but nothing helped. I measured about 3,8V from the ground line (modulator) to the gnd-connection of the power socket. First i thought that this may cause the problem. But then i read that the 7905 is a negative regulator so i think that voltage is normal, right? And could it help to replace the 7805?

This weekend i'll make some more photos of the modifications on the solder side of the mainboard. Maybe the problem was caused by them. I hope you can help me to bring my Atmos back to life ...

[iss]

Welcome here! It's very interesting to see the German version of Atmos .
The 3.8V voltage difference as you measured is normal - and this is exactly because of 7905.
Replacing 7905 with 7805 will require changes on board which IMO is not worth.
You can check all electrolytic capacitors and replace them especially C21. Check this old post.
Just be careful - there is no common ground between power jack and board's ground track, common is the positive power line (i.e. 9V).
Check the power jack - is it good? Does restart occur when you move the power wire from adapter?
Additionally you can clean the soldering points of 7905 and/or re-solder them (on picture they are looking bit ugly ).

[Godzil]

Before everything else remove that C8 capacitor which is badly damaged and put another one in place. C8 is the filter on the power supply entry (just before the 7905)

C21 could also be a problem, but due to the bad shape of C8, that will be the first thing I do before everything else (Dumping the ROM could be done before of course )


For your ground problem what happen if you run your atmos without connecting it to a monitor?
Be carefull, because of this shitty design as the ground is floating on the Oric, it could led to big problems when connected to other devices with major ground level differences.

Also the 7905 can't be replaced by a 7805 without a major change in the PCB layout for the regulator.

If your test was with a monitor connected, you may have two device with floating ground, this could be the problem.

Also, look at the oric services manual and check which reworks has been done on your device, there are a few related with problem on boot that you may hit on your device.

[TomPhobos]

Well, replacing the electrolytic capacitors was the first thing that i did when i opened the computer (especially when i noticed the damage at C8 ). I replaced all 3 capacitors.

The problem also occurs with nothing connected than the PSU. Then i also get keyboard-clicking when i connect the modulator with the ground (e.g. by touching the modulator). The power connector seems to be ok, it has good solder points and the computer even restarts when i don't touch it.

The Atmos has strange modifications on the solder side of the board. There are many cables around, i'll make a photo of them when i'm back home. But e.g. there is no modification 63b.

Well, i also think it's a good idea to resolder the 7905 (or replace it with another one). Of course i don't replace it with a 7805

[Godzil]

What is the voltage between the Oric ground and the +5V line?


You're problem is quite strange, please post a clear photo on the bottom of the pcb when you can it may help to understand.


Edit: what is behind the black tape on the extension connector?

[TomPhobos]

OK, here are some more photos of the back side of the PCB and the modifications. Sorry, some of the photos are a bit dark, i'll make some more photos in the daylight

Here is an overview of the board (small forum version)


And here are some close-ups of the modifications:















And a strage mod (filter?) at the modulator



Well, and at the connector there is only some tape, cause the first thing i read about the Oric was to protect the connector from an accidently touch with the power supply



You can find some full resolution-photos of the board here:
http://www.bloss-nicht.de/Fotos/Oric/DSC00748.JPG
http://www.bloss-nicht.de/Fotos/Oric/DSC00749.JPG
http://www.bloss-nicht.de/Fotos/Oric/DSC00750.JPG

Well, i'll measure the 5V-line this weekend when i connect the computer again.

[Dbug]

The modulator modification is probably to convert the signal to British format to something that a German tv would accept.
I know that my British atmos, when used on channel UHF 36 would have kind of inverted intensities, and there was a small mod to apply, involving few components, to make it appear ok on my tv.

Perhaps it's the same thing.

[iss]

IMO there is nothing modified on your modulator. This is just tantalum capacitor for filtering power lines - it's connected between metal case(GND) and +5V.
Although what Dbug said about channel 36 is very interesting.
I have an Atmos which I bought from UK and on its modulator label is printed: UM1233 E36. Is this E36 just coincidence with channel #36?

I have an Oric-1 48K which I bought from France labeled only: UM1233 but it has the 1 transistor inverter - exactly what Dbug said:

Else, on the bottom side are 2 modifications:
a. One normally unused element from the buffer IC22 (74LS365) is used as "amplifier" for F2 signal.
b. F2 signal is used in the logic for enabling the ROM chips.

I think both modification are not related to the reset issue, but who knows... long wires can be source of noises etc.

[TomPhobos]

Ah, thanks for the explanation
Well, so it seems that the mods on the modulator and the board are ok. Today i'll replace the 7905, but i don't think that this will solve the problems. Maybe i should replace the cables on the board mod, maybe they are really too long. Or i could connect another cable from a GND-signal of the mod to the GND of the modulator to have another connected GND.

[Godzil]

The transistor is just to amplify the signal to the modulator, it can't change anything, just that the level may be too low for the modulator. And for the tantalum capacitor, if you look closely, it's C29

Thanks Tom for your pictures, this are not just small changes o_O. I will take the time to understand what they've done this is kinda strange.. and it's not only about Phi2.
Are there any cut trace on the component side?

Ho and the resistor/capacitor mod on the bottom left side is part of the official reworks.

iss: Yes E36, canal 36 is not a coincidence

[iss]

The transistor is NOT just to amplify the signal - it inverts it too and this is for purpose.
You can take a look here - some devices with CVBS inputs really require the signal to be inverted, else the picture is as Dbug said: with inverted intensities.

Else, no doubt the capacitor is C29 - this is for boards issue 3.
Obviously they found that soldering it to the box gives better results, and than in issue 4 the pads for C29 are removed.

[TomPhobos]

Yesterday i replaced the 7905 with a new one. As expected this didn't solved the problem. The Atmos still resets after a few seconds.
Hmm ... actually i'm getting out of ideas what i can do next.

@Godzil: At the first sight i don't see any cut traces on the component side. I'll check it out when i open the Atmos next time (next weekend, i work in another city ). Then i'll also dump the ROMs.

[Dbug]

Just a stupid question: Is the reset button propertly working

[TomPhobos]

Hehe, i had the same thought yesterday Well, the reset button works, when i press it the Atmos resets. When i slightly touch it nothing happens, so it should be ok. But i also think about to temporarily disabling it

[Chema]

Another silly question from a non-hardware guy, have you checked the tracks for continuity? Something like that could be explained if the power or GND lines are broken somewhere... You could check continuity and voltage on the +5 and GND pins of the main chips...