forum.defence-force.org

Hi,

When testing Novalight a while ago, I noticed that about 8 machinnes required the sound volume at maximum to load the programs, but 2 other machines failed if I did so... But it worked if I put the volume at a very low level for them!

I never understood why, and Dbug suggested that I took some photos of those machines. One of them being still unopened, I opened the other and here are the pictures.
Just in case some hardware wizard could find a reason why the volume has to be very low to load tapes on this machine.

No surprise that loading was complicated if settings did depend from the tape, but also from the Oric model

One possibility -- check the Official "Oric Service Manual" for the various "mods" (resistors, capacitors) that need strapping on/moving around, some of which are to do with tape loading reliability. Maybe the difference is with/without these mods?

From what I see on pics and from my memories (because I'm sure I have an Oric with this mod) here is the "secret":
Both diodes limit the input signal to about +/- 0.7V - this was made as kind of protection for the LM358's input but it changes/distorts heavily the signal too. In the far past I made such limiter but using zener diodes and opamp (recently I saw very similar solution on forums.oric.org - Convertisseur de signal Oric). Here I don't think the diodes are zeners.

So practically, removing this mod (i.e. desoldering both flying diodes) will make the Oric to work like all other normally loading Orics.

BTW, recently I made some measurements about the highest frequency which can pass through unmodified Oric input, and how different mods affect it. I'll post the results but I want to repeat them following some "scientific" order and trying to be more precise. It's interesting!

Thanks for the replies and explanations!
I would be unable to make the link between a drawing and actual components on the borad, so I'll avoid doing anything

Looking forward to see the results of your tests, ISS!
For the moment I gave up updating Novalight - I still have a mystery to solve with the latest version I wrote.

This are the added diodes:
using small sharp cutting pliers you can cut their wrapped together leads (as on picture) and test.
It's easy later to solder the place again if needed.

Symoon wrote: ↑Sat Dec 28, 2019 9:33 pm For the moment I gave up updating Novalight - I still have a mystery to solve with the latest version I wrote.

The same by me
I'm very close to give-up with Jasmin reading - so far no problem with track 0 but after seek to any other track byte start to drop by reading individual sectors. Probably the guys who wrote FT-DOS had somthing in mind to select IRQ to read bytes instead of direct polling... mystery and magic....

iss wrote: ↑Sun Dec 29, 2019 12:11 am This are the added diodes:

Wow thanks, would never have spotted this.
I think I'll keep it as a test machine

iss wrote: ↑Sun Dec 29, 2019 12:11 am I'm very close to give-up with Jasmin reading - so far no problem with track 0 but after seek to any other track byte start to drop by reading individual sectors. Probably the guys who wrote FT-DOS had somthing in mind to select IRQ to read bytes instead of direct polling... mystery and magic....

Could be worth digging magazines for articles about this. I really don't know FT-DOS - BTW there were several versions of it!
Maybe something with disk geometry, or single / double-sided disks?

Regarding mystery, magic and polling... I got something similar with the Oric speech synthesizer: In theory you can poll the status register to know when the phoneme is done being played, in practice on my Oric the value is correct only 75% of the time :-/

Is that a French Atmos? No modulator!

The ROM is also the later one - not sure if that makes any difference?

I have no idea where it comes from in the 1st place, but I'd say a French one, yes. No modulator as you said, and self-powered SCART through the DIN plug.
Still, it could have been modified by a hardware magician, too.

Anyone aware of specific details that could answer this question for sure?