To shield, or not to shield?

[+mytek]

I took Doc's advice and did a search on AtariAge with the search term "MyIDE RF Shields" and immediately got a few hits. I think the first 2 were to this thread, but there were 2 others to the topic of replacement cases which I had made a few posts in as well. And just like Doc had mentioned there were a few statements being made about possible issues with the various versions of MyIDE, and even some references to early SIDE problems, all which seemed to be associated with a need to leave in at least the bottom shield to obtain stability.

 

@flashjazzcat made this post in that topic.

 

 

My brain is likely suffering from faulty refresh since this post I made in that same topic was only a year ago, but I had no recollection of doing so prior to seeing it again 🤪

 

 

So it would seem that a few aftermarket devices did indeed seem to require at least the bottom shield to remain in place for reliable operation. Unfortunately it was never determined as to why, and in which specific cases this was actually required (e.g., what Atari 8-bit Computer series). So because the reason for such never got nailed down, we're left with something of a mystery that might only affect a fringe number of products, and no idea of where or when that might happen. The safe bet would appear to be to always leave in the bottom shield, whereas the top shield can be left out if so desired where RF radiation is not a concern. As mentioned earlier, when wishing to use the RF modulator all shields should be retained, especially if using the original TV switchbox which has exposed leads that can pick up that radiated RF (might not be an issue when using an RCA-to-Coaxial Cable TV adapter).

 

Jon also made the case, that he no longer needed to leave the RF shields in place with the more modern line of upgrades he's encountered and/or installed.

 

Now the only challenge remaining is, can I retain this rediscovered information? I just replaced my brain's leaking backup battery so I should be good for the next 10 years or so

[+kheller2]

As mentioned previously in this thread, XLs should have a rubber support for keyboard types 1 though 3.  Weather  or not that happened or was only for a certain version of those is unknown.  And yes many NTSC XL had that rubber keyboard support to shield. 
 

I am curious about spraying the plastic with a copper shield substance. Wasn’t that done on Apples?

[_The Doctor__]

Spray on, paint on and mesh shielding materials can still be had and applied today. Most upgrades made in 2023 are low enough profile and have layered boards with large power and ground plains that can act as mass and shields not only for themselves but underlying PCB's.

I've used stand offs to raise shields, bent or bulged shields and have gone the tin snips route for some things. If the upgrade is in and of itself a shield some have cut the hole and left the flap on or placed the cut out pieces under the bottom shield so they could do what they want with it later.

I also recollect some people using the backing metal on some keyboards as a shield by running a wire to it and in fact Atari did so on some models with a wire to the cartridge doors and keyboard backing plate as well. the bottom shield with tabs bent locking in the PCB affords shielding across the bottom and some structural support to the PCB bracing the outer edges fully and on some models the shield has internal stand offs punched into the shield so their added support will be in place for the large PCB.

 

The keyboard itself still needs support. for those not making full contact or lacking/having been omitted when they came in to ERC ACS, AFARC or ARSC- The foam or rubber supports would get installed. The keyboards showing too great of a degree of deflection would get replaced. The reason was to prevent a return for failed keyboards in the near future as traces would wear thin or stress fractures from fatigue at the pads and switches depending on model.

 

In modern times the XE's really do see a greater failure rate on keyboards etc, I'd suggest once they have their replacement mylar to ensure the keyboard is supported by not only the case standoffs but the shield tangs that perform a dual role of support and static protection. Some XE's had the same rubber feet you see on the bottom on the case/disk drives etc. on the shield in various points for support and have had a mention when noticed from time to time. I was asked about that at some point on one of the forums or online meetings. It was interesting in that some one thought they were spare replacements for when a foot was lost etc like the extra button you get with a suit/jacket or some other article of clothing these days. If you see tale tell circle and the keyboard moves a bit etc. they might have been peeled off or fell off etc.

 

On the bottom half of shields that had two of those in them, a person might consider trimming any legs that might touch them or ensure they are straight and will go into the already pierced holes. It would be best if they were placed to only contact the mask but that isn't always the situation.

 

Just some more random memories that surfaced.

Edited February 1 by _The Doctor__

[+remowilliams]

XLs ever having anything from Atari on the shield to support a keyboard is interesting.  I've seen the inside of quite a few XLs and have never seen that.

[flashjazzcat]

The rubber blocks have commonly fallen off, but I've seen many which haven't.

[Beeblebrox]

52 minutes ago, remowilliams said:

XLs ever having anything from Atari on the shield to support a keyboard is interesting.  I've seen the inside of quite a few XLs and have never seen that.

It might be a PAL machine thing? Nearly all 800XLs and 600XL's I come across have one or more of these rectangular blocks. I have quite a pile of them now as I said in my earlier post. Perhaps I'll list em on Ebay for £50 each (geniune Atari rubber block....)  Kidding!!!!

Edited February 1 by Beeblebrox

[JagChris]

The shielding probably adds some sturdiness to the structure. I wish now I hadn't let Sal discard my 600XL shielding.

[woj]

I keep the bottom shield on my upgraded 130XE for extra stiffness and weight. I used to on my 800XL too, but with recent upgrades "on the front line" the keyboard clearance became an issue so I got rid of it. Causes some discomfort to be honest, when I for example swap around the GTIA or ANTIC chips, the board really bends and I worry it's gonna snap one day. I also got the impression that the SIDE3 stability got worse when I removed it, but don't quote me on this. For sure it got worse when I converted the 800XL from PAL to NTSC (but it might be as well that the NOS ANTIC NTSC chip I got is dodgy...).

[Marius]

@mytek

 

The problems I did run into years ago went in some cases indeed away thanks to the shield. But not because of the shielding-function but simply because the shield added a huge load of mass which (indirectly) helped just enough for having no problems. That is my theory, perhaps it is BS. Those were the times where things were extremely critical. I also remember that having a non working CF card in my Myide 1 interface, that suddenly perfectly worked when I used a CF card to IDE adapter with 2 slots where the second CF slot was used by another CF card. So with the problematic CF card in slot 1 and no CF card in slot 2, the problematic CF card failed. But with another CF card in slot 2 suddenly the problematic CF card worked fine. Another of those "magical" solutions was putting the shield back. People with knowledge stated that these solutions are simply coincidental (or just luck).

 

I am not an expert based on knowledge, but I have a lot of experience with virtually countless Atari setups and weird solutions. I have seen enough of these magical and funny solutions working, but I indeed can not explain them. My best hypothesis is that some PSU's are not always very clean or stable and that somehow this might disturb in some situations the operation because of the PSU and its "signal" currently being slightly out of spec. Adding mass might help a slightly out of spec PSU just enough to let the disturbed process work again without issue. It is nothing more than just my hypothesis. 

[WOPRpwJOSHUA]

Interesting topic and read, all great comments - I really enjoy when a 'grey' area like this gets explored and experiences mingle with theory.

 

As another data point, my 800XL came w/o any rubber blocks nor signs they were ever there (was purchased as new in 1989 AFAIK, NTSC). Latest date (component IIRC) on mine is '95. Shielding on mine has no tabs, the 2 pieces sandwich the PCB, all fastened together, and that sandwich fastened to the bottom case. Type 4 KB fastened to top case. As a whole unit, it's pretty rigid, getting some from the case and some the guts - the KB alone contributes much to the top casing and the pcb sandwich to the bottom. W/that type 4 KB I'm personally not at all concerned about removing the shielding afa structure, though I left it on as I was able to fit the few mods I've done pretty snugly under it (so far), but even with only a (now-socketed) POKEY + AKI USB, just the (very tiny) wire connections to the panel keys prevent the top shield from seating completely fully to the PCB (thickness of a wire basically), so I can easily see why many leave these off when modding begins - there's just not any way around it. And perhaps those w/non-plate KB's, the shield rigidity might become a greater consideration.

 

Regardless, I suspect age/condition/handling/usage probably have a more outsized influence than has been discussed - these plastics, some sitting in hostile locations, becoming brittle or mushy depending, doesn't bode well for any future mistreatment or (typing, eg.) pressures...just a thought as this is commonly encountered in old AV (etc) equipment of similar era.

 

As well, on the signal side, there's (IMO) room for quirky behavior given age, condition, addition of new signal/EMF sources and our updated understanding of interactions at the working scale(s) of these components/interactions. Stuffing more (new) components under that steel (plated?) box may complicate those behaviors (the closer they are to that plane)?

 

Anyway, thanks for the contributions