Geneve seating in PEB, is there a fix?

[jedimatt42]

Ok, so my Geneve is virtually unusable for so many reasons. But one that seems like it should be solvable is the seating in the PEB.

 

It rocks in the slot.

 

If it is down firmly in the rear of the slot, it will not boot.

 

If it is down firmly in the front, it will not boot.

 

If it has been subject to gravity for a week it will not boot.

 

I have heard others acknowledge the poor fit.

 

Do you guys have a common practice shimming or filing or something to correct this? Mine has no case.

 

Come to think of it, I didn't have the issue with the enclosed Geneve loaned to me for some other testing... Is that the issue? It needs to be supported by a case and how that fits into the slot bay?

 

Looking for advice.

 

[email protected]

[jens-eike]

My Geneve has heat sinks on the regulators, hence it cannot use the case - but the back part of the case helps with the correct orientation and seating. When you close the PEBox's lid, the foam will press the Geneve into its slot.

 

As you can see, I've done it with the Geneve and HFDC. The box is upright (drives on top).

[+Ksarul]

Jens-Eike's solution is probably the best one for open-case Geneve stability.

[+9640News]

When I had the Rave PEBox, I used a layer of packing foam that used the weight of the closure to push down on the card to keep cards from moving.  I did not use bubble wrap, and it was not Styrofoam.  Rather, that flexible packing.

 

Beery

[jedimatt42]

So keeping it from moving is a bit different from getting it to seat correctly.  I should back up... getting mine to seat correctly is an act of patience, that requires repeating. 

 

People have commonly cried that the Geneve and other Myarc cards tend to rock in the peb.  Sounds like filling the space properly with a shell orients the card would solve that problem.  Sounds like no-one has ever bothered to root cause this?  All the other bare boards I've experienced seem to fit square in the edge connector. 

 

I've got a hyperbolic expression's worth of heat sinks and extra 7805s on mine as well, and the genmod board on the back. So a traditional shell, that I don't possess isn't an option. 

 

hmm... I've never noticed the minor impressions in the PEB lid foam padding shaped like TI expansion cards.  I'll try printing a couple card-edge-feet that this existing foam can apply pressure against.

 

[email protected]

[+InsaneMultitasker]

All of my repair/testing is done with the cards out of their clamshell/case.  You should be able to place the card in the PEB then press downward on the top edge of the card - both front and back at the same time  - to level it in the slot.  If you later tug the keyboard, monitor, mouse, or keyboard cables attached to a case-less Geneve, then yes, it can rock the card even with the PEB cover on and locked down.   The case inhibits the rocking.  Many Myarc cards came with a case; after-market changes and heat-related  problems caused people to get rid of their cases in favor of other options, which often led to more repair work.

 

Check the underside of the PEB cover and you should see contact foam or similar. A cheap albeit short-term method to securing the card is to add another layer of foam, 1/2" or so wide,  to the cover of the PEB where the card will hit it. Don't use material that will flake away or you'll have a mess on your hands.  A better solution would be a plastic molded guide or two clips that could be affixed to the top of the card (Geneve, HFDC, and others).  It would add the height needed to meet the foam, and the top of the guide/clip could be that 1/2" or so wide to offer stability. 

 

Anyway, based on experience and little else, the problem with your card sounds like it could be a Genmod board issue, a socket contact issue or a cold solder joint, provided your Geneve edge connector is not warped or malformed.  I would also consider the HCT04 and LS245 (possibly 244s) as common logic culprits if the only issue is that your Geneve can't see the controller card, though in these cases the failure is usually not self-healing.

 

If you feel comfortable doing so you can perform rudimentary tap-testing of the Genmod and components to check for a bad connection. 

 

Take all other cards out of the slot then place the Geneve in slot 4 or 5.  Be sure the card is inserted as squarely as possible. Turn on the PEB.   Use one hand to press down on the card's top edge roughly at he midpoint; with the other hand, reach toward the back of the Genmod daughterboard.  Gently use a finger to tap the board and its components.  You may also want to tap the wires leading away from the board.  Sometimes flexing the board from mid-point to front (or back) can show a failure point. 

 

I like to test with MDOS loaded, when possible. For this I usually use a floppy controller in slot 8.  I'll run a simple batch file from the internal ramdisk that does some things like copy files and type a text file, that way I can know which tap locks the system.  In a pinch, running a directory in graphics mode 6 (MODE 60) is often telling for both VRAM and OS stability. 

 

Ksarul's PEB extender comes in handy for checking cards though I caution against trying to hold and tap-test unless you have a very steady hand.  Rocking a card out of the extender is much easier (and faster) than can be done in a PEB.

 

A wood pencil with eraser is often gentle enough to tap components on the front side.

 

I have no idea if any of this will actually help your situation.  I do know it is not normal for gravity and side glances to stop a Geneve from booting with one notable exception: you are using your system during the official hours of the Chicago TI Faire.

 

(this weekend I have been working on someone's Geneve with a problem similar to yours.  It booted just fine, worked for hours, then the next time around I got a black screen and no swan.  Some tapping and gentle flexing has proved there is a problem most likely with a bad solder joint or bad  single-wipe socket.  I have not been able to isolate it and quite honestly, do not know that I will be able to do so if it is an inner layer trace issue)