teaguecl wrote:
If not, you could short GND to VCC, and your cart and/or your console will release it's magical smoke.
If you look the Gnd and Vcc pins are actually on completely separate sides of the board so that's not possible. But the bus conflicts of PRG A11 tied to ground wouldn't be nice either.
teaguecl wrote:
So, by sliding your cart back and forth you can shift around where the contention happens - but you cannot avoid it.
I've done it with ease.
teaguecl wrote:
This will work for some carts,others it won't. It certainly won't work for anything that uses the expansion pins.
The issue is NOT in the center expansion pins. They are near perfectly aligned. The issue only arises with the pins on the far edge. When the cart sides back and forth.
Memblers wrote:
I'm making an NES cartridge that does use the expansion pins, and I will be a little peeved if that stops it from working in clone systems. It's hard to believe, seems like there must be some kind of consequence. Maybe I can make the pin thinner or in a different position or something, would be kind of a lame solution though.
If it doesn't work in clones due to a pitch issue the expansion pins will have nothing to do with it, the killer will be the edge pins. If the clone's sockets have more free play than mine and thier 2.50mm pitch you may have issues. One thing that may help is to make the traces slightly thinner if you're able.
teaguecl wrote:
Only having 0.19" (0.48mm) of play is good.
No I said 0.019" but I think that was a typo because it is 0.48mm.
teaguecl wrote:
You still seem to be claiming that your connector doesn't have any contentions. That I do not believe, the math proves it to be impossible, regardless of what your multimeter says
. I certainly trust the math more than your experimentation. I don't say this to be rude, it's just that math doesn't lie and therefore I cannot trust your experiment.
While I can believe that you may not trust my experiment, that's a pretty bold thing to say that math proves what I've done IMPOSSIBLE. Perhaps it's just the engineer in me speaking but if you go through life solely trusting math and not physical occurrences you're in for a rude awakening...
The main problem with relying completely on math is it's easy to make erroneous assumptions that throw all your math down the drain.
Your statement is especially surprising to me when you have made ZERO measurements yourself (or at least not very accurately). I'm really not sure where all your numbers are coming from. Those 1.651 and 1.618mm measurements correspond to nothing physical with NES carts or these connectors. I measure cart contacts as around 1.9mm wide, it's not the easiest thing to measure but I'm not off by .4mm, also my pins on the connector are 1.37mm wide not your 1.618mm
teaguecl wrote:
However, I readily admit that it is possible that my math is flawed. My diagram shows two pcb edges, it does NOT show the connector itself. The metal on the connector is probably a little narrower than the trace on the PCB, which might impact the calculation.
You're right about this...
Your picture with pins 1 perfectly aligned and pins 36 and 35 halfway overlapping is PHYSICALLY impossible with this connector and standard NES carts. With only 0.019" (0.48mm) of free play means that the cart can only fall .24mm off center.
You're drawing shows the cart having 2.88mm of free play
72*(2.54-2.50mm)
You're allowing the cart to fall 1.44mm off center which is a "mathematical" error of 500%
You can draw everything much math as you want but really the measurement comes so close that you can't say without doubt what will happen. Mainly because having the slack, trace, or pin widths off by a 0.01mm makes or breaks the final answer. Especially since the pins have some curvature to them so they aren't making physical contact on their extreme edges.
Now I will point out some assumptions that I have made that may cause issues.
*Width off cart traces and connector pin width. If these were to be different for things like the powerpak, unofficial carts, or other female connectors you may very well have issues.
*Cart wear. I did some more testing with other carts. Turns out I was actually able to ground out PRG A11 with one of my carts. It's a heavily worn SMB3. It took me awhile to realize why that cart had issues the others didn't. The reason was because of wear. The cart actually had enough wear to slide far enough to cause that issue. But I placed the board inside the case and inserted it like it normall would be into the NES. The only way I was able to get the pins to short was prying (and holding) it over with a screw driver. The lessened clearance caused by using the case forced the cart back over when I released my screw driver. I had to be actively applying the force to cause problems.
I could see if a cart were in REALLY bad shape you could have more issues.
* proper alignment on the NES motherboard. Like I've stated before there is more slack with this connection and it's easier to cause problems. With my homemade 72 pin connector properly aligned during install has no issues.
In conclusion, I agree these connectors aren't perfect. However in my opinion they're a pretty damn good solution considering what other options are out there. Especially when doing a project like a cart dumper or programmer when you'd rather not sacrifice a toploader to get a easy to manage female cart connector. Ben Heck's solution is even better if you're concerned, and you can't argue, Ben's a genius.