smbaker

Just wanted to let folks know that I do currently have plenty of bare pcboards on hand. You can find them from me on eBay, or you can order directly from here on the forum ($10 per bare pcboard plus shipping here on the forum). Yes I am talking to DrBenkman about supplying him with a quantity of boards that he could use to fabricate completed controllers for people. This would be a great service to the community as I know there is a demand. Scott

I have plenty of blank pcboards on hand now, so anyone needing a blank pcboard, just send me a PM and we can make arrangements. My usual price is $10/board plus shipping. As a reminder, these are blank boards only, you still need to source all the components yourself. Scott

Yeah, it came highly recommended to me by several people. So far I'm pretty happy. Posted a new video today, showing off the 3D case! Scott

One thing to note is that the fire buttons and the keypad are connected directly to the 5200. The only thing that uses the microcontroller is the analog stick portion of the controller. So anything in the thumbstick, microcontroller, or digital pot could be causing the issue. One other thing to check is that you programmed the fuses into the microcontroller. One of the fuses sets the clock speed multiplier and perhaps too slow a clock (I have not tried it) could cause an issue. I ended up getting a Prusa I3 MK3. So far it's been a pretty good experience, I've printed several things. I might try the case in PLA without supports. In PETG there was quite a bit of stringing, and some deformed layers when I tried it without support material.

@chicgamer: Also, check and make sure that you have installed the jumper at JP2, as this is the chip select for the digital pot. Somewhere around here I did make a test firmware that cycles the pots back and forth, ignoring the thumbsticks entirely, and that could be one way to diagnose whether your issue resides with the digital pots, or with the thumbstick. If you're interested, I can dig that up. @animan99: I just got a 3D printer and printed your case design yesterday, in PETG. A little bit of an issue with nastiness around the supports on the beveled edge (I'm trying a PLA print to see if it looks nicer), but nothing some sanding couldn't clear up. Great design, and thanks for sharing it! Scott

I don't have any good suggestions left at this point. I did just try my prototype and verified that removing the IC gives me an open circuit indication on the 5200. I tried putting a 10K resistor in parallel with the 10M, and was able to bring it down to a 5 reading, which is consistent with what you're seeing. Looking at the schematic, I don't see an obvious place where a bridged solder joint would be responsible (and it would be surprising for the same problem to occur on both axis). Can you post a picture of the board? Maybe another set of eyes will help and I can spot something. Scott

Rayik, I'm still concerned that you're getting a low value when the IC is not installed. Can you triple-check that the 10M resistor is 10M (and not, for example, 10K which is a similar color code) and that the diode polarity is correct? Scott

That is interesting. Without the IC installed, the only thing in circuit on each axis should be the 10M resistor and the diode. It should read full right (I think that's what DrVenkman's did when he tried it with the IC removed), not full left. Scott

From my notes earlier, with the power off and the IC removed, using the two-pin pads adjacent to each pot, I measured 265K on the top one and 263K on the bottom one (the difference between 265K and 263K is probably not significant). So I would start by verifying your pots are in the ballpark of those values. If that still doesn't get performance within reasonable limits, can you use a traditional controller on port 1 of the 5200, and the masterplay clone on port 2, to navigate into Pete's test cartridge? I don't have an atarimax, so I'm not entirely sure how the menu system works. Scott

One of the things I was thinking about doing was to define a dead zone around the center of the stick, so that if the stick is moved around inside that dead zone, it doesn't change the pot value sent to the 5200. As soon as you start to move out of the dead zone, then it would start changing the pot value like normal. I think this can entirely be done in software. My concern is that the dead zone might not be alike for all PS2 controllers, or even alike for the two thumbsticks on a single controller. Scott

Great to hear you have it working, Rayik. It should be possible to adjust the code to tune the joysticks. Right now they're tuned to my 5200 and my PS2 sticks. There are some variances with the PS2 thumbsticks. I've noticed in one of my controllers, the right and left tumbsticks don't center up to the same value. Not as bad as what you're experiencing, and not leading to playability issues as far as I'm aware, but far less than perfect. I will try berzerk when I get a chance. Scott

Grr.... This thing seems to be causing more problems for folks than one would expect. Can you use Pete's Test Cartridge to verify the values when the joystick isn't pressed? Also, to verify, you have adjusted the two pots for centering?

Yeah, I have plenty on hand now, $10 each plus shipping. Scott

If you solder the connector on the wrong side of the board, you'll actually reverse the numbering of pins in each row of the connector. For example the rightmost pin on the 9-pin connector if flipped over becomes the leftmost pin. This almost certainly won't work right, and in the worst case could damage something. Scott

U3, U4, and U29 are all "test points". In the case of U3 and U4 these are to accommodate larger transistor packages. U29 is to break out an unused pin, which isn't really needed. I tend to break out interesting unused pins when I have room. U5 and U28 are tactile switch footprints in the center of each thumbstick footprint. I tended to add lots of extra tactile switch footprints, just in case someone wanted to put a trigger button in some weird place. Those bigger "momentary" switches are large size tactile switches that you can buy from sparkfun. They're like normal tactile switches, but bigger. Advantage of the sparkfun big tactile switches is that they come with plastic button tops already. Header SV2 is an ICSP programming header, for use with USBASP programmers and similar, using the 6-pin programming cable. It lets you program the microcontroller in circuit, without having to remove it. Scott

I've thought about getting a VCS adapter (I don't own one now) and installing a separate s-video/composite mod into it. Then use a conventional composite or s-video switch to switch between the two devices. But every time I go down that route, I ask myself why I'm pretending the VCS adapter is anything other than a 2600 sitting on top of a 5200, and why I don't just buy a separate 2600.. Scott

Bare pcboards are $10 each plus shipping. I would suggest lining up someone to build it before you order one. In addition to the circuit board, you (or your builder) would also have to buy all the components (ICs, resistors, sockets, etc). Scott

I received a board revision over the weekend, and had a chance to build one out. New board features a footprint that can be used to directly mount the robotshop ps2 connector.

Electrically it makes no difference, but mechanically I find it easier to terminate one wire per pin on a 0.100" header. My general policy is that having extra grounds usually doesn't hurt, it gives the builder options. Scott

I have a ground for each pin, so the pinout is: GND COMP GND CHROMA GND LUMA I figured if someone wants to run individual coaxial cables for chroma and luma, it's easier if the grounds terminate separately than having to tie them together. I haven't checked the overall height, and I never reinstalled my original RF shield. Unfortunately, I just don't have the time. Typically I make things for myself and sell off the spare pcboards. If I was to start manufacturing something, it would probably be my handheld controller, from a risk standpoint (much more could go wrong with people opening up their consoles). Scott

Multiple bad ICs from digikey would surprise me, especially considering that I ordered the same part number myself. That the same board seemed to work fine in another 5200 suggests to me that it's some as yet not understood compatibility issue between the 5200 you're currently working with and the circuit. I just can't imagine what it is. We checked obvious things like voltage. Absent an oscilloscope, it's hard to think of what else we could check. Scott

It should support both composite and s-video. The s-video pins are brought out to the header. I just have never tried the s-video output as I don't have an s-video capable TV anymore. Scott

Okay, that's as it should be. Still thinking about this, but as of yet no good ideas... Scott

No kits, but I do have some pcboards for a composite mod that sits between the GTIA and its socket. This was based on low_budget's design in one of the threads here. I believe Shawn has built one of my video mods here, and might have some experience to share with it. Scott

Well that's interesting, and shows the Pete's cartridge has provided some useful info. If you remove the DG413 IC, what happens? Scott