4-AXIS 8-Button USB-Adapter Schematics

[Guest MikePowell]

The max numbers of interest in the powered down state are -0.3 < Vi < +0.3 (because Vdd = 0) and -20 ma < Iclamp < 20 ma . The voltage limits are pretty standard for silicon chips. It's low enough to prevent internal channel isolation layers from forward biasing and conducting. The current limit is more device specific.With an active input signal going to a cold input pin, a current limiting resistor is a conservative approach. 1K should do. A resistor alone will not limit the voltage swing on the pin. The chip's input clamping diodes will conduct and limit the voltage, but perhaps not enough to rigorously meet the max specs. (Although, it's probably good enough for casual designs and use.)To be very conservative, external clamping diodes should be used. Schottky power diodes designed for switching power supplies have very low forward voltage drops and work well in this applciation. In particular, the 1N5817 has only about .25 volt drop at 20 ma. If there are caps that could possibly push a negative voltage to the PIC input during power up or down sequences, a clamp to ground will handle it. Otherwise, just a clamp to the PIC's Vdd pin should do it.Don't know if this is over kill. I haven't managed to toast a PIC yet, and I've done at least one or two fairly stupid things to them. On the other hand...Mikewww.mikesflightdeck.com

[Guest Ramsberg]

Ok, i'll put connectors to RA6-7, RC0-2 and RC6-7 then.Should I group them in any way, which are analogue inputs, which are digital? Are there any obvious groups?/ Olle

[Guest LeoL]

Group RA5 (Pin 4) with the rest of the analog inputs since its the only remaining analog input. RA4 (Pin 6) is not an analog input so don't include this pin with the analog group.You can group the rest anyway you like. I simply don't know how they might be used at this time.-Leo

[Guest LeoL]

Great info Mike. Thanks.-Leo

[Guest Ramsberg]

Making my first PCB of the USB device right now. It is compatible with bot LVDTs and pots on the same connector. No ICSP as it is probably not needed.I'll write back when its done and tested, probably this weekend. If it works, I'll post the PCB layouts.Cheers!/ Olle

[Guest MikePowell]

Great work, Olle!Mikewww.mikesflightdeck.com

[Guest LeoL]

Cool! Be sure to post pics.I experimented on my LVDT but I've had no luck so far. I simply can't be sure the osc works since I have no Oscilloscope to verify what's wrong.I've posted the schematics for a single voltage source osc and LVDT signal conditioner. You'll need to calc the R & C for your osc based on your coils resonance freq for max output signal.Some tips on homemade LVDTs:1) Use a primary coil with fewer turns than the secondary coils. 2) Use some kind of freq generator and find the resonance freq of your set up. Really helps boost that output signal. I used a simple 555 with variable pot and a voltmeter to read the output voltage. I then calculated the freq from the position of the pot.3) I used coils ripped out of cheap relays for less that a $1 each. Could save time winding those coils by hand if you

[Guest MikePowell]

Using those relay coils is a neat idea, Leo. You should be able to measure the AC voltages around your LVDT with a DVM, though I agree, an oscilloscope is very nice to have. I wonder if the forward voltage drop on the two detector diodes is eating all your signal.What TI USB chip are you working with now, and what development tools have you found?Mikewww.mikesflightdeck.com

[Guest LeoL]

Yeah, the coils are great and were easy to use. I was getting a rough output with the 555 timer as an osc, but the "proper" osc doesn't seem to oscillate? I seem to get a very low AC output, but I suspect its too low to drive the primary and probably need to connect a power transistor to feed the coil...something I used to do back in my Tesla days with TV flyback transformers. With only a 9V battery I was (inadvertently) interfering with local TV and radio signals. :-batAnyway, it would be nice to see the sine wave its supposed to produce on an oscilloscope just to make sure.FYI - Follow this link to the TI site. USB Keyboard dev kit...Everyone PLEASE stop "hacking" those poor keyboards and build this!:http://focus.ti.com/docs/tool/toolfolder.j...=TUSB3210KBDPDKUSB general peripheral dev kit:http://focus.ti.com/docs/tool/toolfolder.j...=TUSB3210GENPDKYou can order samples and download all the gerber and schematics for their dev kit to build your own. You'll need an 8052 (or 8051?) compiler for it and its Flash based so no OTP/UV headaches like PICs and Cypress.-Leo

[Guest MikePowell]

Slick! Thanks.Mikewww.mikesflightdeck.com

[Guest LeoL]

The kit sells for $200 USD from TI so I'm regenerating the gerber files to use non SMC components and will be ordering some PCBs.In case you're interested, let me know (or anyone else) and we can all split the cost on the PCBs.-Leo

[Guest MikePowell]

I'm interested, Leo. Mikewww.mikesflightdeck.com

[Guest Ramsberg]

Ok, the device is built! I had to make a JDM programmer also, to be able to program the 745. It programmed and verified OK.I plugged it on the USB PCB and plugged it into the computer. No explosions, no fires and no burning smell. The PC (Windows 2000) recognizes a new USB device, but it ends up unknown. Aarrgg! :-(Read your troubleshooting guide Leo, and added a ground wire to the shield of the USB connector. (Red in the picture). Here are the schematics, layouts and 2 poor quality Mobile phone photos of my work.http://forums.avsim.net/user_files/44372.jpgSchematics... Made from Leos schematicshttp://forums.avsim.net/user_files/44373.jpghttp://forums.avsim.net/user_files/44374.jpghttp://forums.avsim.net/user_files/44376.jpgHer you see the added ground connector (RED) to shield the USB cable. No bad solderings found.Any idea what to check next? I will recheck the schematics and verify every wire and connection on the PCB... But then, im out of ideas....Suggestions?/ Olle

[Guest LeoL]

Excellent work Olle.Ok, I suspect 3 possible causes. 1) Your ceramic resonator is not the correct type. I tried 3 different types before finding one that works.2) An incorrect osc setting in IC-Prog. Should come up as

[Guest Ramsberg]

Thanks!I just reconstructed the circuit on a breadboard to verify if it was the PCB. Im proud to say it wasn't, but sad because it didn't work either.... :-/Not sure about the OSC setting, wil check that tonight and do a reprogram. I also noticed 2 other settings which i stood clueless before: WDT (watchdogtimer) and PWRT (power up timer). What do i do with those?Cheers!/ Olle