|
Post by Tom Maneiro on Feb 1, 2011 19:50:02 GMT -5
Well, call that V3 board (no cool codename yet) a Work in Progress. I'm using a different model of breadboard, which will bring much cleaner VCC/GND rails, less wiring madness... and for now, the parallel port plug will be non-removable I will do as you suggested: build the 'LS74 section, test, then wire the 'LS157 section. I haven't checked my stash yet, but so far, the only electronics shop in my town only has those Shitachi/Renesas HD74LS74AP on stock, so.. should i leave pins 1,4,10,13 floating, or ground them?
|
|
|
Post by Tom Maneiro on Feb 2, 2011 15:56:43 GMT -5
OK, testing in progress for the flip-flop section, after 5 minutes of Sonic 1: - Good syncs: A gazillon (well, around 23000 ) - Bad syncs: 94 - - FPS: Solid 60FPS. Tested switching to PAL, and it stays at solid 50FPS - No weird out-of-sync letters. So far, so good... Time to wire the 'LS157 section
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 2, 2011 17:03:32 GMT -5
OK, testing in progress for the flip-flop section, after 5 minutes of Sonic 1: - Good syncs: A gazillon (well, around 23000 ) - Bad syncs: 94 - - FPS: Solid 60FPS. Tested switching to PAL, and it stays at solid 50FPS - No weird out-of-sync letters. So far, so good... Time to wire the 'LS157 section Yes that's good news , that's perfect sync. What does it say for the four letters when you're in the test mode? (e.g. ABCD, DABC,...) Before you do the 74LS157, wire the UP/DOWN first. Verify that everything's still working in the test mode (60FPS, up/down, etc) When you're wiring up the 74LS157, solder VCC, and GND first. Then check that your 74LS74 is still working 60FPS. Then wire the parallel port connections second. Then check your 74LS74 is 60FPS again. Then wire the select signal from the Genesis. Then wire one wire by one, while checking you're still getting good syncs
|
|
|
Post by Tom Maneiro on Feb 2, 2011 18:46:46 GMT -5
Well, bad news: Either my DM74LS157N are FUBAR (i got 2), or something in my setup is also FUBAR, because as soon as i plug the chip, bad syncs happen This is getting ridiculous... Put the LS157 - madness! Remove it - perfect syncs! What the hell!? Anyway, i will do more debugging tomorrow... At least the flip-flop is not a flop
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 2, 2011 22:18:19 GMT -5
I told you, you should have assembled everything slowly, and keep testing the sync thing for every wire you attach I don't think it's your 74LS157, but it's a short. The only thing common between the two circuits is the Select signal from the joypad, and that's input only. Check again for shorts
|
|
|
Post by Tom Maneiro on Feb 3, 2011 15:39:59 GMT -5
Rechecked everything again... found no shorts so far... AGAIN! Even rechecked my cables and plugs... everything is fine. But, i've found something interesting: If i lift pin 1 on the 'LS157, bad syncs go away. But obviously this make the circuit non fully-working, as A/C don't work, only D-pad and START/B (tried the uploader: it successfully moves the cursor, but can't enter the GG codes). So, i think that something is glitching the SELECT line for no good reason... So at least i know where the fault is located. Call it... progress?
|
|
|
Post by Tiido on Feb 3, 2011 15:55:21 GMT -5
Drop the LS part and go for HCT or ACT, or just plain HC or AC Also, are you 100% sure you have counted pins on connectors and chips right ?
|
|
|
Post by Tom Maneiro on Feb 3, 2011 16:16:02 GMT -5
Yup, no errors counting. Continuity tests reveal no problems. Anyway, i will try with HC/AC parts... if i can find them over here I suspect that my Shitachi flip-flop is causing more harm than good (BTW: Can i mix HC and LS parts? I can't remember my Digital Circuits classes, but HCT is TTL-compatible, not sure about plain HC/AC...) EDIT: Replaced the LS157 with a LS257 - no changes at all So my multiplexers were not fried after all
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 3, 2011 17:56:42 GMT -5
I dunno what's happening. Try to insert a small value resistor (somewhere around 100 - 1k ohm) between select and pin 1 of LS157. I'm suspecting that your Select line from the Genesis is very very noisy, and it's causing the flip flop to prematurely flip. I'd recommend building a low-pass filter (to only allow signals < 1MHz on sync) from the Select line to the flip flop. I'd only recommend you build one (it only takes one capacitor and one resistor) only if you know what you're doing . 1 MHz time constant is like what? 1K resistor, and .1uF capacitor? You can mix the two logic families together in this case
|
|
|
Post by Tom Maneiro on Feb 3, 2011 18:52:12 GMT -5
K, will try that tomorrow. I already know that my house is not EMI-safe So i only need one 1K resistor and one .1uF capacitor, eh? I already have both, but i'm too lazy to solder right now... Should i wire the filter like this one? en.wikipedia.org/wiki/File:RC_Divider.svg
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 3, 2011 19:54:11 GMT -5
I wouldn't try it yet . Wait for me to try it out myself (which might be next week) But if you want to experiment sooner, then feel free to do so
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 5, 2011 15:58:21 GMT -5
Also, I'd be careful of latch-up mode too (latch-up is a bad thing) I think I already covered this in the FAQ for non-working circuits. But try disconnecting the entire all the connectors (from the Genesis, and from the parallel port), and connect everything back together. What I would do is plug everything together, run the test utility, and power on the Genesis with Sonic 1 running. If you get bad syncs, then disconnect the connector from the Genesis, and plug it back in (while the Genesis is still powered on). Keep doing this until bad syncs go away I'm bringing this up right now because I had the same problem this morning , and I did some hardcore debugging, but overlooked the problem of the circuit latching up. The low-pass filter won't correct it either. But yeah, definitely a good idea to read my own tutorial Using AC, ACT series CMOS parts will prevent the IOs form latch-up (well, the newer SCR-latch-up-resistant parts from TI have this in their specifications) There are two ways the circuit will latch-up, and their symptoms. 1) Either the first stage, or second stage flip flop will not be able to change Q or /Q during latch-up, which will cause many bad syncs. 2) The joypad buttons (except Up/Down) will appear to be stuck, this is because LS157 is latched-up, which I've seen with the START button stuck low.
|
|
|
Post by Tom Maneiro on Feb 5, 2011 22:32:47 GMT -5
I've already tried the hotplugging thing, with no positive results, but haven't tried the "plug-unplug-plug" loop I don't experience stuck buttons with my current board, just gazillons of bad syncs. Something simply makes the flip-flop to go wild if all sync lines are connected together. I've also tried the low-pass thingy, without changes... So far i only have a nice massive joypad for my Genesis Anyway, the next thing in my list is getting CMOS parts
|
|
oompa loompa
I AM THE GOVERNATOR
"Git 'Er Dun!"
Posts: 1,301
|
Post by oompa loompa on Feb 6, 2011 1:03:39 GMT -5
When you run the test utility, do you get totally random letters for the sync characters, e.g. ABAB, ABCA, etc? Consistently random? Like, how often do the letters change?
|
|
|
Post by Mairtrus on Feb 7, 2011 16:40:39 GMT -5
It's strange, but I have always the same problem when I plug for first time the cable (I unplug it from both PC and console if I will not use it for a long time), and after running the test mode, both good syncs and bad syncs increases really fast (the bad syncs like 10 times faster than the good syncs), and it says that the console is running at 452 fps, or like. But after unplug and plug it again (only from console, and without turning it off), it starts to works perfectly.
|
|