Alex H

If the goal is to simply share I2C and a joystick, I would offer an easy approach using two NFETs and a resistor. Wire the I2C pins to the left and right joystick inputs. In-line with the joystick itself, wire something like this on the left and right joystick signals [the paddle pin and resistor can be shared for both] Joystick/I2C ----, | | +5 ----------. | | | R | | D Paddle ------+--G S | | `----- From joystick During most code, the joysticks would work exactly as they normally would since the paddle input would charge to +5 and stay there. When it's necessary to access the I2C, discharge the paddle inputs and the joystick will be momentarily disabled allowing unfettered I2C access. Note that it would probably be possible to switch the joystick's ground pin instead of switching the left/right inputs, but that could cause problems with I2C communication if the cable wasn't well shielded. The biggest problem I could see with this approach would be that emulators might think the game used paddles. 878573[/snapback] Ok, if I inderstand this correctly, that will disconnect the joystick when we need to access the I2C chip, but we also need to disconnect the I2C when we want to read the joystick. Otherwise joystick movements are going to send all kinds of crazy signals to the I2C. This will leave the I2C in an unknown state. The I2C will also sometimes respond to joystick movements, pulling the SDA line low which will jam the joystick. Another problem is that the memory card also needs to be compatible with the AtariVox which also has the SpeakJet chip connected to the joystick port. So to make it properly compatible, the same would have to be done for the SpeakJet I/O. Communication with the SpeakJet is frequent (every frame) so we need to be able to switch between devices very quickly - can't be messing with waiting for caps to charge/discharge. It might work, but now we're talking adding 8 transistors + resistors to the AtariVox, and it needs to be glitch free. It also means changing the code in games that are already in cart form and in peoples hands, and that's a BIG problem.

It needs to be plugged into the Vectrex when programming the cart - no power on a serial port. But yes, once programmed the cable can be disconnected. It also functions as a dev. cart so it's good that it works this way.

I meant the Vectrex cart.

I've just opened up one of John Dondzilas modded INTV cases and my VecFlash to see if this is possible. Well, it would fit but the VecFlash PCB is bigger than a normal cart PCB and it doesn't have the screw holes. The two posts would have to be cut away and there'd be nothing to hold the board in place. Also the VecFlash has the socket for the serial connection which would end up 1/2 inch or so inside the sloped end of the cart. I'm not saying it's impossible, just that I wouldn't be 100% confident. You do realise that Richard has to hack out some of the posts in the cart shell so it might not hold the original PCB in place any more? Personally, I would be fairly happy to use a bare VecFlash. My main concern would be causing damage by ESD while handling it. Also, I'm not sure how well it would hold good contact with the cart socket with the weight of the serial cable on the other end. Anyway, as I understand it, Mark Shaker has seen Sean Kelly and 1000 cart shells are now being made, so there should be a supply again soon.

Enjoy!

I've never gone as far as to tell the seller their mistake but have felt really bad before. A couple of years back I bought a "broken Vectrex for spares/repairs". I forget exactly how much it was but I was the only bidder - I remember the shipping was more than the bid amount. Anyhow, turns out it wasn't broken - the brightness was just turned down!

How many SRAM ICs? 74xx138 or 74xx139 would probably be what you're looking for.

Play fine for me in Media Player 10.

Here's one I like: ; unsigned divide by 3 sta temp lsr lsr clc adc temp ror lsr clc adc temp ror lsr clc adc temp ror lsr clc adc temp ror lsr

Ta very much! Just bought the latest issue - not a bad mag at all. I think I might have to subscribe.

Me too! @Mayhem: What issue was it in?

Whilst that is possible, it's not entirely straighforward. Working out a signalling scheme for the switching that doesn't interfere with or get masked by the SpeakJet and I2C signals is trickier than it might seem at first. Plus it adds complexity to the hardware (and software) which increases the development time and cost of the unit. I personally think Richard's decision to keep it as simple as possible was a good one. Besides, the AtariVox already exists and people already have them. Making those sort of alterations now would make it incompatible with existing units.

I reserved/set aside some of the low number carts for friends and for the competition I had planned. The whole reason I was going to run a competition in the first place was to promote it, but that turned out to be completely unnecessary. But I still kept hold of a couple thinking that I still might run a compo at some point. I didn't expect that it would be 2 years before I got round to it though.

Dun-Da-Daaaah! And we have a winner! The Protector Limited Edition up for grabs is... #06! Way to go 8bitclassics!!!! (PM will be sent shortly.) Thanks for taking part everyone!

No one has won it yet! Second chance time!

Second chance time poeple! Please see the update in post #1.

Alex, I'm way ahead of you on this - I have them made already, just waiting for the prototype boards to come back for testing. Cool! The problem is there isn't enough RAM to hold the scores whilst a game is being played - only 128 bytes in the 2600. In MGD the high scores are trashed by the game variables then reloaded when the game ends. I think this is how it would have to be done for most games. In fact, I think might even be reloading the scores every time the high score table is displayed! (They load in a single frame so it's barely noticeable.) The screen that shows the number of items collected needs something like 72 bytes just for display vars (not including the actual counters) so I think I had to do that, IIRC.

Me neither. 47 numbers eliminated so far. I think it might soon be time to start letting everyone have a second chance. Soon...

As Mayhem said, 25.00GBP. At the time that was just under US$50 incl. worldwide shipping. It's Defender, init? The nice box and overlay help I guess.

I would think that in many cases it would be possible just to go up one ROM size and have all the high-score stuff in a separate bank like you suggest. The only real change to the existing code would be a hook at the title/game over. To add speech though, there are 4 main hurdles: 1. The extra ROM space that's needed for the driver and speech strings. Again this could be got around fairly easily in many cases just by increasing the ROM size. 2. The CPU time required to communicate with the AtariVox. For this we are limited to games which don't have a whole heap going on outside the display kernel. 3. Code to trigger the various speech strings upon certain events would have to be strategically inserted into the existing program. This was actually pretty easy to do on the Vectrex version of Berzerk (I did the hack in one evening) but it could be tricky on the 2600 where things are tighter. 4. An extra 2-bytes of RAM are needed to hold the speech string pointer. Games which have a 2 player option are good contenders since we can use the memory where the 2nd player's score would normally be held. Anyway, I might have a try at hacking some games at some point - to add speech or just high score saves. As a side note, does anyone think it might be worth making Atari memory cards? Basically a speechless AtariVox? The EEPROM was tacked onto the AtariVox because it was so easy and cheap to do. A tiny circuit board with just a DB9 connector and I2C EEPROM is all that's required. I'm guessing they could be sold for something in the region of $10-15. As far as the driver is concerned, they are two separate devices so any AtariVox enabled software should be compatible. Likewise, any game written/updated to use a memory card would also work with the AtariVox.

I thought Tremelo is cyclic variation in amplitude. Vibrato is cyclic variation in pitch. No? Cool sounds there Richard. I'm sure I can hear some ring-modulation in that (multiplication of 2 waveforms). Didn't know the SpeakJet did that! I really like the noise at the end - great for thruster sounds.

01? As if! 14? My birthday - no. 42? Answer to Life, the Universe and Everything? Nope. 64? As in herbs64? Too obvious. 69? From the guy who's writing Man Goes Down, you might think so, but no. 97? Now what was the number of that cart that just sold on eBay? 99? Logical, but no. Doesn't look like anyone has got it yet. There's no special significance to the number.

Ok, well I've downloaded the binary and had a good play to check that it isn't corrupt or anything - it's fine, working as it should. Fireballs don't destroy other fireballs. Could that be it? The only other thing I can think of that it might possibly be is that you have downloaded a bad BIOS. Seems unlikely that it would screw up the game in this way, but maybe. (If you want to PM me your email address I'll send you the BIOS I'm using.)

Had them manufactured by these people: www.universalscreenprinters.com Not cheap! The most one has sold for on eBay is that recent one which fetched $300, isn't it? You could probably get $500 or more for your cartridge No. 01 though! Well, I'll just have to hack some different graphics into 3D Minestorm and try and sell it as a one-of-a-kind to that guy who's been bidding OTT for the 3D Imager stuff recently!