Vectrex32

I got my new vectrex32 today, plugged it into my vectrex, but it seems no matter what I do, it's just stuck on the text "waiting" (So I get vectrex bios startup, then "vectrex32 v1.18", then "waiting" - am I missing something dumb, or is there something else I need to do?

Thanks in advance!
--jj

That's not supposed to happen. Make sure the V32 is securely seated in the cartridge connector. When you get "WAITING", try pressing the Vectrex's reset button. Another thing to try is to connect the V32 to a PC via USB, bring up a terminal window, and see if you can get anything (hit Ctrl+C and see if some text appears).

If nothing works, you'll have to return it to me. I'll refund you for return shipping (up to $7) and either fix it or replace it.

I'm sorry for the trouble.

- Bob

Thanks for the quick reply, and no need to be sorry at all - bespoke hardware comes with bespoke hardware expectations :).  I'll try the PC later tonight (have to move the vectrex first!), and see what I can see - I did try the reset button, to no avail.  I have wondered if it could be related to my vectrex as well, since I did have some issues where a couple of games on the Richard H multicart wouldn't start (always the same games, interestingly, and nothing GCE official)

Confirming that the issue was on my end - I guess the PCB for the vectrex32 isn't quick as thick as my other carts, and so it's not _quite_ making contact - if I physically hold the cartridge up (i.e. push on the bottom of it), then everything works fine.  I guess I'll need to look at my cartridge port on my unit and maybe tighten that up.

Thanks for your assistance!

I'm having a similar problem. When I start the 3D demo it will often get stuck on 'waiting', other times it will not start, glitch or even reset to the Mine Storm game whlle using it.

The cartridge is very wobbly in the connector. Insertion seems a bit shallow a well. I think with a USB cable attached the issue would be worse, probably not even usable, seeing how much trouble it is giving already. I think this cart would definitely benefit from a full size housing for stability. I'm worried it will bend the contacts of the cartridge port the way it is now.

When it does work, by propping it up with something, it is pretty cool.

It seems like when you add more than ten tie fighters the 3D demo crashes BTW.

It should not be wobbling and should not require support for the weight. Is it possible the connector in the Vectrex is loose.

- Bob

Could be. Testing it some more, even with propping up the 3D demo is very unstable, while the games don't seem to be giving any trouble. Haven't had any problems with other cartridges.

EDIT: I wonder if it would be possible to bend the cartridge port contacts so the connection would be tighter.

I'm thinking more like whether the solder connections between the connector and the PCB are cracking.

- Bob

That would also be possible. And probably an easier fix.

Yeah, my solution ended up being a pile of books propping it up, which seemed to generally fix it, though is suboptimal - the vectrex32 is the only cart I've had this problem with, so it does seem to be something unique to it.  My _suspicion_, which is only verified by my eye and not real experimentation, is that the card edge on the v32 isn't _quite_ as long as standard, which is usually okay except when it isn't ;). 

I'm thinking about building some kind of cart-extender to fix the problem more permanently - ideally what I want to do, is something like http://store.go4retro.com/x-pander-3-slot-cartridge-port-expander/ but obviously for a 36 pin vectrex rather than a 44 pin c64 - this would also have the advantage of protecting the physical port, which is a property I also appreciate.  I believe it's the same pincount/pitch as the second half of a 16bit isa slot, so maybe i'll find one of those and cut it in half, but otherwise I might have to pull out the soldering iron and build something...

When the V32 first came out, someone else reported wobbling and thought the problem was that the edge connector was too short. I lengthened it in the next rev of the board and since then, no one has mentioned a problem to me.

jaymzjulian, who started this thread, had the theory that the V32's board is on the thin side. It's supposed to be 1.6mm, which should be fine. I haven't measured it though - I just assumed it was manufactured to the spec I ordered.

I've never seen problems like this in my own Vectrex. Every V32 I send out gets plugged into my own Vectrex first.

I'm a bit hesitant to suggest this but ... you could try putting a thin layer of solder on the V32's connectors to add some thickness. Just the far left and far right connector might be enough. You want it to be thin though, because you don't want to damage the Vectrex's connector.

- Bob

Firstly, to reiterate paragraph 2, this is not a fair test, since the v32 clearly _does_ boot just fine and run 6809 code - there's just something about how it does its startup sequence that gets flakey, and so comparing it against other cart designs isn't the most useful, since it already _does_ do the thing they do!  I love my v32, and I'm not throwing shade here at the design :).

That having been said: to confirm my earlier assertion or not, I actually just measured against some working cartridges - I think I was wrong about the thickness being the issue, since it seems to be the same. Where there does seem to be a difference, is how far "into" the port the carts go - because the vectrex32 has that little lip next to the pins, which none of my other carts do, it gets stop at that edge, whereas the other carts actually go a little into the plastic, a few mm past the end of the edge connector.  I'm not going to take a hacksaw to either my v32 or my vectrex to attempt to test _that_ theory though, hence thinking about just getting an external edge connector.  My current inclination is how "strong" the edge connector is on your vectrex affects whether it gets quite the connection it needs, and that mine (and the other posters) is weak.  A pile of books solving this because it pushes the contacts together...

To be fair to yourself though - it's not a fair comparison, since the cartridge clearly _DOES_ boot and load code, it just can't seem to establish whatever communication is requires for it to get past the waiting point, something that other carts don't have to deal with!

jaymz just gave me an idea.

If you look closely at the V32's bottom cover, you'll see that it is not symmetrical. The screws are closer to one edge than they are to the other. The cover is supposed to be put on so that the edge the screws are closer to, is near the edge connector. If I messed up and put the bottom cover on the wrong way, it will butt up against the shell of the Vectrex before the edge card has gone in all the way.

Let me know if this is the problem.

- Bob

Just checked this morning, the bottom plate is oriented correctly.

Then I got my Scramble, VecFlash, Protector/Yasi and Debris cartridges to compare them. They all seemed to have a bit more sturdy connection with a slightly deeper insertion. I also noticed that they still have some play (although with a bit more resistance), so the cartridge housings are not so tight inside the slot that it prevents any wobble.

Upon closer inspection I also think the difference is those lips on the sides of the connector. The difference might just be one or two millimeter though.

The thickness of the connectors looked about the same, though I could measure it with a calipher to make sure.

I've played around with this a bit.

The V32's edge card is 2 mm shorter than an original Vectrex game cartridge's edge card. If I plug the V32 in it works fine, but if I deliberately wiggle it, it will crash.

I can also deliberately wiggle an original game cart and cause it to crash.

With a band saw, I shaved off 1.5 or 2 mm of the V32's board to make the edge card longer. If you do this with the covers on, you won't cut any trace. However, I was still able to wiggle the board and make it crash. The board did not seem to be any more secure than it was before.

I use a retractable USB cable, which has a thin and lightweight wire. As such, it does not pull down the V32 when it's plugged in. Perhaps this is why this problem doesn't happen to me.

- Bob

Thanks for testing this! Really odd how the others cards seem to be inserted more securely, as the cartridge housing doesn't seem to be causing that. Maybe the extra 2mm would make a difference with my Vectrex connector though.

Mine is angled downwards a bit, even without any cable attached. The 3D demo crashes often or gets stuck on waiting.

Okay, for science I opened up my Debris cartridge and inserted only the PCB. The connection was deeper and more tight, I could still tilt it up and down. So I think modifying the PCB for deeper insertion would help. It would also make the casing sit close against the slot for extra stability.

Then I measured the thickness of both PCBs and they both measured 1.63mm.

I've just connected it to my computer and tried a little bit of programming. a Text2Sprite object gave me a parameter error when I tried to rotate it. A LineSprite box worked fine.

The command line interface and trying to figure out the correct Basic syntax, then seeing some results made me feel like I was programming my VIC-20 for the first time as a kid. That was a really nice bonus.

I'm not sure my workflow is correct. I might have overlooked something in the manual. I'm now saving directly to the Vectrex32 mass storage device from my text editor and then issueing the load and run commands in the terminal. I'm a bit worried of wearing out the flash memory, as I save often.

Then I tried creating some 3D graphics and at some point the load command froze. Then my text editor told me the code had changed and asked me to reload. I shouldn't have done that, the source was empty now and that was that. I wonder if I tried to load too quickly after saving.

Still odd how the 3D demo has been giving me so much trouble, I wonder what the difference is.

Anyway, I'm enjoying trying to do some programming on this, can't wait to use the power of the 32 bit CPU properly. Odd how I'm still expecting smoother motion than the Vectrex is capable of.

You could try shaving off some of the board around the edge connector. If you do it with the covers still on, any important traces should be safe. If you break it, you can return it to me for a free replacement.

Can you post the Rotate call that caused the parameter error?

The flash is rated for a minimum of 10,000 writes. That's a lot, and it doesn't mean it will fail at 10,001; it just means it's guaranteed to work until at least 10,000. Estimate how often you save (e.g. you work on it 2 hours a day and save every 10 minutes, for a total of 12 saves a day) and you can figure out how long it's guaranteed for (833 days in this example - over two years of working on the V32 two hours a day!).

You're using the same workflow I do, but as you've found, you can't load a file while it's still saving! Sorry about the lost program. :-(

What trouble is the 3D demo giving? You mention it crashed after creating 10 ships. If you have a terminal window open, it'll probably show you an error message, and I bet it'll be that there are too many vectors to draw in the allotted time.

- Bob

Thanks for the kind offer! I will do that.

The only difference with that command was that the first parameter was a text2sprite object instead of a linesprite. I can try to reproduce it and then send you the exact command.

I save more often than that, but that is good to know. I wonder if it would be possible to add a function to upload from the terminal directly to memory and have it execute.

Load freezing while saving is a bit tricky. I'll start waiting a couple of seconds before trying to load. And I'll try to remember to not allow the editor to refresh if it goes wrong :)

The 3D demo often went black, glitched or jumped into Mine Storm. I'll have a try with the terminal attached to see what happens.

I'm not sure what you mean by "upload from the terminal directly to memory". The editor and the terminal are two different programs, both running on the PC.

On my system, when I save a file to the V32, the Vectrex buzz stutters. That's how I know when the save is complete. Do you see (or hear) the same thing?

Jumping into Mine Storm is a really hard crash. Try it with the PC connected and a terminal window up. Does the PC lose the USB connection?

- Bob

I thought you might be able to upload the file over the serial connection, using the terminal, into the memory of the cartridge. Of course the Vectrex32 would have to be updated to be able to do something like that, if it is even possible.

I have a late no buzz Vectrex model, so I haven't noticed that.

I'll hook up the Vectrex in a moment and will test what happens when it crashes.

Well, I've been trying to run the 3D demo and I can't get it to run at all with the cable connected. The display just glitches out and goes black. Here's what I did:

Connected to the computer, ctrl-c, stop, load demo3d.bas, run, PRESS BUTTON 1 IN THE MENU:

Glitch + short audio -> black screen, no message, just a blinking cursor, terminal doesn't respond anymore.

Reset Vectrex

Still a black screen

Same thing after turning the Vectrex off, re loading etc.

Propping up the cartridge

Same thing happens

Overwrite the 3d demo with a new copy

Same thing

Propping up again

Screen collapses, crashes again

Re-insert cartridge multiple times

Doesn't help

Try it without the cable from the menu, runs first time.
Move three ships outside the display with the camera controls and it crashes.
Reset makes it go to Mine Storm

Second try

Dies right after pressing button 1 again, black screen.

Trying misslebreak outvaders:

Crashes to black screen the first time the ball hits the paddle.

Safe to say it is quite unstable here.

EDIT: I did play this game a couple of times before and it worked fine. And during development yesterday it seemed pretty stable.

Okay, I got my Dremel and removed the protusions on the sides. Before I did that I noticed from the markings on the edge contacts that insertion of the contacts had only been 2mm.

It is now about 4mm, looking at the fresh markings, slightly less than two thirds of the lenght of the contacts. I think the housing is preventing it from going in further now. It does already feel more sturdy.

I will clean things up here and do another test.

First try with the 3D demo from the menu: Glitch and crash
Reset and it works, that is odd

10 is still the maximum amount of ships before it crashes. The demo should probably have a limit.

Reset goes to the waiting message

Second go

Glitch and crash

Reset and again it works, weird

Button 4, back to menu

Missilebreak Outvaders

Crash on first ball contact again

This is all with the cable disconnected.

Well this might have helped somewhat, but not enough.

I vaguely remember later Vectrex models having a difference with the cartridge connections. Could that have anything to do with this?

I'm going to try without the housing to see if that helps at all.

Now that I have removed the housing I thought I'd inspect the board a bit. The soldering of the chips is not really clean, I'm seeing specs of solder near the leads of the chips.

But more importantly, I'm seeing some potential and one most likely short between the leads of the two chips. I guess that could very well be the cause of these problems.

A heads up: Removing the two protusions at the sides of the edge connector does get really close to one of the traces. So be very careful if you do this.

EDIT: Inserting it without the housing gives a really nice solid connection. The connector seems to go all the way in now.
Went over the leads of the chips with a toothbrush and a jewelers flathead screwdriver. Maybe this will help.

EDIT2: Well all of this hasn't helped. The 3D demo still crashed right away.

EDIT3: Vectorgunner and Lunar Lander seem to work fine. Yankee does a very short beep and immediately returns to the main menu.

Am I correct that the situation appears to be deteriorating? I.e. things that worked before are no longer working?

I suggested early on that there might be a problem with the way the Vectrex edge connector is soldered to its board. If things are deteriorating, that would be be consistent with bad solder joints that are cracking.

It does look like that Before this I only had trouble with the 3D demo and then only sometimes. I'm not really looking forward to fixing that, but that would have to be done anyway to keep my Vectrex working correctly.

Messing around with my simple rotating box demo seems to be working fine though, without the cartridge housing.

I just saw something surprising. I ran the 3D demo, pressed button 1 and the Vectrex screen went blank. If I reset the Vectrex, it would restart, re-establish communication with the V32, and display the 3D demo screen with the first Tie fighter visible. After that, I could create more Tie fighters.

So something caused the Vectrex to crash but not the V32. At this point, I don't know what. But it doesn't appear to be a loose connection.

- Bob

Yes, I have that exact same behaviour now. I hope this is something that could be fixed in software then.
EDIT: Does your 3D demo crash when you move everything off screen by moving the camera? And by adding more than ten ships?

Finally had some time to look at this.

I edited Demo3D.bas and added:
call WaitForFrame(JoystickDigital, Controller1, JoystickX + JoystickY)

after line 35. This solved the problem where the Vectrex crashes right after you press button 1 to leave the Instructions screen.

The other two problems, crashing after 10+ ships and crashing after moving the ships off the screen, are BASIC program crashes. They display error messages on the terminal. As such, I'm not worried about them. It's only a demo program. Consider them to be demonstrations of issues that can arise in your code. ;-)

- Bob

Cool, I'll try the fix.

I was working on my demo yesterday and had all the objects that used sprite translation fly off the screen, then the Vectrex crashed into Mine Storm which was corrupted and really odd things were happening on the screen.

I thought when you use translation the line sprites are clipped? Or do I have to set the display boundaries?

Sometimes when the Vectrex 32 crashes, or when it freezes while loading (this can happen even if saving was finished), the basic file will stay locked and can't be overwritten or even deleted anymore. I have to do a windows drive check to get the flash drive to be able to remove the files.

While working on the demo I've experienced some incorrect error messages (for instance about a missing '=', when the call command is missing).
And my session yesterday ended when I kept getting an error message about endfunction being an illegal command, even though the same command was used to end a function above it without any problems.

I've started keeping a list of odd things happening while developing, which I will share later on.

I really have to get used to developing in basic again. For instance adding a 2D velocity to a 2D position, you can't just add two arrays together.

Clipping does not happen automatically. You need to set it up.

If a file is getting locked, it sounds like the crash must have happened during a save. Be careful not to do anything else during a save.

When the Call command is missing, BASIC assumes you meant to assign a variable to the result of a function, rather than assuming you were trying to call a subroutine. That's why you get the missing equal sign error.

If EndFunction is giving an error, it's probably because you're not currently in a function or because the EndFunction statement is inside a loop or IF statement. You can't have more than one EndFunction in a function. Function...EndFunction is a compile-time block.

- Bob

Okay. This demo won't need any clipping, but that is good to know.
I'll keep an eye on what is going on exactly that causes files to get locked here.

That makes sense. Being used to other programming languages I tend to forget to add the call command.

I'll have a look tonight at what could be causing that, it is not one of those. It is just a normal function. Maybe the system got confused because I didn't get the syntax of one of the commands inside the function right.

Well, the obscure error went away after letting windows do another drive check on the Vectrex32 storage device. I find I have to do this every time the Vectrex32 crashes now, which is pretty annoying, as it crashes easily. I have been doing drive checks every few minutes.

I'm starting to think my Vectrex32 might have some kind of defect after all.

edit: Loading even often freezes when waiting a good amount of time for saving to finish. Then I have to recheck the drive again. All this is making development slow and frustrating.

It sounds like the flash drive has taken a beating. I'd copy all the files off it, reformat it, and copy the files back.

- Bob

Thanks for the advice, I will try that. Will see tomorrow if it helped. It is time to stop programming now :)

Missilebreak Outvaders still crashes on first ball contact, found out it then works fine after a reset. Yankee crashes after briefly flashing a display.

Come to think of it, a reset button on the Vectrex32 would really help. Then if it crashes, I don't have to turn my Vectrex off and on. I'm starting to worry about wearing the power switch out.

I don't know if one could be added to the current version somehow, or if that would also need a redesign of the board.

The 3D demo off-screen crash and similar crashes I've had seem to be caused by all sprites being off-screen. So there's still a sprite list, but no actual lines to be drawn.

I no longer think that adding a WaitForFrame is the solution. There's something messier going on. I'm looking into it.

- Bob

Cool. I hope you'll be able to fix it.

I'm still getting freezes when loading my demo, even after waiting minutes after saving. It happens maybe one in five times, if I wait a long time before loading every time. If I then reset the Vectrex it is stuck on waiting. This doesn't seem to corrupt the drive.

EDIT: Strike that, it happens much more often. Trying to develop anything is taking ages this way.
EDIT2: And now the drive got corrupted again.
EDIT3: And now I've lost the demo I've been working on for a lot of hours the last three days. I'm giving up.

I believe I've found a fix for the problems where Demo3D, MissileBreak, and Yankee crashed. I'm still having some problems with sound.

I've haven't been seeing the problems you describe with the disk. Is your USB cable plugged directly into your PC or is it plugged into a hub? Are you certain it's reliable?

I'm really sorry about the trouble you're having. If you want to return the V32, I'll certainly give you a full refund.

- Bob

Good to hear you've found a fix.

My cable runs directly from the computer to the cart. I haven't had any other troubles with this port. I could try another cable, but I would need to find one first, because most of my cables are micro USB. Would that even explain the freezes while loading?

I lost my program because I thought of trying to switch between programmer's notepad and notepad++, during that switch the file got erased by the Vectrex32 after the drive got corrupted again. I should have made a back up on my PC.

I've tried undeleting the file, but there doesn't seem to be anything that could be undeleted.

I really just want this thing to work, because I liked the whole concept of it a lot. I was very much looking forward to completing my demo, it was starting to look pretty interesting.
At the moment I'm not looking forward to starting all over again, as development has been so frustrating.

Ooh, good news. I found some obscure software that was able to recover the deleted file. Wow, what a relief, I was so disappointed. Going to find another cable and use Notepad++ to see if that improves things.
EDIT: That doesn't change anything. Still stuck at loading the second time I try to load my demo.
EDIT2: Wow, actually the drive got corrupted again with this load freeze and now after a drive check all the files are gone. This made me lose my earlier demos, but they weren't as much work to make. It is also unrecoverable now. Something seems to be very wrong with my cart.

Can you send me your demo program? I want to try loading it on my system.

- Bob

I will do that.

There's still a weird bug in there that makes everything fly off the screen after a short while (quit before that happens or it will crash). Maybe I'm using wrong syntax somewhere.

Using FAT to format is correct, right? And I'm using a cluster size of 512 to save some space.