New! Dark mode!

Browsing Pinside at night? Getting tired of all the white? Switch to dark mode using the button in the top right (or CTRL-B)!

(Topic ID: 257420)

Opening Flight 2000 new software up to the community


By slochar

10 months ago



Topic Stats

  • 461 posts
  • 51 Pinsiders participating
  • Latest reply 1 day ago by Zzap
  • Topic is favorited by 39 Pinsiders

You

Linked Games

Topic Gallery

There have been 20 images uploaded to this topic. (View topic image gallery).

Roms_Photo (resized).JPG
2732_Remark.jpg
IMG_5443 (resized).JPG
20200918_091216 (resized).jpg
20200918_091101 (resized).jpg
walker decision tree.png
flight 2000-speech board modification.png
F2KLOCK4 (resized).jpg
F2KLOCK1 (resized).jpg
F2KLOCK2 (resized).jpg
F2KLOCK3 (resized).jpg
FLIGHT2000-New-INSTRUCTIONS-FLAT.png
f2k-beta-2-19-20 (resized).png
Untitled (resized).png
20200408_125028 (resized).jpg
20200408_125107 (resized).jpg

There are 461 posts in this topic. You are on page 9 of 10.
#401 36 days ago
Quoted from Stretch7:

Game play has been working fine for me so far but haven't had a ton of games on it yet. One thing i do notice though, if the game is left on it will randomly start repeating the walker non stop... i will have to physically turn it off and on to stop and then it will just do it again. There doesn't seem to be a perfect timing for this to happen ....it can be after half hour, 45, hour etc. Weather my game did this with the old code I'm not to sure. Is this maybe an attract mode but only supposed to do it a few times?

That's strange but that gives me something to do on the weekend I'll leave the game on and see if I get anything like that. So when you turn it back on it keeps doing the walking, or you let it sit some more and it does it again?

#402 36 days ago

Yes when i turn it off its fine again until it starts it again usually with 30-90 minutes. I have seen this happen about half a dozen times and also with one of my employees numerous times as I had this game on location at the arcade. I have since brought it back home. Its really not an issue in a home environment where game isn't left on. Maybe before you look at it I will throw old code back on and see if it happens.

#403 35 days ago

Did you have the attract mode speech on? I am testing now just tossed 2 balls in the walker in a game then let it drain down to game over - that way if it does it the balls will be somewhere else so I can tell.

#404 35 days ago

Wow, you guys are amazing! So many great Sterns really suffer due to "uninspiring" game code. Did you guys get a Flight 2000 game ROM finished and released or are you still working on it? Always liked this game but the lack of more interesting rules always kept me from purchasing.

Now if someone could tweak the simple, but fun, Wild Fyre code.........

Great job guys!

#405 35 days ago

Yes, it's done pending results of the bugs reported so far. Links are back in the thread.

I think most of the later stern games actually have pretty good coding and don't really need too much changed.

Wild Fyre uses the same code as Dracula, zacaj was doing mods on that you might want to look for his threads on that (it might have just been on his website which it think was zacaj.com) - they should carry over to WF as well. I have the wild fyre recreated source code somewhere I think if you want a stab at it.

#406 35 days ago
Quoted from slochar:

mode speech on?

I’m not sure...will have to check

#407 35 days ago

I ran it for about 8 hours today put 2 balls in game in lock, let drain to game over (so I could tell if the balls moved) - no reboot or walk, and I have the attract mode speech turned on. I'm looking at the attract mode speech routine, if you had it on, try it with it off, or vice versa.

That doesn't necessarily mean that it's not some kind of software issue with the new software - what actually happens when it wigs out? Does it reboot, or does it just start walking the walker and lock it on or does it actually 'walk'? Does the rest of the attract mode (lamps, scores cycling still work) or does it freeze compeltely?

Normally with this type of issue I'd be looking at the stack hitting something in lower memory - the first $10 bytes of the 5101's are used for various things (stock, as well, although it does go a tad higher with the new software - which if the stack got low enough, would definitely cause issues if it happened to hit at the correct moment). At one point I did have a routine that ran at bootup that told me how low the stack got in heavy gameplay testing, and there was a $20 byte margin all of time (unlike Dragonfist, which dips REALLY low down, due to incorrect usage of certain commands by the programmer).

The other thing I'm thinking of would be if the display interrupt on your mpu200 happens to be a little slower than normal, or if you happen to have the clock speed set to bally instead of stern (doubtful, you would have noticed many more issues in the past) - stern and bally software in general do not recover nicely from a duplicate display interrupt, it usually crashes the machine, although your case is pretty specific if it's walking the walker.

Have you ever replaced the switch cap that's on the walker entry that makes it more sensitive to hits?

Did you try the normal software?

Anyone else that's running the v47 happen to try a long interval?

#408 34 days ago
Quoted from slochar:

Yes, it's done pending results of the bugs reported so far. Links are back in the thread.
I think most of the later stern games actually have pretty good coding and don't really need too much changed.
Wild Fyre uses the same code as Dracula, zacaj was doing mods on that you might want to look for his threads on that (it might have just been on his website which it think was zacaj.com) - they should carry over to WF as well. I have the wild fyre recreated source code somewhere I think if you want a stab at it.

Great, I'll check out zacaj. Thanks for the info. I'm soooo not able to reprogram - but I find it fascinating. Great job!! I'll keep following this thread, hope the trouble-shooting goes well.

#409 33 days ago

Finally got some time to troubleshoot again, after the photo of the switch was posted.

That photo was helpful, as I could see that the wireform trigger extended right from the top arc of the switch slot. On my game, the wireform was installed (screws for the plate drilled in) about 3/8" lower... so there was *absolutely no way* I could bend that wire to clear the ball, short of maybe a banana shape which is obviously incorrect.

Anyway. I moved the switches, so now the ball will clear the wire trigger - bent in a proper, OEM-looking "straight line" - such that the switch closure is momentary (closes as ball passes over, and opens when ball comes to rest against walker finger). Still only just barely! But it clears.

The good new is, the ball walker works properly in catching and locking balls now!

The bad news is, when multiball starts, the walker only advances ball 1 out... and never triggers again, so balls 2 and 3 stay in the lock area. The "static / launch" sound effect locks on, as the game apparently waits for the next balls to advance. But they never do, so if the single ball ejected drains at this time, that ball "ends" and ball-in-play advances to two. (This seems to be the same "2nd part" of the bug as I described it before, except I was faking/finagling the balls out of there owing to the weird switch behavior so I didn't quite understand what was going on).

My buyers are working on their arrangements to pick up this week, so I don't know if/how long I'll be able to test before I have to revert to stock roms (or maybe v.33 which was stable but didn't have all the speech & adjustment add ons).

#410 32 days ago

Make sure your kickbig switches do not have switch caps on them. Check your launch next rocket switch that's between the kickbigs - that's the trigger for the walker to start again.

The ball ended because the lack of physical switches in the locks means that it's a virtual switch - and when multiball starts, it clears the register because the balls are all supposed to come out - anything where you remove a ball when the game thinks it's there is going to mess up a lot of things, although I think there was some error correction on the trough (i.e. 3 in trough means there can't be any in the lock).

#411 32 days ago

You might be onto something; I noticed if I pressed the kickbig (? I never understood that name... kick*big*? Not kickback, or kickout? But I digress....) switches the balls would feed out but by that point the game was already confused. I'll look for the caps this afternoon/evening. They'll be easy enough to disconnect if there.

#412 32 days ago

It's from F-14 they call the yagov kicker a kickbig.

Probably not 100% accurate to apply it to an older game. At least it's not the other 'new' thing where people are calling saucers 'scoops'. That's just.... wrong.

#413 32 days ago
Quoted from slochar:

It's from F-14 they call the yagov kicker a kickbig.

AHAHAHA Is that really the first use of the term?? Say it with a Russian accent and it almost makes perfect sense!!

#414 32 days ago

OK I disconnected the caps on the kickbigs... and there's a bit of improvement... for multiball, the walker still won't advance balls 2 and 3 on its own. But when "multiball 1" drains, the ball-in-play doesn't end anymore.

So! How do I get balls 2 and 3 out of the walker? I'm not sure. Pressing the kickbig switches didn't seem to do it, at least not immediately. Nor did pressing the entry switch again. But at some point of reaction it finally advanced both balls properly. And those balls counted properly for "scoring x" and drains. When the 3rd multiball drained, bonus counted down and ball in play advanced. So... that's an improvement but the walker is still acting odd.

Is the "advance next rocket when lit" switch somehow key to this? I can't recall if pressing this finally woke it up or not as I only had a few quick moments to sneak a test in. Seems to me that's one of the switches that should stay capped. I did try the full gameplay lock sequence twice and the balky walker repeated both times. Both times I somehow got it to "wake up" but again, it was kind of a "surprise" so it wasn't yet clear to me how. It may become clear if/when I have more time to test later.

#415 32 days ago

Can you post a video?

#416 32 days ago

ALL SYSTEMS GO!

Turns out that indeed, the "rocket" rollover switch in the maze was not as reliable as needed. It triggered with a finger but only momentarily with the ball. When the multiball launch was hung, if I pressed that switch it would eventually trigger the walker. A ball wouldn't do it. I confirmed this performance in the switch test mode - it closed but only for the briefest barest flicker with a ball. No bueno.

So on closer look, wouldn't you know *of course* that might have been the single, only switch I overlooked when resurrecting this game! Its gap was wide and it had some powdered zinc oxidation on the keeper blade. Cleaned & adjusted it and everything worked fine the next test! I played two full games with 4 multiballs (and random "X scoring" walkouts) between them, and zero issues!

It nonetheless seems weird to me that this particular flaky switch never presented an issue on all the other code versions, but hey... I'm not complaining, and now it's known.

So for those keeping score, I had a confluence of issues to resolve:

1) The walker maze entry switch MUST open after the ball passes over, without the ball holding it closed. On my game I had to physically move the rollover and switch to allow this.
2) Clip the caps off the "kigbig" switches
3) Make sure the "rocket" rollover switch in the maze is well clean and adjusted. Mine apparently worked fine on other code revs (OEM; v.25; v.33...) but apparently that switch's closure timing is more critical on v.47...?

FWIW, I did record a video of the weird behavior and can upload that if you're still interested slochar , but I figured out the issue while doing that so I went ahead and posted here. PM me if you want a vid link for reference.

BTW, this updated code is WELL WORTH the aggravation. WOW but all your little enhancements and details really make this a whole new game! I love all the new callouts, and how it blinks "shoot again" when locking balls, etc. Superb work. I'm kinda sad to be selling it so soon now but the buyers should be thrilled.

#417 32 days ago

The original code blinks shoot again on ball locks as well but thanks for giving me kudos for that! (I probably would not have added that.... the game actually 'cheats' IIRC to get the shoot again blinking, it's linked into the normal shoot again blinking when you don't get a score and drain, which is not possible in f2k. Try getting a ball locked, and instead of shooting the ball, tilt it - see if you get that ball back. If I get a chance tomorrow I'll try that as well)

The crux of why that switch became important is because of the choreography of how I re-wrote the entire lock mech code - I had to do this because the original code had lots of redundancies, and a primitive form of error correction in it - needed the space. Original routine was almost 500 bytes, to add the other features in and fit the mpu200 rom footprint I had to cut something.

The older version 33 was still using the old original code. It also looks at that release next rocket switch, but also has something in there to proceed if it doesn't see it when it should. I could probably add that in a maintenance release, because the right kickbig should also trigger that flag..... there's still 16 bytes in the high rom to play with.

Glad you're enjoying the changes, however briefly. I hope when shows start back up that someone drags a modified one to the show and people are like What the heck??? Since when does f2k play this way???

Oh, and can you do me a favor since the other bug reported I was unable to duplicate - lock 2 balls and end the game and leave the machine on for 6-7 hours for a burn in and make sure it doesn't freeze up/reboot. You can tell if it rebooted because the balls will no longer be in the lock.

Did you add the knocker too?

One of my favorite features I'm back coding to other stern games (and probably bally, too) is to hold the slam switch in adjustments to go backwards instead of forwards when setting replay levels. I *always* overshoot by 10k, and have to restart.

#418 32 days ago
Quoted from slochar:

Oh, and can you do me a favor since the other bug reported I was unable to duplicate

I went to go play with this the other day and noticed my walker coil burnt which was is weird ...so I'm going to have to replace it and test again.

Quoted from slochar:

One of my favorite features I'm back coding to other stern games (and probably bally, too)

Very cool

#419 30 days ago

I've got my 3347 roms burnt and ready to roll. I'm planning to rig up something to record some video while testing so hopefully if I see something quirky I can post videos. Machine has a stock MPU board, and I won't be rewiring the walker coil in the first instance, but most likely will after a bit of testing.

IMG_5443 (resized).JPG
#420 30 days ago

Thanks for explaining the difference in switch timings. These glimpses into the inner workings of the pinball machine code are fascinating; there's a lot more going on than meets the eye, or even a first assumptive logic breakdown. Pretty awesome that you're able to suss and parse it all out years later!

Yesterday got a bit away from me and I didn't get a chance to leave the game on with balls in the walker for that length of time. I did leave it on for a few hours and heard attract speech throughout; when I went to play a game afterward nothing seemed amiss.

I was interested in the knocker mod but turns out I don't have a suitable mech handy (an early Sys11 knocker doesn't seem to mount or modify appropriately without running a lot of wire). And with the appropriate cautions about fusing the mech and such, coupled to a pinball newbie's sudden interest in purchasing this game from me, I decided not to add that possible failure mode for now. Were I keeping the game I'd probably go ahead figure out the Sys11 knocker mod.

And yes, I really wanted to drop this at LAX 2021 next March and have people go "What the...?" but the way things are going I'll be surprised if the show happens at all... and I might be foolish to refute a surprise buyer out of the blue. But the update is awesome! So much added speech; I *love* the spinner count and tilt warning features... the skill shot was sorely needed... as was multiball X scoring and walker-eject... everything is well thought and executed. I really got the sense this is what Stern would have released with a bit more dev time back in 1980. Wife and I played several more games yesterday without issue since I found my machine's unique "hardware bugs". I'm glad to have experienced the update and hopefully help out a bit.

#421 30 days ago

Oh, I've had a lot of 'going too fars' that ruined gameplay in some cases on other games. As for able to reverse engineer code, this is about all I'm good for. Modern programming languages mostly just annoy me, and I don't really "know" any of them.

For instance, 10 years ago I would have made the bonus exponential, going up to 55k like some games do. (quicksilver) - that would have ruined it, especially at 15x.

I'm a sucker for the 2x/3x style multiplied scoring (black knight is one of my favorite games) so it's no surprise it's made it here as well into my custom system 7 firepower.... there were a couple one-offs in the past where I added that as well.

I didn't think I'd like the spinner counters as much as I do, thank cottonm4 for that suggestion. I liked them well enough on Dragonfist but didn't think it was that big of a deal on other games (although I did add a score counter on firepower for the spinner when lit, but never really equated that to just a raw counter.)

Hopefully the bug regarding the reboot/freeze up from stretch7 is a one-off vs. something systemic but we will see.

#422 26 days ago

I've been following this thread with delight. So many great changes. I am completely new to burning roms. Is there a recommended overview somewhere on how to do it. I would need to know what burner to buy and what chips. Never done that before.

#423 25 days ago

Ok, it's taken me a bit longer to get around to testing than I would have hoped, but a few steps closer today. Got the kickbig switch caps cut out and tested on stock code - all good. Also got a camera rig setup to record play with the new roms, so should have some footage over the next few days.

#424 25 days ago
Quoted from ArcadeRaid:

I've been following this thread with delight. So many great changes. I am completely new to burning roms. Is there a recommended overview somewhere on how to do it. I would need to know what burner to buy and what chips. Never done that before.

If you have an original mpu200 in your game jumpered for 4 chips 2716's, if you want to get a burner you can get the minipro 866 or even a minipro II, really cheap, and 2816 eeproms. You'd need 6 of those, 4 for the mpu and 2 for the speech. The II is a newer version of the 866 which is not as versatile, but will still be able to do the 2816's. Actual 2716's would not be able to be burned by either of these models though so you can't just erase what you have in there and reuse.

Sometimes you can get the chips and burner for less than $50. Ebay is a good source for these, amazon works too but tends to be more expensive for the same thing. Look for a USA shipper to get quicker, stuff from overseas is taking a very long time to arrive now.

#425 25 days ago
Quoted from slochar:

If you have an original mpu200 in your game jumpered for 4 chips 2716's, if you want to get a burner you can get the minipro 866 or even a minipro II, really cheap, and 2816 eeproms. You'd need 6 of those, 4 for the mpu and 2 for the speech. The II is a newer version of the 866 which is not as versatile, but will still be able to do the 2816's. Actual 2716's would not be able to be burned by either of these models though so you can't just erase what you have in there and reuse.
Sometimes you can get the chips and burner for less than $50. Ebay is a good source for these, amazon works too but tends to be more expensive for the same thing. Look for a USA shipper to get quicker, stuff from overseas is taking a very long time to arrive now.

Slochar is an expert in these things, but I was also in the same boat and had some interesting experiences to cross over the hump.

I used a MiniPro 866 as well. Be aware there are some firmware version changes for it, and apparently there are now knock-offs which don't have all the capabilities. The one I use came from a friend who got one of the last made by the OEM before they discontinued it and knock-offs appeared, so I didn't have to stress much about the possible gotchas... yet the software & drivers for it were tricky to find which is how I came to hear about them.

That said, the MiniPro *can* burn authentic 2716 chips. I know this because my Stern MPU-200 had OEM stickered 1980 2716's installed... which I erased (with a UV eraser) and re-used in the speech board. The MPU-200 can be jumpered to use 2x 2732 chips instead of 4x 2716. I recommend this as 2732's are easier to find. Be warned, avoid TI (Texas Instruments) 2732A's as these seem to be very prone to faults and don't like being re-burned reliably, whereas the ST 2732's I had burned, and re-burned, and re-burned without grief.

To Slochar's warning though, the MiniPro will NOT reliably "erase" or over-write existing ROMs with different bits, so if you make a mistake, have a problem, or want to salvage & re-use some OEM roms, you'll need a UV eraser. I got one from AliExpress for $15 shipped, it took about a month to arrive. Note this will only work for ROMS with the window under a sticker; masked or sealed ROMs cannot be erased by this method (nor by the MiniPro).

The speech board is the biggest confusion I had in this project and why I had to re-use the 2716's from the MPU. The readme for the new code states the speech board can only use 1x 2532 (*not 2732*), or 2x 2716. 2532's are hard to find and I had none on hand. Hence, I modded the MPU to claim and re-use those 2716's. (I've been told a single 2732 might work if the 2716 images are combined but did not try that. Yet, I also discovered an earlier version that someone burned for me before I had all the gear, put the speech image on a 2732 and it sorta worked, so... maybe???).

#426 25 days ago
Quoted from goingincirclez:

That said, the MiniPro *can* burn authentic 2716 chips. I know this because my Stern MPU-200 had OEM stickered 1980 2716's installed... which I erased (with a UV eraser) and re-used in the speech board. The MPU-200 can be jumpered to use 2x 2732 chips instead of 4x 2716. I recommend this as 2732's are easier to find. Be warned, avoid TI (Texas Instruments) 2732A's as these seem to be very prone to faults and don't like being re-burned reliably, whereas the ST 2732's I had burned, and re-burned, and re-burned without grief.

What definition did you use with the 2716's? Most 2716's I've seen need 25v to program which no version of the minipro can supply, tops out at 21v.

To avoid the issues with requiring erasers that's why I recommended the 2816's they are an eeprom version of the 2716.

Quoted from goingincirclez:

To Slochar's warning though, the MiniPro will NOT reliably "erase" or over-write existing ROMs with different bits, so if you make a mistake, have a problem, or want to salvage & re-use some OEM roms, you'll need a UV eraser.

No burner will. You can overwrite set bits with clears but the only way to set it back is the UV eraser. That's not the intended use of eproms though, they always need erasing via the window and UV light. EEproms on the other hand, can be erased and have to be, the minipro automatically erases them. If you are using the EEprom variety (so far only the 2816 and the 27512 varieties are known to work on this era board - the 256 one might, and apparently other sizes just aren't available) you would only burn it once, since erasing it and reburning isn't going to change anything.

Regular UV erased eproms you can double/triple burn, and some stubborn older ones you have to do this. Any UV eproms I just always burn twice, remove from the zif socket on the burner, clear the buffer, wait ten seconds or so, put the chip back in, then read it in, save, and verify it. Nothing is a worse head scratcher than putting a newly burned eprom in a game and it not working, which in my case leads me to think I made a mistake on writing the software.

Quoted from goingincirclez:

2532's are hard to find and I had none on hand.

This is a myth At this point in time, all vintage eproms are harder to find, but 2532's have been "hard to find" for over 20 years now, according to people in the pinball/arcade communities - but lots of them are out there. None of these older eproms have been made for decades. The myth likely started because Texas Instruments went ahead and screwed up their numbering for a variant on the 2716 - the 2516. The TI labeled 2716 are all goofed up, triple voltages needed to burn or something or a pin in a different place (I forget what....) - but they did go back to the correct numbering scheme on later chips (including the 2732A, which does work... unless you get a bunch of counterfeit ones!). Their reputation was probably mud by then though which is why the 2532 gets a bad rap.

I believed the myth too, and did a lot of hacking and conversion to get 2732's into everything. Then I just built the Bob Robert's adapter and burn them just fine so I don't have to change a lot of jumpers. True, the minipro will not do a 2532 out of the box (there's no reason it can't, a definition file just needs to be added by the maker, all of the pins on the minipro are programmable by the software) - other burners like the QX4 will.

The easiest solution by far if you have stock mpu200 and stock vsu100 is to just use 2816's. No jumper changes, no eraser needed. Chips were a buck or so each. I did have a couple dud ones so I ordered extra (and obviously, I go through a lot of burns anyway so I have to have extras....)

Everyone should be aware that there are LOTS of counterfeit and remarked chips out there on the used market. I always buy from people that will certify that the chips are genuine pulls. Last large buy I bought was 256 eproms from a music warehouse that had pulled them from old synthesizers. Yeah, I had to strip labels and scrape off 200 chip pulls, but they were in great shape (no bent pins). Then I erased them in batches of 40 on my eraser, and I'm good to go for a long time on that size chip.

The reason I'm listing that the 2732 is not recommended for the vsu100 is that the board factory stock is set up for either 2x2716's, or 1x2532 - no trace cuts and jumpers needed. It is strange because normally you can pop a doubled up 2732 into a 2716 socket and it will work, but I believe I tried this and it just didn't. Since I can do either the 2716 equivalent in the 2816 now or 2532's, I just did that rather than hack up my vsu100 trying to get that combo to work.

#427 24 days ago
Quoted from slochar:

What definition did you use with the 2716's? Most 2716's I've seen need 25v to program which no version of the minipro can supply, tops out at 21v.
To avoid the issues with requiring erasers that's why I recommended the 2816's they are an eeprom version of the 2716.

I don't recall for certain, but I'm pretty sure it was a Fujitsu profile (oddly, two of the OEM Roms were Fuji but the other two were a different make!) and yes, I did dial the voltage back to 21v, but otherwise these were very trouble-free ROMs, I was kinda surprised!

Quoted from slochar:

No burner will. You can overwrite set bits with clears but the only way to set it back is the UV eraser. That's not the intended use of eproms though, they always need erasing via the window and UV light. EEproms on the other hand, can be erased and have to be, the minipro automatically erases them... ...Regular UV erased eproms you can double/triple burn, and some stubborn older ones you have to do this. Any UV eproms I just always burn twice, remove from the zif socket on the burner, clear the buffer, wait ten seconds or so, put the chip back in, then read it in, save, and verify it.

I'd seen that advice a few times, and learned the "triple-burn trick" the hard way earlier this year when you were helping me out with my Time Fantasy UV/code project and all I had on hand were some stupid "blank" TI 2732's and the MiniPro. It's a good general habit to get into, and only takes a few extra seconds.

Quoted from slochar:

This is a myth At this point in time, all vintage eproms are harder to find, but 2532's have been "hard to find" for over 20 years now, according to people in the pinball/arcade communities - but lots of them are out there....

Ha, point taken... Still, in relative terms I've not seen "2532" nearly as much as the other types and I didn't have any on hand so, I stand my my statement But yes, one might say they're all difficult to various degrees, and that could include more and more thru-hole logic ICs with each passing month. It's good that we have widely available resources and alternatives to troubleshoot, re-use, and re-engineer in many cases.

Quoted from slochar:

The reason I'm listing that the 2732 is not recommended for the vsu100 is that the board factory stock is set up for either 2x2716's, or 1x2532 - no trace cuts and jumpers needed. It is strange because normally you can pop a doubled up 2732 into a 2716 socket and it will work, but I believe I tried this and it just didn't. Since I can do either the 2716 equivalent in the 2816 now or 2532's, I just did that rather than hack up my vsu100 trying to get that combo to work.

I can understand that. Had I kept the game longer I would have tried the 2732 myself to put this debate to rest, and get my HTF 2716's back :/ I think Quench has gone on record somewhere saying the 2732 *will* work. But I totally respect that you can only assure what you yourself have tested and confirmed.

#428 24 days ago
Quoted from goingincirclez:

I think Quench has gone on record somewhere saying the 2732 *will* work.

You can use two 2732 inplace of two 2716 if that's all you have, like pinsider ZZap wanted to do here:
https://pinside.com/pinball/forum/topic/opening-flight-2000-new-software-up-to-the-community/page/8#post-5804464

But if you want to use a single 2732 on the VSU-100 speech board, you need to modify (cut/run wire(s)) at the U9 socket.
I thought I posted this mod about two months ago but can't find it now :-/
Personally I'd just use drop in ROMs rather than modifying the board.

I have a genuine TL866A MiniPro, and a bootleg knockoff unit.
They program 2716/2732 with too short a program pulse. Programming 2716 multiple times usually gets the data to stick.
I modified the software for the knockoff unit I have to extend the programming pulses and it improved the 2716 writes, but still wasn't perfect. Since the schematics are out there it looks easy enough to modify the unit so it can write at 25 volts. It just uses a buck converter with resistors to set the programming voltage. Should be able to bump up the voltage with a resistor change - I just need to get around to try it..

#429 24 days ago

Once I got those 2816's I'm all about not re-jumpering. I'm actually relatively lazy.... so never went beyond the first test of the expanded vocab chip, where it most certainly did not work as doubled up 2732's. If I still have ZIFs on the sound board I might try in the future once any maintenance releases come out if needed.

Here's a 2532 modern replacement project someone came up with:
https://8bitshardway.blogspot.com/2020/04/building-2532-replacement-4k-rom.html

The board should be able to be modified to support 2732's, for people that want a nice eeprom version of either 2532's or 2732's.

#430 24 days ago

Ok, thanks for all the input. Does the exact 2816 matter or will they all work? For example, a quick ebay search turns up all these options and more.

X2816BPI
X2816BP
PQ2816A-300
DQ2816A-250
X-2816CP-12

Are those just different brands of the same thing or are the other letters and numbers critical?

Thanks.

#431 24 days ago

As long as it's an eeprom in 24 pin dip package it should work. I got atmel ones through Amazon.

#432 23 days ago

Thanks for the awesome work on the software slochar! I got some testing done yesterday on the 3347M software, recorded the below video and made some notes. The gameplay improvements are really neat, so I was keen to test as thoroughly as possible and get things running smoothly (not there yet!). I’ll continue to record and document as I make any adjustments so others learn about any tweaks required and and potentially spot any other bugs in the videos.

Modifications From Stock:
• Clipped out capacitors on Left Kickout and Right Kickout switches (switches 15 and 16)
• 3347M software - 716 images doubled and burnt to 6 M2732A EPROMS

Dip Switch Settings:
32: On
31: Off
30: On
29: Off
28: Off
27: Off
26: Off
25: Off
24: Off
23: On
22: Off
21: On
20: On
19: On
18: On
17: On
16: On
15: On
14: On
13: On
12: On
11: Off
10: Off
9: Off
8: On
7: Off
6: On
5: On
4: On
3: On
2: On
1: On

Issues Found:
• Very rarely detects switch Ball Launcher switch in walker (39) (4:09 is a rare time it does, also 6:02, then doesn't again at 6:30). This seems to be similar to previous reported issues where the switch needs to be moved up slightly? In video I lift the ball slightly and drop it into place when solenoid actuates. Are you sure the stock firmware detects the ball being off the switch, not when it's down?
• At 6:56 I complete BLAST OFF, with both stages Go, should it have announced Commence Countdown?

Speech issues:
• 0:35 - Scratchy sound - Player ###### #### for mission
• 2:56 - Scratchy sound - Player ###### #### for mission - followed by Skill Short - does Skill sound right?
• 4:56 - Scratchy sound - Player One Brief? for mission - Note 'One' is audible at this time but was scratchy above
• 5:27 - Skill Shot - Skill is quite scratchy
• At 7:16, hitting the 3 drop target during the Countdown, it doesn't sound like it says 3 - Clear? (also 11:51)
• 11:21 after locking the second ball - the word Second doesn't seem right for Second Stage Go?

#433 23 days ago

Just a follow up, I pulled the sound roms and successfully verified their data against the published 3347M version.

After a bit of research, is it possible that these EPROMs are too slow to run in the speech board? They are labelled on them as ST M2732A-2F1 21V FAST, however I believe that they aren't genuine as I couldn't burn them at 21V and had to use a 12.7V profile (all burnt at least 3 times and verified, GQ-4X programmer)...

#434 23 days ago
Quoted from Zzap:

ST M2732A-2F1 21V FAST, however I believe that they aren't genuine as I couldn't burn them at 21V and had to use a 12.7V profile (all burnt at least 3 times and verified, GQ-4X programmer)...

Got a picture showing the silicon die inside the EPROM?

Was recently helping someone with fake ST M2732A-2F1 remarked chips:
Compare his image against the various 2732 die images in the web link below:
It doesn't match any of the genuine ST 2732 dies.
It does match the NSC "NMC27C32BQ-150" which is indeed a 12.75V program chip.

http://cpucollection.ca/2732.htm

2732_Remark.jpg

#435 23 days ago
Quoted from Zzap:

Speech issues:
• 0:35 - Scratchy sound - Player ###### #### for mission
• 2:56 - Scratchy sound - Player ###### #### for mission - followed by Skill Short - does Skill sound right?
• 4:56 - Scratchy sound - Player One Brief? for mission - Note 'One' is audible at this time but was scratchy above
• 5:27 - Skill Shot - Skill is quite scratchy
• At 7:16, hitting the 3 drop target during the Countdown, it doesn't sound like it says 3 - Clear? (also 11:51)
• 11:21 after locking the second ball - the word Second doesn't seem right for Second Stage Go?

This is the same behavior I had with a (single) 2732 ROM that someone had burned for me (I'd asked for a 2532 and assumed that's what it was, but when I pulled the sticker off, d'oh! Apparently I'm not the only one confused by this speech board going against "conventional typical practice"). Did you enable the tilt warnings and try tilting the game? I got the same scratchiness. BUT all of the "stock" sounds work! So I wasn't sure what the problem was (I had other issues going on too as posted above).

I think the key differentiating point I've only now come to understand is that, unlike the MPU, the speech board won't allow you to *combine* two separate 2716 images onto a single 2732. But you can (maybe) *duplicate* each 2716 image onto a discreet 2732... in which case you would still need to use two chips for the speech board. The only single-chip option remains a lone 2532, due to wiring / pinout differences.

Can you confirm what you were trying?

#436 23 days ago
Quoted from Zzap:

I’ll continue to record and document as I make any adjustments so others learn about any tweaks required and and potentially spot any other bugs in the videos.

You definitely have something going on with the sound roms, there's even more in there you didn't note, you're getting static for some of it when you shouldn't. Almost as if BOTH eproms are feeding data at the same time to the speech processor. I was not able to double up the 2732 and get it work but as I said previously, I didn't put too much testing into that after I got the 2816's. It stands to reason for me though that since the stock speech board is set up to handle 2x2716s, OR a single 2532 (which normally has different pinout, so it's surprising that this setup works, but that's what they designed it for). I think the autoswitching 2716 selection is messing up having 2x2732s in there.

Do you have a good cap on the walker entry switch? Is the gold plating still on the switch?

It's very possible that the stock software can react to a stuck switch on the walker, neither it nor I are checking for the ball necessarily coming OFF the switch before processing the rest of the routine (you can't do that because in multiball if you shoot one ball up there, it's waiting for the spinner to finish, and if you get another ball up there during this, the 2 balls would always just sit there and do nothing, since it can't clear the switch in that situation.)

The previous discussion on the non-reactance of the walker switch was caused by an additional flag I'd added to the spinner routine, to try and ensure that the spinner truly was done before moving the walker (since you lose score if you're walking with the stock transistor). That I will likely eliminate, since that case was the machine not processing the spinner timeout correctly. That's on the maintenance fix list already for that edge case)

#437 23 days ago
Quoted from Quench:

Got a picture showing the silicon die inside the EPROM?
[quoted image]

Thanks for the detail on the fake chips, seems like I've got two different types in the batch I received! Here's a pic, my burnt sound roms were both of the type that has U9 written on it in the pic.

Looks like the U9 one might be NMC27C32QE-45? and the U6 one might be the same one you identified as NMC27C32BQ-150?

Roms_Photo (resized).JPG

Quoted from goingincirclez:

Can you confirm what you were trying?

I am using two 2732's, one with a duplicated version of U9, the other with a duplicated version of U10

Quoted from slochar:

You definitely have something going on with the sound roms, there's even more in there you didn't note

I wasn't 100% sure with some of the other noises around, and also relatively new to F2k, so pointed out the most obvious ones

Quoted from slochar:

Do you have a good cap on the walker entry switch? Is the gold plating still on the switch?

I'll have to inspect and test it later today. I never had an issue with the ball getting stuck there with stock code, so it definitely seemed to be registering 100% of the time.

#438 22 days ago
Quoted from Zzap:

Looks like the U9 one might be NMC27C32QE-45? and the U6 one might be the same one you identified as NMC27C32BQ-150?

If U9 is really a NMC27C32QE-45 chip, those are 25V program chips. The large size of the die is early gen so points this way. If you can't program at this voltage they may be bad. If they seemingly program and verify ok at 12.7V I wouldn't trust them.
BTW change the programming speed to 0. A deep program of these chips should take over 3 minutes.

Your U6 chip looks the same as the NMC27C32BQ-150, those are 12.75V program chips.
The 'B' in the part number specifies 12.75V
2732 is 25V
2732A is 21V
2732B is 12.75V

#439 22 days ago
Quoted from Quench:

If U9 is really a NMC27C32QE-45 chip, those are 25V program chips. The large size of the die is early gen so points this way. If you can't program at this voltage they may be bad. If they seemingly program and verify ok at 12.7V I wouldn't trust them.
BTW change the programming speed to 0. A deep program of these chips should take over 3 minutes.
Your U6 chip looks the same as the NMC27C32BQ-150, those are 12.75V program chips.
The 'B' in the part number specifies 12.75V
2732 is 25V
2732A is 21V
2732B is 12.75V

The "NMC27C32QE-45" chips did program, and verify (including removing, and testing several days later) at 12.7V, so really I'm not sure what they are...

I’ll have a go programming the sound roms onto the -150 type chips

#440 22 days ago

Progress! Got programmed the sound roms onto the EPROMs with the smaller silicon die - what seems like the -150 versions, and they work perfectly! I think at least on the principal that the smaller die must be a newer and faster rom, that the other one was too slow to be used as sound roms.

With the switch I haven't been able to pull the switch apart yet to check the gold and capacitor, but I did a test where I held the switch close after spinning the spinner, with no reaction from the walker. Whenever I lifted the switch the walker solenoids would then fire.

#441 22 days ago
Quoted from Zzap:

what seems like the -150 versions, and they work perfectly! I think at least on the principal that the smaller die must be a newer and faster rom, that the other one was too slow to be used as sound roms.

Great to hear!

The problem isn't the EPROM speed. The speech processor is only running around 32kHz.
The MPU board CPU is running nearly 900kHz and the ROMs on it are 450ns speed, same as those "NMC27C32QE-45".
Normal practice with a decent EPROM programmer is they perform verification at +/- 10% of voltage. The GQ-4X doesn't do this so it might verify ok in the programmer, but voltage being a little different in the game might not get good reads out of the EPROM.
I've seen Pinsiders program old EPROMs that verify ok on the GQ-4X but don't work in game. It's rare but does happen. The MiniPro TL866 also sometimes suffers from this.

#442 21 days ago

Here's some video of stock software on my machine and its behavior locking balls at:

0:33 (first stage - 2 balls previously locked from previous game)
1:19 (second stage - 2 balls previously locked)
2:16 (start multiball)
3:29 (first stage)

#443 21 days ago

The delay in processing the walker is intentional to allow the spinner to stop so you don't get screwed out of whatever score you would have (which can be substantial after multiball).

The way the new code is set up is there is both a flag and a timer set when the spinner is going (showing the spins on the match display, as well). Walker switch activates, this starts the walker routine, the first part of which checks both the spinner timer and the spinner flag, and sets another timer to re-enter the walker routine. (There are 16 time delayed virtual switches on stern mpu200 games).

When your walker doesn't activate on the new code, does the spinner count remain on the display? That means that the spinner routine timeout (another time delayed virtual switch) is not timing out. The addition of the flag was a late addition that probably shouldn't have been. There's no harm in running the spinner timeout code even when the spinner hadn't been going so I will probably remove the flag, or repurpose it to goose the timers involved.

When you held down the walker switch and had the spinner going it should not have started to process anything moving the walker or locking balls until that spinner timer expired (visualized by the ball in play reappearing on that display). If it hit the walker switch first before reacting to the spinner (on a very fast shot) it would process the walker while the spinner was still going - this did happen on my game because sometimes the ball just rockets into that area and the spinner is going too fast to react.

So to sum up.... what does the spinner count do when you go into the walker lane that will tell where the issue may lie. Stays on display=likely a software issue regarding the timeout which I plan on making more robust or at least retriggering it. Leaves display, no walk=either a software issue where the walker thinks it's already completed the routine or more likely, another timeout issue. I can probably reuse the redundant spinner flag to use for the walker re-trigger instead, as a double check that the time delayed switches do run.

There's nothing in the code stock or modified that prevents switches from running so that would point to a head scratcher in that they should ALWAYS run. I already know the game doesn't come anywhere near to running out of stack space (everyone would have reported that, even on the Weebly board, because the game uses the lowest 10 bytes there for sound) - also I had some developmental diagnostic code that told me how low the stack dipped on the previous boot. It never went below $230.

#444 21 days ago

Ok, it will take me a few more days to run some further tests. I have taken the new code out and put stock back in for a few days - got my nephews coming around to play a few games.

What I was able to discover is that 3 out of the 4 MPU roms I burnt were of the type causing issues in the sound board. So there's a chance the issue is coming from there too... I've got a UV eraser coming in the post, and 5 of the good type that need erasing, so in the meantime, I have them sitting in the sun and we'll see whether nature or the postal service are quicker at delivering erasure to the chips

#445 17 days ago

2816 chips ordered! Can't wait to try this out.

#446 15 days ago

Ok, the postal system beat the sun and I got my UV eraser! I reprogrammed the CPU roms onto my "good" chips, and they haven't made any difference in the running of the system, so exactly the same as previous with the ball gets stuck at the first switch in the walker. I guess it's interesting to know that the CPU is less picky than the sound board for these roms?

slochar - to answer your question, the spinner counter does stay on the screen indefinitely when the ball gets around to the switch, and doesn't come off until the switch is lifted. As you can see in this video, the counter stays on the screen until I move the ball upwards, disengaging the switch. The ball then falls back and hits the switch again, so that it thinks that's the second ball locking.

#447 15 days ago

Do you have it adjusted so the switch isn't down when the ball is sitting in the top of the walker?

You're getting second ball locked because you are activating the switch again. The walker code is 'in progress' - the switch basically sets a timer which eventually processes the switch. There are 3 switch events connected with the walker, one is the physical switch, another is the entry point of the walker, and the third is the progression point of the walker. The progression point can be re-entered multiple times as its waiting for the spinner counter to finish.

I forget if you've tried to just toss a ball in there without activating the spinner - what does it do there. Also, when it gets stuck again, activate the spinner for a couple spins and see if it 'wakes up' after the spinners stop.

The no-spinner timeout is definitely an issue, too, because that should never happen. The game should not be missing timed switch hits - like I said in a previous message either here or in the meteor or big game thread, it shouldn't clear pending switch events in the timed column until they're actually activated, but it is. Since it's happening on all 3 games as well as sometimes on stock roms, I'll have to look at the switch code to see what's changing parameters it shouldn't.

#448 15 days ago
Quoted from slochar:

Do you have it adjusted so the switch isn't down when the ball is sitting in the top of the walker?

The switch is definitely down when the ball is sitting at the top of the walker. I questioned this here:

Quoted from Zzap:

This seems to be similar to previous reported issues where the switch needs to be moved up slightly? In video I lift the ball slightly and drop it into place when solenoid actuates. Are you sure the stock firmware detects the ball being off the switch, not when it's down?

You answered:

Quoted from slochar:

It's very possible that the stock software can react to a stuck switch on the walker, neither it nor I are checking for the ball necessarily coming OFF the switch before processing the rest of the routine (you can't do that because in multiball if you shoot one ball up there, it's waiting for the spinner to finish, and if you get another ball up there during this, the 2 balls would always just sit there and do nothing, since it can't clear the switch in that situation.)

The previous discussion on the non-reactance of the walker switch was caused by an additional flag I'd added to the spinner routine, to try and ensure that the spinner truly was done before moving the walker (since you lose score if you're walking with the stock transistor). That I will likely eliminate, since that case was the machine not processing the spinner timeout correctly. That's on the maintenance fix list already for that edge case)

--------------------------------------

Quoted from slochar:

I forget if you've tried to just toss a ball in there without activating the spinner - what does it do there. Also, when it gets stuck again, activate the spinner for a couple spins and see if it 'wakes up' after the spinners stop.

Retested, with video below. In both scenarios, the machine just sits there, just the same as when the spinner spins and it's sitting there. Final part of the video shows me pushing the switch down, waiting a period, then lifting the switch. When I lift the switch it detects it and proceeds as it should.

#449 14 days ago

Do I see a playfield protector on there? That's possibly going to interfere with getting the switch to come back up, it looks like the top of the switch just can't come up over it. Can the slot be cut slightly bigger to allow the switch to come up? This is pretty much the crux of the issue - regardless of the original software's behavior. Going back to the original activation removes having the spinner work as the modified software's intention, which is to award the spinner as long as its' spinning, instead of robbing you of any score that would have happened while it's walking.

#450 14 days ago
Quoted from slochar:

Do I see a playfield protector on there? That's possibly going to interfere with getting the switch to come back up, it looks like the top of the switch just can't come up over it. Can the slot be cut slightly bigger to allow the switch to come up? This is pretty much the crux of the issue - regardless of the original software's behavior. Going back to the original activation removes having the spinner work as the modified software's intention, which is to award the spinner as long as its' spinning, instead of robbing you of any score that would have happened while it's walking.

Ok, so are you now saying the switch does have to come up, and you aren’t detecting on the switch being pressed? What you stated previously, which I quoted again yesterday to try and get an answer from you was that you aren’t checking for it to come off.

Quoted from slochar:

neither it nor I are checking for the ball necessarily coming OFF the switch before processing the rest of the routine (you can't do that because in multiball if you shoot one ball up there, it's waiting for the spinner to finish, and if you get another ball up there during this, the 2 balls would always just sit there and do nothing, since it can't clear the switch in that situation.)

Promoted items from the Pinside Marketplace
From: $ 5.00
Playfield - Toys/Add-ons
UpKick Pinball
$ 6.00
Cabinet - Other
Siegecraft Electronics
1,600
Machine - For Sale
Grand Rapids, MI
From: $ 14.00
Electronics
Third Coast Pinball
$ 149.95
Boards
Allteksystems
$ 12.00
Electronics
Third Coast Pinball
$ 149.00
$ 5.00
Playfield - Decals
Doc's Pinball Shop
From: $ 155.00
From: $ 10.00
There are 461 posts in this topic. You are on page 9 of 10.

Hey there! Got a moment?

Great to see you're enjoying Pinside! Did you know Pinside is able to run thanks to donations from our visitors? Please donate to Pinside, support the site and get anext to your username to show for it! Donate to Pinside