I'm in the club! Thanks, @pinbackpacker. I see you got the magnet fixed.

There's a good number of things I want to add (plastic protectors, Titan rubber, etc.) and some tweaks to make (update the code, slingshot adjustments, scoop power), but I'm excited to really explore this game.

20170804_064255 (resized).jpg

Is this something that he sells or a one-off? It doesn't look too bad to do myself, but if it's available for a reasonable price, I would much rather save the time. I was planning to put an OCD board in that area, but I could print a different bracket for it. I did get the topper with mine and it would be nice to have it synchronized. I'm not a big fan of the random cycling, and I don't even power the one on my RZSSI.

Yeah, I'm not a huge fan of toppers, but the AMH one is pretty awesome. I say that, but I now have them on 5 of my 8 games.

Thanks. For some reason, I thought you owned one in the past.

Awesome. I hope you get to play yours more than I do. I need to invite people over more often because that's the only time I play it.

Code upgraded from v15 to v23 - check
Fixed optos with the help of an email from Christopher at Spooky Support at 11:07 PM - check
LED OCD board installed - check

I was getting a lot of theater shots bricking off the end of the ramp, so I printed my take on Betelgeuse clothes pin.

20170820_072723 (resized).jpg

I do also have a question. Once in a while, the hellevator starts going up and down on it's own, and sometimes it scores a lock when the ball rolls under the ramp. I've taken the ramp loose, and the switch seems to work correctly when worked by hand. I've tried to go into switch test right after the hellevator goes nuts, and the switch is not locked closed. I can also time a shot so that it fits into the hellevator and the up and down will stop for a while, but reaching under the ramp and working the switch by hand does not cause it to stop. Any suggestions on what to check? Maybe the nuts holding the switch are too tight. Maybe it's nothing to do with that switch.

20170820_073007 (resized).jpg

I think I have another issue solved. I could light Prison Lock, but not actually complete the lock. What I found was that the upper subway switch wasn't being triggered. The switch worked, but wasn't being pressed.

At least on my game, the subway is much larger than the ball, and the switch slot is off center. Mine is AMH #1, so maybe it was improved later in the run. Mine looks something like this:
Fullscreen capture 8212017 33858 PM (resized).jpg

If the ball was riding the right edge of the subway, it wouldn't trigger the switch. I messed with adjusting the switch, but eventually ended up printing some ramps that clip on the edge of the subway. Like this:
Fullscreen capture 8212017 34135 PM (resized).jpg

Now, I can start the Prison Ghost.

I still have to figure out why my hellavator sometimes starts going up and down repeatedly. I added some test cables that would let me play until the issue happened, the disconnect the switch, making sure the switch wasn't still closed. This did not stop the cycling. I'm wondering now if it's just binding a bit and slowing it down to the point that it times out.

I feel like I'm talking to myself here, but if it helps someone in the future, it will be worth it.

The up and down elevator was being caused by a ball hitting the back side of the switch actuator under the ramp. This made the game think a ball was in the elevator, but it never hit the subway switch. I made yet another 3D printed piece to block the back side of that switch.

20170822_200453 (resized).jpg

The ball would normally just trickle out of the scoop, but occasionally it would rocket out so fast you had no chance to do anything with it. I measured the scoop and found that the vertical channel was a good bit deeper (front wall to back wall) than necessary for the ball to fit. The ball was riding close to the front edge and hitting very near the top of the curve, killing it's momentum and breaking the weld on the scoop.

Fullscreen capture 8232017 104329 AM (resized).jpg

As a test, I printed a shim that I stuck to the inside of the scoop with VHB tape. This shim forces the ball toward the back of the scoop where it rides the curve better and makes the kickout much more consistent. I'm not sure how I feel about the tape, but it's tough stuff and it has stayed in place for a couple of weeks now. What I'd really love to see is a redesigned scoop (Kerry?). This seems to be working for now, though, and since it's a home use game, I can keep an eye on it.

Fullscreen capture 8232017 104742 AM (resized).jpg

I get that occasionally as well, but not often enough to make it a high priority. I think I read that putting two rubbers on that post, one that fits correctly and another that is larger and fits over the fatter part of the post, would help. I plan to give that a try when I get my plastic protectors to install.

So, I tried this and on the second game my wife had a ball stick there. It's easy to shake loose, but still a bit annoying. During the same game, she had a ball stick on the right inlane switch, and it machine gunned a sound. My wife and son both looked at me like they were afraid the game was going to explode. I just told her to shake it loose.

I also added a mod that is really only a proof of concept at this point, mostly because I don't have access to v23 code. Using the v22 code, I added knocker support and installed a knocker.
20170821_201309 (resized).jpg
20170821_201503 (resized).jpg

It actually works in solenoid test in v23 as well if you choose the correct "UNUSED" selection. I sent Ben a link to my code, but he says he doesn't plan to release new code at this time. I'm really hoping he'll make an exception for a small feature addition.

Does anyone have suggestions on how to avoid airballs from the theater ramp? I have set my flipper strength to 0, and it still plays well, but if I get a really good shot on the theater ramp, the ball will completely jump over the landing ramp and hit the side of the cabinet. Anyone put a roof on the landing?

I could 3D print one as well, but I think that clear PETG would be better if it could be shaped right. That said, if a 3D printable was available now, I'd use it until something better came along. Every couple of games, I get one that hits the cabinet. Often they end up riding the gap between the ramp and cabinet wall, fall off the end onto the plastics, then roll all the way out the outlane.

I'm surprised you let that bother you. It was obvious to me that you were not changing the game from how it was supposed to play.

I have most of them whipped now, and the airball is the last big issue I have. I can usually get by with it, but when it drains out the left outlane, it really is a bummer. I agree that it is a great game. I traded a Scared Stiff for it, and I don't regret the decision.

It's not pretty, but here's my attempt at getting something to work. I'm curious what mixed results you had. This one seems to work well, but I haven't put a lot of games on it yet.

IMG_7967 (resized).JPG
IMG_7968 (resized).JPG

One other ramp related problem I was having was at the transition from the sheet metal section to habitrail. The habitrail is inset, and if the ball was traveling fast along the left wall, it would hit the rail and cause a big hop. Often, it would jump completely out of the habitrail and go out the outlane. Another ugly attempt at a fix. Hopefully, if these work well, they can be made nicer looking.

IMG_7969 (resized).JPG
IMG_7970 (resized).JPG

All of the AMH pictures I looked at, as well as my RZ, have a different shaped bend at the upper end of the habitrail than my AMH shown above. I decided to bend mine to match, and it appears to have fixed the hop I had previously. I was able to get rid of one of those ugly 3D printed parts. I have no idea of whether this will help anyone else, but here's a pic, just in case.

20170903_190228 (resized).jpg

Still working on this thing, trying to make it perfect. The scoop really fires the ball out super fast, so I tried to drop the coil power. This resulted in balls getting stuck if two or more were in there. I thought about the geometry a while and came up with this. I dropped the subway to make the balls sit nicer so that the coil could get the bottom one out easier.

20170924_071459 (resized).jpg

Here's a diagram to show the before and after. I couldn't get a good picture since the playfield would have needed to be down.

Untitled (resized).png

With this change, I can now set my scoop at 1 and still not have trapped balls. 0 works for a single ball, but still won't work if there's two balls in there.

I did write some code using v22 that increases the scoop power for one hit after a failed attempt, and that works nicely. Unfortunately, that's not an option for v23, and I'd rather have the v23 improvements.

The Hellavator servo burned up during a game last evening; what a stink! Luckily, I had an extra that PinBackpacker included with the game. I swapped it out and promptly set a new GC score. Woohoo!

I ran into a pretty aggravating bug last night. I shot a ball that glanced off the target bank, starting a minion battle, then it went into the VUK, starting the Doctor Ghost. The target bank stayed down, and I couldn't progress through the Doctor Ghost. After I failed Doctor Ghost, the bank went back up on it's own, and all was well.

I bought some through ebay. The seller I bought from is almost out, though.

ebay.com link: NEW 1 10 AXIAL YETI SCORE TACTIC TSX45 METAL GEAR STEERING SERVO WRAITH RUBICON

Unfortunately, there is no distinction between speech and sfx in AMH. There IS a setting for music vs speech/sfx.

There is also code in the AV kernel (at least in v22) that automatically decreases the volume when multiple sounds are playing simultaneously. It only does this for the later played sounds, though. There is nothing that you can change about this short of building new AV code.

Most likely the connectors for the optos are not making good contact. Those 0.1" connectors are very prone to this.

My scoop was not working very well at all, and I was tired of tweaking on it. I saw in Crile1 's FS post that he had upgraded the scoop, and that later in the production run, Spooky had upgraded them as well.

So, I bought the scoop from Pinball Life (https://www.pinballlife.com/index.php?p=product&id=4579), and it is working great mechanically. The one issue I have is how to connect the switch. The new scoop comes with optos. Do newer games use optos, or do they still use a mechanical switch? Anyone willing to take a picture showing how it attaches?

I ended up removing the opto and printing a bracket to mount to the same holes that held the opto. The holes in my print were not quite the right size for the screws, so I used glue for now. When this falls apart, I will adjust the hole sizes. For now, though, it is working great, and the scoop is like night and day better than the original.

20180428_095221 (resized).jpg

20180428_095305 (resized).jpg

Hey, guys. It's me again. I just thought I'd share what work I had done on the colorization of the DMD files in case it was useful to anyone.

To use these with a real game, you'd have to contact Spooky, buy a display, hack it into your game, and load my hacked up firmware. It's not an easy project, and it's certainly not a finished product.

For virtual pinball, I believe the animated GIFs will work by just dropping them into the project, but I don't really do virtual pinball, so I can't say for sure.

I created a really crappy webpage to host the files:
http://ledocd.com/amh_color/index.html

Someone correct me if I'm wrong, but aren't they just standard blue caps?
https://www.pinballlife.com/index.php?p=product&id=327

I tried to post a few samples yesterday, but I just figured out that I had to post original size to keep the animation.

K0B.gifH2A.gifH4A.gifBAA.gifB0W.gifQA3.gifSAB.gif

I re-crimped the connections on mine, and it has been flawless since.

You can use an IDC insertion tool to make sure the wires are pressed into the back side of the connector well. You can often get by using a small screwdriver, but it's better if you have the tool. You could also choose to use to replace them with Molex crimp connectors rather than IDC, or even direct wire like Tallon mentioned. Direct wiring them, however, it gets very difficult to work on because everything is tied together, unless you also add connectors further down the wires like Tallon also mentions.

Not likely. It requires new display hardware that I bought from Spooky. It's the same display that they used in RZ, but since they are no longer making games like that, they probably won't have many of those displays around any more. I would also think that there are not enough AMH games to make another run of controller boards worthwhile.

Once you get past the hardware, I had to modify the AV firmware to make it work. Charlie has said that he's OK with that, but there is absolutely no support for any of this, so you'd be completely on your own if you run into problems. The firmware also still has minor bugs in it.

I had to know, so I looked. It's actually worse than that. One is standard blue. One is smoked acrylic. The third has no cap.

20180708_143420 (resized).jpg

It's hard to see from the picture, but the screw is in the normal location. The cap is not a full circle, though. It has a cutout for the ramp.

Probably a good idea. He has been very helpful with that sort of thing in the past, and I'm pretty sure they do their own plastic cutting. If they don't already have one laying around, he could probably cut one for you.

You will need the 0.100" size, but other than that, yes. I'm not sure if Pinball Life sells one or not. I have the one in the link below, but it's overkill if you're only going to use it a few times. I think Pinbits used to sell a cheaper one that was a metal tip with wooden handle.
https://www.digikey.com/product-detail/en/molex-llc/0638133504/WM5151-ND/2327986

It's hard to tell from the picture, but that connector doesn't immediately jump out at me as being bad. The plastic look at bit dinged up, but the plastic is only there to hold the contacts in place.

The red wire does look like it's stripped back a bit far, so I'd probably pull it from the connector, cut the wire flush and recrimp it. I don't think it's very likely to fix your problem, though.

If you really want to replace it, you can use this IDC connector:
https://www.digikey.com/product-detail/en/te-connectivity-amp-connectors/3-640440-2/A31080-ND/698278
or this crimp and stuff connector plus contacts:
https://www.digikey.com/product-detail/en/molex-llc/22-01-2027/WM2011-ND/171991
https://www.digikey.com/product-detail/en/molex-llc/08-50-0114/WM1114-ND/26475

Assuming it is consistently triggering, my suggestions would be:
1. Some cameras can see the IR light coming from the IR emitter. My Samsung S5, for example, can. The emitter should be lit all the time. If you know your camera can see IR, that helps, but even if you don't know, you can try. If you do see the light, you know that half is working.
2. If you have an IR flashlight, you can shine it into the receiver to try to get it to stop triggering. If that does get it to stop, you know the receiver half is working.
3. Be sure to check all involved connectors. There are connectors that run up to the backbox as well, if I remember correctly, so be sure to check those. If it is consistently triggering, wiggle one connector at a time to see if that single connector can get it to stop. If so, concentrate on making that connection better.

Yes, the target bank is up at the start of the game. If you are getting false hits on the optos, it can cause the target bank to act weird. I believe the target bank is a symptom, not the cause.

You asked what I would do, so I would start by using my phone to look for IR light coming out of the emitter. Some phones won't see it, but my Samsung S5 does. Look for the faint purplish light.

Next, turn on switch test to see if the switch is triggered. Mine is not, so it shows as an empty dot until I block it with my finger.

I'm not necessarily suggesting you do this next step, but it's what I would do. I used a small screwdriver to short across the opto receiver contacts to simulate the opto being closed.

Any of these steps can help to cut the problem in half by eliminating the emitter (or not), showing the opto working in situations (or not), or by showing that the receiver circuit works (or not).

I sent this as a response to PM, but thought it could be useful to someone in the future, so I'm posting it here also.

In switch test, there is a dot that corresponds to each switch. You can see in my video, there is one that is displayed when I block the loop with my finger. If that dot is always displayed, the loop is always triggered. If it's never displayed, it's not being triggered at all.

You won't hear any audio unless you turn it on using the enter button. So, I guess it's not clear which case you have. I would guess always triggered, as if a ball is always there.

That could be caused by either the emitter or receiver. As Ben mentioned, if your door opto works, it's much less likely that the emitter is the problem. Both emitters are on the same board and share a power source. It's still possible you could have a failed solder joint on the ghost emitter, though.

The emitter is always on if the game is on. In my video, you can see it come on a soon as I power the machine on.

I haven't had any problems with it recognizing balls. I did have an issue where balls under the ramp would hit the back side of the switch and cause false hits, but I fixed that long ago. Make sure the lever piece that sits in the bottom of it can move freely and that the switch doesn't have to move far to trigger.

I thought it just picked one at random at the start of the game. I could be wrong, though.

Probably a wire has come loose. Look up the lamp matrix to see whether those lamps have a row or column in common. Track the common wire back toward the backbox until you find the disconnect.

I got tired of having a ball stuck in the pops every few games. It's too early to say for sure if this will fix it, but I made a wonky star post to try. It makes the ring slope to the left where the ball sits.

20180922_102117 (resized).jpg
20180922_115558 (resized).jpg

We'll see. There's not much of a moment arm, and the post resists turning more than you'd think, but it is in the pops.

I'll certainly post the file. I guess I could add it to my website, but it would be more to ship than the post would be. I'd really want to make sure it works first.

Yeah, it's the Prusa i3 MK3, built from a kit. Much nicer than the Maker Select v2 I used before.

Pretty well. My bed was off by up to 0.6mm from one corner to the other. The leveling was good enough to turn that into a good first layer. Since it only tests 9 locations, however, you can end up with high or low spots between the test points, and there's nothing the leveling can do about it.

I'm not usually concerned with the surface finish of the bottom as long as it stays stuck to the bed. If you were, you might be better off using a glass bed.

I've played about 30 games, and have had no trapped balls. I'm still not convinced it's perfect, nor that it won't spin eventually, but I think it's good enough to share the file. If you have a 3D printer, you can print this:
http://ledocd.com/files/AMH_Wonky_Star_Post.zip

If you don't have a printer, I would prefer to get some more time playing with it installed before printing and shipping them.

I'd estimate I've put 50 or so games on AMH since installing the wonky star post, and I've had zero stuck balls. The post has not turned out of position. So, if anyone wants one and doesn't have access to a printer, I'll print you one. Just pay shipping.

http://ledocd.com/cart/index.php?main_page=product_info&products_id=35

Anyone who wants one, please use the link below. It automates things like calculating shipping and printing labels. Thanks.

http://ledocd.com/cart/index.php?main_page=product_info&products_id=35

In multi-player, does anyone else's game give players 2-4 5010 points when the ball is kicked into the shooter lane? I'm wondering if I need to check for a touchy switch or if there's a bug.

It's actually doing it on every ball after player 1, ball 1. I found some code related to GLIR handling that I think might be causing it, but I'm curious whether other games do it, or if it's only mine.

Confirmed, it is the GLIR handling code. It's supposed to score when the ball rolls over one of the GLIR switches. The same function is used to check whether GLIR is complete, but it scores 5010, even if it wasn't ran due to a switch hit.

This really makes me sad. I hate to see hard feelings between you guys. AMH is one of my favorite games, and while the code is great, there are a few bug fixes that could be released.

I've posted earlier about my color display project. Using the stock Chroma-Color board, it was limited to 256 colors. I got some motivation, and modified the FPGA and AV firmware to allow indexed color. It's still a 256 color palette, but I can choose that palette for each video scene instead of being limited to a single 256 color palette that was predetermined and locked into the FPGA.

Here are a few samples of the improvement:

AT0.gifAT0.gifB0Z.gifB0Z.gifL6A.gifL6A.gifL9A.gifL9A.gif

I have had people ask, and the biggest hurdle to making this available to other AMH owners is availability of the Chroma-Color boards. Charlie says he only has a couple left, and he's keeping them for RZ warranty claims. I have gotten quotes on the boards, both from Macrofab and the factory that builds LED OCD, but the board is licensed Creative Commons, and I can't resell them unless Parker were to OK it.

The FPGA firmware, I think, is the same situation. The AV firmware was open source at one point, and I'm not even sure what the license is on it. So, in short, it would need a lot of approvals before I could even share the code.

It's working well for you? I know I haven't had a single ball trapped there since installing mine, but that's only one machine.

If it were only up to me, I'd gladly push forward and make this happen. It's not, though. The original designs are not mine, and I don't have permission to use them to make a batch. Specifically, the board and FPGA code says the design can't be used commercially. Even if I were to coordinate a group buy, it's still a gray area. Unless Parker says it's OK, I'm not going any further than my own machine.

Also, it's not really a "done" project. It works, but I've only colored about 75% of the videos. I got totally burned out doing that part of it. The technical parts, fun. The coloring, not so much.

It depends on the quantity. The quote for 25 boards from my normal fabricator was about $70 each plus shipping, etc. If part of the deal was that I use Macrofab, those were a bit more. It would probably have to be in a near kit form, including the display panel and bracket. Charlie (Spooky) sold me the one I have for $250. I also had to change the DMD protector from green to clear and build a power cable.

I would love for it to work out that way.

Who hasn't played Bumps in the Night? I hadn't until today. I normally go for defeating ghosts, and try to avoid minion shots since they're so dangerous. Today, I concentrated on getting to BITN for a few games. Wow, what a fun mode!

I know boustrophedonic has seen the thread, but for anyone following this thread that may miss the other:
https://pinside.com/pinball/forum/topic/color-display-for-america-s-most-haunted-it-s-happening

Basically, I spoke with Charlie, and I have the green light for a single run of boards. This is the only time I'm going to mention it in this thread, so please try to keep responses there.

If you want one, I would suggest getting on the pre-order list as mentioned in the other thread.
https://pinside.com/pinball/forum/topic/color-display-for-america-s-most-haunted-it-s-happening

I may order a few more than the pre-order quantity, but getting on the list would guarantee you have one with your name on it.

Open the coin door. Look on the right side. There should be a sticker. The last three digits are the production number.

20190316_093216 (resized).jpg

That's cool!

Does your ghost move? If not, the servo is likely dead. Servos burning up is a common fault. If it DOES move, there is not a center position adjustment, only a min and max. You would have to disassemble and center the ghost mounting adapter on the servo hub.

Likewise for the other servo controlled stuff. Only the endpoints are adjustable.

I'm really glad it helped you, but I can't take credit for the scoop replacement idea. Crile1 did it before I did. It really does make a great game work as well as it deserves.

It's a mini-wizard mode. It's also VERY cool, so try to get there if you can.

« Spoiler! Click to reveal.Light Multiball Complete, Photo Ace, and Minion Master to start it. »

That's weird. Your AV firmware is different than nearly everyone else since you're running my Pin2DMD custom build. What PIC version are you running? I really don't know if an older PIC code with new AV would cause scores to not save. If you don't figure it out by this weekend, I can try installing my Pin2DMD and test that code. There are really no changes involving high scores vs AV v23, though.

Maybe try resetting high scores via the menu. There isn't any setting to prevent them from saving.

Yeah, firmware version is the PIC. Since you have the Pin2DMD display, you should have v23 and v91 for firmware and AV.

High scores are stored in the EEPROM.

JeffF

Oops! The high scores not being saved WAS my fault. I have put an updated version of firmware on my website that will get it working for you. You won't see the firmware version anymore, and the first video sequence will have messed up color, but your high scores will work.
http://ledocd.com/amh_color/

Long version:
The Pin2DMD splash screen takes longer than the Chroma-Color startup. I didn't want to ask lucky1 to change that.

By the time the splash screen is done, the firmware version screen has already finished being displayed, and the first video sequence has started. The Pin2DMD firmware missed the first palette, so the first video will have incorrect colors.

So, in the v91 firmware, I made the firmware version screen display longer. This fixed the above issues, but it also meant that when the PIC tried to read the high scores from the EEPROM, the AV firmware wasn't ready. I will spend a bit more time to try to properly fix this, but for now, v92 will get you up and running. Sorry for the trouble.

...and the version posted now (v93) does fix this. It's still possible that the weird palette could happen for that first video sequence, but I tried 10 startups and it worked fine.

Sure. I'll post it this evening.

http://ledocd.com/files/AMH_Jump_Shim.zip

Whoa! That post on the left of the scoop must make it really tough to hit. Lots of times mine bounces in from the side.

I do think it makes a big difference, but I don't have any of the Spooky boards left, and I don't plan to build more. What I do have is some custom code that increases the PWM frequency by changing how the rows and columns are ordered, call it OCD-lite, if you want. It doesn't use any extra CPU time because it's only reordering the row/column transitions.

The custom code that I run on my game has this alternate driver as an option, among other enhancements. I promised Ben I wouldn't distribute it, though. The code for the lamp driver is posted somewhere around here, and you could compile it into the v22 code if you wanted. The v23 code was never officially released.

The firmware release IS Ben's official release. It's just that the source wasn't released along with it.

It was really designed more for RZ than for AMH. It plugs directly into the board, and the connectors are placed differently between AMH and RZ. You would have to add cables and Z-connectors to get it to install properly.

I'm glad to be able to. I agree that the ramp is no problem with the "clothespin" installed. When the shot is hit solid, it makes the jump. I even had to put a roof on my landing area because the ball would sometimes take flight and hit the left cabinet wall.

Betelgeuse did it first using an actual clothespin. I copied his idea using a 3D print. It adjusts the end of the ball guide and also extends it a bit to avoid having the ball hit the end of the ramp wall.

fc08c24b103b4137aff1ca658d85b145d4219157 (resized).jpg

Good luck. If you are willing and able to assemble a board with SMD parts, I'll give you a bare board. It would still need programmed, though.

Nice, creative install. It looks much nicer than mine.

If you've already used the software to tweak the lower end brightness, I'd say the difference is minimal. I almost always change the bulbs, myself, but I also like to play with color choices.

Granted, this is the REV7 design, but the lamp matrix really did not change between the two:
https://github.com/LonghornEngineer/Pinheck_Pinball_System/tree/master/Hardware/Pinheck%20Main%20Board

Which bulbs are the problem? This is what I have for the column that includes the scoop flasher:
61,Upper Left Flasher
62,Hell Flasher
63,Scoop Flasher
64,Photo Hunt
65,Army Ghost
66,Bar Ghost
67,Spirit Guide
68,Basement Camera

Have you tried single lamp test within the game or test mode using the LED OCD software?

I also have a Siegecraft lamp matrix tester and harness that is great for determining whether the driver or wiring is at fault. If I could be assured of getting it back, I could lend it to you.
http://www.siegecraft.us/presta/index.php?id_product=11&controller=product&id_lang=1

It's a great tool and well worth the money. I've used mine to test thousands of boards, so I've definitely gotten my money's worth, but even for tracking down a single lamp matrix issue, it will pay for itself in time saved. The only issue is having to build your own harness for it, but judging by your installation, you've got that whipped.

Hmmm... Something's wrong there. If a lamp matrix diode is installed properly, it needs the lamp to complete a circuit. It should look like this:

Column wire ------- ->|- Diode--------Lamp-------- Row wire

The diode needs to be used to prevent the reverse flow of current through the lamp matrix. If you install it like you would for "normal" flasher/coil circuit, it won't work correctly for this lamp matrix flasher configuration.

I was going to take a picture of mine, but when I opened it up, I found that mine is attached to GI4 rather than the lamp matrix.

Then, I looked at the code, and sure enough, it is using GI4 to control that flasher.

Anyway, here's a picture of how mine connects.

20191027_110204 (resized).jpg

That is correct. GI4 lights my scoop flasher.
Lamp #42 lights nothing.

Mine is production #1, and I have found some weird wire colors on mine as well.

I spoke with him a few times when he was planning to start on it. He wanted to send me a sample scene to confirm the process, but I haven't heard anything from him since then. I suspect life happened.

If I knew it would get the project moving again, I'd buy a drive and ship it to Ben. Hell, this one is only $48.
https://www.amazon.com/dp/B06VVS7S94/ref=twister_B081ZV4WBZ

Eh, I feel bad about running out of steam on the coloring, plus I want it for my own game. A drive is on its way now.

Whether Malenko decides to take over or not, I figure it's more likely to happen, regardless of the artist, if I help it along.

Getting involved in this as a courier has also rekindled interest in a project I had been thinking about. The audio in the game has always seemed a bit "off" and had a good bit of noise in the background. I think I even heard Ben mention once the possibility of using an external DAC to improve over the Propeller's internal one. Well, that's what I did.

I bought a cheap PCM5102 DAC board from Amazon, hacked up an old SD card adapter to get access to enough I/O to control it, and wrote some firmware. I'm pretty happy with the results. It's still a hackjob, and I'm trying to decide how much work to put into cleaning it up.

20200319_183507 (resized).jpg

You can kinda tell from video how much it has cleaned up the audio, but I think the stock audio sounds worse here than in person.

...and here's the audio from the external DAC.

Looking at prices, I can barely buy the parts for what the DAC board costs. I decided to just make an adapter board and use the DAC board as-is.

Starting with an open source Sparkfun design and modifying for our needs, these are probably the easiest board I've ever designed. They're also really cheap at 3 for $12 at OSHPark. I should have this order in a couple of weeks. After testing, I'll have a couple of extra, and I'll share the design on OSHPark so that others can order direct from them.

dc2e4a26e901665e0b4698b6b3a535f8 (resized).png

e2f1f154227b9dbc5113657cd6224e32 (resized).png

Ah, crap! I was just looking at some of the files on the drive from Ben, and I noticed in a picture that the second SD card slot wasn't populated on the board.

At first, I thought the picture might have been an early board, but it seems like all of them in this thread, other than mine, are also missing this card slot. The board in the previous post won't be as easy for you guys as I hoped. It should be great for me though.

The DAC should still be workable, but I don't see a way for you guys to use it without having to solder something to your board. It could be the SD slot, or wires attached to the pads and staked in place.

Yeah. Basically, I have a "left justified" driver running on one of the unused COGs, and ROY_DAC feeds samples to both the Propeller DAC and the other COG. I was just barely able to get it to run at 22kHz by highly optimizing the code. I'll send it to you.

Agreed. It would not be hard to solder wires to those pads, you just want to make sure to stake (glue) the wires so they don't end up pulling the pads off the board. I only used 3 of the 4 signals, and I'm getting 5V from the 5V connector on the upper right of the PinHeck board.

I do not plan to sell anything, but I will post instructions on how to do it. Keep in mind that it is not as easy as plug and play. As mentioned above, you're going to have to solder directly to the PinHeck board. That is, unless you're lucky enough to have the unused second SD card slot like mine does.

I hope you got this one. It will make things much easier for you since it will match my instructions.
https://www.amazon.com/HiLetgo-Lossless-Digital-Converter-Raspberry/dp/B07Q9K5MT8/ref=sr_1_1_sspa

Mine is AMH #1, and I've run into a number of minor differences with it.

There are two different code files used in AMH. One is game rules, the other is the AV kernel. What we're talking about here is the AV kernel.

Neither of the latest is actually open source, but Ben has been gracious enough to allow me to change the AV kernel, previously for the colored DMD mod. Now that I'm writing it, I realize he hasn't blessed the DAC mod, so I probably shouldn't post a build unless he does.

To make things a bit more convoluted, there will have to be separate builds based on which DMD is in your game; the stock XPin, my colored DMD mod, or a Pin2DMD.

Ben has not approved of any changes to the the game code, so you would continue to use the released v23 for that.

Ben did tell me "You should totally make a kit of this.", so I took that as permission to share the custom AV kernel builds. JeffF installed his DAC today, and he seems happy with it. His game has a Pin2DMD, so that tests that firmware build. The one remaining untested is for a stock XPin DMD. If anyone would like to install the DAC in their game and it still has a stock display, send me a PM.

Now that we know my game is likely one of only a few with the SD slot populated, this may be of no interest. It's gotta be better than more bad news, though.

Because the only option from OSHPark was 62 or 31 mils thick, I had to get a 31 mil board and shim the thickness a bit. It works great, though.

20200402_155944 (resized).jpg

20200402_161905 (resized).jpg

20200402_163139 (resized).jpg

I've noticed a lot of little things that I couldn't hear through the game previously. I have listened to the files on a PC before, and that's just as good, maybe better. It is still all 22kHz audio, so there are things that are a bit rough. I notice it most with S's. Sounds like ssstatic or a sssnake hissssing.

Yeah, I was thinking about it yesterday while installing my faux-SD. I've made castellated boards a few times before, mostly to use an IC in a smaller package on a larger footprint. Luckily, I'm still working a regular schedule from home, so not a lot of free time. That, and there's not as much motivation with my game modded and cleaned up already.

Now that all three cases have been tested, I suppose it's time to share with the masses. I threw together a page on my website with instructions and the necessary files.

http://ledocd.com/amh_external_dac

Not readily available, as far as I know, but I do at least have the files to remake them. It wouldn't be cheap, but it would be better than a dead game.

I just couldn't get this idea out of my head, so here it is. It's untested, so use at your own risk.

https://oshpark.com/shared_projects/GgMSjTo4

AMH_Top (resized).png

Must be the same compulsion that made me design it, even though I can't use it.

The new board is just a slightly easier way to install the DAC. You would have to solder this board directly to the main board, rather than wires. If you're nervous about it, don't do it. This is not a case where you want to be unsure of your abilities because the PinHeck board is not easily replaceable.

Pinheck_Carrier (resized).PNG

It would look nicer with an integrated board. Why I decided to stick with an interposer board rather than putting the ICs on it:

-The design would have taken a little more time.
-Assembly would have been more complex. I can hand solder TSSOP parts, but most people would not be able to.
-Using someone like Macrofab for assembly would make it more expensive. You'd also want to gang all of them into a single order for quantity discounts.
-The DAC board is available for less already assembled than I can get the DAC IC for.
-No matter how it's designed, it will still require soldering directly to the Pinheck board.

Doing it this way means that anyone who is capable of installing it is also capable of assembling it, the cost isn't increased unnecessarily, and I don't have to get involved in buying and reselling anything for others.

Using the files that Ben sent, I added some color to the War Fort videos. I still hope that someone better at coloring than I am will be able to finish the other modes and improve on what I've done, but at least these are something.

Update files for owners with color displays are here:
http://ledocd.com/amh_color/

WGC.gif

There's certainly nothing all that impressive here, but in trying to make the most of my Adobe trial (ends tomorrow), I sprinkled color on Prison mode as well.

New files are posted on the website.
http://ledocd.com/amh_color/
P9A.gif

...and just one more. The main attract video was bugging me. If you downloaded last night, you have the option of getting it again.

AT0.gif

Sigh. It really seems like updates fail way more often than they should.

After emailing you, I downloaded the files from my website (didn't use the local ones on my PC), purposefully corrupted the AV firmware on my game, and used the recovery program with a newly built SD card. All the files from the website, including the recovery program are confirmed working.

I wish you hadn't jumped into removing the board, since it's very unlikely that would be the cause of any problems. It attempted to update, so it was still able to read the card on the first pass.

Yes, that is the correct board. It says it arrives in August, though.

If you're not very methodical about building the SD card, you can easily trash the contents of the EEPROM that stores your AV firmware. Worse, you need a working firmware to be able to read the contents of the SD card. Once you corrupt it's contents, you can no longer recover it by flashing code from the SD card.

There are instructions on my website that describe how to load code over USB that will allow you to reprogram your EEPROM. See the bottom of this page:
http://ledocd.com/amh_color/

You still have to get the SD card right, though. Format the card first. I've seen a number of people corrupt their AV firmware, and it seems like it's always the same story. I tried to update, it didn't work a few times, then it finally did, and now my game is dead. My suspicion is that people don't get the files in the right places, they keep writing files to the card until they do, but they don't reformat the card each time. If there is ANY fragmentation on that card, you are screwed. Format the card EVERY time you copy files to it.

Worst case, if you can't recover it, I can program a replacement EEPROM and mail it to you. It really shouldn't come to that, though.

If you turn your game on now, does it say SFX ERROR, or does it still show the green dots of static? If it says SFX ERROR, your EEPROM is confirmed fixed, and your card still has issues.

There's nothing magical about a new card, and I would still format it right out of the package. If your old card is truly dead, then yeah, a new card would be necessary.

I'm starting to wonder if using full card images rather than individual files would be more reliable. It's a higher barrier to entry, but it would prevent getting files in the wrong places. This problem is far too common.

Well, I'm just going to have to man up and take my lumps on this one. It was bugging me, so I tried to reproduce what you did, and I ended up finding that I had posted the wrong file for XPin games. Try this one instead. My apologies.

http://ledocd.com/files/AMH_AVKERNEL_PCM5102_DAC_20200827.zip

Do you think it's possible to get a picture or video of this? It might be a bug related to the color display. Occasionally, it does have glitches due to SD card reads failing. I've tweaked that as much as I can, I think. That shouldn't be mode specific though, so if BitN is causing it repeatedly, it might be something else.

There is a known issue where an SD card read can fail and cause garbled fonts and/or colors until the next scene caused a reload. Luckily, it is rare. That also seems to line up with what you're talking about. I've tweaked that code so much that I don't think I can really do any better. The Extreme Pro card does it less frequently than the stock card, but the color code is reading the card so much more often than stock code that it does run into issues, even with a fast card.

Are those sold as AMH art blades? ...and they didn't pre-cut for the LEDs?