Seems like it really depends if they're being used without the foam tape, which most of the time I'd think the foam would be desired since the light from the backbox is otherwise visible. The lines are there initially since they're the exact dimension of the digit & help when lining things up. Then they're just left there after pcb design to help show where the digit goes & also slightly just for the "artwork" since they then make the blank pcb look like something.

I could do something about them, I was messing around with just having a slight corner marking at each of the corners. But to me it takes away from the artwork on the pcb then And I'd really think the majority of people are going to want to use *something* around the digits to block the light, so the lines wouldn't be visible then. Not trying to create any more work than needed for anyone, just looking at what the majority of people might be doing and basing decisions on that.

Tried this, even changing the pad to an oval and minimally increasing the width is creating clearance issues with the previously routed traces. With the number of digits on the board, you have to watch trace size and pad size otherwise it starts getting heavy on the vias pretty quick.

I'm really not against this change at all sometime in the future, just don't want to mess around with something that is going to cause a complete re-route of all the traces right now, especially with PCB orders in the $500-$1000 range. It's just too much to risk without doing a smaller run of pcbs first, but that then delays availability and gets costly as well. I'm still trying to get out of red on this project.. and it looks like I'll be doing that for a bit yet, so it really comes down to not wanting to take the risk for something that is at this point a nice-to-have item and hasn't caused anyone problems yet. If it becomes a problem or I need to re-route for other reasons, I'll definitely keep this in mind. Of course if I could just resize without rerouting all the traces, I'd already have done it It's mainly because of the complete reroute this would entail that I'm going to leave it as-is right now.

I've created a pre-order sign-up (well, really an "expression of interest list") for the 2nd run of PCBs. If you're interested in trying a kit & want to be reasonably sure you get in on the 2nd run of PCBs, please get yourself added to the list. Not asking for payment at this time, but just be sure if you add yourself you're pretty certain you'll be ordering when the materials come in.

Sign-up is at the bottom of this page:
http://www.pinitech.com/products/ballystern_conversion_led_display.php

I'm limited to about a dozen sets of each color right now since led digits are in short supply. This should work better than separate PMs or people commenting on the thread & help me figure out if there's going to be much demand after a few small runs of these

Happy to announce I'm receiving some submissions via the form and I don't need anyone to resubmit lol Thanks guys! Nice to see people putting their Pinside username in there, that makes it easier to contact via Pinside when the materials come in.

Going to be short on WHITE digits for the 2ND round from the response so far. In that sense it's good I just put in an order for more WHITE as there's a 40-50 day lead time on getting new digits in. Unfortunately for me the min order quantity jumped twice now since I ordered the digits.. from 1000pcs to 2000pcs and now 5000pcs. And these digits aren't cheap! So to at least keep things rolling, I just decided to order the most-versatile of digits at this point.. but that's why on the form I'm asking for opinions on native BLUE/AMBER digits going forward. On that note, it's good to see WHITE is popular.. I'm happy with the colors achieved with the filters for sure and it's neat to be able to change them to anything.

Now I just need to figure out how to streamline things and actually make a bit of a profit to keep this going These kits are for sure costly in materials and the materials go pretty quick with the number of digits/components being used.

Thanks again to everyone for posting some pictures of their completed displays. It's always great to see a project or product come to life! And it's pretty cool seeing what different colored displays do for many of these games!

IMO sticking with just the higher ohm value for BLUE & lower ohm value for WHITE is the way to go. But there's a caveat in that some people may want to use the WHITE digits as WHITE (ie. in Centaur) and in that case a higher ohm value resistor is better (or to offer a neutral density filter).

It's not so much a cost savings on not including two sets, as much as it probably is that "what the heck do I do with these extra parts?" scenario.

Yep.. seems the 2N5401s aren't really necessary, nice to know others had the same experience I did. Rarely do those have to be swapped out because of a dim digit. I could omit them entirely or just include 1-2x per kit.

PCBs are one of the more expensive parts -- along with the digits. Now that I'll be ordering a few hundred PCBs at a time, the costs will be slightly less so that'll help some.

Pricing is like fractions of a cent cheaper between 1000pcs and 5000pcs unfortunately. Supplier left me no choice though, it was 5000pcs or nothing. I chose WHITE for right now. These aren't custom digits at this point.. really not sure what the capabilities are there with having a black edge.

What I'm using was bought in bulk. It's not ideal but it was cheap and works okay. I may look into custom-molded foam at some point, but initial quotes I got were PRICEY.

Yep, once the 500x or so WHITE digits I have are gone.. it'll probably be November time-frame when I have more WHITE digits in.

Is there interest in a plug-in PCB (at the J1 header) to test the converted displays? I can either offer wire harnesses that would just light an "8" up.. or do a small PCB that would have more functionality. Probably toggling of the BCD data so you could set any digit.. and an alligator clip (like Cheddar showed with his harness) to connect to the various digit lines.

I'd just like to know it's going to sell more than 5-10x if I do it though It's not like it'll take a huge amount of development effort, just that I don't need to be ordering in yet more Molex parts or SIP resistors that only get used for a single board design.

Amen to that. I got side-tracked into diagnostic equipment products for *years* and although it's been neat to build that stuff, I screwed myself pretty big in not dumping more time into figuring out faster assembly techniques of nvram (the only thing I've done that stood a chance of selling in high volume). Those adapters just took too long to build, thus the $25 per adapter price per adapter for a while. I didn't focus on the assembly "problem" and I should have. My lesson is to spend more time on figuring out how to be stream-lined with products that stand a chance in selling in volume.. and less time creating or stream-lining the really low volume stuff.

I'm going to be pretty choosy with diagnostic tools going forward, I've niche'd myself out of funding over the years just doing that stuff. It's like reinventing the wheel over and over.. and in the end you're stuck with a lot of pcbs and materials purchased in bulk because the market taps out quick. Better I spend some time on once-per-game products versus once-per-individual. That said, if I can find a few products that can provide some consistent funding, I'll continue doing *some* level of diagnostic equipment, but mostly the stuff I need first and foremost.

Here's one, lower amperage on the input.. but also only 30mA available on the output. Not sure how much the plasma displays need, I don't think much though.

ebay.com link: DC DC High Efficiency Voltage Boost Board for 6E2 6E1 6E5 Tube Pre AMP 12V 180V

Another, output adjustable..
ebay.com link: DC DC 45 390V High Voltage Boost Converter Module Constant Current Adjustable

Anyway, I'll not be doing something for testing plasma displays. Not worth the headache... potential exposure to high voltage to the end user anywhere on the boost converter PCB or diagnostic PCB. The market would be very small. I'm also aware of someone that was using a boost converter successfully to test Bally displays until it at some point went up in smoke for no reason.

From current counts it looks like I'll be out of WHITE digits. Still have plenty of BLUE digits at the moment. Doesn't hurt to get on the list regardless, if someone drops out or people order less I'll contact the next person in-line.

I'm pretty sure the 2nd round of these kits will happen around mid-October. I have PCBs due in at the end of Sept, but some other personal commitments in early Oct... and these will take a few days to get ready, I also want to update the instructions a bit if possible. So at this point that looks like the time-frame I'd have kits ready and begin contacting people on the list.. I won't be asking for payment until I have kits ready to ship.

Can you add yourself via the form at the bottom of this page?

http://www.pinitech.com/products/ballystern_conversion_led_display.php

This makes it easier for me to keep track of things & to contact people in-order of submissions received.

Thanks!

If anyone's had issues with submitting via the form, let me know -- there's a SUBMIT button at the bottom that may have been hidden on some web browsers since the pop-up is CSS window and wasn't really made for a longer form. As long as you clicked SUBMIT and saw the confirmation message you should be ok. If you didn't see a submit button or part of the form was hidden from view, then I probably didn't receive anything. I've made some modifications so it hopefully works better.

Try searching for part CD4543 -- should be widely available from electronics distributors, eBay, etc

Normal functioning of modified displays won't exhibit this issue. In a dark room, no segments should be lit (even ever so faintly). Sounds like the parallel resistor method may have been used here, which as far as I'm aware would be the first person that used that method -- so could be a step that was missed or a few resistors were soldered in the wrong place since some of the silkscreen markings on these boards make it hard to know which resistor is which (unless you're looking at color codes of the original resistor).

My suggestion would be to compare resistors on the front for each value resistance. So for a group of 0k resistors, flip the board back and forth and compare the resistor color of the original resistor with the 7x or so other original resistors where a 0k was soldered. Do that for each value resistance. It's likely on the segment driver circuit, so you should only need to compare the zero ohm & then whichever value resistance you used to set segment brightness.

It's also quite likely to be the 4543 though.. those can partially die and do some weird things.

Yep, you're on the list

I wouldn't be surprised if the new 4543 fixes the issue. There's a *chance* it's a solder bridge somewhere, but more likely you'd have a segment lock on or not light at all.. or maybe light when another segment turns on. I had a decimal lighting when it shouldn't have because I had a solder bridge on one board. If you want, email me a picture of the front & back of the board & I can compare quick to my own displays I did with parallel resistors. But yeah otherwise I'd not spend too much time on it until you get a new 4543 in there since that chip can fail all kinds of odd ways and might be your only issue.

Vid, thanks for posting these, I like the format of them! I had plans to do something with showing pictures of the different model boards and identifying components, but you went a step further than what I was going to do by showing with the components depopulated & then writing in the values. Makes it very easy to identify what needs to be removed & dead-simple to repopulate! You're welcome to create that other thread we discussed any time over the next month as is convenient if that makes it easier to organize your pictures, steps, tips, techniques, etc. No rush on that at all though, since I won't be shipping much for the 2nd batch of pcbs until around mid-October due to other commitments I have between now and then.

Couple of notes..

• I'll be changing over to a 47k resistor (instead of the 100k). No reason other than I had to order more resistors, and that 100k value was initially chosen because I had a higher base resistor value at one point. Instructions will mention both as possible values included with the kit for a bit.
• I'm going to go with a single value for the brightness resistors dependent on the led digit color since I'd bet most people went with the recommended higher ohm value for BLUE and lower ohm value for WHITE. Really has nothing to do with cost savings so much that feedback seemed to indicate the secondary value wasn't necessary & when it came time to order resistors, that value turned out to be OOS anyway.

Rev A,B,C should not make a difference there -- at least in any of the other 6-digit and 7-digit model displays it has not. I've mostly seen Rev C DA-300s in pictures posted & with the DA-300's I have (not many).

List of components to remove/replace have been standardized so there's not a massive list of display models, revisions & what to replace. It's simplified to a list for 7-digit Stern, 7-digit Bally and 6-digit Stern/Bally.

No comma/decimal on the DA-300's as the original boards did not include comma circuitry on the component board.

---
http://www.pinitech.com - "Pinball Inspired Technology"
Kits, upgrades and test equipment for pinball machines