Quoted from barakandl:I asked the same thing. Ace indicated he was only interested in the green aspect of recycling the old boards. The post where I ask is buried back in this thread somewhere.
Well, I'm green with the conversion boards. But I am and had always planned to work on a FULL DISPLAY KIT since I started screwing with displays back in 2010. Just posted a sneak peek not to long ago here of what I'm working on. Targeting end of April on a full display kit if all pans out. They'll be priced right, they'll look nice and work well.
Quoted from barakandl:A pcb that mounts like Hudson Arcade's idea would be simple to make. But I guess there is enough repro options out already the market is pretty tight. I havent added up the BOM or looked at what kind of off the shelf LED digits are available.
BOM and component sourcing is one thing, but the absolute easiest thing to do with PCBs is just to take what's out there that's already been done and assume it's a good design -- throw it into CAD and pump out products. That's not my intent here. I spent a lot of time load-testing on Bally machines before I did these conversion kits because there was speculation on whether the Bally machines with 3A regulators could handle the additional load. Along with helping me work toward an efficient circuit with old-school tech for the conversion displays, it opened my eyes to some things -- mainly that even popular aftermarket displays sold for years have some design issues or just aren't as energy-efficient as marketing hype claims.
Some examples..
BRAND "A" DISPLAY
Transistor gain issue that causes very high current usage per display (over 250mA per display). Brightness resistors also very low to pump brightness to the max since the displays are multiplexed.. again adding to high current usage. Best to not connect this one at the bench with signals faked without a current-limiting supply or you'll burn the leds out. Need the aid of the multiplexing while in the machine to pulse the current and prevent burn-out.
BRAND "B" DISPLAY
Brightness resistors set more reasonable here, around 24mA per segment, but the source driver IC can supply only 70mA max across its rails. This means the more segments that are lit, the less current available per segment. 7 segments lit = 10mA max per segment. Display is pretty energy efficient because it's limited by the IC max current, but also dimmer because of the current-limit occurring.
BRAND "C" DISPLAY
Energy efficient if set at a reasonable brightness, but brightness turned fully up and it's measuring 125-150mA per display. Depends on the user to adjust display brightness down a bit to achieve better efficiency.
Even some of the kits out there aren't that great IMO. Recently noticed that one of the kits seems to be an exact-match for another company's board layout/design.. so unless there's private labeling going on there it explains a few things. Interesting though in that it appears there are no current-limiting resistors for the led segments & if that's the case, the design relies on lowering voltage to the leds & multiplexing to "regulate" current to the leds. I've seen discussion of these types of circuits where people want to eliminate extra resistors & unless brightness is set around the forward voltage drop of the led, it'll be over-driving components and/or the leds. And voltage is not locked just above forward voltage drop, since it's a "feature" to allow the end-user to adjust brightness. Hard to say how that type of design holds up. Another kit just uses too low a resistor value to remain all that efficient, but I suppose the linear regulator can handle an extra 700-800mA as many displays out there are adding somewhere around 0.75A-1A load to machines.
My point is, people are building what they see enough of the time.. not questioning it, not trying their own designs, not understanding circuits & the variables involved. If someone designing displays is asking why WHITE/BLUE leds will not light as bright as AMBER, that's "LEDS 101" -- basic stuff, that if you designed an entire display by yourself you would know right off the bat. There's voltage drops over a number of components and it's all part of the calculations for an efficient design or to explain why setting brightness a certain way isn't resulting in the expected current usage. It's pretty clear to me enough designs in PCBs are just getting copied and thrown into CAD without much thought. And again, that's the easy part. Reminds me of bad code that gets copied over and over, pulled off Google and used in mission-critical apps inside companies, because enough people just grab code and don't try to understand what it's doing.