Quoted from mikeflan:
Hi folks -
I know, I know - we're talking pinball here, there are no permanent fixes. And if you aren't getting a solid 5 volts to the MPU, you're bound for problems. That said, with the Kahr daughterboard, my problem has gone away. I was getting around 4.7 volts on the MPU without, and a steady 5.013 with.
My question is, rather than spend the $100+ to get the board repaired the way it should, along with the time involved of it being out of the line up and/or off location, will the Kahr keep working? Or is it a stop gap measure/time bomb that will fail sooner rather than later. I'd sure rather spend the $25 bucks on the daughterboard than the c-note plus shipping back and forth.
All opinions are welcome. FWIW, it's a beater of a WCS 94 that has had a rough life. If it can limp along on location making $40 a week for a year or two with the Kahr board - I'll be happy.
I first was alerted to your post late yesterday but watched from the sidelines to let others chime in, but now I'll offer a few thoughts... If you haven't already done so, I'd recommend a review of the technical details on my website: http://www.kahr.us/index2.html . Note in particular the last sentence in the summary section that includes: "The power supplies still need to be maintained..."
To start, the WPC Power Fix is a permanent fix. But remember, my Daughterboard isn't called the "reset fix", so it is most decidedly not a permanent fix for the reset symptom experienced in WPC era games. When installed, it realizes an improved balance of power consumption loads across two of the available power supplies (the 12v regulated and the 5v) and, as long as the Daughterboard remains installed your machine will operate in this improved state. The rebalaning of the loads is necessary because the original designers did not optimally distribute the loads across these supplies - the 5v supply is very heavily loaded in WPC while the 12v digital is minimally used.
The reset symptom is caused when the MPU is starved of voltage because an undervoltage sensor triggers an MPU shutdown. In the machine's original configuration with a heavily loaded 5 volt supply, the resulting reduction in load brought about by MPU shutdown typically allows the 5v supply to sufficiently recover such that the undervoltage sensor releases the shutdown trigger and the MPU boots up again. So, the reset is a combination of two events - a shutdown and a subsequent booting. If the 5 volt does not recover the MPU will not boot and the machine will stay in a shutdown state.
When you install the Daughterboard (or the PSU5, or a computer's supply, or the resistor hack, whatever) that undervoltage sensor is still on your MPU monitoring the available voltage level and ready to shut down the MPU if the supply falls low. In the case of the Daughterboard solution, if your 12v digital supply were to be unable to deliver power the game would shut down. So why does the Daughterboard so consistently eliminate the reset symptom? This brings up the concept of "headroom" with the supplies. The 12v digital is derived from the 18 volt supply. The stepdown from 18v to 12v seems to offer 6 volts of headroom before the 12v would drop from lack of power upstream. I say "seems to" because the math is not quite that simple thanks to the built-in voltage drop in the LM7812 that regulates the 12v and the two worthless diodes upstream of the LM7812, but for argument's sake a 33% headroom offers a robust buffer to fluctuations. Next, in the Daughterboard we are deriving a unique 5 volt supply for the MPU from the 12 v digital. As with the 18 to 12, there is a lot of headroom between the 12 to 5 that gives significant buffer to fluctuations. So, the net effect is a very stable supply for the MPU derived from an under-used supply that is built with a lot of margin for imperfection.
Comparatively the PDB 5v supply construction has very little headroom to accommodate imperfection. The designers could cold have specified transformers with a couple more windings on the 5 volt's secondary connection. That would have raised the upstream side of the LM323 slightly and offered additional error tolerance. But, that comes with a tradeoff because the LM323 is a linear regulator so it would have likely needed a larger heat sink to operate with the higher source voltage and it's already hot in the backbox so you really want to minimize waste heat. Anything you do to the LM323 (hack it, replace it, etc.) has to live within the boundaries created by what the WPC design offers that LM323. So, there just isn't a lot of room to correct for error... and error invariably is introduced as the power supplies age (meaning the typical problems in these supplies - capacitors, connectors, and such).
So, no matter what you do to your machine you are going to need to maintain your power supplies eventually. If you continue to drive your MPU via the original PDB 5 volt supply (modified or unmodified) you are going to need to keep your supply in a very healthy state - like 95%+ of original capacity to keep the reset symptom at bay. With my Daugherboard installed there is a lot more tolerance for imperfections in the supplies... and that can translate into much longer cycles before you have to revisit power supply maintenance. This is what drives the second half of that last sentence I referenced at the beginning of this writeup: "...but no longer do we need to fret over every hundredth of a volt."
The Daughterboard should be in every WPC machine. Can't close without a sales pitch, right? Come get yours at Allentown next month!
-visit http://www.kahr.us to get my daugtherboard that helps fix WPC pinball resets or for my new Skee Display Board Set