Gave this a shot this morning. Sometimes it seemed to change things, but I couldn't tell if it was me pushing on the socket or because I was moving the board and jiggling my clips. I'm going to order some more components today so I'll make sure I toss a 9-pin connector on there to build a better power connection.

I also did a static diode test on U31-U37 and the only "bad" reading I got was on pins 3 and 11 of U33. Looking at the pinout for the 74LS20 it would seem those pins aren't connected and this is to be expected.

One odd thing, with the RAM uninstalled I can more consistently get the Leon test to run. In some of the cases where I could get it to run this morning, I was seeing pin 31 (D1) on the PIAs as low. I traced that back to the chip I replaced at U16 to pin 3 and read it as low, but read a pulse at pin 17, SR3-3, and U15-32. Similar to the problem I was having with D3 a few days ago. However, sometimes that data line DOES pulse at all the PIAs and everything is fine. The only two factors that really seem to affect this are my connection at 1J7 and board flex/position. I can't figure out a reliable way of changing these symptoms, it just seems random.

Given the randomness, should I start considering more seriously that my problem is alkaline corrosion related? Or is it something simpler like a cracked solder joint on one of the address decoder ICs?

EDIT: I also still need to test the 6802s I have to make sure they're good and rule those out as a problem. I will do that as soon as I can.

While I recognize the rarity of the oscillator being an issue, is there any other test I can perform with my DMM or logic probe to make sure the clock is working? When I probe pin 39, in any board state, I see a steady pulse. The guide at techniek.flipperwinkel.nl only mentions I should get a square wave which I believe the probe is indicating with a stead HIGH/LOW beat. I believe the same friend that has a Space Shuttle has an oscilloscope so when I go to check my 6802s perhaps we can hook it up to check on that.

Ha. I think I came across a thread somewhere here on pinside where someone forgot to do that and their machine went haywire.

This reminds me of something I thought about. I ran several verify runs on the EPROM I burned the Leon test onto and they all passed. The EPROM I used was just a random brand I found on Amazon, where I was ordering the program. I have since ordered some EPROMs from Mouser that are Atmel branded. Would it be worth it to burn a new Leon ROM on one of these chips? Probably a long shot, but I thought I'd ask.

Another data point that crossed my mind. The reason I have this power supply is because the 12v line died on me. It was previously in a 3D printer I had and ran the heat bed. The 5v still tests good so that's why I use it. I say all this because I've never hooked up the 12v to this board when bench testing and it's somehow able to boot up, but it can take it a few moments. As it's had the reset section replaced with a MCP120, I believe that's why it no longer needs the 12v. Is there anything else on the board that requires 12v?

Good to know.

Man, I really thought at this point I'd have this thing licked but it appears to be a real bear of a problem. I know I still have a few avenues to go down, but it feels like I'm reaching the limits of my ability and tools.

I've poked around on eBay for replacement boards without much luck and there's also the Rottendog board of course but that feels like a super last resort. I'd really like to make the board I have work, but at least I know there's a mostly surefire way to get the game working.

I've seen Chris's name passed around on here. I hesitate to send this board out to a repair guru considering all the prior work that would be a headache for whomever worked on it, but I'll keep that in mind if I give up.

Is that the red board in the picture he submitted recently? I was wondering what that was.

I have some extra single row sockets I can use to start replacing these as well as some 40 pin sockets for the CPU so I'll start addressing those.

Regarding the corrosion, am I looking for components that have dull solder or pins? The D2 diode and a caps near the 1J7 header are the two things I've noticed that look corroded, i.e. a greenish paste on them. I've seen pictures of very advanced corrosion and seen it first hand on the original board that came with my Trident, but I don't see anything that bad on this board like lifting traces and what not. The crazy thing is, the battery holder, which I've since removed, is very clean. But it's possible someone cleaned it up when they addressed other things.

Huh, actually now that I'm looking at my board and the schematic, D1 looks kinda jacked up too. If I'm reading the schematic correctly, the 5v logic power flows through that diode to U25. I wonder if that could be the cause of my intermittent issues. I'll try cleaning up that area and soldering in a new diode along with all the other stuff I need to do.

And just to derail the conversation for a sec and inject some positivity in this thread, I received my CPR backglass today! They had a BK2k on clearance due to some smudging in the black, but it's very subtle and doesn't bother me. I'm very happy with it.

IMG_7774 (resized).png

For D1, the manual specifies "Diode, 1N5817, 1.0A". I know I have some 1N4007's, which are 1A 1000v. I may have others. You say the 1N400X series is borderline, I presume 1000v may be too much?

Ah, forgive me. I see now that 1n5817 is a Schottky diode. I don't have any of those. The lowest I have is a 1N4004 which is a 400v 1A diode.

Holy cow! That is a MOUNTAIN of impressive work, DumbAss ! If I can’t get this board working I would be very interested in one of your blank boards to populate.

Duly noted. I read up on this after I made a bad assumption and understand the difference now. I added them to my next order and got it just under the cut off time for delivery tomorrow. I’m very glad slochar chimed in before I made a mistake.

Update!

I replaced the electrolytic caps at C24, C26, and C29, the 1k pf caps at C22, C23, C25, C27, and C28, and the Schottky diode at D1. I didn’t get much luck with my CPU power supply but if I plug the board into my machine, Leon ROM installed and without the RAM installed, I CONSISTENTLY get a blinking diagnostics light. Every time. With the new RAM or old RAM installed I get no blinking at the diagnostics light.

This seems to narrow down my problem to either the RAM socket or related circuitry. I will replace the socket as well as the 1N4148 at D2, in case it’s dragging voltage to the RAM down somehow, later tonight and report back.

More progress. I went to start desoldering the RAM socket and noticed a flake of solder bridging a couple pins. I think this may have been from my solder sucker earlier when doing the work on the caps. Usually I brush off the back but I was moving a little too fast.

Anyhow, after removing the flake I decided to try installing the RAM and adding power with the machine instead of my power supply. The diagnostic LED lit! But it didn’t blink with the Leon ROM installed. I can’t get it to happen consistently like I can with the RAM uninstalled so I still think maybe the socket needs to be swapped out even though it looks brand new.

Right after this, for fun I installed the game ROM and got a lit diagnostic LED! No blanking LED though. I’ve only gotten that to happen once.

Another observation. I’ve never tried it before, but the system 11 guide mentions checking the PIAs to see if they’re hot. I’ve done that and never felt any to even be warm. However while probing around and checking the voltages with the board connected to the machine I rested my fingers on the 6802 at U15 and it was very warm. Not enough to burn but enough to make me question it. Should the 6802s get warm or is this more evidence that they are also a problem and need to be tested?

I tried both 6802s from the board in U15 and I also tried the 6802 I got from Marco the other day and they all feel about the same. All of them get warm, borderline hot even, but I don’t feel like I need to take my finger off. I have a temperature probe for my DMM I need to dig out and I’ll try to get an accurate reading.

I also finally built a little 9 pin connector for my computer power supply I’ve been trying to use and boy howdy did that make a difference! I get a solid 5v at all the chips and TP2. The Leon ROM runs perfectly, all the PIAs check out, and the diagnostic LED blinks as it should in time with the pins. All this of course with the RAM uninstalled.

Yeah! It feels good. I think I’m zoning in on the problem but I’m not sure what the right solution is. I will swap the socket but given how new it is I just don’t feel confident it’ll change unless there’s something going on underneath.

On the bench I can’t get the diagnostic LED to light like it did in the machine with the RAM installed even with my new connector and a rock solid 5v. I’ll search the forum to see if anyone has encountered a similar issue.

I did check for continuity between pins and their neighbors at the chip and on the solder side and didn’t detect anything odd. I’ve tried three different RAM chips at this point with no change. I also verified continuity on the jumper in case that was being weird. By the time I did all that it was a little too late to start trying to desolder that socket so I moved on to finishing up the playfield reassembly. I did pop in the 1N4148 with no change.

I just had a thought. I’ve never done any of this testing with batteries installed. I believed the board would boot and run fine without them, but throw an “Adjustment Failure” if it were connected to the screen. Would my RAM issue be caused by not having batteries installed?

Scratch that last post. Didn’t make a difference if batteries were connected.

This morning I swapped in a new socket for the RAM and CPU. I replaced the ‘B’ capacitor and the 4.7k resistor at R52 next to the RAM. Doing all this changed the voltage I get at pin 28 from 4.7v to 4.8v but still made no change to the diagnostic LED not lighting with the RAM installed. I’ve once again tried all three RAM chips I have. Leon still runs perfect with no RAM.

I also measured the temperature of the CPU with the probe on my DMM. Measuring over the course of a minute from the center of the 6802 at U15 I read 41C and still climbing, but very slowly. Seems wrong.

I might just go ahead and replace all the ROM sockets to be sure but I’m not confident that’ll change much.

Looking around on eBay I came across a seller in Arizona selling some Hitachi 6802s. The listing claims in the description:

“All of my parts are surplus Military and Aerospace unused devices that were
purchased at employee auctions here in the USA. They were purchased between
1982 and 1998 and have been in controlled storage since then. They all meet the
full manufactures specifications. I am a recently retired Aerospace Electrical
Engineer.”

Which sounds reasonable. I bought a couple and they should come next week. I believe I read with the Hitachi chips I need to remove the jumper at W16 and W17. I’ll double check before installing.

One more data point in case it’s relevant. In doing a continuity check on adjacent pins on the CPU at U15, I noticed I get about 560 ohms of resistance between pins 39 and 38.

According to the schematic, pin 39 is the EXTAL line that connects the two 6802s to the clock at CR1. Pin 38 just connects to ground. Could this be a potential short that’s causing issues or is it expected?

8E45897D-6780-403A-B375-EFB501C96248 (resized).jpeg

EDIT: I should do my homework before posting. Looking at the data sheet for the 6802 the EXTAL and XTAL lines are used when using an external oscillator so I suppose I should expect to see a connection there. Ignore my grasping at straws

I’ve never gotten the blanking LED to light but I’ll try slapping it all together and seeing what happens.

Ordered!

You’re right he did! Oi. Sometimes my brain doesn’t work.

Bummer about the Hitachi chips. Do you know a good source for 6802s if mine check out bad?

No dice. I popped RAM in with the game ROM, plugged it all together and I just get a 5v LED. No diagnostic or blanking LED. Nothing on the display. I can’t test sound because I’m missing the ribbon cable that runs from the audio board to the CPU board.

Same here. At the time it happened I turned it off and back on to try to reproduce it and I couldn’t and haven’t since. So either it was a fluke or there’s something loose on the board. I’m going to try redoing all the sockets for the ROMs to see what happens.

I have not done any of the ROM sockets, i.e. U27, U26, U21, or U22, or the CPU at U24. I realize the last three are audio related and won’t affect what I’m seeing.

The last IC I pushed in was the game ROM at U27.

Progress and setbacks.

I swapped out the U27 socket with no change. Ugh.

After doing some head scratching I went back to the Leon guide and worked through it with the RAM installed, treating it as if I couldn't get it to work (even though I actually CAN get it to work without the RAM). I came back to this step, Chip Selection, which I hit early on in this thread:

"We continue to work without program chip. Because the cpu goes through all addresses, it will also come across every selection address of every PIA and program chip. The selection pulses come in at these pins: for PIA's check pins 23, 24, 25, 35 and 36 which should give alternating signals (around 3 volt, with a testlight both leds are on, one a bit more then the other but always both on). For the test-ic in U27 check pins 20 and 27 (2 to 3 volt). If there are pins which don't get pulses but always +5 or ground then the selection has a problem. Check the schematics to find the ic where this signal comes from."

Before this step you remove the test ROM at U27. Doing this check I saw pin 23 low on all PIAs. Tracing that pin on the diagram led me back to the 74LS138 at U37. Now I know DumbAss said he rarely, if ever, sees these fail, but I figured what the hell, I've tried everything else! So I clipped off U37, installed a socket, and popped in one of the new chips. Turning on the board with just the CPU installed I checked pins above again and lo-and-behold pin 23 was moving! Hot damn I thought, and turned it off. I put the Leon ROM in and turned the board back on and... no diagnostic light. No movement on the PIA pins 2-17 as before.

If I learned anything from y'all, I know it's to suspect the last thing I did if something breaks! Here's what I think went wrong:

Before I clipped off U37 I inspected the three chips Marco sent me a few days ago. Two of them say 74F138N and the other says SN74S157N. I looked up these parts on Mouser and the 74S157 is a "Quadruple 2-line to 1-line Data Selector/Multiplexer". Wrong part. The 74F138N came back as a "3-line to 8-line Decoder/Multiplexer", same as the 74LS138, I thought. The schematic on the data sheet looked the same! Wrong again!

Oi, I can be dense sometimes and make too many assumptions. It's right there in the name! LS vs F are the logic families. The F is short for FAST (Fairchild Advanced Schottky TTL) which is faster than the LS in terms of the "Propagation Delay Time" 9ns on the F vs 41ns on the LS. This might explain the way faster pulsing I saw and seems like it could affect functionality of the rest of the board. I'll order the LS and hopefully that gets me at least back to a working Leon test, but hopefully to a board that boots correctly.

Fun note, I was looking at a TI 74F138N but what I actually have is a Fairchild Semiconductor 74F138N which the data sheet says is a "1-of-8 Decoder/Demultiplexer", but I think that's just a different way of saying "3-line to 8-line Decoder/Multiplexer". The schematics between the Fairchild 74F138N and an 74LS138 look the same as far as inputs and outputs go.

Anyhow, work continues. I will post an update early next week when I have the correct chip in hand.

Correct impression! It may take me four or five read throughs to understand, but I think I can wrap my mind around it. I really appreciate your detailed replies and @pinballmaniac40's help as well. You both have gotten me very far in this process. I hope I haven't worn out y'alls welcome.

Yes I do. I kinda worked my way through it after posting last night. Just a different way of saying the same thing.

I spent some time reading about Propagation Delay Time on wikipedia last night: https://en.wikipedia.org/wiki/Propagation_delay

In the entry, it mentions: "The difference in propagation delays of logic elements is the major contributor to glitches in asynchronous circuits as a result of race conditions." Is it possible the new issue I'm seeing is because the circuit is designed for and expecting the U37 to execute at a certain speed and I just put a faster chip in? It seems plausible, as all the other chips involved in IC select are of the LS family. It's also plausible that I screwed something up like bridging or cutting a trace

Is it safe to say, or rather is it plausible, that had I replaced the U37 correctly that I might have fixed my issue with the Leon test not running with the RAM installed? I suppose going by that logic, I could also suspect U33, U34, and U31. To cover my bases I had ordered the chips that could replace all these so I have them on hand if needed. I'm going to start with getting ahold of a LS138 and installing that to see if my theory on the prop. delay time is correct. If that doesn't change things I'll go back over my address and data bus to see if I broke something.

When I removed U37 all the solder pads seemed bright and shiny. No evidence of corrosion, at least to my eye. If I have to move on to the others I will be sure to pay close attention for signs of damage.

Always, always, ALWAYS check prior work!

I received the LS138 today and swapped it in with no change (Leon test still not working). Ugh. So I pulled the test rom out and started over again with just the CPU installed and saw all the address lines moving as they have been. After more head scratching I got out the magnifying glass and flash light and looked over the back of U37. No bridges, no busted traces.

Now I can't tell you why I didn't think to check this prior to today, but I moved over to the socket I replaced at U27 prior to U37 and lo and behold I had left two pins unsoldered! I blame my habit of working in my sweltering garage at midnight for the oversight. Anyhow, soldered the two pins, popped the Leon ROM back in and got the board back to where it was. It's still not running with RAM, but at least I got it back to its prior state. I also learned that you can in fact use an F138 in place of an LS138, at least at U37.

Now that I've patched up the CS2 signal (if that was really ever a problem) I believe my next steps include:

– Verifying my 6802s, which I should be able to do this week as my friend with Space Shuttle is back in town.
– Working down the chain of ICs tied to the R/W pin of the RAM: U31, U33, and U34.

EDIT: After posting this, I ran out to the garage and checked pin 27 of the RAM socket both with and without RAM installed. Installed, pin 27 is stuck high, uninstalled it pulses in time with the diagnostic light. I would expect the R/W pin to pulse as data went in and out of RAM so perhaps one of the IC select chips involved is malfunctioning?

Another data point. With the RAM installed and checking pin 20, the CE pin I believe, the probe has the high light on and shows a pulse which I believe is "Logic 1 with negative single pulses". I had thought maybe the CE wasn't firing and thus the RAM wasn't doing anything. It's also still possible, though improbable, that all three of the RAM chips I have are dead. I plan on starting my search at U31 and working backward to see if something is amiss with the R/W line in the IC select chips.

Apologies for doing this in spurts. Life is distracting!

We have a boot with the game rom! I hadn't yet tried the other RAM chips so I installed the one that came with the game, Sharp brand, and I got diagnosis led pulsing with the Leon ROM!!! Turned it off, installed the game ROM and... 11 diagnosis LED flashes and then the 50% flashing it's supposed to do, so it looks like my new problem is U26, according to the table here: http://techniek.flipperwinkel.nl/wms11/index2.htm#cpuled.

I swear I ran a verify on it with my new EPROM programmer! I suppose it could also be the socket there, it's the original one.

Anyhow, HOLY COW! I guess it really was that LS138. I'll verify some more things this evening and see if I can move past the last error and then I think we can consider this thread SOLVED. I've got a mountain of other things to solve next, but it feels really good to at least having a booting CPU.

Special thanks to DumbAss and PinballManiac40 for their tremendous insight and help as well as slochar for making sure I used the right diode and Tuukka for pointing out my sound roms being swapped. I'll give a final update as soon as I can get the whole board doing what it's supposed to.

Right now I have the original 6802 that came with the game (one was missing and a 6821 was in its place) and the one I got from Marco in the game. I just swapped them and still got it to boot. I haven’t tried the one I got from pinball life yet. I also haven’t tried the Hitachi 6802s either.

Anyhow, I swapped the socket at U26 with no change. Turns out the ROM I had was burned on a 27256 and U26 takes a 27128! So I doubled up the bin for U26 on a new eprom (I don’t have a UV eraser yet) and popped that in. SUCCESS! No initial status blinks just straight into the normal 50% cycle blinking on the diagnosis LED. The blanking led was off but it turns out that led was busted because I saw it low on the cathode side. I swapped in a new LED and confirmed. Everything is working!!! Picture attached!

35F2F299-4A4F-44F0-B938-E980A0F6469D (resized).jpeg