Oh following this thread! I’m interested in knowing about this as well.

Sorry, I got nothing. I wish I was better at system 1 stuff. Just keep asking and one of the masters will respond.

Do the ones you haven't done so far. If you can't find the right housings, ask here.

In no particular order...

The original low power switches on System 1's are gold plated. If someone has used sandpaper or a file on them, they should be replaced. The fingers on the board connectors get worn. Pull the boards and look at fingers. If you can borrow a Ni-Whupf or other aftermarket CPU board, definitely try that to rule out the CPU board.

So what happens if you manually only give the spinner a decent flick or only tap each target/scoring switch one at a time 10 times very quickly? Do they all fail registering about the same % of score increments when dealing with just one switch? Might help narrow it down if it is really only a problem with certain switches or certain group(s) of switches that are common to the same strobe lines, chip, connector etc.

Not sure it is relevent but:
I have owned system 1 thru 80b and have always notice a lag when multiple scoring occurs.
My Volcano that i presently own has "oiled" spinner. Hitting that spinner and continued fast scoring requires the mpu to "catch up" when it can.
I remember it was the same with basically all of the SS gottliebs i have owned over the years. Some had original mpu and at least one had a ni- wumph.

That makes sense to me. I have 2 Sys1 games, and if you do happen to trigger multiple events, it will catch up.. even over the span of many seconds if needed. Sinbad is a great example where you could collect the 5x, and have that count down, while scoring more bonus.. and have the machine literally play catch up for a minute.

Perhaps this memory is stored or buffered in ram or one of the spiders? I really don't know, but I can say the ram is definitely suspect to go bad on many original boards.

I've repaired a couple sys1's from the ashes, but never had this symptom.

If you want a less expensive test, the lisy.dev sys1 boards are EXCELLENT, and very inexpensive to build..

Thanks-

Which diodes did you replace, and what part numbers did you replace them with? Switches take 1N270 diodes, not your typical 1N4004's.

4148 were used by other manufacturer's at the time to replace 1n4004 on the switch matrix, but Gottlieb is fussy and wants the 1n270's for some reason. While it may seemingly work, the 270's should be used instead.

Just repeating long ago wisdom from RGP and other places. Never tried it myself.

1N270 are germanium diodes which typically have a lower forward bias voltage drop of around 0.3 volts compared to silicon diodes which is said to be around 0.6 volts. Most 1N4148 silicon diodes I measure at low current are around 0.7 volts.

Gottlieb switch strobe lines are active low and the return lines connect to a 7405 TTL chip.
The logic "Low" level for a TTL chip is considered between 0 and 0.8 volts. Using a 1N4148 leaves next to no room for margin with only 0.1 volt difference which can get lost in connectors. This can result in active switch return voltages not reaching an acceptable low level and becoming "indeterminate".
Germanium diodes at 0.3 voltage drop leaves more margin of 0.5 volts at the input of the CPU boards 7405 chips switch return lines.

If good 1N270 diodes can't be sourced, try 1N5819 schottky diodes which have a low voltage drop. I've measured some here at low current to be 0.25 volts. It's also a reason they're used on pinball MPU boards as blocking diodes for non-rechargeable batteries because of their low voltage drop.

This may or may not have anything to do with the OP's problem though.

The shorting of that cap with the wire is kind of normal, it's to bypass the slam switch circuit if it's causing problems.

I'm still thinking you replaced the wrong diodes with the wrong diodes.....But what are the chances that the ones you removed had been replaced previously with the wrong ones?

I'm certainly no expert, but I can tell your that Quench (see above) directed me towards fixing my System 1 problem with the tilt bob being unresponsive by switching to the 1N5819 diode.

All system 1 games' displays blink when switches are hit, normal. What it decides to do with them later, that's a different story.

I had replaced a Sinbad's spinner with a Black Hole one for better spin response and it was the same thing, 5-6 registrations. I think it's just crappy. That was my only game with an original system 1 board and it's been moved out a long time ago, so I can't even try the pascal replacement in it (my other system 1 games do not have spinners..... countdown even with the pascal board misses a drop target sometimes if you get a LOT of them down in a very quick interval, kind of frustrating when that happens.)

I actually recommend 1N5817, BAT42 or BAT49 diodes over the 1N5819s.

More info on diodes than most people could ever use...
For a System 1 - typical switch current per row = 5V / 2700 ohms = about 2mA.
Rounding up and comparing a few devices at 10mA:
BAT42 (Schottky diode):
VF = about 0.35V typical at 10mA (voltage drop, lower is better)
VF Max = 0.40V

BAT49 (Schottky diode):
VF = about 0.28V typical at 10mA
VF Max = 0.32V

1N5817 (Schottky diode):
VF = about 0.2V typical at 10mA
VF Max = 0.45V

1N5818 (Schottky diode):
VF = about 0.2V typical at 10mA
VF Max = 0.55V

1N5819 (Schottky diode):
VF = about 0.2V typical at 10mA
VF Max = 0.60V

1N270 (Germanium diode, VERY old diode design):
VF typical at 10mA = unknown but supposedly quite low (nobody publishes full data any longer)
VF Max = 1V

1N4148 (Small Signal or Switching diode):
VF = about 0.7V typical at 10mA
VF Max = 1V

1N4004 (1N4001 through 1N4008)(Rectifier)
VF = about 0.7V at 10mA but they aren't designed to run at these low currents
VF = 1.1V max

All of the above diodes are more than sufficient to work with a 10mA current load.
The problem with Germanium diodes is ... well, they're germanium. Old, obsolete technology and quite expensive.

The problem with Schottky diodes is high leakage current. But leakage current doesn't have much effect on this circuit as the active components in this circuit overpower it.
BAT42/BAT49 works well but is about the same size as a standard 1N4148 switching diode -- tiny glass and frail
1N5817 will work and is larger, beefier.

Switching diodes are often used by other manufacturers in their switch matrix.
The problem with switching diodes such as 1N4148 in a Gottlieb is they have a higher forward voltage drop.
Good side is they have far less leakage current... which really doesn't matter here.

Next item - what do we need for maximum forward voltage drop rating?:
Receiver = 7405 (cheap version of 7406).
Two levels of operation -- high input and low input.
Vin high (min value to guarantee valid output) = 2.0V min but we don't care as this is provided by 2.7K pullups.
Vin low = 0.8V max (this is the value we care about)

Driver = 7404 inverters:
Two levels of operation -- high output and low output.
Vout high = 2.4V min
Vout low = 0.4V max (this is the one we care about).

Allowed voltage drop between driver and receiver:
Maximum Vdrop = Vin (low) - Vout (low) = 0.8V - 0.4V = 0.4V (minus tiny drop through switch).
So in theory, we cannot use any diode with a VF greater than 0.4V.

The bottom line -
The BAT42 and BAT49 Schottky diodes are the only ones guaranteed to work. And of course, BAT49 is now obsolete.
1N5817 diodes will most probably work since their "typical" voltage drop is less than 0.4V but not guaranteed.
Switching diode might work.. might not.
1N270 -- apparently works but nobody publishes full data any longer so I don't know. These are also known *only to the state of California* to cause cancer if you consume large quantities of them.
1N4004 -- not recommended.

Thanks Ed.
I'm looking at their STMicro Schottky datasheets and for context I've added the current details pertaining to maximum spec'd VF (Forward Voltage drop).

BAT42 (Schottky diode):
VF Max = 0.65V @ 50mA

BAT49 (Schottky diode):
VF Max = 0.42V @ 100mA
VF Max = 1.0V @ 1A

1N5817 (Schottky diode):
VF Max = 0.45V @ 1.0A

1N5818 (Schottky diode):
VF Max = 0.55V @ 1.0A

1N5819 (Schottky diode):
VF Max = 0.60V @ 1.0A

The listed VF Max specs for 1N581x are at a much higher forward current than the BAT4x so it's not a fair comparison. Looking at their graphs below of typical values over the range, it seems to me the 1N581x are the better choice and have lower VF than the BAT4x (highlighted 10mA @ 25°C). Even your listed VF @ 10ma are lower for the 1N581x devices.
Or am I missing something?

I have some paranoia that makes me choose the 1N5819 over the 1N5817 because of it's higher rated voltage, but that's just me

BAT42.png
BAT49.png
1N5817.png
1N5819.png

Different manufacturers = different specs but not a lot of difference.
These should really be compared at the operating point in which they will really be used - 2mA (although I arbitrarily rounded up to 10mA for simplicity sake). If you see ever any currents in the 20mA or more range than you have other problems to contend with - this is more current than the 7404 driver can possibly sink and can take out the 7404. Likewise, if you ever subject the voltage to anything higher than 7V - you will also lose ICs.

The part you used for specs shows BAT42 with about 0.38V at 10mA and about 0.3V at 2mA.
The BAT49 specs shows pretty close to the same as the BAT42 but a hair lower in forward voltage at 2mA.
The 1N581x diodes show pretty low (about 0.25V) at the bottom end of the current draw. But these are not intended to operate below the 10mA range (particularly in the 2mA range of the Gottlieb matrix) and are intended for use in the 100mA through 1000mA range.

Only manufacturer datasheet I can find for BAT49 is from STMicro, so to be fair I chose STMicro specs for all of them.

Agreed, at the moment all we have are some values at 10mA.
If we visualise the 1N581x graphs projection below 10mA, I think its safe to say their VF will continue to decrease.

I use 1N581x Schottkys as blocking diodes for CR2032 batteries on MPU boards. A board I have here with a Philips PCD5101P battery backed RAM chip whose standby current spec (max 5.0 uA) measures 0.015V voltage drop across a 1N5817 diode. Granted it's not switching but is operating at extremely low current.

Anyway, suffice to say I wouldn't be using 1N4148 diodes in Gottliebs switch matrix

Why not just install the diode recommended here rather than ask what can go wrong?
Cheap and easy change.

As i said before, every SS gottlieb i have owned seemed to be slow in adding when compared to my DE JP. For that matter, they seem slower than my 1979 Stern Trident.

The troublesome niggly things to me are:
1) Multiple switches tested showing it's scoring fine for five closures in a row but then the five after that, zippo, nothing? Of course from the 10 closures test you can't really tell if it's missing the last 5 because it isn't holding more in memory or whatever or if it's simply not registering around half of the hits on average.
2) A switch being hit once and scoring but another switch quickly after that not registering at all BUT the display blipped for both as if both closures 'reached' the mpu but something on the mpu is somehow at fault. How/why could it blip the display but not score if it's the diodes?
3) The spinner test must have done a lot more than 10 closures, still only 5 register so that doesn't sound like a 50/50 switch thing.

Just wondering if it's mpu and not the diodes so much though they may be causing problems too. As for Gottliebs being slow in adding, right, but this one seems to stop adding, period.

"Must consume mass quantities of switching diodes. Mebs! Mebs!"

giphy (2) (resized).jpg