Defender stuck in book keeping

[jehuie]

It's not dead, it's just resting.

Check the Continuity between pins, don't trust your eyes. Redoing the pins, you might have created a new cold solder joint, so check your soldering for consistency. Post a picture of what you did as well.

Check for excess solder that might have fallen on to a different part of the board.

The connector itself could be bad. Are the pins on the header clean?

What are the voltages now? Did they change?

I checked to be sure there is no continuity between pins that should not be and that was good. Checked for continuity from each pin to the back of the board where I soldered it. That was good. Here's some pics. I also re-soldered the 6 pins on the reset switch as well as a few other minor components. But stayed away from chips. I'm frustrated as heck and about to give up. I appreciate all of you guys help but crap.

I'm going to take a break for a bit and decide what to do later. Thanks guys.

EDIT: The voltage at the ram chips is -12, 4.6 and 11.7. The header pins look clean. Can't think what else you told me to look for right now.

 

[YellowDog]

Those voltages sound suspect. Check the voltages at the power connector on the CPU board.

From Pin 1 (side away from the reset switch):
1 - N/C
2 - GROUND
3 - GROUND
4 - key
5 - +12V regulated
6 - N/C
7 - +5V regulated
8 - +12V unregulated
9 - -5V regulated

There is no way with a properly operating power supply to get -12V on the board. Also if the +5V is actually +4.7V that is too low for reliable operation. The voltages may be the reason for the reboots. If the +5V is low, when the sound board kicks in a lot of sounds it may pull the voltage down just enough to cause a reboot.

ken

 

[jehuie]

Sorry, that was supposed to say +12, not -12. And yes, I'll check the voltages at the board when I get back home later. Thanks again. But even if they are not right....it still doesn't explain what I did to hose the whole thing so bad.

 

[Arcadenut]

Sorry, that was supposed to say +12, not -12. And yes, I'll check the voltages at the board when I get back home later. Thanks again. But even if they are not right....it still doesn't explain what I did to hose the whole thing so bad.

It's hard to tell from the photo, but I would touch up the pins I have circled, they look kind of funny (might just be the angle/light, etc...)

 

[Arcadenut]

One more...

 

[splattergatz]

The voltage at the ram chips is 4.6

4.6 is too low
did you install the 5k Ohm pot at R2 and dial in the +5v to 5.1 ?

i would measure for the +5v everywhere along its route to the chips to detect where you are getting the voltage drop assuming the power supply is outputting the correct +5v

might have have a voltage drop due to weak pins/oxidizing traces, wires, header pins, cold solder joints, some kind of problem creating resistance so you dont have the correct voltage at the chips. assuming again you have the correct +5.1 out of the PS

if you dont have a 5k ohm pot hooked up, might hook up a switching power supply just to the +5v where it comes out of the PS so i can turn the +5v up and measure it at the chips and see what i have to turn the switching power supply up to, to get the CPU to boot. if the switcher needs to be turned up a lot to get it to boot then you know there is a weak spot along the way creating resistance and no booty booty

 

[jehuie]

4.6 is too low
did you install the 5k Ohm pot at R2 and dial in the +5v to 5.1 ?

i would measure for the +5v everywhere along its route to the chips to detect where you are getting the voltage drop assuming the power supply is outputting the correct +5v

might have have a voltage drop due to weak pins/oxidizing traces, wires, header pins, cold solder joints, some kind of problem creating resistance so you dont have the correct voltage at the chips. assuming again you have the correct +5.1 out of the PS

if you dont have a 5k ohm pot hooked up, might hook up a switching power supply just to the +5v where it comes out of the PS so i can turn the +5v up and measure it at the chips and see what i have to turn the switching power supply up to, to get the CPU to boot. if the switcher needs to be turned up a lot to get it to boot then you know there is a weak spot along the way creating resistance and no booty booty

I re-soldered the pins in question above and it still does the same thing. The voltage is still low at the ram so I swapped out the other rebuilt power supply and got the same reading or close. I get 5.7 at the PS test point...but only 4.9 at the header pins on the PS and 4.6 or something at the ram still.

I don't have a 5k pot to throw in there but I can pick one up. So I just solder 2 of the legs of the pot in where R2 is? What about the third leg?

 

[jawhn]

... What about the third leg?

Really?



Sorry, had to.

 

[YellowDog]

One of the more common causes of voltage drop are the filter capacitors right by the power header. They are 100mfd at 25V (stupid forum software wants to make an email address when you use the 'at' sign ). They are cheap enough that I swap them out a lot of the time when I am working on Williams boards.

ken

 

[jehuie]

One of the more common causes of voltage drop are the filter capacitors right by the power header. They are 100mfd at 25V (stupid forum software wants to make an email address when you use the 'at' sign ). They are cheap enough that I swap them out a lot of the time when I am working on Williams boards.

ken

Cool info. Thanks for that! I'll pick some up while I'm at Fry's getting a 5k ohm pot or two.

 

[YellowDog]

You should get 50K pots. The original R10 is a 51K resistor.

ken

 

[jehuie]

You should get 50K pots. The original R10 is a 51K resistor.

ken

Really? Drat, I already made the drive over and got 5K ones based on this pic. So what would happen if I used the 5k ones then like this one?

 

[tb lilley]

Do these boards still work in your other Defender cab? If so, then I'm guessing your problem is power related.

You say that the power board has been newly capped, but there's a load of other stuff on there that could be causing you problems if they've not been replaced. I've seen this before where power boards have some but not all parts replaced - can make things worse.

You might want to consider replacing the 3 different rectifiers, the -5 voltage regulator, 2 x LM723s, and the 2 transistors. Which is a lot of work ...... The other way to go might be to switch to 4164 RAM which only needs +5 to run. That would mean that you'd only have to replace the one rectifier (BR1), the 2N3055, and just one of the LM723s.

Even that is a lot of work, but in return you should have a very solid and reliable power base which would give you reliable power for years to come.

 

[YellowDog]

There are several different ways to make the power supply adjustable. I normally don't suggest it just because I've had to fix too many that somebody slipped while turning the pot and fried the overvoltage circuit. If you have the old Defender power supply that won't be an issue as there isn't one. You can just fry the boards.

The 50K pots are for the newer (Joust style) power supplies (the ones with the seperate heat sinks).

On the Defender power supplies you need to replace R2 with the 5K pot after setting the dial to the middle of the range. The two resistors (R2 & R1) act as a voltage divider to reduce the reference voltage from 7.3V to 5.1V. By replacing R2, you are making that side of the voltage divider variable. Never adjust the voltage by watching the LED, always use a meter. The LED will show that the minimum voltage has been reached, but it will not tell you when the voltage has exceeded the operating range. Always adjust the voltage with the boards unplugged.

ken

 

[jehuie]

So it sounds like maybe I should really be just replacing R2 with a new 5k resistor since I'm getting slightly low voltage at the PS, right? My first order of business will be to replace the 100uf caps and see what that does for me.

If I do go ahead and put the 5k pot on there, should I set the voltage to 5.1 at the PS with the game board disconnected? Or should I measure at the chip and turn it up until it reaches 5.1 there?

There are several different ways to make the power supply adjustable. I normally don't suggest it just because I've had to fix too many that somebody slipped while turning the pot and fried the overvoltage circuit. If you have the old Defender power supply that won't be an issue as there isn't one. You can just fry the boards.

The 50K pots are for the newer (Joust style) power supplies (the ones with the seperate heat sinks).

On the Defender power supplies you need to replace R2 with the 5K pot after setting the dial to the middle of the range. The two resistors (R2 & R1) act as a voltage divider to reduce the reference voltage from 7.3V to 5.1V. By replacing R2, you are making that side of the voltage divider variable. Never adjust the voltage by watching the LED, always use a meter. The LED will show that the minimum voltage has been reached, but it will not tell you when the voltage has exceeded the operating range. Always adjust the voltage with the boards unplugged.

ken

 

[tb lilley]

Is any of that going to make a difference if you've got a dodgy LM723?

 

[jehuie]

Is any of that going to make a difference if you've got a dodgy LM723?

I think you are referring to the other thread called "Dead Defender"? That's actually a different board. I've got 2 screwed up Defender games actually.

 

[tb lilley]

I'm suggesting that you think about replacing the LM723. It's the +5VDC regulator on the power board that sits quite close to the resistor that you're going to change.

Dunno. From what you've posted so far on the history of this particular game, you seem to have had, at one point, a working board set (when sat in the other cab). You've also said that the PS board has been recapped, and that someone has fitted new ribbon connectors. You've also reflowed some of the header pins. But the problems remain.

I'm just suggesting that there may be an underlying issue here with the power supply that has not been properly addressed. The symptoms you described (resetting board) could be any of a number of things. But it could be as simple as the LM723 occasionally playing up resulting in a dip in the +5 = reset. Could also be any of the 2 other voltages dipping. But a short cut would be to fit 4164 RAM that only uses the +5. if that doesn't fix your problem then try the LM723.

 

[jehuie]

Oh I follow you now. I didn't know what that chip was and I thought you were referring to the one from the other thread. That's a good thought as well. The reason i was thinking I would swap that resistor out is because it appears that I'm getting good voltage at the +5 test point....but by the time it reaches the pin on the connector it's dropped. So I was thinking it could be just a bad resistor. But now I see what you are saying.

 

[tb lilley]

I'm no expert on any of this, but there's not much that can go wrong with a resistor. I think all you'd be doing, by fitting the pot, would be to throw a higher voltage at the LM723. It might solve the problem, it might not.

The problem that you described of occasional resets might be because the amount of voltage getting to the RAM was so incredibly borderline that it was enough to run the game, but not enough to do so reliably. Which is possible. Might also just be that the LM723 is flaky. That chip will be 30 years old, so could do with a replacement.

I think Defender boards have a bad rep. But I know quite a few people that run Defenders without any problems. Those cabs usually have a rock solid power supply. Defender boards, like any other thats 30 years old, are sensitive aging bits of electronics. Treat them well to a solid clean constant reliable power supply and you might find a bit more reliability.