Uh oh...he's done it again...geez...

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Uh oh...he's done it again...geez...

Post by ModRob »

I've got a 1982 model Ms. PacMan video game that WAS working fairly well. It has a slight hum in the speaker output, but the problem I attempted to remedy was a very light moving bar or two about 1 inch wide rolling across the screen (you could barely see two moving across at one time). I took out the main board and replaced all caps-the farads were right, but for most of them I had to go with a little higher voltage rating, since I could not find the correct ones. WEll, after that, I put it back in; on power up, the monitor flashes on briefly and then out. I have an area on the board near two 10,000 mF 16v. that contains a round,ceramic looking diode. It's getting very hot--enough that if left for any time, it unsolders itself. Now on quick checks after various times of checking things out on the board (yes, all caps' polarity are right, and no evident traces of solder bridges) the area around the diode still gets terribly hot, and there is an edgecard connector right above it with a pin or two that goes to this and another paired diode that gets hot as well.<p>I know it's impossible to remedy this online, but where else can I start looking? I have all the schematics that are in the service manual, but I'm just a novice at this, and don't have much knowledge in the fundamentals of test points, and am a bit afraid to go too far into experimenting in testing....haha..<p>Again, any thoughts are appreciated.<p>Timbo
toejam
Posts: 253
Joined: Wed Dec 05, 2001 1:01 am
Location: n.c.
Contact:

Re: Uh oh...he's done it again...geez...

Post by toejam »

it sounds like you have a short in the power supply That was not there before you replaced the caps. There is the remote possability that one of them is shorted or mislabled, but I would go over the board again looking for a solder bridge. The things can be hard to find, so clean the flux off the board and inspect it very closley. Sometimes the solder can actually come out of the holes and short the componant side of the board in the smaller caps. A vom would be helpful. Do you have one?
User avatar
Edd
Posts: 885
Joined: Sun Jan 27, 2002 1:01 am
Location: Dallas Tx
Contact:

Re: Uh oh...he's done it again...geez...

Post by Edd »

Assuming this to be your machine or the table model variant :
http://www.klov.com/game_detail.php?gam ... 2&letter=M
and not an Atari book size scale game variant.
With the units initial effect being a shaded line/hum bar across the screen, starting at the bottom and slowly working its way up until it is almost all the way up then a second line is replicated at the bottom and starts upward also just as the initial line is about to disappear at the screen top.
I also concur with tj’s analysis, with your initial problem being a deterioration of the filtering within your power supply section. Typically the 2 high value caps you mentioned…...possibly even with polarity reversal on the replacement(s). Your description is of what I would guess to be a round glass bead [usually just a bit larger than 1/8 in dia] encased power diode or a zener diode. Check across it in diode test mode and see if its junction is still intact or if it has avalanched and left it in a dead short state. Expect its cathode end to be marked with an encircling black ring. However, somewhere else I would expect a full wave bridge rectifier package being incorporated for primary power rectification.. Those old games used a lot of TTL and were very current hungry on the 5 Vdc buss..
With current enough to puddle solder , certainly you must have created a dead short on the power supply in your repair/replacement procedure. Ohm out across the filters and then start pulling the parts replaced , or area you worked in, until the circuit no longer indicates a short on the power buss.<p>
73's de Edd
[email protected] ........(Interstellar~~~~Warp~~~Speed)
[email protected]........(Firewalled-Spam*Cookies*Crumbs)
;) ;)<p>[ January 10, 2005: Message edited by: Edd Whatley ]</p>
Enzo
Posts: 276
Joined: Tue Sep 14, 2004 1:01 am
Location: Lansing, Michigan, USA
Contact:

Re: Uh oh...he's done it again...geez...

Post by Enzo »

If you have a real Midway Pac Man board, your problem may have been a burnt edge connector trace. They were under rated for the job and burned up on all of them eventually. The power supply was on the board, so it was AC coming on and the rectifiers were the class bead things on the board. There are two sided traces for the AC and two sided double traces at either end for common.<p>If it worked until you capped it, it is almost certain your caps are either defective or you miswired something. When you do somethig to it and it then works different, the change is something you did. That is the law of the shop.<p>If the diode is getting hot, it is either shorted, or the current flowing through it is being shunted somehow.<p>Loss of ful AC or full common on those will indeed give audio hum as well as a humbar in the video. It is also quite common on games of that age to need caps in the monitor. All manner of picture distortion cam result, including humbars.<p>Measure for resistance across the 5V busses on the board - pick a spot - and see if it measures shorted.<p>I have not services one of those in some time, but if I recall the trace around the edges was common on one side of the board and +5VDC on the other side. So do not assume the trace around the edges is always ground. One of the smaller caps on the 5v might be backwards for that reason.
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

My goodness...some great information there guys...a lot of it is still over my head, but it gives me much more ways to get deeper into it.<p>I have looked over and over for solder bridges, and improper orientation, but nothing found yet. And all caps appear to be the right ones, except for the fact I had mentioned before where the voltage rating on some of them was actually higher, as I could not find exact matches (i.e. 25v instead of 16v.) <p>The edge card situation is right on--there is one tab that is burnt, and the adjacent one next to it (looks like a double-wide tab) has two what looks like rusty spots on it as well. The burnt tab looks like it wraps around the edge so I assumed it's ok, even though only the top half of it nearest the edge is actually missing, with the other half near the inside of the board still there.<p>The description of the diodes is right on as well. They look exactly as you have described.<p>I'll keep on looking and using your comments to see if I can get lucky and zero in on what may have happened...<p>Oh I should have left well enough alone...ha ha..<p>Thanks again guys.
tim
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

Edd:<p>Great site for the info on the game...if you'll notice on the image of the PCB board, the sections I'm talking about are:<p>the diode in question is the very first one directly under the large heatsink at the right hand top. It's labeled as D8 and looks like a glass bead. The edge card damage is near the top of the right hand unit..
toejam
Posts: 253
Joined: Wed Dec 05, 2001 1:01 am
Location: n.c.
Contact:

Re: Uh oh...he's done it again...geez...

Post by toejam »

assuming the diodes are still good and those guy were pretty hefty, hook a 40 watt light bulb in series with one oh the ac input wires. whwn you plug it in the light shoulg glow brightly. than ramove the caps one by one till the light gets dim
(unpluging is when you pull the caps).the one you pulled out just before the bulb dims is the culprit. if you had a vom, things would be much easier. If it still glows brightly after all caps were removed, suspect a shorted diode.Didany of the caps feel warm after the diodes melted the solder?
rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: Uh oh...he's done it again...geez...

Post by rshayes »

The interesting point is that you have not mentioned the transistor on the heat sink overheating. According to the information Edd supplied (schematic), the power supply is a full wave rectifier (not a bridge) supplied by a 15 VCT winding. The two 10,000 uF capacitors are the filter capacitors. The transistor on the heat sink is a series regulator that drops the rectified and filtered power down to 5 volts. If you had a short on the 5 volt power output, this transistor would also get very hot. So would one of those power resistors nearby.<p>If the transistor and heatsink are not getting hot, it indicates that the fault current is not flowing through the transistor. This leaves the two filter capacitors. The ground trace is on the top layer running around the edge of the board. In the PCB photograph, you can clearly see the negative lead of the filter capacitors connected to this trace. If one of these capacitors is installed backwards, it will act like a short circuit. Another possibility is that the one of the filter capacitors is shorted. A very small fraction of new parts are bad, maybe one in a thousand or one in ten thousand. It does happen.<p>A higher voltage rating on those filter capacitors would not do any harm. The PCB photo shows them as being rated at 10 volts. The peak voltage of the 7.5 volt transformer winding is 10.6 volts in theory, so the 10 volt rating is very marginal.<p>Another possibility is that one of the rectifier diodes is open. This would make the other diode a half wave rectifier, which would dissipate at least twice the power in the diode.<p>The diodes look like the ones that Semtech used to make. The case was some type of moulded ceramic rather than moulded plastic. The current rating is probably more than 1 amp, possibly 3 or 5 amps.<p>If one of the diodes is shorted, it will cause the other one to overheat, since this would pass current from one end of the transformer winding to the other on one half cycle. Getting one of the diodes in backwards will do the same thing.<p>The overall schematic shows 5 amp fuses in the transformer leads. A real heavy fault current should have blown these fuses.
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

Hmmm...once again, you guys are amazing at how you can tackle this kind of stuff with such gusto. I only wish that I had more knowledge that I could really grasp what you advise...<p>Let's see...I do have a digital multimeter, and have a fair knowledge of it.<p>I don't remember any caps feeling warm after the diode loosened itself from its pad. If memory serves me right, I don't recall any heat at all around the large heat sink, nor the transistor in it. The only heat was concentrated around the diode, and just above it to the area of the edge card connector.<p>Earlier it was mentioned to "ohm out" the caps to check...how exactly is this done? And I would gather that it would be best to check the components out of the circuit? As in desoldered away from the board?<p>I didn't find any fuses blown at the bottom of the cabinet, near the big transformers.<p>How can I go about checking things out with the board away from the game? Is it as simple as getting some kind of variable power supply, and bringing in the particular power to a particular point on the board (like a tab on the edge card?)
It's fundamentals like these that stump me, as I'm afraid to go to far in experimenting for fear of damaging other components.<p>Again, it's so fascinating to find a group of folks like yourselves that love to help others like us...amazing in this day-and-time. I thank you...<p>More to come later...
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

Toejam: In reading again, I notice you make a point of solder flowing through to the other side of the board at a pad....I think this could be something. There were a couple of places on a cap that I was not sure of a good connection, so I added a dab of solder to the lead and to the pad, even though the component side had clearly defined circles for a pad. I took care not to stray from that circle. So maybe I should not have done that?
rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: Uh oh...he's done it again...geez...

Post by rshayes »

If the joint is well heated on a printed circuit board, it is normal for the molton solder to wick up through the hole and form a fillet on the component side of the board. As long as the solder doesn't bridge between pads there shouldn't be a problem.<p>The significant thing is that the dog isn't barking in the night. For the diode to unsolder itself requires it to get hot enough to melt the solder. It normally doesn't do this even though it normally carries an amp or more. This implies that the current was much higher, possibly 5 or 10 amps. This much current flowing through an electrolytic capacitor for practically any reason is usually enough to make it explode or at least vent some gas. So your capacitors may be ok.<p>Your digital multimeter probably has a diode test range on it. This will be near the ohms ranges and be indicated by a diode symbol. This range provides a current source of about 1 ma and measures the resulting voltage up to a volt or so. A normal silicon diode will read about .5 to .7 volts with a milliamp flowing through it in the forward direction. In the reverse direction it will read overrange. With the edge connector off, you won't even have to remove one lead of the diode from the board.<p>With the positive lead of the multimeter on the anode of one of the diodes (no stripe) and the negative lead on the cathode (stripe) the meter should read about .5 to .7 volts. With the leads reversed, the diode should read full scale. Sometimes this is indicated by a row of dashes. Test both diodes.<p>If a diode reads below a couple of tenths of a volt in both directions, it is shorted. This will probably cause the other diode to overheat. If a diode reads full scale in both directions, it is open. This will convert the circuit into a half wave rectifier, causing the other diode to conduct more than twice its normal current, and probably to overheat.<p>An open connection at the connector will have a similar effect. It appears that there is some damage in this area. Using the lowest ohmmeter scale, try taking a reading from the anodes of the diodes back through the connector. This should show a direct short, or a few hundredths of an ohm at most. If the resistance is much higher, the connector is bad. Cleaning the contacts on the circuit board might help, if they are not too badly burned. From what Enzo said, one of the contacts could be burned entirely off the board. This would cause all of the current to flow through the good contact and its associated diode.<p>If the connector is bad, I would get a different type of connector for the power leads. The large pin Molex connectors that Radio Shack sells will handle several amperes. I would get a 5 or 6 pin connector and bring the other leads from the power transformer through it also. This will guarentee that they are all connected or disconnected at the same time. The regulator circuit uses a second set of rectifiers on different transformer taps to form a 12 volt supply. This provides the base current that the pass transistor needs to operate properly when the 7 volt supply drops down close to 5 volts.
User avatar
jollyrgr
Posts: 1289
Joined: Thu Jan 03, 2002 1:01 am
Location: Northern Illinois
Contact:

Re: Uh oh...he's done it again...geez...

Post by jollyrgr »

Let me ask this. Are you SURE none of the caps are in backwards? <p>With power removed and the caps discharged, check each cap for shorting, especially around the area where the solder melts on the diode. Check and see if the caps are all on a common ground. Then connect the negative lead of the DMM meter to this point. Set the DMM to measure resistance on one of the higher ranges. Connect the positive lead to the capacitors' positive lead. The meter should go from a dead short/low reading to an open/infinity reading. Eventually you will find a cap that stays shorted. While it may not be the cap itself, something near it will be.
No trees were harmed in the creation of this message. But billions of electrons, photons, and electromagnetic waves were terribly inconvenienced!
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

Roger...I am sure that no caps are backwards...I've looked over and over....and over and over. <p>I'm taking the board to work with me tonight, and hopefully I can get some time to look into some of the tips you guys have given me.<p>Toejam...thanks so much for the in-depth instructions. Very understandable.<p>Thanks again guys. More results to come tomorrow.
Enzo
Posts: 276
Joined: Tue Sep 14, 2004 1:01 am
Location: Lansing, Michigan, USA
Contact:

Re: Uh oh...he's done it again...geez...

Post by Enzo »

In the field, many game operators have installed external wiring to the AC supply. The edge traces to the rectifiers burn up, so folks solder wires to the board and wire around the edge connector. A small Molex in that wiring makes it convenient to remove the board.<p>The ground or common traces on the board - two sides and two spaces at either end - tend to burn a bit on the wires from the power transformer CT. By wiring all the end trace wires together and even running a wire from the ones at one end to the ones at the other, we can spread the current and get some relief for the burned spots.<p>If the rectifier is melting off the board, it woul;d be a good idea to replace it even if it tests good, but it burnt up for a reason. Find that reason first. Measure resistance to ground from either end of the diode. If it end away from the current source is grounded, that explains the draw.<p>Same with the caps. Measure across each for resistance. A good cap will charge up from the meter and show an ever increasing resistance. If it shows negative resistance, that means there is voltage already stored in the part. If you measure a low resistance, then either the cap is shorted or the circuit it is in is.
User avatar
ModRob
Posts: 195
Joined: Sat Jan 11, 2003 1:01 am
Location: Chattanooga, TN
Contact:

Re: Uh oh...he's done it again...geez...

Post by ModRob »

Hmmm...I did more testing today...retouched the solder areas of the caps again. And the hot diode. I even replaced the feed from the edge connector with a detachable wire soldered directly to the associated diode pad. Upon last testing under power in the machine, nothing much new, but I noticed both diodes in question did not get extremely hot as before--they were warm to the point of almost not being able to hold onto them--and the area above them just under the edgecard connector was cooler this time. So I take it that helped. I'm going to replace the feed wires that go to the next tabs also.<p>I had the PCB on my bench, and tested those "glass bead-looking" diodes with my DMM on the diode setting and got a reading on both. I then checked each cap on the "ohm" setting and on most got an increasing reading. A couple did not show anything, but when I shorted the ends a couple of time, and then tested, I got a reading. One was very slow in increasing; the other was very fast. (not sure how to interpret that)<p>Again, none of the caps exhibit any warmth to my fingers. The large heatsink that surrounds a transister did have a little warmth to it in this last round.<p>Earlier, I was noticing that upon powering off the machine, the vido monitor would flash the play screen for a split second before shutting down. The other day it would flash on quickly and then go dark. Now it does nothing. And then once, while letting it sit in a powered state, I noticed a small, white dot about half the size of a pencil, sitting on the display. But after more tinkering, I don't see it anymore.<p>Enzo said...."Measure resistance to ground from either end of the diode. If it end away from the current source is grounded, that explains the draw." Can you explain this process of testing just a bit more, to make sure I'm not missing something? I'm sure it's simple, but I want to make sure my testing methods are fundamentally correct. And when I measured the caps, I did so on the board, and out of the board. I held each lead of my meter on the correct lead to get my readings. To short, I just connected a wire to both ends, assuming this would drain anything stored properly. But in some cases, after doing that, I still got no reading (00.0) A couple of more shorts, and then I'd get a reading....<p>Thanks again.
Post Reply

Who is online

Users browsing this forum: Ahrefs [Bot] and 26 guests