Atari Vector Monitor Repair/Upgrade

Contents:


9.8) TESTING TRANSISTORS WITH AN DIGITAL MULTI-METER

Set your meter to the diode test. Connect the red meter lead to the base of the transistor. Connect the black meter lead to the emitter. A good NPN transistor will read a JUNCTION DROP voltage of between .45v and .9v. A good PNP transistor will read OPEN. Leave the red meter lead on the base and move the black lead to the collector. The reading should be the same as the previous test. Reverse the meter leads in your hands and repeat the test. This time, connect the black meter lead to the base of the transistor. Connect the red meter lead to the emitter. A good PNP transistor will read a JUNCTION DROP voltage of between .45v and .9v. A good NPN transistor will read OPEN. Leave the black meter lead on the base and move the red lead to the collector. The reading should be the same as the previous test. Place one meter lead on the collector, the other on the emitter. The meter should read OPEN. Reverse your meter leads. The meter should read OPEN. This is the same for both NPN and PNP transistors. Thanks to Randy Fromm <YMNP18A@prodigy.com> for this excellent summary of the diode test method.

Anyway, there are 2 types of the large chassis transistors:

SPECIAL NOTE: Q705 and Q102 lie in such close proximity that it is not uncommon for their respective transistor sockets to be switched at some point which obviously swaps the corresponding functionalities (symptoms).

This gives reference frames for when the monitor is mounted horizontally (for games like Space Duel and Major Havoc) and vertically (for games like Tempest and Quantum).

For those of you without a manual, here is how to identify which one is which. The following descriptions are relative to the monitor oriented horizontally on it's base (like a TV would sit and like it is inside most games) and viewing it from the back side (so that you see the components and not the picture). 2 of them are on the outside of the chassis on the right side. The one nearest the back (component) side is Q606; the one nearest the front (viewing) side is Q605. On the inside of the bottom of the chassis are the other 4. The one on the far left nearest the HV unit is Q706. The one on the far right nearest the first 2 is Q103. The other of the 2 remaining middle ones which is closest to the back is Q102 and the last one which is in the same area but is closer to the front of the monitor is Q705.

An easy way to remember which kind of transistor goes where is to know that each connector has one of each kind; the 2N3716s are on pins 1, 3, and 4 (key at pin 2) and the 2N3792s are on pins 5, 6, and 7. If the wiring has not been altered, all 2N3716s are connected to the yellow, green, and red wires and all 2N3792s are connected to the purple, blue, and white wires. Here are 2 diagrams of the layout as viewed from the top:


+---------------------------------------+---------------------+
|          (front; Picture Tube)        |                    /
|               2   _                   |  2   _            /
|               N  / \                  9  N  / \          /
|               3 / Q \                 0  3 / Q \        /
|               7| 705 |                |  7| 605 |      /
|               1 \   /                 d  1 \   /      /
|               6  \_/                  e  6  \_/      /
|   2   _                    2   _      g 2   _       /
|   N  / \                   N  / \     r N  / \     /
|   3 / Q \                  3 / Q \    e 3 / Q \   /
|   7| 706 |                 7| 103 |   e 7| 606 | /
|   9 \   /                  9 \   /    | 9 \   / /
|   2  \_/                   2  \_/     b 2  \_/ /
|               2   _                   e       /
|               N  / \                  n      /
|               3 / Q \                 d     /
|               7| 102 |                |    /
|               1 \   /                 u   /
|               6  \_/                  p  /
|                                       | /
|                (back)                 |/
+---------------------------------------+

or, more simply (where 2=2N3792 and 6=2N3716):


            _________________
           /      FRONT      \
          |   Picture Tube    |
          |-------------------|
          +-----+   6  +-----+6
          |HV   | 2    |Def.2||
          |Cage |   6  |PCB  |2
          +-----+------+-----+-

These transistors often go bad and here is a quick lesson on how to check a transistor with a meter. Unplug the red plugs from the deflection board to isolate the transistor from the circuit. From the bottom of the transistor, the configuration is (Oh boy; time for a picture!):


          _
         / \
base -> /o o\ <- emitter

       |     |
        \   /<- the case is the collector
         \_/    

Pretty crude but you get the idea I hope. Test the transistors as described earlier, one by one, by placing the multi-meter leads on the tops of the pins of the red connectors or on the transistors themselves. Neither method requires the removal of the transistor from its socket so don't bother. You do, however, need to disconnect the associated connector from the deflection board prior to testing.

So what symptoms go with what? Well if only one of the deflection amplifying transistors goes then you will lose the corresponding half of the screen (for example if Q705 went bad in a Tempest game, you would be missing the top half of your screen). It won't actually be gone; it will be collapsed into a line across a 0 axis of the screen. If more than one goes or either of the power transistors (Q103 or Q102) goes you will get no picture because the spot killer turns on (and the LED on the deflection board lights) and shuts down the electron beam.

Checking these transistors is one of the first things you should do if parts of your screen are missing or your spot killer is on. Exact replacements are nice but I've successfully substituted others especially if you put the nonstandard transistors in the place of Q102 & Q103 as those two are for the + & - power supplies and not as critical (image-wise) as the ones that drive the deflection coils.

When replacing these transistors, make sure the socket is centered. Don't forget the little rubber or clear mica (plastic) insulator that was stuck to the chassis or to the bad transistor you just removed. You need to grease this insulator on both sides with a special grease called Silicone Heat Sink Compound. It is not strictly necessary but helps transfer the heat from the transistor to the chassis so the transistors last longer. The grease is pretty expensive so you may want to just buy new insulators (they should be about 10 cents each) because most come "pre-greased".

Don't be nervous about remembering the orientation; the pins are off centered in order to make the connection somewhat idiot-proof (see earlier ASCII picture. They are obviously "handed" but can be put in upside-down; it is a tight fit and hard to do but possible). Also, after you replace these and before you power the game up, use your meter to make sure none of the cases of the transistors are shorted to the chassis. This is very common and will cause fuses to blow and will probably destroy the transistor. If any part of the transistor is touching the chassis you are asking for trouble.

On the deflection board, the most common failures are Q100 and Q101 and when these go they take R100 and R101 (respectively) with them. These transistors are part of the + & - power supply circuit and are often bad with the resistors really burnt. To properly test these, they should be desoldered and removed from the PCB. Even in-circuit and not isolated from other components you can still get a pretty good idea with a multi-meter (analog meters work best since digital meters show infinite resistance most of the time) if the transistors are bad as transistors tend to fail catastrophically. In other words, they usually completely short (0 Ohms) or open. If you see 0 Ohms where there should be an open circuit or 2.5K Ohms, then the transistor is probably shorted. If you see greater than 2.5K Ohms when the reading should be in that range the transistor is probably open. When these transistors are bad, they usually show a large crack in the case if you look closely at them.

Replace Q100 with the same type transistor but if Q101 is bad (and even if it is not), I STRONGLY suggest you upgrade it to a larger transistors that will handle more current. I ALWAYS replace Q101 with a TO-202 package instead of the much smaller TO-92 package that the board comes with. In fact, the P327, and P339 versions were manufactured with this upgrade. The 3 labels I have seen are 119;NSDU57 (National Semiconductor) and M152;126-1A ("brandless"?), and NTE50 (NTE) [semi-colons denote a new line of text found underneath the previous line]. I always use NTE50s as they are relatively cheap and very easy to find.

Even though these transistors (and resistors) are the most common failures on the deflection board, I have NEVER seen them go bad after Q101 was upgraded to a NTE50. If you can't find the generic parts, a common modern day replacement for MPSA06 (Q100) is NTE287 and a common modern day replacement for MPSA56 (Q101) is NTE159 (but I cannot stress enough the utility of going with the larger replacement for Q101). Upgrading Q101 is always the first thing I do to any deflection board I get; replacing it before it fails saves me from having to replace the other parts that go bad when it does fail.

Once in a while you will see D105 or D104 open or shorted, too. Also watch for broken solder joints at the base of the connector pins for all the major connectors (especially on the deflection boards of black and white Wells-Gardner units which almost always are bad). You tend to rock the plugs back and forth when you pull the connectors and this often cracks the solder joints to the circuit boards which were poorly soldered to begin with. You may also want to check each pin for continuity with the next component on its trace line (and also with adjacent pins), and remove and resolder any dubious connections. While you are at it, resolder the 3 pins of the PTC thermistor (in the middle along the left edge of the deflection board) as they are almost always loose for some reason.

It is unusual but sometimes some of the 4 heat sinked (or is that "heat sunk"?) transistors on the deflection board will die. If any of these is bad, you will usually get no picture at all but you will see "background brightness" that lets you know some electrons are being thrown at the tube. Be sure to check the resistors and diodes around any bad transistors you find.

Usually, the electrolytic capacitors are still OK (though always be suspicious of electrolytics drying up and loosing micro-Farads). If the transistors in the chassis are OK, most missing pictures are due to problems with this board (whereas most distorted pictures are due to bad capacitors in the HV unit). I frequently see HV units with the protective shield removed from the case and discarded in attempt to keep it cooler. I like to keep mine on because I rate dust and foreign object attraction as a more severe problem than heat for this board but like I have said, I don't let my stuff run hot.

Very infrequently, you may have problems with some other transistors in the X/Y amp section of the board. Most of the rest of the transistors that populate the deflection board (Q600-602,700-702) are type TPS98 and are not easily found anymore. The good news is that TPS98 is equivalent to the PN3569 and the ECG or NTE 194. All Electronics (800-826-5432) has PN3569 transistors at 5 for $0.50. NOTE also that All has just dropped their minimum order requirement (still $5 S&H on most orders, though).

The neck board very seldom has problems. The few I've seen are from mishandling where someone has broken some of the pots that control the RGB drives. Check the pots if you are missing a color. If you are blowing your 5A fuses and your other boards are OK, check C503 (33 uF @ 250V). If this is open, shorted or cold soldered, it will cause the 5 (or 6.25 depending on your board) fuses to blow.

Last is the HV supply. I've worked on lots of these and have only seen one of the infamous bad HV transformers. Normally HV failures are due to a semiconductor or capacitor failure. I must mention that the HV unit can put out very weak X-ray radiation and and creates lots of ions in the air but both are in sufficiently small quantities that there is little risk of injury. Nonetheless, many people feel a tad queasy during or after working near high voltage. Like anything else, it gets less noticeable with successive exposures. Many people are quite worried the first time they experience the sensation but it is harmless. Please note that I am not a doctor nor a radiation specialist so refer back to the CYA clause at the beginning of this document. Certain unlikely faults that I have never run across could cause large amounts of X-rays to be produced which would definitely be harmful.

Before you work on this beast, discharge the tube as it can really zap you even when turned off (it is extremely inadvisable to work on the HV section while the machine is on or even plugged in; always unplug the game before you go in here). Connect a clip lead between the chassis and the shaft of a long narrow plastic-handled screwdriver. Work the end of the screwdriver under the big suction cup on the top of the tube until you hit metal. There will often be a snap (from the spark) as the HV runs at around 20 thousand volts. Just go slowly and use only one hand. It won't bite as long as you are careful. You may want to leave the game plugged in BUT TURNED OFF for this step so that you can ground to the earth instead of just the chassis. If you plan to remove the HV unit, you must disconnect the secondary anode from the tube which is a little tricky. Under that suction cup is a double-barb. The barbs extend perpendicular to the wire as it enters the suction cup. Simply pinch the suction cup as best as you can and wiggle it back and forth while pushing in at the edges and pulling out at the center. It will eventually come off. Here is a cut-away depiction of the anode; the wire will run perpendicular to this view (i.e. towards/away from the reader):


                         |    /\   /\    |
                          \     \ /     /
                           \____|_|____/

When I was in the habit, I discharged the tube with my HV probe so I could watch the voltage go down as the internal resistance of the probe bled the voltage off slowly. If the snap bothers you, put a resistor in series with your clip lead to drain off the voltage more slowly. If you don't have this equipment available, a 1 or 2 hour wait after unplugging the game should be sufficient for most of the excess charge to bleed off naturally. NOTE: The anode is designed to hold onto charge so it will NEVER discharge completely without being deliberately grounded. In fact, if you wait a couple of days after grounding it, it will actually build up another (small) charge! I suggest that if you go through the trouble of discharging it, you should keep it grounded by using an alligator clip on the end of your wire and leaving it connected to your ground. This way you have no chance of getting shocked (just be ABSOLUTELY CERTAIN to remove this ground wire before you turn the power back on again).

Personally, I never worry about any of this anymore since the shock (when received properly) is harmless and mostly painless (I have gotten zapped dozens of times). Just make sure the game is unplugged first. If you aren't going to wait, follow the "1 hand rule"; only use one hand when disconnecting the cable and keep the other one in your pocket or in the air. The only likely way for the charge on an unplugged machine to hurt you is for it to form a circuit between your 2 hands (one grounded and one on the tube) sending current directly across your heart; these levels could easily interfere with your heartbeat and even cause it to stop! I should also point out that black and white monitors use significantly lower high voltage than do color monitors and raster monitors have high voltages that are and order of magnitude higher than vector monitors use. I have been shocked by them all and assure you that rasters DEFINITELY have a bigger sting than do vectors so beware!


Go to [Next] segment
Go to [Previous] segment

Go to [Table 'O Contents]


Please check attribution for Author. Processed by filipg@paranoia.com [Feedback Form] [mailto]. The most recent version is available on the WWW server http://www.paranoia.com/~filipg/ [Copyright] [Disclaimer]