Just thought of something regarding my K6100...

TheDrewster

Well-known member

Donor 2012
Joined
Oct 5, 2008
Messages
7,668
Reaction score
314
Location
Reno, Nevada
I originally replaced all the frame transistors but left the old spacers the way they were. When I went to install my newly rebuilt K6100 chassis, I ran to radio shack to get some heat sink compound and replace the spacers with ones I got from Bob Roberts. After doing so, I lost my high voltage completely.

Here is what I want to know, straightforward. How do I correctly apply compound to these transistors, and how to I test them to make sure I didnt short them or completely blow them?
 
Do you still have deflection? Is the spot killer on? If your getting deflection, probably your answer lies in a bad connection somewhere else, that just happened to rare its ugly head.. Other then that, make sure nothing on any of the bottle caps are shorted to the frame. If you have doide test on your meter, check collector to emitter and base to collector and base to emitter. The meter should read open in one direction and .7ish in the other.

Andrew
 
Transistor Testing

These are the same transistors on Asteroids machines as well. You would apply the compound to the spacer and then place the transistor onto the freshly applied compound.
Secure to the chassis. Set meter to continuity. Either lead can go onto the transistor outer case an the other lead to the chassis. If beeping, it's shorted to the chassis. Re-install.

As for testing the transistor to ensure they are all ok, there are two types of transistors mounted on the chassis. There is a base, emitter and collector. These are diodes so place the meter in diode test. Then place one lead on the base and the other lead on each lug. You should get 4 - 6 ohms. If under 4, it's most likely bad. You should also touch both legs, red on one and black on the other. You should also get a reading. Check Marvin pinball website for TO-3 diode testing.

John
 
Do I apply the compound to both sides of the spacer? Does it matter how much I apply?
 
Definitely white. Should it matter if I got compound on the terminals of the transistor?
 
Definitely white. Should it matter if I got compound on the terminals of the transistor?

I don't suppose getting any on the legs of the transistor would be bad, I definitely wipe the excess off if I do happen to get it on the legs. I use rubbing alcohol and Q-tips/soft-ish paper towel for wiping old paste/excess paste off, as it's the only way that really cleans it off.

considering you're putting it on the base of the transistor though, is it bad? I don't know. you didn't get any in the solder joints right?

testing transistors is interesting. you put one probe on the base (the middle) leg, then probe the legs on each side of it. then flip the probes around and do it again. on an NPN transistor, if I'm understanding correctly, you should get negative readings on the outer 2, and positive in the middle. then when you go the opposite way, they should read open.

don't quote me on that though, google it first, cause I suck with electronics. :)

best bet is to take a known working good transistor and use that for your values to compare against.

as for the proper way to apply the thermal paste, you just put a little dab (some may call this the size of a grain of rice..?) on the transistor... affixing it to the heatsink again will smear it around rather effectively. you never want to put too much, otherwise it won't actually transfer the heat away, it'll kind of get stuck at the heat source (the transistor in this case) and burn it out faster.
 
Yes, put a very small dab of white goo (BB size or smaller) and smear it very thin on the transistor. Put the mica insulator on. Put another very small dab on the now installed mica and smear it all around. Grab a rag and rub off all of the old heat sink compound that is on the metal frame of the monitor....both sides. Should look nice and shiny. Place the transistor in the holes and place the socket over the pins. Carefully line up the brown socket "lugs" with the holes at each extreme end of the openings. Push the transistor and socket together so they are a snug fit. Install the screws and tighten snugly. Don't torque them down hard. Just snug and then a bit more.

Put your multimeter on continuity or "beep" test and touch the black probe to the frame and the red probe to each pin on the back of the socket and then the outside top of the transistor case. If you get a beep, you don't have the transistor installed properly because it is grounding out somewhere. The sockets frequently get the lugs mashed and the screws then make contact with the frame. The only thing you can do is order replacement sockets from Mouser.

The easiest way I have found to test the transistors is to do it when they are installed. I do this test before removing them and after as well. There are six different measurements you must take on the three soldered pins on the back of the socket. Set your multimeter on diode test/transistor test. These instructions are for a digital MM. Put the black probe on one pin (I start at the leftmost) and the red probe on the pin next to it. There should be a reading of 0 or .547 or something close to that (maybe .457...depends on your meter). Keep the black probe in place and move the red probe over one pin. Again 0 or .547. Now keep the red probe in place and move the black probe over to the pin the red probe first started out on. 0 or .547.

Whew, you are halfway done.

Now....put the red probe where the black probe originally started on and the black probe where the red probe started on. Read 0 or .547. Move the black probe over and read 0 or .547. Now move the red probe to the spot the black probe just left and read 0 or .547.

Right about now there are 60 guys screaming WRONG WRONG WRONG! and typing a post.

Now let me finish.

You should have 4 total readings of 0 and 2 readings of .547. If you have anything other than that, then the transistor is bad or you touched the probes to each other when you did the test. I have found that .001 is a frequent reading and that is for sure a sign the transistor is bad. You really don't care. As long as you have those readings you are good. (Assuming you have the transistors in their proper spots)

If you are sharp, you will notice that the readings between the 3716 and 3792 transistors are opposite each other.

To test the 4 transistors on the bottom of the frame, I put a towel under the face of the tube and tip it over so that I have access to the bottom. You are going to have to get there anyway to get the old comoound that oozed thru the holes 30 years ago off the frame anyway. I also have a rubber mat I put the tube on to protect it from damage but you could use foam rubber or cardboard. Just don't neck the tube. You can do this with the monitor in a Tempest without any difficulty as the 4 transistors are sticking right in your face and the other 2 are easily accessible. Finally.........You have to do it with the transistor's connectors disconnected from the Deflection board or you will get a wrong reading.

I have been hoping to make a video showing people how easy this is but just haven't had the time or a helper to hold the camera. Maybe soon.

Good luck.
 
Wrong, wrong, wrong! ;) j/k

One thing I like to do is to test the transistors from the connector that plugs into the deflection board. This way you are testing the transistor, cable, and connector. If something is wrong you have to troubleshoot, but it is much easier to test this way. I have posted the pin by pin instructions a number of times, I will see if I can find them.


Ok, here they are.


You should unplug all of the chassis mount transistor cables form the deflection PCB and test them with a multi-meter set on the diode function. It is really easy to test them this way since you can just put the probe tips on the little bit of metal that you can see exposed on the side of the connector that attaches them to the PCB.

If you get any out of the ordinary readings, you will need to check the mounting of the transistor or swap them out.



NPN (2N3716) Test on connector at Pin positions 1, 3, 4 of J100, J600, J700
===========================================================================

Test #1
-------------------
Red -> Pin 3
Black -> Pin 4

Result: .45v to .9v

Test #2
-------------------
Red -> Pin 3
Black -> Pin 1

Result: .45v to .9v

Test #3
-------------------
Red -> Pin 4
Black -> Pin 3

Result: Open

Test #4
-------------------
Red -> Pin 1
Black -> Pin 3

Result: Open

Test #5
-------------------
Red -> Pin 1
Black -> Pin 4

Result: Open

Test #6
-------------------
Red -> Pin 4
Black -> Pin 1

Result: Open



PNP (2N3792) Test on connector at Pin positions 5, 6, 7 of J100, J600, J700
===========================================================================

Test #1
-------------------
Red -> Pin 6
Black -> Pin 5

Result: Open

Test #2
-------------------
Red -> Pin 6
Black -> Pin 7

Result: Open

Test #3
-------------------
Red -> Pin 5
Black -> Pin 6

Result: .45v to .9v

Test #4
-------------------
Red -> Pin 7
Black -> Pin 6

Result: .45v to .9v

Test #5
-------------------
Red -> Pin 7
Black -> Pin 5

Result: Open

Test #6
-------------------
Red -> Pin 5
Black -> Pin 7

Result: Open


Yes, put a very small dab of white goo (BB size or smaller) and smear it very thin on the transistor. Put the mica insulator on. Put another very small dab on the now installed mica and smear it all around. Grab a rag and rub off all of the old heat sink compound that is on the metal frame of the monitor....both sides. Should look nice and shiny. Place the transistor in the holes and place the socket over the pins. Carefully line up the brown socket "lugs" with the holes at each extreme end of the openings. Push the transistor and socket together so they are a snug fit. Install the screws and tighten snugly. Don't torque them down hard. Just snug and then a bit more.

Put your multimeter on continuity or "beep" test and touch the black probe to the frame and the red probe to each pin on the back of the socket and then the outside top of the transistor case. If you get a beep, you don't have the transistor installed properly because it is grounding out somewhere. The sockets frequently get the lugs mashed and the screws then make contact with the frame. The only thing you can do is order replacement sockets from Mouser.

The easiest way I have found to test the transistors is to do it when they are installed. I do this test before removing them and after as well. There are six different measurements you must take on the three soldered pins on the back of the socket. Set your multimeter on diode test/transistor test. These instructions are for a digital MM. Put the black probe on one pin (I start at the leftmost) and the red probe on the pin next to it. There should be a reading of 0 or .547 or something close to that (maybe .457...depends on your meter). Keep the black probe in place and move the red probe over one pin. Again 0 or .547. Now keep the red probe in place and move the black probe over to the pin the red probe first started out on. 0 or .547.

Whew, you are halfway done.

Now....put the red probe where the black probe originally started on and the black probe where the red probe started on. Read 0 or .547. Move the black probe over and read 0 or .547. Now move the red probe to the spot the black probe just left and read 0 or .547.

Right about now there are 60 guys screaming WRONG WRONG WRONG! and typing a post.

Now let me finish.

You should have 4 total readings of 0 and 2 readings of .547. If you have anything other than that, then the transistor is bad or you touched the probes to each other when you did the test. I have found that .001 is a frequent reading and that is for sure a sign the transistor is bad. You really don't care. As long as you have those readings you are good. (Assuming you have the transistors in their proper spots)

If you are sharp, you will notice that the readings between the 3716 and 3792 transistors are opposite each other.

To test the 4 transistors on the bottom of the frame, I put a towel under the face of the tube and tip it over so that I have access to the bottom. You are going to have to get there anyway to get the old comoound that oozed thru the holes 30 years ago off the frame anyway. I also have a rubber mat I put the tube on to protect it from damage but you could use foam rubber or cardboard. Just don't neck the tube. You can do this with the monitor in a Tempest without any difficulty as the 4 transistors are sticking right in your face and the other 2 are easily accessible. Finally.........You have to do it with the transistor's connectors disconnected from the Deflection board or you will get a wrong reading.

I have been hoping to make a video showing people how easy this is but just haven't had the time or a helper to hold the camera. Maybe soon.

Good luck.
 
Last edited:
Also, did you reflow the solder on all of the header pins on the deflection board? Plugging and unplugging will agitate cold solder joints and make them go from working to not and vice versa. Best to reflow them if you have not.
 
Good post on the connector cables as well. Yes, I do that too after I check the transistors as 90% of the time it is the transistor that will be bad. I have found several broken solder joints on pins/sockets. Most of the red connectors and wires that plug into the deflection board have been good unless some yahoo really jerked on the wires when unplugging them. Nevertheless, it is a good idea to make sure you have good readings so that you check the transistor, connector, cable and socket. Again, some day I hope to make a set of short videos about 6100 vector monitor repair to help the newbs out. Once you do a dozen or so, it gets a little easier. With that being said, I still have 2 deflection boards that are giving me the fits. Currently out of town but can't wait to get back to them. I love a challenge.
 
Back
Top Bottom