Creaky Old Award Winning Bastard Technologist Infinitus Excellentia Ion Laser Dominatus
Join Date
May 2007
Location
A lab with some dripping water on the floor.
Posts
10,046
Originally Posted by mophead
Not on hand (maybe in the junk box), but can easily get some.
How many?
I'd get two. wait, as I think this out, big Rs as loads are the wrong way to go...
Then some bias resistors for the gate current, I'd get 4.7 K , 2.2 K, 1 k, 560.
one or two of each, and 1/4th watt metal film.
And a 1k, .25 watt pot to adjust the bias.
Idea is to load test the transistors carefully using junk parts and a car battery or other source of dc power.
your gonna have to unsolder each base and emitter and clip the circuit in.
Digikey has waved their minimum order, so this should cost more to ship then its worth...
and a few 15 amp fuses in case some transistor goes nuts
Idea is to put two or four ohms in the emitter of each transistor to ground, and tie the collector to 12V PSU aka a car battery, which is 14.6 volts when charged. Then tie the base to each of the resistors in the series above for a few seconds and see how much current is passed.
So if you have a two ohm power resistor on a shorted car battery, your gonna see 7.3 amps for as long as its on. P=ie so 14.6 * 7 = 100 watts , so maybe we better do say 4 ohms or a car light bulb of some sort.
I'd use a small low wattage resistor and pulse it, but them I'm A. Nutz, and B. Used to doing this sort of thing by pulsing things so the resistor doesnt become a smoke emitting, light emitting, explosion. Hum, light, now there is a idea.
Lets see, do they still make 1157s? Yes, autozone says they do. A 1157 draws 2.10 amps, so that wont overstress the 2n6259s at 28 watts , that gives you 15-20 seconds without a heatsink, and if the transistor shorts, no kaboom.
Make a rig up with some allegator clips, so you can drive the base, starting with the highest resistor and working your way down.
If you want to get sophisticated you can sweep the base current with a pot and the 560 ohm resistor.
Idea is each transistor is a current amplifier,so beta for one of these things is 20-35 or so, so as you increase the base current by going to smaller and smaller bias resistors, you see more glow at the lamp.
If you ground the base you should see NO current.
See attached pic.
One of your passbank transistors is leaky and is biasing the supply on. If not, you have control card issues. This rig elimiates the leakers.
In reality you need a "curve tracer", but they are not common or cheap.
A 1157, a bunch of gator clips, and a few parts sets you back 20$ and rules out half the possible problems.
A 44 pin (aka a 22 edge ) connector for the control card is next on the parts list if this doesnt work, along with a bicolor led and some 10K pots and other parts.
ie a control card test rig.
There is another way before we go through all this first, and I'll try to type it up tonight,
Steve
Last edited by mixedgas; 10-27-2009 at 05:44.
Qui habet Christos, habet Vitam!
I should have rented the space under my name for advertising.
When I still could have...
Creaky Old Award Winning Bastard Technologist Infinitus Excellentia Ion Laser Dominatus
Join Date
May 2007
Location
A lab with some dripping water on the floor.
Posts
10,046
You were not there when we showed how to do this.
But you were cookin so you have a excuse.
This is the method usually used to find a shorted passbank transistor, but you may see the leaky one this way. This does NOT work for the two driver transistors.
Power down the laser. Unplug it.
Take a Ohm Meter, set it on the 200K ohm scale. Some work with 200K, some work with 20K, it depends on the bias current the meter's maker designed it with. We're depending on the fact that the passbank self biases on a little to protect itself from ignite pulses.
Clip one lead to the collector rail. This is the big fat silver wire that goes to the outer case of all the transistor. Pick your first transistor. Most people start at the top left and work their way left to right, then do the bottom row. Place the other lead to the emitter/ballast resistor junction of your first passbank transistor. This is the lead that goes from the 2 or 3 ohm resistor direct to the transistor. Each transistor has its own resistor, they electrically force the transistors to share current. Note the reading of the meter. Write it down. You should see quite a pattern. Do the remaining transistors in a like manner.
Reverse the meter leads. Do the same thing. Write down the readings. Bad transistors usually stand out like a sore thumb with a much different resistance rating. . Dead ones usually show up as shorted. Most have the same value within 1-2% unless people replaced a few transistors with a differnent lot.
Then check each ballast resistor for the proper reading. Then check the base resistors of each passbank transistor, usually they are around 150 ohms.
Make a sketch of the section around the two driver transistors. Unsolder the leads of one driver. Check it with the LBT for reverse leakage and make sure it works.Reinstall its two leads and then do the next one.
R60 is the zener diode for the passbank if I remember correctly. Its usually a 50V diode in a metal package mounted on the heatsink. Unsolder its lead.
Put your voltmeter on diode check. One way it should be .62V or so. the other way should be infinity. If you have a way of making 70V or so current limited to about 30 mA, you can make sure it shows zener action. But usually if it shows diode action its OK. It does not normally conduct when running.
With the laser running, observing proper HV safety precautions and having a safety person with you... Use a isolated or battery powered voltmeter, With the tube lit and the current up a bit, measure the voltage across the zener.
If it is the rated voltage of the zener, something may be wrong.
remember to disconnect both leads of your meter when hitting start!!!!!
Thats a beginning.
Steve
Qui habet Christos, habet Vitam!
I should have rented the space under my name for advertising.
When I still could have...
Hey, while its running, measure the voltage across each emitter resistor and see if they are equal. Might save you some time.
Would this save more time than checking each transistor again?
After I returned from SELEM, I went through each transistor in the passbank and none seemed out-of-line resistance wise with any of the others. I checked those when I was having trouble with overcurrent shutdown (discovered and corrected the tap setting was the problem). I found, at that time, either a cold solder joint OR a joint that had melted and wasn't making good connection on one of those resistors. I took that transistor and it's nearest neighbors and test them out of circuit. They tested fine with LBT.
Like I said in a PM awhile back; I'm glad that i wasn't getting spoon-fed because I have honest interest in learning this vs. blindly following directions.
Creaky Old Award Winning Bastard Technologist Infinitus Excellentia Ion Laser Dominatus
Join Date
May 2007
Location
A lab with some dripping water on the floor.
Posts
10,046
It may not find the problem but its a crucial clue if there is one.
Hey, make sure the AC volts test points are 130V or less. That might mean you have a mistapped passbank.
when I get back to akron I'll find the tap chart (nov 7 or so)
yes, it could save time vs testing each one. If one emitter resistor has blatently more voltage its a leaker. But it may not show if one has a C to B leak or short.
Steve
Qui habet Christos, habet Vitam!
I should have rented the space under my name for advertising.
When I still could have...
OK. Finally had some free time to do some checking. All these readings were taken in current mode with the pot all the way down. Laser had been running for about 30 mins to stabilize. LPM on the power supply indicated 1-1.25W
Originally Posted by mophead
Reply With Quote
