Amplifier Agony: More Vector Graphic 3 CRT Debugging (some questions for experienced TV techs)
Вставка
- Опубліковано 28 чер 2024
- The video information signal is runs through a common emitter amplifier on its way to the cathode. We see the signal at the base, but not the emitter or the collector. This meant the transmitter is suspect. A diode test on the BE junction of the transistor reads 530 mV (although that alone doesn't guarantee the whole transistor is working). We thought that the 680 ohm collector resistor had gone bad, since we were measuring infinite resistance on the back of the PCB, but I fixed that by just reflowing those joints. (Maybe other places on the board need reflowed?)
The power supply for the amplifier is 35 V, but we see a 300 mV gunk signal at the horizontal sync frequency (we think) on top of that -- we don't know if that's acceptable or not (if not, maybe a filter capacitor is becoming dodgy?)
This is the first CRT I've tried to debug, so any advice is appreciated! Normally I'd just get a new monitor, but since this CRT is integrated as part of the Vector Graphic 3 I'd really like to get it to work.
Note that when I say the anode is at 7 kV, I just took what we read on the meter and multiplied it my 100. I couldn't find any information on the Simpson high voltage probe we were using to know for sure what the scaling factor is (I assumed it was 100), and more to the point what input impedance it expected the meter to have. I measured the ancient Simpson meter we were using (it was the only one we found that had connectors that matched the probe) as having 5 Mohm input impedance. If the resistor in the probe was was chosen according to expecting a higher input impedance, then the actual voltage is higher. The voltage marked on the schematic was 11 kV. Since we are seeing traces on the screen we didn't pursue this issue any further.
Support this channel via a special purpose donation to the Georgia Tech Foundation (GTF210000920), earmarked for my work: • Support This Channel v...
0:00 -- Getting to the PCB
1:36 -- Grid voltages
3:03 -- Intensity adjustment
3:57 -- Rolling flicker
4:55 -- Input signals
6:15 -- Signal amplifier
8:30 -- Resistor is suss
9:19 -- Resistor is exonerated
9:33 -- Disappointment
9:59 -- Cathode & collector signals
12:06 -- Base & emitter signals
13:05 -- Is this ripple a problem?
15:25 -- BE junction test
15:50 -- Reflow all the joints?
A few tips from an old TV repair guy: use two probes on your scope, sync the sweep the H-sync pulse. Use the other channel to probe video. Then you can see the waveform of a line of scan (or more than one line if you slow the sweep). The "brightness" is too high, the diagonal white lines are re-trace. With the grid set so high, it won't actually display any video.
Also, don't overuse the Autoscale button. You'll confuse yourselves when the timescale changes. It's all over the place during your measurements. Just adjust gain and vertical position (or AC-couple the input) if you measure signals with different amplitudes.
1. Keep going!
2. Stop using autoscale, and learn to use the vertical scale and offset controls.
3. Trigger scope from either Hsync or Vsync, and measure using a second channel.
4. Clean the PCB, with isopropanol. That flux residue might not be the issue, but it isnt helping.
5. Don't guess. Unsolder the transistor and measure out of circuit.
6. Keep going! The tube is working, and most other parts can be replaced.
Hey all. The noise you're seeing on the power rail and elsewhere is probably due to the location of the ground clip on the oscilloscope probe. There's a lot of EMI being emitted by the flyback and it can easily be picked up. Chances are if you connect the probe ground to the ground trace on one part of the PCB and measure ground elsewhere, you'll see noise, even if they're the same trace.
I'm starting to wonder if there's a problem with the brightness pot on the back of the unit? If you compare the signal coming from the computer, the signal going into the brightness pot, and the signal coming off the wiper, and see if there's a major difference? Or even just measure it with an ohm meter. The signal on the base of the transistor seems very weak. Also, don't forget the CB junction! If it's shorted, you'll see a lot of voltage across the emitter resistor (and a lot of smoke), but if it's decided to go open, the emitter resistor will be loading the brightness pot and the signal will be weak like you're seeing.
The numbers on the CRT symbol represent the pin number on the neck. Same as a tube, except for the anode cap which has a voltage measurement. Confusingly, G1 is internally connected to pins 1 and 5, which is also labelled on the diagram. Best of luck.
From the schematic, there is no horizontal or vertical blanking implemented in the monitor. The brightness is normally set so that the raster is just below visability. I suspect that the video amplifier is open collector-base. There should be a large (10 to 30V) negative going signal at the collector.
Those aren’t G1/G2/G3 test points, they are the connection points for grids 1, 2, & 3.
Something going on in the flywheel sink circuit, you shouldn’t see the Fly back lines, the video output is stopped when the scan flies back to start a new line
I thought of that too, but it may just be the brightness is turned up too high. Worth double-checking. I know if you turn up the brightness too high, you will see the retrace lines.
@@davidjh7 that’s right when the brightness is too high, you see the fly black lines I’ve not worked on such a circuit since 1976!
Not in this particular monitor. On a TV, the blanking is often handled at G1 during retrace, since the input is composite video. This monitor relies entirely upon the signal being blank during retrace intervals. If the screen voltage is set too high, you'll see the retrace lines. Hence, it needs to be turned down more. But it is very good to confirm raster!!!
The numbers next to the crt are pin numbers, not voltages 3,4 are the heater nominally 12.6V, 2 is cathode (this also has the video modulation) about 20-30V, 1 and 5 are G1 about -5 to -20V depending on brightness control, 6 is G2 at about 500-600V, and 7 is G4 at -300 to +500 for focus (the tube has an Einzel gun where the focus electrode is near ground).
Dial the G1 (bright) down to where you just barely see the raster. Q101 is a common emitter amplifier, but it in this context, it functions more like a switch. A positive voltage on the base, biases the transistor on, which pulls the cathode negative, and current flows in the CRT. Remember, CRT works like any other tube, cathode high, no current, no cathode beam, no light on the screen. Cathode pulled towards ground, current flows, electrons beam from the cathode, the screen lights. The train of pulses get painted on the screen as the beam sweeps. Given the emitter resistor is 47-ohms, and the base has a 22-ohm resistor in front of it, Q101 gets biased on pretty hard for every lit pixel and completely off for a dark pixel.
The emitter of the transistor should show DC, due to the decoupling caps on it. Any signal other than DC should be assumed to be magnetic noise induced in your probe loop.
Also back in the day, we'd physically tug on suspect capacitors and bad ones frequently just pulled right off their leads. Low ohm resistors go up in resistance, and high-ohm resistors go down sometimes when they fail.
I admit I clapped to myself when you turned up the brightness control as I had recommended in the last video and you got raster. Since all the main supply voltages are derived from the fly back, transformer, the ground reference needs to be the signal ground, which may be at a different potential than the chassis ground---measure first.to make sure you don't have a substantial voltage difference using your meter. It would likely be worthwhile to use your scope in differential mode, two probes, add and invert the second channel (I know you know that routine) to isolate your ground reference from mains ground.. It is very possible that electrolytic filter on that supply is bad---heat in CRT monitors dries out electrolytic caps. There will always be some 115 KHz noise present, thus why you need to reference to the signal ground...If you are still seeing the 15 KHz ripples using differential mode, check the cap, or caps, including the emitter bypass caps. That ripple may be enough that it is swamping out the video signal is my initial feeling. Mire measurements are in order to get a clearer picture. Please let me know if I can offer anything else. or answer any other questions. I truly enjoy your videos, and your work a great deal, so I appreciate the chance to give back a little. You are doing great, BTW for your first foray into CRT circuitry. Nice work!
I could only fix it if i was there.
Just a matter of comparing the output signal to input signals of each stage and visualizing how the circuit should work. Sync circuits and oscillators need a bit of visualizing too.
Don't bother reflowing everything. ;) Just look for solder joints that are discolored, "crystalized", look like they've been hot or have obvious cracks around the lead.
When you measure video signal at the base of Q101 the vertical scale is 100mV/div and the video peak amplitude looks only about maybe 300mV? Q101 does not have DC bias so I suspect the input video amplitude is too low to open Q101. On the schematic at the video input to the Contrast potentiometer does it say 9V peak? If so, the contrast pot is either (almost) all the way down or shot. Might be the wiring around it or solder joints that are broken.
Is the signal at the base of that transistor biased high enough to turn on the transistor? At one point, it seemed to be just a few hundred mV from ground. I agree with the other posts about not using autoscale. I think it's better to learn how to adjust the controls properly yourself.