• Please review our updated Terms and Rules here

Commodore 3016 scrambled screen

So the key statement there is "brightness has no effect".

Does the brightness control have an effect on the monitor image when the PET is working though?

Dave
 
Which of the PET monitors have you actually got inside your machine?

Let's find the correct schematic...

Dave
 
TheMatrix said:
Checked C1 and C27 both ok. Replaced C22 by 330uF but no effect.
Brightness has no effect.
It's standard power supply.
With other mainboard same effect.
Click to expand...
As an experiment, can you disconnect one leg of the diode CR5 and the Collector of transistor Q2.

Turn the brightness control to minimum. The screen should be completely blacked out because the CRT's cathode will be sitting around +33v and its grid around -30v (making the relative voltage grid negative with respect to cathode about 63v). You could also check that with the meter.

Then turn the VDU off and see if the bright flash has gone away, or still appears at turn off ?

(I'm wondering if there might be some leakage in the diode CR5 or Q2, but if the effect is still there, it will require some investigations of the way in which the grid-cathode voltages are falling at turn off)

This is the same VDU unit which had problems with Q1 and some diodes in the area, and a failure of Q2 to switch correctly ? if this is the case there could be more problems in this area. It might be worth replacing Q2 and CR5, but first lets see what the experiment shows.
 
Last edited:
Hugo Holden said:
As an experiment, can you disconnect one leg of the diode CR5 and the Collector of transistor Q2.

Turn the brightness control to minimum. The screen should be completely blacked out because the CRT's cathode will be sitting around +33v and its grid around -30v (making the relative voltage grid negative with respect to cathode about 63v). You could also check that with the meter.

Then turn the VDU off and see if the bright flash has gone away, or still appears at turn off ?

(I'm wondering if there might be some leakage in the diode CR5 or Q2, but if the effect is still there, it will require some investigations of the way in which the grid-cathode voltages are falling at turn off)

This is the same VDU unit which had problems with Q1 and some diodes in the area, and a failure of Q2 to switch correctly ? if this is the case there could be more problems in this area. It might be worth replacing Q2 and CR5, but first lets see what the experiment shows.
Click to expand...
I did the steps and if i turn it on it's black and no flash when i turn it off
 
TheMatrix said:
I did the steps and if i turn it on it's black and no flash when i turn it off
Click to expand...
Hmmm, there must be something going on with either CR5 or Q2. Possibly Q1 too.



As a start, replace CR5 and Q2 with the correct parts. With everything connected as normal and test again.
 
TheMatrix said:
I replaced CR5 (1N4148) But no change. Q2 is a 2N4401, i don't have this part now. So i need to order this one.
2N4401 can be replaced by 2N2222?
Click to expand...
Yes, in theory that should be a reasonable replacement. If it is a genuine 2N2222 like the ones I suggested from National Semiconductor and not a fake part from elsewhere.

Probably though, if I could not get the original 2N4401 part, I would replace it with a Fairchild MPSA42.

If replacing CR5 and this Q2 doesn't fix it, then we will have to move to Q1 and the other diodes.

The experiment confirmed that in the case where the VDU was de-powered and the video voltage, at that moment on the cathode, was in the region of +30 to 35v, that there was no bright flash and at that time CR5 and Q2 had no contribution to the problem, because they were disconnected. Meaning that, in your VDU at the moment the VDU is switched off, there is something odd about the video stages, Q1, Q2 driving the CRT's cathode.
 
We know so far that if Q2's collector is disconnected at the time the VDU Is switch off, the initial bright flash is gone because the CRT cathode stays at about +30 to + 35V for long enough to prevent it. This also appears to be the case in most PET VDU's at least when the collector is connected and the unit is in the process of powering down, Q1 must be conducting for long enough and robbing Q2 of base current at least initially until the current flow is depleted. Q1 gets its base current via the +12V supply and the two series diodes.

It is almost as though the +12V rail in the VDU is collapsing more rapidly than usual.

As an experiment can you tack another large electrolytic capacitor across the main power supply filter capacitor in the VDU, of about the same value, to slow the rate of fall of the supply at power off and see if that has any effect at all.

Otherwise the next move would be to fit the exact original transistors for Q1 and Q2 and see if that had any effect.
 
Last edited:
Hugo Holden said:
We know so far that if Q2's collector is disconnected at the time the VDU Is switch off, the initial bright flash is gone because the CRT cathode stays at about +30 to + 35V for long enough to prevent it. This also appears to be the case in most PET VDU's at least when the collector is connected and the unit is in the process of powering down, Q1 must be conducting for long enough and robbing Q2 of base current at least initially until the current flow is depleted. Q1 gets its base current via the +12V supply and the two series diodes.

It is almost as though the +12V rail in the VDU is collapsing more rapidly than usual.

As an experiment can you tack another large electrolytic capacitor across the main power supply filter capacitor in the VDU, of about the same value, to slow the rate of fall of the supply at power off and see if that has any effect at all.

Otherwise the next move would be to fit the exact original transistors for Q1 and Q2 and see if that had any effect.
Click to expand...
I did this:
As an experiment can you tack another large electrolytic capacitor across the main power supply filter capacitor in the VDU, of about the same value, to slow the rate of fall of the supply at power off and see if that has any effect at all.
And also no effect
 
OK. Get and fit the original type transistors, double check all of the diodes in the area and try again.

After that I'm not sure what to do, except if you have a working VDU that doesn't do it, you could make a scope recording of the CRT's cathode voltage in the two units, and confirm that the defective one is falling faster at turn off. Then try to figure out why.
 
I went to one of our power station sites last week - and one of the interesting things that they suspected about a transistor that they had replaced in a faulty card that sometimes worked and in other cards not is that the original manufacturer had selected the specific transistors they used for this particular location based upon test (a specific maximum leakage limit).

If the above is true, then a replacement transistor from the same manufacturer and part number may work - but also may not - dependent upon the leakage current parameter.

The particular fault on our site did not materialise until 2 to 3 weeks of operation (presumably the transistor needed to 'bed in').

I wonder if Commodore did a similar 'trick' here?

You may have to use a specific transistor and select 'by test' against some acceptance criteria that was never published.

Dave
 
I feel such a idiot :oops:. I was going through it al again. Frist C22.... but when i was working on it the first time for the spot repair, i had everyting in reverse in front of me.
I thought i replaced C22, but instead i replaced C28....
Now i put the original back in C28 and replaced the correct C22 :oops: and now the spot/dot is gone.
 
Of note, C28 a 10uF special bipolar capacitor, it is a very special capacitor with a very low ESR to couple the H yoke coils , it does pay to keep the original one in there.

It pays to replace C22 with a 220uF capacitor and that will avoid the "late" turn off spot. Make sure it is installed correctly with the + terminal of the capacitor connected to ground.
 
Hugo Holden said:
Of note, C28 a 10uF special bipolar capacitor, it is a very special capacitor with a very low ESR to couple the H yoke coils , it does pay to keep the original one in there.

It pays to replace C22 with a 220uF capacitor and that will avoid the "late" turn off spot. Make sure it is installed correctly with the + terminal of the capacitor connected to ground.
Click to expand...
It's installed correctly. I always look twice for + and - ;)
 
You must log in or register to reply here.
Back
Top