• Please review our updated Terms and Rules here

Commodore PET 2001-8 text wobbling on screen

The only "PET power supply" that is actually connected to the monitor PCB itself is an AC supply directly from the transformer - so no active or passive components at all... The rectification, smoothing and voltage regulation is all done on the monitor PCB. Yes, you can feed an external AC or DC supply into the monitor PCB itself.

Yes, you can separate the PET monitor from the PET case and extend the power and signal cables if you like. In fact, this is a good way of making sure that it is not magnetic interference from the transformer affecting the monitor image.

However, as Hugo has already stated, this can usually be done by viewing the monitor with the PET case closed and then open - as you have changed the distance and angle between the transformer and the CRT.

Can you post a video that I can actually view? I can't look at your previously posted videos (unfortunately). The office IT blocks them...

If you think it is power supply related - it is likely to be a component on the monitor PCB itself, and a simple probe with an oscilloscope should identify this. Set the timebase to view a 50/60 or 100/120 Hz signal and attach the oscilloscope probes to the DC voltage rail after the monitor voltage regulator.

I would also view the stability of the HDRIVE signal on the oscilloscope.

R20 is a cheap commodity item. If you have already removed it - and are having problems with it - replace it. Simple as that and then we can move on...

Basically, the AC supply arrives at the monitor PCB from the transformer.

Diodes CR19-CR22 are a full-wave bridge rectifier to convert the AC voltage into a DC voltage.

Capacitor C1 smooths the pulsed DC from the bridge rectifier and feeds the input to the 12V voltage regulator (VR1).

The 12V output from VR1 is smoothed via capacitor C3.

Basically, measure the DC voltage across first C1 and then C3.

We are looking to see any sign of 50/60/100/120 Hz ripple on both the input and output side of the VR1 12V voltage regulator.

EDIT: The Horizontal drive also takes the +12V rail and has a series resistor (R34 10 Ohms) and smoothing capacitor C17 all of its own. So it would be worth measuring the DC voltage across C17 - also looking for ripple.

hi Dave,
I have swapped R20 potentiometer and no change. I will look for other advices.thanks
Last edited:
I have measured DC voltage across C17. This is how it looks like. Is it normal?
I also did 3 tests to check magnetic interference dependency, because something was not clear to me. The results really confused me and surprised me.
  • Test 1 - I put monitor in its oryginal position (on PET case) ant turn power on. I was able to see whole text on the monitor wobbling a little.
  • Test 2 - I connected monitor outside PET case with longer cable adapter (0,5m) to further distance between monitor and power transformer. I was able to see only first line (****Commodore Basic****) wobbling a little. other lines weren't.
  • Test 3 - This is Test 2 scenario + I powered on bench DC power supply that was standing next to CRT (10 cm from the front of the monitor). I saw text wiggling on monitor too.

The question is, can we eliminate it somehow? Is this a result of one of the monitor PCB compenents that is likely sensitive and vurnelable for such interference and do such bad job?

When I turn on PET I can hear a litte humming sound from main transformer.
Have you done the same with C1 and C3 before we start delving inside the circuitry

A story has a beginning, a middle and and end.

C1 is the beginning, C3 is the middle and C17 is the end.

A small amount of transformer hum would be normal.

Some new measurements:
1.AC voltage across CR19,CR21 (before rectification):
2.DC voltage across CR19,CR21 (after rectification) and that goes to C1 (+):
3. DC voltage across C3 (after regulation from VR1):


  • SDS00008.png
    28.3 KB · Views: 1
1. I don't think the AC voltage measurement is correct. On the negative half cycle, the voltage only goes to -1.4V. I suspect this to be a measurement error issue with the oscilloscope probe ground clip. This clip will be connected to mains earth internally...

Using a multimeter set to AC Volts would be a much better way to measure this.

If you want to measure it with an oscilloscope, make sure the transformer winding is not connected to mains earth at either end, and measure the voltage on the oscilloscope with it NOT connected to the monitor.

2. Looks fine. You can see that one pair of diodes has a slightly lower voltage drop than the other pair - not that this is a problem of course.

3. Looks well fine.

Hi Dave,

AD. 1. You are right I connected ground probe clip to PCB ground.

Do you have any other ideas what to check? Is there any other suspicious component I need to check for this behavior? Thanks
I have asked for a video of the problem that I can view...

Post #9 had a 'wobble' on your external monitor using the RGBtoHDMI converter. This was 'fixed' by post #10.

Unfortunately, I can't view the video in post #12. My laptop and phone just wants to download the MOV file - not view it... Any chance of posting a little video elsewhere (YouTube?) so that I can see what your problem looks like on the PET monitor?

I also asked to see what the stability of the HDRIVE signal is like.

To do this, measure VDRIVE on one channel of your oscilloscope. Set this channel to be the trigger for the oscilloscope.

Measure HDRIVE using the second channel of your oscilloscope.

Adjust the oscilloscope timebase to see a number of HDRIVE pulses on the screen - and see how stable they are with respect to the VDRIVE signal.

If this HDRIVE signal is 'solid' wrt the VDRIVE signal - the problem must be in the monitor.

In this case, the interaction should either be power-supply related (ruled that out) or magnetic interference.

You stated that moving the monitor outside the casing did have an effect (post #26 test 2). This does imply magnetic coupling...

Your metal case (of the PET) is actually earthed to the mains earth isn't it?

I hate these types of problems :)...

I also asked to see what the stability of the HDRIVE signal is like.

To do this, measure VDRIVE on one channel of your oscilloscope. Set this channel to be the trigger for the oscilloscope.

Measure HDRIVE using the second channel of your oscilloscope.

Adjust the oscilloscope timebase to see a number of HDRIVE pulses on the screen - and see how stable they are with respect to the VDRIVE signal.

If this HDRIVE signal is 'solid' wrt the VDRIVE signal - the problem must be in the monitor.
Hi Dave,

Thank you for all advices. I really appreciate.
I apologize for not answering. I was on holidays, so I had no time to make further tests.
Please see video with my situation (I uploaded it to YouTube):
Video of the PET text wobbling

During this video the PET monitor was connected outside the PET the casing. The monitor metal case was connected to casing via separate wire. The casing was earthed - I check continuity test on my multimeter between earth pin in the power plug and the monitor metal case.

I also did a test with an oscilloscope on the HDRIVE and VDRIVE according to your suggestions.
- CH1(yellow) probe is connected to VDRIVE signal on monitor PCB pin 5 on J1 connector.
- CH2(purple) probe is connected to HDRIVE signal on monitor PCB pin 7 on J1 connector.
- GND of both probes are connected to brightness R9 potentiometer metal legs.
Here are some screens with results (different versions with different timebase):


You have shown static photographs of the oscilloscope traces.

A static photograph does not tell us whether or not the signals are static or varying.

To be honest, I can't see any (excessive) wobbling in the video you posted either.

Hi Dave,

Thanks for quick answer.On the video you can see text shaking a little (pulsing). It is normal? Maybe I'm looking for something that is normal in all PET from these years? But please zoom the video on youtube and you can notice this pulsing.
I will have a look on the iMac tomorrow rather than my phone. My phone does not zoom in very much...

I can see 'something' happening in the video - but I also see the text 'colour' changing gradually from white to saturated cyan, so I am wondering what is real and what is induced by the camera taking a photograph of the PET screen itself. IT is very slight, nonetheless...

If I remember correctly, the effect was worse when the monitor was installed into the PET itself. Yes?

However, can you first take a video of the HDRIVE/VDRIVE signals on the oscilloscope in the configuration above and a separate video of the DC power rail at the monitor itself (as displayed on the oscilloscope screen)?

We need to find out whether the instability is being induced via a slight instability of the VDRIVE or HDRIVE signals from the PET logic board, or via the power supply in the monitor, or externally via magnetic interference.

Hi Dave, Thanks and very appreciate any advices.
So do you think that changing this R20 will not help with text wobbling? When I was touching this (with finger or plastic screwdriver )I noticed screen pulsing, and when changing potentiometer value the screen changed vertical size. But during that it the text was waving. It was also waving when I touched it with a little pressure or tap it with plastic screwdriver.

Is there any way to power PCB outside the PET power supply? I am wondering with a such test scenario when I connect external power supply to the E2 and E3 wires of the monitor PCB with 11V DC 0,5A power supply current and connect HSYNC, VIDEO and VSYNC to this PCB connector from PET mainboard only. I would like just to eliminate any problems with PET Power supply here (I know, PET mainboard is connected to it). I am looking for a best case scenario to test these component and eliminate any potential problem.
In the PET VDU, there is positive feedback around the vertical scan amplifier, it is a common technique to control the linearity of the vertical scan. In the case of the PET it is only just over a critically damped situation, so on these VDU's if you adjust the vertical height, it induces a transient voltage disturbance into the vertical scan amplifier, and there will be a few cycles of oscillation in the vertical scan current. This will make the screen image bob up and down vertically, and that is normal for these VDUs.

But as Daver2 points out, this is not related to any horizontal wobbling at all.

There are only a few likely causes of the horizontal text wobbling:

If there is a phase shift error in the H drive signal or in the video signal. These signals would have to be wobbling with respect to each other this alters the relative position of the video information with respect to the horizontal component of the raster scan. Testing each signal separately might not make it easy to see on the scope. So you can lock one scope channel to the H drive signal and the other channel to examine the video signal and see if you can see the same timing error that you see on the VDU's screen, if so this rules out the VDU.

Or direct magnetic radiation to the CRT's beam, typically this comes for the power transformer's magnetic field. That one is usually easy to detect because it disappears as the VDU is moved physically away. Or possibly from other stray fields in the area.
On a related topic (if it turned out to be magnetic interference) I was once told by a Physics teacher, that if you were inside a vessel of some kind and couldn't see outside and it was travelling at a constant speed, that you would and could never know you were moving, even if it was moving at many thousands of miles per hour. You would only detect anything if you were accelerating or decelerating.

Later in life I was riding in a lift in a very tall building. The lift had a vintage CRT VDU as the display for the floor numbers. I could feel the lift initially accelerate then nothing as it a acquired a constant speed heading toward the top floors of the building. But I noticed that each time the lift passed by a floor, there was significant magnetic disturbances that caused the raster scan on the VDU to wobble . Presumably from the extra iron structures at each floor around the lift door systems or other ironwork.

So the remark the Physics teacher made was not entirely qualified, because it will depend on the medium you are travelling through. If the medium had a perfectly uniform gravitational and electromagnetic field, way out in deep space maybe, the remark about "not knowing you were moving" might well be true, but if the vessel you were in traveled at a high speed and went through a large gravitational, or electromagnetic field non-uniformity, it could well be ripped to pieces, or at a bare minimum deflect the beam of a CRT.

A CRT and its beam makes for a very good magnetic field detector, as it only takes a small angle change of the beam near the gun to get a significant movement of the beam on the face plate. This is why most scopes have MU-metal shields around their CRT's..