• Please review our updated Terms and Rules here

Screen Wobble on Commodore PET 4032 12" CRT

J

jtjacques

Member
Joined
Feb 7, 2025
Messages
10
Hi all,

I am working on a PET 4032. Wondering if you could help with the above issue.

Video of screen wobble:
(looks even more severe in person).

Board is a universal board, -03 marked (pg. 3 onwards). CRT board appears to match pg. 41
https://www.zimmers.net/anonftp/pub...4000_Series_4016-4032_Technical_Reference.pdf

I've done the basic checks, transformer voltages all as expected, DC mainboard voltages all as expected.

CRT board: DC (i.e. rectifier) in to regulator shows ~2V peak-to-peak, but totally smooth 17.98V on the output of the regulator.

CRT board checks: all test points match the waveforms in the service manual (pp. 42-51).

Suspected transformer (US import to UK, using step-down at 50 Hz), but operation when open shows same wobble with no discernible change.

Does anyone have any suggestions on where I should look next (or what I should double check)?

Thanks,

Jason
 
Welcome to VCFED (as this is your first post).

2V pk-pk seems a bit high before the 18V regulator. Can you check the large smoothing capacitor after the bridge rectifier.

Can you check the electrical earthing of your setup - in particular mains earth to case to transformer core.

One way round is more compatible than the other (60 Hz vs 50 Hz). @Hugo Holden should have that information at his finger tips...

What EDIT ROM is fitted to the machine (the ROM number) and what type of lighting are you using to view the screen?

After this, it will be looking at magnetic shielding of the transformer using a single loop of copper.

Dave
 
Hi Dave,

Thanks for the welcome and reply.

I'm attaching a picture of the waveform across C901, showing the same ~2V mentioned. Regulated output is smooth, none the less.

IMG_1262.jpg

Earthing looks good to me. 0.5 ohm from pretty much everywhere, including the transformer, to the earth pin on the cable (and the step down passes the earth through).

The machine has the expected 901499-01 (0683) ROM (i.e. 60 Hz US model). I'm unclear, to be honest, why they have different ROMs. Everything gets rectified and the system doesn't use the line frequency for driving anything (as far as I can see). Is it a software hack to hide the magnetic/line frequency interference?

The previous video is under artificial light. I think they are florescent (commercial). See following video for daylight (window) with the same effect.

I'm understand that 60 Hz transformers are less happy on 50 Hz (than the other way around), hence my earlier suspicions about the magnetic interference. Based on your thoughts, I decided to recheck that with a more objective reference. I've stuck on some post-it notes, with camera pressed against the bezel. You're looking at BASIC 4.0 *** and comparing closed and on the prop.


So, on reflection, it looks to me like it is the magnetic interference. Even when open, it is significantly worse across other parts of the screen - but I zoomed on this so you can see clearly the effect.

I had tried a very thin copper band (previously) but no stark difference, so I removed it. I'll have to see if I can get some thicker copper. In the mean time I may try running it off some bench supplies to take the transformer entirely out of the equation before I look at the options.

Thanks for the prompts/help so far. Happy to take any further advice!

Jason
 
When you said : "suspected transformer (US import to UK, using step-down at 50 Hz), but operation when open shows same wobble with no discernible change"

By that when you say "open" do you mean that if you lift the lid of the pet (with the VDU on it) up and away from the transformer, that it has no effect at all on the wobble ?

If that was the case, then it is not a magnetic interference issue from the field of the transformer to the CRT, and instead represents an electrical interaction between the frame rate of the VDU and the line power frequency..
 
Hi Hugo,

Yes, that is what I meant: with the lid propped open. Hence the head scratching. However...

I have since been a bit more objective about the display and noted that magnetic interference is variable and observable (see new video); though it is only significantly variable in the upper corner (furthest from the transformer, go figure!), hence not noticing this earlier.

(My reply admitting this to Dave was awaiting moderation when you replied; hopefully visible before this one!)

Thanks for the input. Happy to take further advice.

Jason
 
I wonder if the screen editor ROM should be 50 or 60hz version on this PET? It looks awfully like one of mine where I had the wrong one in.

Colin.
 
It looks to me like magnetic interference of the raster scan from the transformer field because of the way it is moving.

I'm still not 100% clear if there is even thesmallest effect at all (or absolutely zero) moving te VDU further away from the transformer ? The wobble that I see does have many features of magnetic interference. If that is true, then you should see some change or at least some reduction when the VDU is moved further from the transformer.

It could be that you have very high radiated fields from the transformer. One thing you can do as an experiment (apart from unbolting the VDU and moving it well away), if you have a Variac, is to wind down the Line voltage. What you would find if it is high peak flux in the transformer starting to saturate its core, is that the wobble would get less as you dropped the voltage. At some point though ripple will breakthrough the regulators and the screen image will start getting interfered with again for a different reason. Typically for an American transformer will provide unsatisfactory secondary voltages under about 85 to 90V input.

To give an example of how this happens with American transformers operating in the UK, lets say you are running one from a 240V source via a step-down, often these give a 2:1.05 ratio, or about 5% extra on the secondary output to allow for a voltage drop under high or full loads power. So if you had 240V in your locality, the voltage applied to the primary of the American transformer could be in the region of 1.05/2 x 240 = 126V.

The peak flux in the core is inversely proportional to the applied voltage and inversely proportional to the frequency, so another factor of 60/50 or 1.2 steps in.

So it is as though you are back in America with the transformer, but the transformer core's peak flux is now 1.05 x 1.2 = 1.26 times higher than its design center. Which would be equivalent over in the USA, to running the transformer primary from 110v x 1.26 = 138 volts line power (in terms of the transformers peak core flux value, but not in terms of the voltage transformation ratio) so this is why there are troubles.

Also if there is any interference, its also a lot more obvious if the vertical scan rate doesn't match the Line power frequency, or is not close to it.

To get to the situation in the UK, where the peak core flux in the American transformer matches that of the transformer running off 110V 60 Hz in the USA, the applied primary voltage on 50Hz line power would have to be 110/1.2 = about 92 Volts.

Of course that 92V may or may not be enough to provide satisfactory pre-regulator voltages, which for the 5V regulators needs to be around 7.5 to 8.5v. There needs to be a window for line voltage fluctuations. I generally run most American power transformers here (on 50Hz) at around 95V Line voltage for the above reasons and usually that is ok for the pre-regulator voltages.

The other things that happen when the New York transformer is on Holiday in London, is that with the core being pushed further up its B-H curve, the hysteresis cycle enlarges giving more core hysteresis losses and this heats the core. But also, the primary winding's incremental inductance drops as the B-H curve flattens out and the peak and average primary magnetization current increases significantly (this is independent of the load current) so the primary winding heats up more too due to the I^2R losses (copper losses). In other words the American transformer in a 50Hz country runs hot.

So overall, the hallmark of American power transformers in 50Hz countries, they tend to run hot and have high radiated magnetic fields. In some cases where this is extreme, I have had to replace the transformers.
 
Last edited:
Hugo Holden said:
I'm still not 100% clear if there is even thesmallest effect at all (or absolutely zero) moving te VDU further away from the transformer ? The wobble that I see does have many features of magnetic interference. If that is true, then you should see some change or at least some reduction when the VDU is moved further from the transformer.
Click to expand...
It seems pretty clear to me now that there is a reduced effect when propped open. Compare 0:12 and 0:25 in the video:
https://forum.vcfed.org/index.php?t...ommodore-pet-4032-12-crt.1251664/post-1431079
Would be happy to get a second opinion.

Hugo Holden said:
if you have a Variac
Click to expand...
Not to hand, unfortunately. I am hoping I can sidestep the transformer entirely, at least for verification.

Hugo Holden said:
if there is any interference, its also a lot more obvious if the vertical scan rate doesn't match the Line power frequency
Click to expand...
As noted, this was what I suspected might motivate variations in the timing and thus the ROM.

Hugo Holden said:
for the 5V regulators needs to be around 7.5 to 8.5v.
Click to expand...
Good to know.

Hugo Holden said:
So overall, the hallmark of American power transformers in 50Hz countries, they tend to run hot and have high radiated magnetic fields.
Click to expand...
I've not noticed any appreciable warming. I am reasonably familiar with the theory, hence my initial suspicion (since renewed). However, the lack of a more noticeable reduction, especially with a copper band, sent me to seek out people who have more experience with the PET in particular.

Hugo Holden said:
In some cases where this is extreme, I have had to replace the transformers.
Click to expand...
If further testing indicates that it would be better replaced, then that is worth considering.

Thanks again for your thoughts.

Jason
 
For a copper flux band to work well, it needs to be a good thickness, almost to the point it is had to bend, maybe 0.6 to 0.8mm thick and properly soldered too. Thin copper foil is pretty useless. It relies on having a very low electrical resistance.

Also, is your PET a plastic body one with a plastic VDU shell ? Some PETs had metal shield plates in these cases as well as the copper flux band.

In cases of zero magnetic interference there is no reason to have any issue with the VDU's vertical scan rate being different from the line power frequency, so you might not have to change that if the basic issue is fixed.
 
Thicker copper was the plan. At that stage I was just looking for observable difference in the output.

Yes, unfortunately, the body is plastic and without a metal plate.

It seems sensible to try and resolve/contain the interference first.
 
A nice thick complete copper band around transformer of about 1mm thickness soldered into continous band sorted my wobble.
Please search for post -

Commodore PET 2001-8 text wobbling on screen​

 
Hugo Holden said:
So overall, the hallmark of American power transformers in 50Hz countries, they tend to run hot and have high radiated magnetic fields. In some cases where this is extreme, I have had to replace the transformers.
Click to expand...

I have the same problem in my PET. How can I know if the transformer is 50Hz o 60Hz ? Also, is it possible to replace the transformer by a toroidal one/ones ?
 
The part number on the transformer should effectively identify the frequency.

Also, if the transformer has a 110V primary, it is likely to be 60 Hz. Not a foolproof method though!

Yes, you can replace the transformer with toroidal type(s).

Dave
 
As far as I know the original output voltages are:

pins 4 and 5 : 9 V
pins 4 and 6 : 18 V
pins 5 and 6 : 9 V
pins 7 and 8 : 18 V
pins 9 and 10 : 18 V

Then, can I use a 18 volts transformer and a 9 volts transformer to replace these transformer outputs ? If so, which current is advisable?
 
You require a 9-0-9 transformer (one with a centre tap).

The 18V should be for the monitor.

I will look up the current rating for you later.

Dave
 
Just wanted to confirm that magnetic interference was the cause, confirmed using an external supply. Currently considering both a more substantive band or simply replacing the transformer.

Thanks all, for the insights and comments.
 
daver2 said:
I would go with the more substantive band first - as it keeps the authenticity of the unit.

But that is only my view.

Dave
Click to expand...
I agree, but I tried to set a band and it's hard to bend a 0,8mm cooper band giving the shape of the transformer.
 
Papalapa said:
I agree, but I tried to set a band and it's hard to bend a 0,8mm cooper band giving the shape of the transformer.
Click to expand...
That is what I said. Once the copper gets to about 0.8mm thick, it is harder to manipulate and bend. Try copper at about 0.6mm or close, and you will have more luck. I know, I have made many flux bands for troublesome transformers. At worst, use 0.5mm thick material.
 
You must log in or register to reply here.
Back
Top