• Please review our updated Terms and Rules here

Early PET VDU Restoration Article

Hugo Holden

Veteran Member
Joined
Dec 23, 2015
Messages
4,647
Location
Australia
I have written up an article about early PET VDU's.

One aspect of it is the general restoration of the VDU.

Another aspect is about modifications to the VDU. These are not absolutely necessary, but they can improve the performance and help protect the VDU from damage in the event it is fed with abnormal signals from the computer board.

The article essentially has 13 sections and as usual got bigger than I thought it would when I started it.

One aspect is about the analysis of the LOPT (Line output transformer, aka Horizontal output transformer). There is not a lot of good specific detail on this in the literature, so I have tried to put it all in one place, how to test these with the generator & scope.

Also how one might adapt in LOPT's from other VDU's. The LOPT's on these early VDU's are not available as replacement parts.

One interesting result, is that it is dead easy to check the LOPT in circuit, without even having to remove it from the pcb, with a dual resonance test as explained in the article.

Anyway, here is the article first draft.

Aspects of it might help those planning to restore one and re-cap the pcb:

 
That's an extremely impressive piece of work. Thank you so much for sharing your knowledge and experience with everyone here.

Alan
 
That's an extremely impressive piece of work. Thank you so much for sharing your knowledge and experience with everyone here.

Alan
Thanks, I hope aspects of it can help keep the PET VDU's running well into this century.

As is the case with vintage TV's that use CRT's, the CRT itself remains the major "non renewable resource".

The last CRT re-gunning plants in the world closed many years ago (last one was in France). There have been some attempts at re-building a CRT plant in the vintage television community, but as yet no new regular service exists that can fully re-build a full variety of monochrome CRT's (yet alone a color one) completely to a new standard, with the ability to re-screen them with new phosphor too. The glasswork, metallurgy, phosphor technology, industrial processing & controls , like many processes, that took many decades to fully evolve, was much more complicated & difficult than it looks.

One of the most mind blowing CRT's ever made, when the evolution of the CRT was complete, is the oscilloscope tube used in the Tek 2465B scope. It is a masterpiece of construction & Electron Optics and CRT technology, the likes of which the World will never ever see again. It was just too expensive compared to flat panel displays.

There are still some locations of NOS CRT's in the the USA, UK & Europe which I buy CRT's from, but I have watched over the last 25 years as the supplies have dwindled to low values. I was able to acquire some new old stock CRT's for the 9" PET VDU's to put in my CRT farm, which has become somewhat of a Noah's Ark for vintage CRT's.
 
Thanks for the kind remarks. There are a few minor typos here and there in this first draft so I will clean those up soon.
 
Fantastic bit of work and its got me thinking

The 40 column PET then had a horizontal sync of ~15kHz, vertical 50hz which is basically PAL ?

So to fit a C64 into a PET case (al la Pet64) I would need to strip out the sync pulses, then burn the Pet64 ROM to only display B&W as the input amp of the video stage is bistable ?

I just have an empty PET case and C64 boards are far cheaper than PET boards.
 
Fantastic bit of work and its got me thinking

The 40 column PET then had a horizontal sync of ~15kHz, vertical 50hz which is basically PAL ?

So to fit a C64 into a PET case (al la Pet64) I would need to strip out the sync pulses, then burn the Pet64 ROM to only display B&W as the input amp of the video stage is bistable ?

I just have an empty PET case and C64 boards are far cheaper than PET boards.
It sounds like that plan would work. Are the keyboards compatible or would you have to put the C64 keyboard in the PET case too ? I wonder if it would fit the cut out. I don't have a C64 yet.
 
Fantastic bit of work and its got me thinking

The 40 column PET then had a horizontal sync of ~15kHz, vertical 50hz which is basically PAL ?

I just have an empty PET case and C64 boards are far cheaper than PET boards.
Gary,
Only the later PETs (4032/ 8032 with 12" monitor) with the 6545 CRT Controller support 50 Hz with 20 KHz horizontal. However Steve Gray has programmed EDIT ROMs for these PETs that can support NTSC or PAL:

http://www.6502.org/users/sjgray/projects/editrom/

The older 2001-8 and 2001-N and 3008 PETs (9" monitor) have hard wired timing and are at 60 Hz vertical with ~15 KHz horizontal.
-dave_m
 
Even in the UK with a 50hz mains ?
This is one of the typos I'm correcting in the article. My PET (with the 9" VDU) which came from the UK, has a 60Hz vertical scan rate (not 50 Hz which has a 20mS period) and this matches up with the information on the Zimmers schematics, which show the vertical interval (period) as 16.7 or 16.66 mS, depending on the schematic. (1/60Hz =16.6666 mS)

It really doesn't matter if the line power frequency is different to the vertical scan rate, unless there is excessive ripple on the power supplies. Since these are well regulated in the computer an in the VDU no trouble is seen. If the VDU though was exposed to significant magnetic external radiation from a say a power transformer's magnetic field it could possibly cause some crawling ripple in the raster scan. But in my PET running it here in AU, on 50Hz line power, I cannot detect any trace of interference. A transformer's radiated flux field can be significantly reduced, if required, by putting a copper flux band around the transformer, and re-positioning the transformer physically can help when a transformer's field interferes with a CRT. Probably I think the PET's metal CRT cabinet helps a lot here. But I have a feeling there might have been some plastic cased pet VDU's ?

Years ago I was told by a Physics teacher, that if you were inside a sealed enclosure or capsule and it had accelerated up to a speed, which then became constant, and stayed constant, then inside that capsule you would never know you were moving. (there was no mention that the craft might have to be shielded). In principle it is correct but that assumes of course that the electromagnetic medium, outside the capsule was uniform.

I once took a ride in a lift in a very tall building, from near the top to the ground floor. It was easy to feel the lift initially accelerating and then, for most of the ride, it reached a constant speed and within a few floors of ground started to decelerate. But the interesting thing was, it had a CRT VDU for the display inside the lift. As the lift passed each floor, due to the additional metalwork there and alteration in external fields, the CRT's entire raster was modulated around the CRT's faceplate. So, with the CRT as a magnetic field detector tool, I could tell that the lift was moving from inside the lift travelling at a constant speed.

It made me realize that if you were travelling very fast in a space capsule and it past through a strong local magnetic field (Perhaps put there as a force field to deflect projectiles) it could destroy everything in the ship. If the rate of change of magnetic flux with time, was high enough, very large voltages would be induced in pcb tracks and other conductors and at a bare minimum, all the onboard computers would get fried.
 
Last edited:
Gary,
Only the later PETs (4032/ 8032 with 12" monitor) with the 6545 CRT Controller support 50 Hz with 20 KHz horizontal. However Steve Gray has programmed EDIT ROMs for these PETs that can support NTSC or PAL:

http://www.6502.org/users/sjgray/projects/editrom/

The older 2001-8 and 2001-N and 3008 PETs (9" monitor) have hard wired timing and are at 60 Hz vertical with ~15 KHz horizontal.
-dave_m
So the sync range of the 9" PET screens means an NTSC C64 would possibly fit the bill as suggested that the output from the first PETs would display on an NTSC monitor.
 
The original PET has hard wired logic that creates a 15.625 KHz Horizontal clock that was counted down from the 8 MHz crystal. But the video board in the monitor compartment does not use a composite video signal. The Horizontal sync, and Vertical sync are separate inputs along with the video data. All TTL levels. What is the video output of the C64 to a TV? Are there separate timing signals available?
 
Also the PET timing signals in the early units have wide pulse widths that drive the video electronics. If the C64 signals meet NTSC specs with its skinny negative pulses, one may need a pulse stretching circuit like a one-shot multivibrator (74121) to meet the requirements of the PET monitor.
 
Also the PET timing signals in the early units have wide pulse widths that drive the video electronics. If the C64 signals meet NTSC specs with its skinny negative pulses, one may need a pulse stretching circuit like a one-shot multivibrator (74121) to meet the requirements of the PET monitor.
That would be a good solution to acquire the correct duty cycle for the H drive signal for the 9" PT VDU, when you were starting out with relatively narrow H sync pulses.
 
Back
Top