• Please review our updated Terms and Rules here

8032-SK Restoration

As I have two 8032-SK's to diagnose, I am flipping a bit between them so forgive me if it gets confusing :)

Have nothing on the screen of the second unit, I note that I dont have the 400V signal at the brightness controls. Now looking at the circuit diagram for the monitor and checking the loop I found that R252 was open circuit.

Its listed as a 56 ohm RN 1/4 and is of a form I'm not familiar with. It looks like an externally etched metal film, which would be for accuracy and stability ?. Hugo, is there anything special about these resistors ? I note that they have a standoff on one end to lift them from the PCB, Heat or prone to failure ?

I have replaced it temporarily with a 1/4W carbon film and the monitor has now sprung into glorious life but I dont want to leave it like that if I really need to put the right type of resistor in.
 
R252 (and, I think, there is another one) are really fusible resistors. There is a high failure rate of these resistors as the machine gets old. However, if you don't use the correct type, then (say) a faulty smoothing electrolytic downstream of the resistor could cause a more serious failure upstream.

I managed to get home at 2 AM this morning - a flat tyre on the M6 just past Preston!

Dave
 
Damn, thats a pain !

Hope you get today off because of it. No one wants to get stuck near Preston ;)

Yes, there are three of these resistors. Do you know where I can source some ? the designation on the drawing doesn't seem to match with anything.

Machine is now running PETTESTER and staying at page 0/1 test and page 0 is showing random and changing errors.
 
Gary C said:
As I have two 8032-SK's to diagnose, I am flipping a bit between them so forgive me if it gets confusing :)

Have nothing on the screen of the second unit, I note that I dont have the 400V signal at the brightness controls. Now looking at the circuit diagram for the monitor and checking the loop I found that R252 was open circuit.

Its listed as a 56 ohm RN 1/4 and is of a form I'm not familiar with. It looks like an externally etched metal film, which would be for accuracy and stability ?. Hugo, is there anything special about these resistors ? I note that they have a standoff on one end to lift them from the PCB, Heat or prone to failure ?

I have replaced it temporarily with a 1/4W carbon film and the monitor has now sprung into glorious life but I dont want to leave it like that if I really need to put the right type of resistor in.
Click to expand...

Generally I would not be too concerned about these resistors.

Most VDU's and TV sets don't use these as protection devices on the auxiliary voltages produced by the LOPT. I cannot even think off hand of many other VDU circuits where they are used in this location. So they are not "standard practice" here in all VDU's.

They do have some theoretical advantage if there happened to be a short.

The power dissipation, in use, will be very low, due to the low R value and the low currents.

It can be a dual edge sword with devices made to fuse, or fuses selected to be sailing close to the wind, or degradation of the fuse over time, in terms of nuisance fuse blows.

Probably these failed due to exposure and microscopic damage of the surface film over years, so they go open at a lower current than intended.

I think in all the cases I have heard of with these going open in these particular VDU's, it was not because a short had developed, and when just the resistor was replaced everything started working again. So they appear to have more nuisance value than protective function.

One could argue that you could have these sprinkled everywhere in all power supply circuits "just in case of a short" but the utility value of this can be limited as other factors, voltage regulators etc limit the current. The main utility value of a fuse is to protect the wiring around it and the best place for one is at the inputs of the main power supply. Unless fuses are super fast acting, like semiconductor fuses, they have limited value in protecting semiconductors or other components on a pcb. Fusible resistors are very slow acting. There might be a better argument for a 100mA fast blow glass fuses or even slow blow in this location instead of resistors. The initial currents charging the filter capacitors could fatigue a fast blow type, over time.

If you were concerned about the fusing function one option would be to replace them with small bodied 1/8 w film resistor types, which would be more likely to fail open if a genuine short ever happened to occur and it would vaporize the film. Or use a glass fuse.

Mouser sell fusible resistors, one example:


 
Last edited:
Cheers.

I know in some instrument applications we used etched metal film resistors as they were tested as they were laser cut to provide a very accurate and stable resistor. These just looked odd and I wasn't certain of their design function.

Replaced with a tested carbon and its working fine and no need to fish under the cap (still can't find my Fluke HV probe)

But this one has very obvious Page 0/1 faults. Its stopping at the first test after the SRAM test and the first 256 characters are showing g's and b's that change randomly, the second not really showing a decisive pattern and the next two banks showing static errors.
 
Gary C said:
Cheers.

I know in some instrument applications we used etched metal film resistors as they were tested as they were laser cut to provide a very accurate and stable resistor. These just looked odd and I wasn't certain of their design function.

Replaced with a tested carbon and its working fine and no need to fish under the cap (still can't find my Fluke HV probe)

But this one has very obvious Page 0/1 faults. Its stopping at the first test after the SRAM test and the first 256 characters are showing g's and b's that change randomly, the second not really showing a decisive pattern and the next two banks showing static errors.
Click to expand...

When it became common to make resistors by applying a conductive film to a cylindrical surface, the way they were calibrated was to cut a spiral groove into the surface.

Typically these types of resistors had a push on metal end cap to attach the lead wires at each end of the ceramic rod. This leaves the bump at each end of the resistor.

Back in the early 1960's a company in the UK started to make a new design. They fitted the lead wires into holes in either end of the ceramic rod. It created a streamlined looking resistor with no end caps. However, 40 to 60 years later something interesting happened. Due to corrosion/oxidation where the leads were fitted, the expanding oxides split the ceramic and the resistors go suddenly open circuit. I call these "self cracking resistors". I wrote this up once, on page 4-5 of this article (if you see any of these resistors in mission critical equipment replace them ASAP):

www.worldphaco.com/uploads/ELECTROLYTIC_CAPACITOR_FAILURE_MODES_IN_TEKTRONIX_TM_SERIES_POWER_MAIN_FRAMES.pdf

Generally oxides of metals expand in volume, rust under paint etc. An analogous thing happens with oxides around Amalgam fillings in teeth and it can result in the tooth cracking, typically the cusp or side wall cracks off.
 
Hugo, I need to replace the 400V supply diode and its specified as a V08E. Now I can't find this diode anywhere but I note the data sheet specifies it as an avalanche diode with a reverse breakdown of 1600V.

So it operates in the 400V line forward conducting, but is the reverse breakdown required to allow it to reverse conduct without failure if necessary ? Struggling to find an equivalent.

Humm, thinking a BYW53-TR
 
Last edited:
Gary C said:
Hugo, I need to replace the 400V supply diode and its specified as a V08E. Now I can't find this diode anywhere but I note the data sheet specifies it as an avalanche diode with a reverse breakdown of 1600V.

So it operates in the 400V line forward conducting, but is the reverse breakdown required to allow it to reverse conduct without failure if necessary ? Struggling to find an equivalent.

Humm, thinking a BYW53-TR
Click to expand...
My go to diode for that sort of part would be the BY228 sinterglass diode. They are pretty rugged but very compact being about a 4mm spherical shape. They probably are physically and electrically similar , if not identical to the V08E and they are easy to get too.They have quite thick lead wires that are about 1.3mm diameter.

 
For interest, why was it specified as an avalanche diode ? from my electronic days they were diodes designed to survive reverse current flow. In this application, is it designed to survive reverse spikes ?

Annoyingly, my go to supplier has the by228 as a discontinued product with no stock :(

did find a job lot of BYW53's for £2/10 which I think will do.

The failed V08 just fell apart from corrosion when I tried to test it.
 
:), thats who I was looking at , doh !

No idea why it didn't come up when I searched. I get free delivery from RS too.
 
A little bit of testing today and there is certainly an intermittent data problem somewhere. If I power it up, I get the normal BASIC 4.0 message and all looks good. Then if I type in 10 PRINT A and then LIST I often get 23.33400 PRINT A back which is really odd, then a further LIST command gives me the proper 10 PRINT A list. If I try to add a second line I often get out of memory error.

I managed to find one of the Tynemouth boards and with ROM & RAM replaced it all works fine, but I noted that this board uses chunky pins and as I didn't want to damage it I used an additional pressed pin socket, and it didnt work !, pushed it in with the chunky pins and it works.

So I need to replace the CPU socket, but I wonder if thats the problem.

Need to enable/disable ROM then RAM and see if its down to either of these. Don't want to end up with a machine with an add on inside it, but at least I have a useable machine at the moment.
 
Right, with the motherboard RAM enabled it all works fine, but when the onboard ROMS are enabled it does the odd numbered list thing.

The checksums seem to be right under PETTEST, but I think the best bet is to socket all the ROMS and replace them with a complete set in 2532's and see where I go from there.
 
Something is clearly on the edge somewhere isn't it.

You seem to be narrowing it down though...

Keep at it, you'll get there!

Dave
 
Yes, something is marginal. Disabled the tynemouth ROM and RAM and its working fine for now with the original ROM/RAM.

CPU base first as the turned pins of the tynemouth board have knackered it.
 
So CPU socket replaced and I'm now getting some repeatable results

With the system RAM mapped out and the Tynmouth board working, it runs fine, but map the system RAM back in, it continually crashes in very strange ways. The way its crashing suggests zero page is being corrupted but when PETTEST is used, it gives it a clean bill of health.

Only other thing of note, even when the RAM is mapped out, and its all running ok, the IEEE port fails to respond. Now thats possibly a totally separate fault, but not sure.

So adapting PETTEST to give a delay between writing to page 0 & 1, then checking and repeating in a continuous loop shows no errors.

Probably next mod will be to copy memory from Static to Dynamic and back again and look for errors there.

Really confusing me this one.
 
Program written to write to VDU memory, then write it to low memory, then write it back to vdu memory and it works with no errors at all.

I dont seem to have a problem with the memories themselves. How odd.
 
Finally had time to do some more work on this machine.

Took it all to bits and laid out the motherboard and PSU on the desk top to make it much simpler to work on. The monitor rotates and unclips (though only in one direction) and now stands free. Its so much better not to have to work with the monitor jacked up.

So, with a Tynmouth board installed and set to replace the RAM, the machine works fine, which is good, but PETTESTER tests all the dynamic RAM as fine. Had a niggle that they aren't being refreshed properly with PETTESTER working so fast that problems with refresh aren't revealed and just noticed that half of UD3 isn't clocking which means the refresh address lines aren't operating !

Fingers crossed.
 
Probably ought to add in my defence that I'm sure I had checked RA0 to RA7 before and believe they were good so I'm thinking the counter was flaky and has now failed totally.
 
You must log in or register to reply here.
Back
Top