Converting a monitor from 625 to 405

[ppppenguin]

Thanks for the explanation. Since one reading will be at 15kHz and the other at 10kHz there will be an unknown difference in the meter error between them. Dummy resistance not important. The CRT will do just fine, it can stand short term overvoltage.

Not sure what CRT is in the example I'm working on as the label has gone. All I know is that it has a 6.3V 150mA heater. It's quite possible that other examples have a 12.6V heater. The W suffix refers to the phosphor colour which looks like ordinary 9300K monochrome white on this example. This monitor was certainly supplied with D6500 phosphor if the customer wanted it. This would have been preferable in a mixed colour/mono stack.

When the time comes to deal with heater voltage properly I will probably use my Rigol scope as suggested by Mark. It's quick and easy. Otherwise Amie's suggestion looks easier than the dual lamp method I proposed.

The subject of accurate AC measurements is large and complex. For the highest frequencies and most awkward waveforms then thermal methods are the only ones available. This obsolete chip implements thermal based measurement on the chip: https://www.analog.com/media/en/technica...lt1088.pdf There are many possible analogue RMS to DC converters, with differences in frequency and crest factor capability. Finally there are digital methods, basically sample the waveform and do the sums. That's what my scope will do.

[Nick]

...must get around to mounting some in boxes for this sort of job...

Some years ago I picked up a s/h ISB-060W medical isolation transformer for about 40 quid - I reckoned that by the time I'd spent several hours messing around and getting the bits I didn't have, it was a good deal - they crop up on eBay in the UK from time to time.

There's a few around at the moment, but not in the UK - they are worth waiting for... Note that you need the "W" version as that has the IEC 60320 type "F" outlets that you need for the UK.

[pwdrive]

Thanks for the explanation. Since one reading will be at 15kHz and the other at 10kHz there will be an unknown difference in the meter error between them. Dummy resistance not important. The CRT will do just fine, it can stand short term overvoltage.

Sorry, I thought you had a meter to measure the correct voltage, apologies for any confusion.

Lawrence.

[Mark Hennessy]

...must get around to mounting some in boxes for this sort of job...

Some years ago I picked up a s/h ISB-060W medical isolation transformer for about 40 quid

Ah, but that's a mains isolation transformer. I'm OK for those. I was talking about signal isolation. I've got loads of really nice ones - all different types, but all full-bandwidth and designed for professional audio. The one I was thinking about has a bandwidth of many 100s of kHz, and can take huge levels. The isolation is tested to 4kV IIRC, but obviously I'd have to double-check that before deciding how to rate the finished thing.

I already have various audio transformers in boxes for studio/stage use, and have used them for this sort of thing in the past, but it's a lot of faffing around with adaptors to make the required connections. So I'd probably have BNCs on both sides, as this is for the bench. If I ever find the bench again

Of course, the modern way of dealing with all this is something like this: https://www.eevblog.com/product/hvp70/

But quite "spendy"...

[ppppenguin]

Floating the monitor would be no problem. Either by lifting the earth or using an isolating TX. The Aurora providing the 405 signal is powered from a wall wart so that floats unless I give it a video input. Its internal Test Card C is fine for the work I'm doing.

The pictures show the earlier and later versions of the DU1-12. All the ones I've seen have BNC connectors though the manual says UHF/PL259 There's evidence on the earlier chassis of holes for those.

The later version has easier access, especially to the line timebase PCB, because there isn't a wide fixed metal bar from front to back at the top of the chassis.

[Murphyv310]

Hi.
Well I'm not offering theory here but as we know reducing the line frequency increases EHT, and any secondary from the scan transformer due to longer ramp up time. One rough and ready thing to do during tests is to fit a damping resistor across one of the higher impedance secondary windings. Some non regulated D/S sets had this idea early on. I'm not advocating it as a permanent fitment. It's quite possible that a simple reduction of supply to the scan transformer and fitting alternative tuning cap values may all that is needed but beware of horizontal non linearity.

[ppppenguin]

This set has some clever stabilising loops so it's hard to predict how it will behave. The basic rule about reducing the HT must be true.

[ppppenguin]

Got back to it today.

First I soldered extra capacitors across the line oscillator timing caps to bring down the frequency. While playing with the line period control the picture disappeared leaving a blank raster. Didn't take long to find that the -25V supply, taken from the LOP, had disappeared. The rectifier D5 (1N4820 on circuit, SF4M1 in reality) had gone open circuit. Replaced with BYV95B and it came back to life. No idea why it failed. OK the LOP voltages were all over the place but it was a 400V diode running with only 250V peak and very little current. Must have died of old age and fright.

Decided to wind down the width/EHT control which brings down the main HT.  Gave a small but perfect picture on 625. Nominally +27V it was nearer 16V. Thought this would be safer for the LOP. Played with the line period pot and the picture shows the result. Horrible but things will get better. Winding up the width/EHT stopped it locking, no idea why yet.

Incidentally, I was wrong about the CRT heater current. I misread the analogue meter on my bench supply that's doing the 6.3V for now.  It's 300mA, not 150mA.

[ppppenguin]

The nasty kink in the middle is due to inadequate drive to the LOP. This in turn is because I turned down the main stabilised power rail to minimum to minimise chances of damage. Just not enough voltage available to give a nice clean drive to the LOP. I powered the line osc and drive from an external PSU which cleared up the kink but it won't lock. Decided to stop for the day and have a think about it.

The next plan is to run the LOP from an external PSU so the main PSU can be wound back up to its normal voltage. Once I've established correct operating conditions in the LOP I can return to making the whole thing work on internal supplies. To a first approximation the LOP supply voltage will need to be about 70% of that needed on 625. Since that will store the same amount of energy in the LOP on each cycle as on 625 the EHT and aux supplies should all be about right. LOP tuning and S correction will certainly need attention but I want a nicely locked picture before dealing with 2nd order problems.

[ppppenguin]

That's looking better. Main 27V rail at about 20V (won't go any higher at the moment for some odd reason) and about 15V from external PSU on LOP. Takes just over an amp which feels right. The moire patterning that's making the scan lines look like twisted rope are an artefact of compressing the JPG file. Looks a lot better on the full size file and better still on the screen. EHT is a comfortable 13kV or so.

I'll need to look at how to make an approx 15V supply from the 27V rail. I think they will need to track to some extent. Crudest method is a watty resistor. Might just work. Next would be a chunky zener or two. Or even a long string of silicon diodes. Another way is a power transistor with its base dirven by a potential divider from the 27V rail. That's probably the nicest method. Possibly a darlington so it's easy to use a pot to set the LOP supply voltage. The transistor will be dissipating around 10W so it will need to be attached to the case, along with several other power transistors.