If all the supply voltages check out then it most probably is a fault around the line deflection drive generation stuff. Excessive B+ voltage would cause raster to expand but it would also cause HV to rise and x-ray protection should kick in also...

I've just replaced the horizontal oscillator IC and as expected, nothing has changed.

Tiido wrote:

it most probably is a fault around the line deflection drive generation stuff.

I start to doubt that this fault has anything to do with a horizontal deflection contol/drive area. Horizontal oscillator is OK, small signal transistor that drives isolation transformer for HOT and adjacent area are also OK, HOT is not excessively hot (lol), B+ is in specs and can be adjusted, but it affects only horizontal width, not the wrapped picture.

This odd monitor uses a 555 timer, but I'm not exactly sure for what purpose. Timer's clock is controlled by an external signal (H-SYNC output from a 74LS86) instead of a pin 7 of the timer (discharge), output basically goes into a second H-HOLD pot, that does nothing and into a second B+ adjust pot, that also does nothing when adjusting. I've tested the timer chip in a working application, that uses such timer and it's functional. Why would they even use a 555 timer in a monitor...

Sweep problem like that is 100% definitely having to do with line deflection current. Wraparound happens when the sweep current ramp is wrong (so that image starts or before the peak, or ends in the next peak) and that ramp is directly controlled by interaction between flyback, the primary sweep capacitor, EW correction (if present) and to an extent the deflection coils themselves. If the sweep capacitor (the 1kV+ rated film one) is fine the fault is most probably in the flyback, assuming the drive current freq is correct on the HOT. The extent of the wraparound is so great that it would almost seem that the sawtooth waveform is twice the freq or has really gentle slope for what it is supposed to be. Do you have an oscilloscope to look at the HOT drive waveform + Hsync freq at the same time ? If they line up then the problem is afterwards, if there's a big disconnect the problem is something around the external circuit of the chip that generates the drive pulses.

The 555 I suppose is for generating left-right adjustment, if it is doing a oneshot delay that is retriggered each sync pulse.

Thanks for your feedback.

Your reasoning behind a distorted sawtooth waveform seems legit, it looks as if its period had doubled, causing the horizontal stretch. I believe the HOT cannot be turned on and/or off fast enough, that would explain why monitor loses horizontal synchronization completely when graphics card is set to EGA mode (22KHz H-SYNC).

I've measured H-SYNC signal at the input of the TDA2595 with a DM's frequency counter and it showed ~15.65KHz reading, but at the output, DM couldn't measure freq. properly, probably because of the +12V suppy line that is connected through a resistor to it.

I've tried to replace the horizontal driver transistor (C1213), thinking that it might have become leaky/weaker over the years, but that changed nothing. It could be that the insulating horizontal driver transformer has one or more shorted turns, I could test it with a Dick Smith FBT tester, but I doubt that its reading would be accurate enough to diagnose such marginal fault, same goes for the primary winding of the flyback transformer.

I've pulled out the HOT and couldn't properly test it, because it has a inbuilt damper diode and a 56 Ohm resistor between the base and the emitter, but I've noticed that parasitic capacitance between the base and the collector is equal to 333pF, instead of ~85pF (according to the BU2520DF datasheet).

Unfortunately I don't have a functioning oscilloscope at this moment.

Substituted HOT and horizontal driver transformer with known good parts and monitor produces the same, stretched out, raster. I just can't continue this guessing game, all major parts in horizontal section were checked good or substituted with known good parts. Looks like, that a scope is a must in this situation.

That's too bad, I really wonder what is wrong with the thing...