Deflection and HV generation in general, in particular the switching transistor and usually with it various passives and in worst case flyback itself and stuff connected to its outputs which may be briefly very out of spec while destruction is underway. It can be a significant repair.

As far as destroying the horizontal output transistor goes, my impression is that *running the sync frequency too low* is what kills it. The lower frequency results in the H.O.T. being switched on for too long. Once the horizontal yoke and tuning inductors become fully saturated they essentially become a very low resistance short circuit. In normal operation, the H.O.T. is meant to be switched off well before the yoke saturates. A higher-than-expected hsync frequency just results in a picture that's too narrow. But now your problem is that the higher hsync frequency will produce higher voltages at the flyback transformer output.... which is another way to kill a CRT while also producing bonus X-rays in the process.

Anyway it was a bit ridiculous for IBM to build the 5151 without a horizontal oscillator, and it's a terrible idea to do "experiments" with the 5151 as a result.

Too bad, it would have been interesting to see it's shades of green outside the limitations of the MDA card.
I'm definitely not going to risk destroying my 5151's but if someone would ever come up with a safe converter I'd love to play around.

kdr wrote on 2021-12-18, 03:32:

A higher-than-expected hsync frequency just results in a picture that's too narrow. But now your problem is that the higher hsync frequency will produce higher voltages at the flyback transformer output.... which is another way to kill a CRT while also producing bonus X-rays in the process.

Yes, the voltages increase, including the flyback pulse so the breakdown voltage of the switching transistor is exceeded (along with the various passives) and you'll still get a catastroptic failure still 🤣.

rmay635703 wrote on 2021-12-16, 22:13:

If you had a newer MDA screen (not ibm) with pot adjustments for v/h hold/sync

Remember the not IBM part?
Still would need a scan converter like a Averkey micro

https://m.youtube.com/watch?v=yItzHekCygo