Spent a couple of hours on the Tek 2225 this morning. Determined to finish something off for once.
I had the same mentality. My first Tek1502 TDR's PSU was half working: the device ffunctioned but the mains wasn't charging the battery. I'd had a brief look before, but not found the cause, and then Family intervened.
The PSU is mostly discrete and seriously strange...
The switching PSU half is a fairly standard switcher with an "extra" section that ensure switching half doesn't turn on if either the NiCd voltage is too high (i.e. cells missing) or too low (discharged). That worked.
The mains charger is a 50Hz switcher which always pumps a constant
mean current into the NiCd cells, but also supplies an extra 150% current if the TDR is switched on. The voltage is defined by the NiCd cells (what else?). It is based on an advanced opamp (a 741!), a programmable unijunction, an SCR, a discrete current mirror, and quite a few resistors,
small capacitors and zeners. Suffice it to say that the internal waveforms are "interesting" and undocumented.
I chased my tail for a while: was it the 741, the tant beads? No. Move onto the PUJT what the hell
should the voltages around that be doing? Or the SCR (easier to understand!)? There was one saving grace: I haven't yet flogged my other 1502, so I could compare waveforms.
It turned out to, of all things, be the discrete current mirror: the 2N3406 transistors had decided to become 2NNorwegianBlue transistors. Once the 2N3406s had been whipped out and replaced with BC557s from the
junk box store cupboard, it sprang back into life.
Once I've made a proper 9 cell NiCd battery pack, that ought to be job done.
(until the next time)