oscilloscopes PM3394A vs tektronix 2465

[LuisRene5]

Hi Guys, I wan to buy an oscilloscope and I have 3 options in a store. An Fluke PM3394a, an tektronix 2465 and tektronix 2440. I know the limitations in bandwidth and type of measurements. However, I am worried about future maintenance and repairments for these second-hand oscilloscopes. I googled the PM3394A and there is not so much info about parts and repairments however I like the functional switch between analog and digital modes of this combiscope. However, I see that the tektronix 2465 is already well-known but many sources indicate that its difficult to deal with repairments when custom-tektronix ICs are dead.  The cost of the PM3394A is around $350 and the tektronix 2465 is around $220. My needs of the scope is common analog signals analysis (power supplies, audio signals, digital modulations) and digital analysis of traces , as you can see in this video
(I have a logic analyzer based on FPGA but I like to analyze the continuous waveforms of digital sequences) The only function that i required from digital scopes (that I can't afford) is the data analysis in a stop mode or external triggering function. What do you recommend in terms of maintenance? I hope you can give some recommendations.

[shteii01]

According to this: https://www.eevblog.com/forum/testgear/flukephilips-pm3394-options/ The 3394A had some software issues.

I don't want to sound like an a$$, but Tek 2465 was top of the line scope in its day.  Fluke/Phillips... I don't know.

[tautech]

TBH unless you have specific need for the BW you'd be better off with a modern DSO that will be more reliable than any of these aging scopes. While nobody can doubt their performance and quality, they now are quite old and unless you already have another scope, problematic to keep in service.

If that's no problem, pick the one with no or the least unobtainium parts.

[tggzzz]

My general debugging philosphy: use a scope to ensure signal integrity. When the analogue signals are known to be within the spec for digital signals, switch to debugging digital signals in the digital domain. If protocol analysis is required (e.g. decoding UART/SPI etc) then consider something like a Bus Pirate or printf statements.

There are significant differences between the 2465 and 2465B, both in terms of capabilities (autoscale and measurements) and some repairs (A5 SMD caps, calibration memory) that might be needed. There are also many commonalities (mains filter caps, ICs). All are well-described on this forum and elsewhere. (Std caution: when operating out of the case, have gentle cool air blowing across the ICs on the main board).

Replacing through hole X/Y mains filter caps is easy, and should be done before they do damage.
Replacing A5 caps (n/a 2465) requires basic SMD skills and equipment. It might be necessary to replace a few standard SMD components nearby if the caps have damaged them.
Replacing the ICs is trivial, many are available on fleabay from sphere, qservice and elsewhere. A few are scarcer, but still available.
My main concern would be the calibration memory; if it has already failed then you will have to completely recalibrate the scope. If it hasn't failed then you could copy the contents (via the screen if necessary!) and insert them into the new memory. My attitude is to make a video of the contents as seen on the CRT, and only when necessary will I replace the components.

I have no knowledge of the 2440 and Fluke, but I seriously dislike the early small portable DSOs I've played with.

If you can see the 2465 showing a good display (preferably with the correct risetime!) and the service screen showing that all tests are passed, then I would consider that option. But would that offer you usefully more capability than your 2440?