2465B - a treasure or a time sink

[tekky]

Tek 2465B B062XXX, <3000hrs, unbroken Tek calibration seals.
I pulled it from storage to sell it. In 15 years I never really used it. A year ago it had powered up fine, and knowing about the impending need to replace the NVRAM, I could at least have recorded the cal contents. Of course now that I am serious it shows the no-cal error for the 1st time. A few days later I tried again and it showed no errors on power-up, but shortly after the display started behaving strangely, entire screen turning on and off with the focus going in and out - too blurry and inconsistent to read the cal data.

My guess is the battery is just on the edge, and it may be the piggyback battery hack, if possible, would save the calibration, but I'd still need to fix that display issue first to record the cal. At this age the scope will need a re-cap as well.

This forum has good info for 2465s but it is spread over dozens of posts. It would be nice if there were a sticky or wiki with a list of likely replacement components for each board, so you could order it all in one go. Even though I don't plan on keeping the scope it could be a fun job, like opening King Tut's tomb. Hmm...wasn't there something about a curse?

[tautech]

Welcome to the forum.

a treasure or a time sink ?

I lusted after one and I still think they are both. 

All you need to know about them is here:
https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/

Their A5 board is known to be the root of many problems especially if it has SMD caps that are prone to leak and damage the PCB as they age.

[tkamiya]

As far as I know, NVRAMS either work or don't.  Once onboard battery is below a certain threshold, RAM loses content, and it's no longer "NV".  Powering up and supplying real power doesn't restore what's already gone.

I know there are some success stories copying over someone elses' NVRAM content, but it didn't work for me. 

If it works for whatever reason, my suggestion is to save it off NOW....  I did do the piggy back hack.  It did restore functionality.  BUT, for me, the data was long gone.  (it wasn't a tek scope either)

[tekky]

The 2465B display started behaving well enough to run the diagnostics. It passed the cal data checks Tests 06:RAM battery, and 04:Calibration Data (parity & checksum). The hours are just under 3000 and the power cycle count is only 36. I bought this scope from an equipment reseller and I used it very little. I know I didn't put even 100 hours on it, but I'd think most of that count is my usage. It doesn't make sense unless someone 'turned back the odometer'. The Tek seals and Original owner seals were unbroken.

With the cal data safely archived, I broke the seals on 'King Tut's oscope'. I see no evidence of prior work - very clean inside. The only dust is on the fan blades and on U2935 which is behind a vent.

I see little to no leakage of the A5 caps C2331,C2113,C2011,C2010. No residue on the board but some gunk between pins 19,20 of U2101.

The NVRAM is not the +70 version so probably has the battery on the underside of the module. Good thing I don't need to do the piggyback hack. I see 2 six digit codes, the 9309D3 seems to imply the 9th week (Mar) of 1993. This is confirmed by the 1993 date on the EProm label. So it is around 26yrs old, and I have owned it for the last 15yrs (25 was a typo). 26yrs is a very long time for the NVRAM to still be holding data.

[tekky]

With some help I was able to remove the NVram and read it on a chip programmer. It read fine and matched the data I had captured in video with just one transcription typo. BTW, I had some trouble focusing a cell phone on the sometimes fuzzy screen chars, but attaching the phone to a camera tripod, and periodically tweaking the scope focus helped to insure a legible recording.

I mounted a FM16W08 on an inexpensive adapter, programmed it, and it passed the cal check no problems. Note that there are several sellers on ebay offering these SOP28 to DIP28 adapters, but only a few are selling the one you want (0.600 pitch). The others are selling a 0.700" pitch board which AFAIK was never a DIP standard package. Most of them do not list dimensions so beware.

I noticed that when I pulled the Fram and re-compared data against the original capture the Cal section still matched, but much of the other data did not. I suppose this is expected since stored settings change every time you touch a knob or push a button, not to mention the power count and hours.

With a little x-acto knife digging on the DS1225Y, I exposed pin7 of the embedded DS1218 and measured the battery at 3.26v out of circuit. This is what a new battery would read, but I did see that cal error once so what gives? Applying a 10k load quickly brought the voltage down to 2.5v and still dropping, so the initial high reading was just an illusion. Some combination of powering on the scope after such a long time in storage, or maybe desoldering the NVRAM and letting it sit for two weeks allowed the voltage to recover. I had the impression there was no additional load on the battery even with the chip in-circuit, so what's going on?

[rodd]

Hi,
I have a couple of Tek scopes that should have their Nvram replaced (while there still time for it....).
Do we need to make any other mod when replacing the Dallas NVram with a FM16W08?
I have a faint recall that there are other additions/modifications required.
Would you please help?
Best regards,
Roger.

[tekky]

Some have added a 1k pullup on *WE. This is easy since you can straddle either the IC legs or the adaptor pins with a smt resistor (0402 vs 0603) for *WE and VDD (pins 27,28). I'll probably do that when I get around to it. Limited testing without the pullup shows no issue.