Dead Tek 2465

[elecrst]

I have a Tek 2465 hat has worked fine for several years until last week when I forgot to turn off and left it on all day and overnight.  The next morning the display was blank and all indicators were not lit.  When I turn it on the fan runs, no relays are clicking and the "A SWP TRIG'D" indicator is flashing.  I have the service manual which tells me to refer to LED-coded fault in tables 6-5 through 6-7.  I'm not sure what the charts are telling me.  Anyone have any recommendations on what I should be looking for before I really tear this thing apart?
Thanks.

[tggzzz]

Have you checked all the voltage rails for correct value and correct ripple? There's a connector in the middle of the main board where they are all available.

[tautech]

I have a Tek 2465 hat has worked fine for several years until last week when I forgot to turn off and left it on all day and overnight.  The next morning the display was blank and all indicators were not lit.  When I turn it on the fan runs, no relays are clicking and the "A SWP TRIG'D" indicator is flashing.  I have the service manual which tells me to refer to LED-coded fault in tables 6-5 through 6-7.  I'm not sure what the charts are telling me.  Anyone have any recommendations on what I should be looking for before I really tear this thing apart?
Thanks.

Welcome to the forum.

There's a wealth of info on these in this thread:
https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/

Always check the PSU first just as tggzzz says.

[lordvader88]

I bet there are multiple versions of the service manual, for good and bad, beware of that. But all thanks to those concerned that they are available to everyone.

If u do power it up again I highly recommend using a light bulb to limit the power to something no more than the 2465 should use. A current limiting light bulb in series. And also an isolation transformer and variac for everyone once they can get them.


I'm pretty new to getting serious about electronics.............so this applies to anyone

[Armadillo]

The series light bulb is for suspected shorted equipment initial startup testing purpose. Don't expect it to be hooked up the whole of the time.!

[tggzzz]

Variacs and light bulbs can cause significant problems for switched mode power supplies.

SMPSs will tend to take whatever input current is necessary to provide the internal voltages and currents. If you reduce the input voltage, the input current will go up - which can cause failures.

[Armadillo]

Variacs and light bulbs can cause significant problems for switched mode power supplies.

SMPSs will tend to take whatever input current is necessary to provide the internal voltages and currents. If you reduce the input voltage, the input current will go up - which can cause failures.

Thanks, that is good experience talk there.  Depending on the load and the wattage of the bulbs, I generally observed that the switchmode power supply tends to do the restart loops repeatedly and other visual abnormal stuffs.
But I don't think the current will increase though.     so it should be safe for initial startup testing purpose.

[oldway]

Even if it's a bit off topic, I would like to explain my point of vew for beginners who would like to buy a 2465.

At the time it was still manufactured, this oscilloscope was probably one of the best portable oscilloscopes on the market.

But it is very sophisticated.

At the time, this was not a problem because spare parts were readily available and breakdowns were rare.

But 20 or 30 years later, this sophistication becomes a very negative element, because the number of breakdowns is higher, the breakdowns are more difficult to repair and we do not find any special components such as hybrids.

So if I can give you an advice, avoid this kind of oscilloscope, choose either a simpler analog oscilloscope, or a modern dso, preferably a new one.

For elecrst, welcome to this forum....
The first think to do is to check all the power supplies rails...

Good luck.

[Armadillo]

the number of breakdowns is higher

Plus +1;

Judging from the numerous breakdowns in this forum, if I remember.... something to do with "HEAT"... I tend to agree with the advise given.

[tggzzz]

Variacs and light bulbs can cause significant problems for switched mode power supplies.

SMPSs will tend to take whatever input current is necessary to provide the internal voltages and currents. If you reduce the input voltage, the input current will go up - which can cause failures.

Thanks, that is good experience talk there.  Depending on the load and the wattage of the bulbs, I generally observed that the switchmode power supply tends to do the restart loops repeatedly and other visual abnormal stuffs.
But I don't think the current will increase though.     so it should be safe for initial startup testing purpose.

Ignoring losses, P_acMains = P_scope, and P_acMains = V_acMains*I_acMains. For a constant P_scope, if V_acMains decreases then I_acMains will increase proportionately.

[Armadillo]

Ignoring losses, P_acMains = P_scope, and P_acMains = V_acMains*I_acMains. For a constant P_scope, if V_acMains decreases then I_acMains will increase proportionately.

Thanks, that is Joules/ohms law but in magnetic circuit, you should apply the faradays law. That is voltage divided by inductance of the transformer or inductor is equal to the rate of change of the current with time at operating frequency condition. The power cannot be constant because the secondary side of the transformer will refuse to provide that constant power, meaning the power will be reduced.

Hope that clarifies the issue.   

Edit: you need voltage to push the current through the inductance.  That is also why you want to use higher voltage to drive stepper motor.
Edit: if you design switching transformer, then its easier to understand the application.   

[tggzzz]

Ignoring losses, P_acMains = P_scope, and P_acMains = V_acMains*I_acMains. For a constant P_scope, if V_acMains decreases then I_acMains will increase proportionately.

Thanks, that is Joules/ohms law but in magnetic circuit, you should apply the faradays law. That is voltage divided by inductance of the transformer or inductor is equal to the rate of change of the current with time at operating frequency condition. The power cannot be constant because the secondary side of the transformer will refuse to provide that constant power, meaning the power will be reduced.

Hope that clarifies the issue.   

Edit: you need voltage to push the current through the inductance.  That is also why you want to use higher voltage to drive stepper motor.
Edit: if you design switching transformer, then its easier to understand the application.   

Well, I'm sensible enough not to design SMPSs

However, the currents/voltages I was referring to are not those in the transformer, but in the surrounding non-magnetic components. Although I haven't thought about it in detail - and certainly haven't done any measurements - I would be surprised to find that energy conservation doesn't hold. However, I would expect that the phase relationships will complicate analysis to some extent.

[elecrst]

Thanks for the replies.  Wow! even to how to power up with a light bulb which I have done many times in checking old vintage AM gear.  But power up is not the problem.  It powers up but apparently does not finish the self-check because the only thing that show up is a blinking "A TRG'D SWP" indicator.  I am not a beginner as I have fixed 3 older analog Tek scopes over the years as well as other test equipment and ham equipment.  I looked at the thread on the 2465B tear down referred by tautech.  It has a lot of good and some conflicting information and doesn't address my problem.  After making my way through the service manual,  I checked all the power supply voltages and they are all within spec.  I guess I'm up a creek without a scope until I can borrow one to check this out.  I didn't pay a whole lot for this one several years ago so if one of the hybrids is gone it is no great loss.  Thanks again.

[tggzzz]

Take it slowly, follow the faultfinding flowcharts in the manual corresponding to your particular scope (there are some notable differences). Think about what you can and cannot infer. Look at the info here and on the Yahoo TekScopes group.

The hybrids are usually available on fleabay and from qservice and sphere. But it is best to ensure a specific hybrid is faulty before going down that route.

[MarkL]

A flashing "A TRG'D SWP" indicates that there is a failure in the Kernel (pg 6-12, "Kernel Tests").  This means that the processor has detected a problem with its own operation, RAM, or ROM.  This is before it gets to any self checks on scope-related peripherals.

If you look at tables 6-5, 6-6, and 6-7, it will try to tell you where it has found a problem.  It does this by lighting some combination of Ch1 through Ch4 LEDs, "+" LED, "-" LED, and "READY" LED.

According to the tables, if nothing else is on besides the blinking "A TRG'D SWP", it indicates a probable issue with the A5 control board RAM or associated circuitry.

The manual also says you can press the "A/B TRIG" button to force the scope to run anyway.  This may or may not work depending on the severity of the problem.  If you do this and it starts working, you still need to figure out what's causing the kernel error since it may crash or not operate properly.  Any other issues you might see at this point should not be considered a valid problem until you fix the kernel error.

[elecrst]

Thanks MarkL.  That is exactly the problem I have.  None of the LED's light, and pressing the "A/B TRIG" button does nothing so I guess the problem is somewhere on the A5 board.  I guess I'll have to see if I can borrow another scope to check this one out.  Anyone know of any known issues with the A5 board?

[MarkL]

You are definitely going to need a scope to troubleshoot this problem.

When you get one, I would suggest starting by checking the processor clock to make sure it's ok, and then put the processor in NOP mode by removing jumper P503.  This will allow you to check the address lines and the address decoders as the processor cycles through all 16 bits of its address space.  In particular, see if the address decode to the RAM is working, since that's what it's currently complaining about.

NOP mode is described in the theory section, pg 3-5 under "Data Bus".  And while you're in that part of the manual, it wouldn't hurt to read all the theory section on "Processor and Digital Control".

If you're rarin' to do something until you have a scope available, you could do a quick tour around the A5 board and verify that all chips are getting power.  It's a long shot, but I did have a case once where a hairline crack in a trace cut the power to a chip (but it was not on a 2465).  The dead chip dragged down the data bus to the point where nothing was working.

Did you get this scope already dead?  Or did it just stop working one day?

[Armadillo]

Sounds like the "Photovoltaic Torchlight" Syndrome !!   

Do the obvious first, leaky caps, broken trace, dried joints, etc.....   

[elecrst]

The scope was working when i got it and has worked flawlessly for a couple of years.  As stated in my original post I forgot to shut it off one afternoon and it was on all night and when I looked at it the next morning it was dead.  I'll start checking the A5 board for power to the chips etc.

[elecrst]

Checked all the voltages on the A5 card and all OK.  One strange thing I see is that U2496  on the card is a HM6116 2k X 8 SRAM.  But the schematic shows a 4801 and the parts list for this board doesn't even have a listing for U2496.  The service manual I have is revised Oct. 1984.

[MarkL]

Oh well, it was worth checking.  Time to get a scope on it...

The manual I have is March 1985, and it does list U2496 as an HM6116P-3 in the parts list.  But like yours, in the schematic it shows a 4801.  The 4801 would be the Mostek MK4801A 1k x 8.