开云体育

When I turn this 'scope on it boots up and works for a few minutes.
Then it starts re-booting with relays clicking and all power supplies coming on briefly to correct voltages with no ripple, then dropping to 0v.
While troubleshooting the problem went away and the 'scope ran for hours.
That was two days ago. Today it started the same way and is again cycling.
I am reluctant to replace all capacitors in the power supply without knowing the cause of the problem.
I am even more reluctant to start replacing diodes and possibly introducing a second fault.

Is there a way to 'scope the capacitors for ripple?
Can the power supply be run out of the 'scope?
Troubleshooting assumes the PS is working. The PS is always the first thing I check.

Bonjour cher Monsieur

Very common issue.

The cycle indicated an overload on one PS bus, the time delay indicated heat sensitive component failure.

May be in PSU or on a board in the scope.

First get full service manual for the serial number range of the scope, see tekwiki, W140, Artech, etc.

Follow troubleshooting process.

DO NOT blindly replace caps.

Check threads on similar in EEVBLOG 2465B teardown 75 Pages, and both tekscopes and tekscopes2 in groups.io

Bon chance


Jon

Thanks Jon. I have a paper manual but it does not cover this 2465 CTS model.
It has a board under the top cover which I think is the CTS option.
Can I run the 'scope with this CTS board disconnected?
The manual says the power supply needs a load so it will be a problem taking it out to troubleshoot.
I don't see instructions to try disconnecting one board at a time. Can this be done safely?
It is strange that the problem is not there for a few minutes, then starts but all supplies have correct voltages between very short shutdown, then cures itself after 5 minutes and runs for hours. It does not return until the 'scope is shut down for a long time.
I don't think it is a heating problem or it would be worse after warmup, not get better.

On Sat, Feb 17, 2024 at 10:24 AM, peter bunge wrote:

Then it starts re-booting with relays clicking and all power supplies coming on briefly to correct voltages with no ripple, then dropping to 0v.
While troubleshooting, the problem went away and the 'scope ran for hours.
That was two days ago. Today it started the same way and is again cycling.

I have no experience with 2465s, but to me, this screams bad connection or broken solder joint. Something you disturbed re-made the connection, but it opened again after some random time or possibly a bit of mechanical movement. Inspect the bottom of each board for solder joints
with cracks around them. Pay special attention to connectors or anything else that might get stressed from some outside force, or by heat. Since this has been happening for some time, it may show some signs of overheating that area of the board.

There are very few items I would ever replace on sight. Shotgunnng capacitors without understanding the problem is likely to create new problems without solving any old ones.

Real troubleshooting takes patience and builds understanding. It will pay off in the long run.

One of my favorite troubleshooting memories was a Tek 5000 series dual beam scope. It had 2 vertical amps and one time base, so both traces scanned horizontally together. At times it would stop scanning one have of the screen. The other people who used it had gotten used to pulling one random plugin out and slamming it back in place. That often "cured" the problem, temporarily, but I thought it abusive.

I got out the service manual and found that there were 6 identical transistors that drove the deflection plates, a pair for each plugin. Out of each pair, one pulled one deflection pate to the left (or down) while the other pulled the other plate right (or up.) It didn't seem like there was much chance of a problem other than one of those transistors. These being metal cased transistors, I suspected that one of the weld connections inside the metal case had come loose and was intermittent. So, since they were in sockets, I swapped two of them around.

The next time the problem occurred, it affected the deflection that I'd just moved the suspect transistor to. I checked the Tek parts list, and it said "Replace with xxxxxxx" (some sort of Japanese transistor.)

I ordered two, replaced the now known bad transistor, and put the spare in a zip lock bag stuffed somewhere in the scope chassis.

A year later, a second of the same bunch failed, and I replaced it with the spare. A few months later, another of the OEs failed, so I ordered 4 more of the Japanese replacements. When they came it, I replaced all the remaining OE output transistors. There was clearly something unreliable about this batch of original Tek transistors. It ran without a problem for at least another 15 years.

Never forget that the most common problems are connections. That includes transistor and IC sockets, switch contacts, and board interconnects, and, occasionally, as above, internal transistor connections. And don't forget unintended connections from dropped bits of solder, wire, or hardware, as well as tin whiskers in old Ge transistors.

Good luck with your search,

--
Jim Adney
Madison, WI USA

Thanks Jim; I have a lot of troubleshooting experience with mainly HP test equipment.
The troubleshooting for the 2465 power supply starts with "disconnect the fan" and see if this solves the problem.
Since the fan keeps trying to start every cycle it seemed like a good idea.
However 18 connectors and a handfull of screws later I finally removed the power supply and got to see the famous fan connector.
Now what? I can disconnect the fan then re-install the power supply and 18 connectors and try again.
This is the worst designed piece of equipment I have seen.
How do you guys troubleshoot a 2465? I could make patch cables to try and run it in pieces.

Peter: Got my frst 2465 in 1992. Have six 2645/7/B/CTT so lots of experience...


0/ We use fingers (WITH CAUTION!) and freeze mist (CAREFULLY) to spot overheating culprits.

00/ SMPS has notorious bus caps on + and -15, that are over loded wth ripple I and fail.

BEWARE OF PCB SILKSCREEN ERROR. See Conrad Audio in Isreal and Qservice in Greece for advise on 246X recapping.

1/ Some component is heating or causing current overload. Most likey a tantalum cap. Careful visual inspection of EVERY PCB may show a discolored r distorted tant.

2/ Many threads here and on EEV blog re 246X dummy loads and PSU troubleshoot.

3/ Get exact manul in PDF form. CTS is an option in the options manual.

Disconnect CTS id fine BUT needs certain connectors and cables jumpered to bypass thru connections.

4/ Generally just a single supply needs a load eg the main prereg bus or 15V, I forgot.

Great Scope esp wit the CTS option, very useful.

Gott jam!

Bon Chance


Jon

Thanks Jon; I removed the power supply and disconnected the fan. Put it back together and get the same symptoms. It is not the fan circuit.
I don't think the troubleshooting chart is going to help.
Same symptoms: Boots up and runs normal for a couple of minutes then starts cycling.
I fed a hot air gun into the power supply area through the holes. Cautiously for about 10 seconds (moving access holes) then waited. It started working within 15 seconds.
Previously it took 10 minutes to work reliably. Then it would work for hours.
I just repeated the experiment with hot air after an hour of being off and got it running again within 20 seconds.
I have the top cover removed and identified the horizontal board under the top plate as the A20 buffer board.
The vertical board on the left is the A20 HPIB. I can pull that board safely I think.
There is no A25 TV board in the center slot.
The vertical board on the right is the Counter Timer Trigger board? (CTT) What is CTS then? Can I just pull this board as well.
Where is P302 and P303 options connections on the inverter board? (Troubleshooting chart tab 11).
I will try to find that next but I think there is a heat sensitive component somewhere.
It would be nice to leave the top cover and the options off to direct hot/cold air at specific parts og the power supply.
The GPS, Lemo, and BNC connections just go to external connectors.
That leaves C, D, and E on the right, two cables on the left and two ribbon on the front.
I don't know where they go without spending days searching.
Can someone point me to a link on how to run without this board. Someone must have done this.



.

First of all, I would avoid use of a heat gun!!!! too easy to burn. way too large area heated!!!

Freeze mist is more precise and cooling rather than heating should make it work when suspected components area is found.

VERY unlikely CTT bd is the issue. BEWARE replacement of CTT requires some connectors at back bottom edge in hard to reach easy to miss location.

Strongly suggest you research these topics here, and the other Tekscopes group.io as well as EEVBLOG 2465 Teardown thread.

Tek serve manual and indication on main bd ?1 shows certain jumpers to replace several ?ables from A1 to CTT, when CTT is removed, path for trigger signals into, out of CTT.

Easy to get connector position lost.

Jon

It is nearly impossible to troubleshoot this power supply.
I can see why people shotgun replace all caps.
I cannot find any reference in this forum on a fault being traced to a specific capacitor.
I see a mention of capacitors on the + and - 15 v supplies but that is a bit vague for me to locate.
I will try to find the EEVBLOG 2465 Teardown with a search.
My test shows the scope works 10 seconds after a few seconds blast of hot air in the middle rectangular hole beside the connector to the A20 buffer board connector that plugs into the power supply. Then it works indefinitely. Without that it takes many minutes to start working.
Funny that it come on working in 15 seconds, works for 40 seconds then fails for many minutes before working indefinitely. The heat just speeds up the recovery.

sounds like possible cracked/poor solder connections or occasionally a bond wire issue occurs in a device...
had a scan generator that took 6 months of chasing my tail to find a similar issue. turn it on , it worked for a few minutes then failed for 2 or 3 minutes (or apply some heat)..then came back on and ran fine , no more issues until it sat for 24 hours or more. it was definitely a temperature issue...finally found it - was bond wire in a chip ( guess)..worst part freeze spray did not find it only a gentle heating of the entire board. intermittent are difficult, hard failures are easier. This was a pain to reach too- buried part! (IIRC ua741 or TL081 had a 1 on the end..IDR it was National chip...lol)
btw i only had a couple days of the week when the implanter was down for me to work on it.
搁别苍é别

These power supplies use comb-like structures, connecting PCB's. The comb's square pins reach into hollow square "tubes". Good contact between pins and tubes is dependent on a tiny spring inside each tube. I've seen situations where these springs had disappeared, causing intermittent contact.

Raymond

Thanks Raymond and Renee.
I found the EEVBLOG 2465 Teardown but it just has pictures like I have been taking.
I cannot find a reference to a specific capacitor that has failed. With the difficulty of replacing one I can see why.
I don't see anyone running the power supply out of the 'scope either.
I don't see any damaged components or bad solder joints on my boards.
There is a 10 second lag between applying a few seconds of hot air and the recovery. It seems too short to be heating a capacitor.
Tapping the boards has no effect but I will try to get an insulated rod into the tight spaces.
Prying the boards or removing the PS several times should change the symptoms but it has not.

Are you familiar with the (typical) tantalum capacitor problem in tek equipment of that age?? The manufacturer said that the parts were good working too close to their voltage limit.? i.e. a 6.3 volt part would be fine on a 5 volt rail.

The typical parts that seem to fail are on +/- 15 volt rails. The parts failing can be either on boards or in the supply itself.? If the supply will run without an external load, then check the voltages without the rest of the scope.? If not, you may want to look at building a dummy load.? The service manual may have a section on troubleshooting the power supply.

If possible, you can look at the supply voltages and see what happens when it fails?? 15 volts low?

It's possible that you may be able to figure out what parts on the A20 board (may be suspicious) connect to the 15 volt rail and check them.

If heat fixes it, you might try freeze spray.? It may or may not be a capacitor.

Harvey

See this: and also in my book Tektronix volume 2 you can find several notes on 2465 and similar PSU repair (free on www.k100.biz).

But I agree with Harvey: you should try to investigate what exactly fails and if the inverter runs or what. Remember that the inverter part is with live ground, you need an insulation transformer. And never forget the extra fans if you switch it on outside the box. Another ideas is trying to operate it at lower voltage with a Variac. Remember that the TL594 can be tested also without switching on the main supply, with an external power supply.

Does the unit restart automatically when shut off?

I put a power supply in from another 'scope. It starts right away.1. The problem is definitely in the power supply.
The power supply runs out of the 'scope with a load per the manual.

Installed in the 'scope and running on an isolation transformer so I can attach anothe 'scope:
I shorted the current sense resistor and put a 'scope on the collector of Q1040. The?PS does?not start. No sign of Q1040 turning on.I put an analog meter on the current sense resistor. The voltage is steady at 103 mV (515 mA).2. It is not overcurrent preventing starting.

Heat blown into slots in the top cover does speed up recovery. It does not take much and I don't really know what I'm heating. Something center or more to the front I think.

When running the AC current flickers at 1.1 amps and seems to want to drop to 1 amp. I have seen it running at 1 amp with jumps to 1.1 amps.
My other 2465 runs steady at 0.95 amps. It?has one less option board.
There is?something surging power. If the current surges were caused by any of the regulators after T1060 it would show in the current limit sensing.
Same goes for anything in the Q1060 and Q1070 drivers. It only leaves the Q1062 circuit or the voltage doubler input. I did not change the voltage doubler capacitors.
Any suggestions? Remember I have to disconnect 18 cables, remove the top cover with the CTS and GPIB option boards, and remove the power supply for every test.
All aluminum electrolytic capacitors except the doubler pair have been replaced. CR1022 and Q1040 have been replaced, so has C1024.
Q1050, Q1060, and Q1070 would show up in the current sense if they were the cause.

I posted this once but it did not show up.
I put a power supply in from another 'scope. It starts right away. The problem is definitely in the power supply.
The power supply runs out of the 'scope with a load per the manual.

Installed in the 'scope and running on an isolation transformer so I can attach another 'scope:
I shorted the current sense resistor and put a 'scope on the collector of Q1040. The PS does not start. No sign of Q1040 turning on. I put an analog meter on the current sense resistor. The voltage is steady at 103 mV (515 mA). It is not overcurrent preventing starting.

Heat blown into slots in the top cover does speed up recovery. It does not take much and I don't really know what I'm heating. Something center or more to the front I think.

When running the AC current flickers at 1.1 amps and seems to want to drop to 1 amp. I have seen it running at 1 amp with jumps to 1.1 amps.
My other 2465 runs steady at 0.95 amps. It has one less option board.
There is something surging power. If the current surges were caused by any of the regulators after T1060 it would show in the current limit sensing.
Same goes for anything in the Q1060 and Q1070 drivers. It only leaves the Q1062 circuit or the voltage doubler input. I did not change the voltage doubler capacitors.
Any suggestions? Remember I have to disconnect 18 cables, remove the top cover with the CTS and GPIB option boards, and remove the power supply for every test.
All aluminum electrolytic capacitors except the doubler pair have been replaced. CR1022 and Q1040 have been replaced, so has C1024.
Q1050, Q1060, and Q1070 would show up in the current sense if they were the cause.

I checked the Inverter Drive power supply and it works for seconds until something shuts the drive off.
The power supply draws normal current for up to 5 seconds and the pulses appear on TP75 then stop abruptly.
This repeats faster and faster until the 'scope stays on. It is accelerated by hot air down the slots in the top plate.
It is NOT current limiting.
Maybe something is shutting down U1030. I replaced C1034.

Peter,

R1020 is known to go bad. That should be 2W. Make sure to get a resistor that is about 15mm long. Some 2W resistors are 3.68 x 8.72mm which is smaller than a 1/2W. C1034 is originally a tantalum. I replaced the three on that board with 50V film types to ensure low ESR. C1055 and a few others are the clear amber tinted RIFAs that are known troublemakers. I used 630V film types to replace them. C1051 is another condenser that is a RIFA type. Look to see if those were replaced. CR1110 is known to go bad. It is a Schottky rectifier. I replaced mine with a Si type and it worked with the proper voltage out. Mine is 200V, I think, because I have some in stock. If you decide to replace the doubler condensers, I suggest using Nichicon LGR types. I put in 390mfd in mine. Make sure the diameter is small enough to fit where the originals are. Add wires from the lugs/pins of the LGR type to the board.

Mark

Thanks Mark, I replaced R1020 although it was not bad. The troubleshooting
for that circuit using a power supply says the switch should activate from
20 to 25 volts. Mine latches at 17.5 volts. I don't think it is significant.
I replaced C1024 with no change.
The RIFAs look good so I did not change any. They don't fit with the
symptoms.
I don't know if C1051 would cause my problems. I won't discount it.
CR1110 would show in the current sensing if it was shorting. There is no
evidence of a problem after T1060.
I don't think the doublers would cause my symptoms but I may change them. I
have the ones recommended.
I'm thinking that something in the U1030 circuit is acting up.Maybe U1040
feedback optocoupler? I need to see what its supposed to do.
I have good 'scope pictures of the signals but cannot post them. They show
the drive signals suddenly stopping after running from 1 to 5 seconds. The
recovery to working happens when the interval DECREASES. It seems that
there is something like overvoltage that needs to drift down to a control
range.
Peter

On Wed, Mar 6, 2024 at 8:15?PM Mark Vincent <orangeglowaudio@...>
wrote:

Peter,

R1020 is known to go bad. That should be 2W. Make sure to get a resistor
that is about 15mm long. Some 2W resistors are 3.68 x 8.72mm which is
smaller than a 1/2W. C1034 is originally a tantalum. I replaced the three
on that board with 50V film types to ensure low ESR. C1055 and a few others
are the clear amber tinted RIFAs that are known troublemakers. I used 630V
film types to replace them. C1051 is another condenser that is a RIFA type.
Look to see if those were replaced. CR1110 is known to go bad. It is a
Schottky rectifier. I replaced mine with a Si type and it worked with the
proper voltage out. Mine is 200V, I think, because I have some in stock. If
you decide to replace the doubler condensers, I suggest using Nichicon LGR
types. I put in 390mfd in mine. Make sure the diameter is small enough to
fit where the originals are. Add wires from the lugs/pins of the LGR type
to the board.

Mark

I found out how to send pictures of the 'scope screens
This link shows TP75 on schematic 9 in yellow and the collector of Q1062 in
Purple. 10 uS/div and 5v/div
The 'scope ground is on Ref 1 just marked REF on the 2465 schematic (see
2465B schematic to compare the two reference * and **)

and this shows the same signals at 1 second/div

I hope everyone has access.
I am moving my attention to the U1030 circuit and how it regulates.
This has to be a common fault. Two out of Four 2465 that myself and a
friend own have the problem of clicking for several minutes before settling
down.
Several people have told me it is a common problem. No one says what causes
it. The common repair is to replace all capacitors and a few other
components.




On Wed, Mar 6, 2024 at 10:32?PM peter bunge via groups.io <bunge.pjp=
gmail.com@groups.io> wrote:

Thanks Mark, I replaced R1020 although it was not bad. The troubleshooting
for that circuit using a power supply says the switch should activate from
20 to 25 volts. Mine latches at 17.5 volts. I don't think it is
significant.
I replaced C1024 with no change.
The RIFAs look good so I did not change any. They don't fit with the
symptoms.
I don't know if C1051 would cause my problems. I won't discount it.
CR1110 would show in the current sensing if it was shorting. There is no
evidence of a problem after T1060.
I don't think the doublers would cause my symptoms but I may change them. I
have the ones recommended.
I'm thinking that something in the U1030 circuit is acting up.Maybe U1040
feedback optocoupler? I need to see what its supposed to do.
I have good 'scope pictures of the signals but cannot post them. They show
the drive signals suddenly stopping after running from 1 to 5 seconds. The
recovery to working happens when the interval DECREASES. It seems that
there is something like overvoltage that needs to drift down to a control
range.
Peter

On Wed, Mar 6, 2024 at 8:15?PM Mark Vincent <orangeglowaudio@...>
wrote:

Peter,

R1020 is known to go bad. That should be 2W. Make sure to get a resistor
that is about 15mm long. Some 2W resistors are 3.68 x 8.72mm which is
smaller than a 1/2W. C1034 is originally a tantalum. I replaced the three
on that board with 50V film types to ensure low ESR. C1055 and a few

others

are the clear amber tinted RIFAs that are known troublemakers. I used

630V

film types to replace them. C1051 is another condenser that is a RIFA

type.

Look to see if those were replaced. CR1110 is known to go bad. It is a
Schottky rectifier. I replaced mine with a Si type and it worked with the
proper voltage out. Mine is 200V, I think, because I have some in stock.

If

you decide to replace the doubler condensers, I suggest using Nichicon

LGR

types. I put in 390mfd in mine. Make sure the diameter is small enough to
fit where the originals are. Add wires from the lugs/pins of the LGR type
to the board.

Mark