Tektronix 2215 problem

[MXstarter]

Hello everyone,

I have recently purchased a Tektronix 2215 scope from ebay,
When I tried to switch it on, it does not respond at all. I checked
the fuse and it is blown. 

So I went and bought a new fuse, and it blew that one as well.
The previous owner said he did a test with both probes hooked
on to the testing point and the scope switched off suddenly.

Has any of you experienced the same problem ? If you know how
to fix it, please help me.

Thanks a lot,
have a nice day,
Henry

[poot36]

I would start by checking the line input filter + caps for shorts as well as the bridge rectifier.

[MXstarter]

Thanks for your reply, I checked the input caps on the front panel, nothing unusual or fired.
I disconnected the wires connecting the testing signal port to the PCB. It's too late to buy
another fuse today, I will try tomorrow.

The input is rated at 400 V, I just don't understand how a 500 mV peak to peak signal is going
to kill the whole scope.

I was soo happy when the scope arrived today,  I am sad now. 

[poot36]

Does this unit have a physical power switch or a soft one?  If it is a physical one turn it off and measure the resistance between the power cord pins and see if there is a really low reading (under a couple of ohms).  If it is a soft power switch unplug the power supply from the rest of the scope and do the test above and if that passes plug the scope in and see if the fuse blows again (if you have a dim bulb tester use it to prevent the fuse from blowing even if there is a short, use a 40W light bulb).  I do not think that the problem is at all related to measureing the test signal or the probe input circuit.  With most old Tek scopes the tantalum caps in the power supply short out causing this type of problem and sometimes more problems.

[tautech]

Grab the link from the first post in the Repair board.

Thanks for your reply, I checked the input caps on the front panel, nothing unusual or fired.
I disconnected the wires connecting the testing signal port to the PCB. It's too late to buy
another fuse today, I will try tomorrow.

The input is rated at 400 V, I just don't understand how a 500 mV peak to peak signal is going
to kill the whole scope.

I was soo happy when the scope arrived today,  I am sad now. 

It will be nothing to do with the signal inputs, something has popped inside, commonly in the PSU.
Find a Service manual and use in conjunction with the PDF in the previously mentioned link.
Make a "dim bulb tester" to limit the energy supplied on power up.

[MXstarter]

Thanks for your replies.

I am going to buy another fuse tomorrow and start to test for short caps ( starting from the PSU ).
My first oscilloscope turns into my first repair, who would have thought ? 

Please give me some tips for safety, I don't want to give myself a nasty shock .

This is my first time doing a serious repair attempt. It can give me some precious hand on experience
( I am in my first of master of mechatronics engineering).

Hopefully it will work out.

Like Dave says: don't turn it on, take it apart !!

[David Hess]

Did you get a 2215 or 2215A?  They are significantly different and further there are two versions of the 2215 power supply; the early one uses a TRIAC based preregulator and the later one is like the 2215A with a switching preregulator.

You need a service manual if you want to diagnose and repair this.

[poot36]

Some safety tips use an isolation transformer, plug the unit into a GFCI protected socket, only use one hand to probe the circuit (ie you do not want to be grounding yourself so that if you do accidentally touch something live you will only feel a tingle not a shock) or better yet attach your meter with clips  to the points you want to measure and then only plug the unit in for long enough to get a reading and then unplug.  Not so much a safety tip for yourself but for the equipment is the aforementioned dim bulb tester which is just a light bulb socket in series with one of the power supply input leads to limit the current the device can draw.  I made mine from a old lamp cord a electrical outlet and outlet mounting box and a light bulb socket.

[MXstarter]

WOW, you replied me a second time. Thanks a lot.
Before doing live testing, I am currently just checking individual component in the psu and trying to find a short somewhere.
I don't even have a hot air gun or desolder  tool, just wondering how to get a 2 pin through hole device off the board with a iron.
Hope I can find the failed component soon.

[poot36]

Depending on what the component is I usually take different approaches.  If it is a through hole cap I heat up on pin and then solely push the cap towards the other pin and when it wont go any further I do the same thing for the other pin.  If it is a through hole resistor I try to get a flat head screwdriver under it and heat up one side and then twist the screwdriver to lift the heated side until it come out and then either grab the resistor with my hand and heat up the other side and pull or use pliers, you can also use pliers instead of the screwdriver just grab the leg of the resistor that you are heating up and pull (this method has a higher chance of damaging the resistor but it works great when the board is packed and there is no room to get the screwdriver under the resistor.  If it is a three pin in a row device just try to heat up all three pins at once and wiggle the component across the axis of the three legs.  I would recommend getting a spring powered solder sucker for this work or at least some solder brade.

[Mark Hennessy]

I'm afraid these scopes have notoriously unreliable PSUs. I have 2 squirrelled away, and hope to end up with one working one eventually. Not that I feel especially motivated, as these 'scopes are nothing special at all - despite the badge.

I did test the most promising one by applying a DC voltage to the instrument. The mains PSU produces a single DC supply rail, so applying that from a bench PSU confirmed that the rest of the instrument worked OK.

Fixing switched-mode mains power supplies is a highly specialist activity - one that many highly experienced engineers would not attempt. I completely second the recommendation to use a lamp limiter and an RCD/GFI. All of the circuitry is "live" w.r.t. mains earth, so don't touch! Any attempt to connect earthed test gear - such as another 'scope - will result in a large "bang" unless you have an isolation transformer.

Also, be aware that faults in power supplies usually "cascade". It is doubtful that there is one single failed component - there will be a chain of destruction that will often result in new components instantly being destroyed at power-up (the lamp-limiter will help to an extent here).

In short, switched-mode power supplies had a bad reputation among service engineers at the time this 'scope was released, and for good reason. Today, they are more reliable thanks to better semiconductors and better design, and failures tend to be more about bad capacitors. From what I've read on this thread, it seems that you are a relative beginner to this, and I would suggest that this is not a good first project for you (I mean that completely respectfully ).

How was it described on eBay? Was it "used", or "spares/repair". Clearly it was faulty when the seller shipped it - if it was incorrectly described, I'd be seriously thinking about sending back for a refund. There are countless good oscilloscopes out there for almost no money - it might be wiser to move on. Again, the 2215 really is nothing special.

Good luck

[David Hess]

I'm afraid these scopes have notoriously unreliable PSUs. I have 2 squirrelled away, and hope to end up with one working one eventually. Not that I feel especially motivated, as these 'scopes are nothing special at all - despite the badge.

Fixing switched-mode mains power supplies is a highly specialist activity - one that many highly experienced engineers would not attempt. I completely second the recommendation to use a lamp limiter and an RCD/GFI. All of the circuitry is "live" w.r.t. mains earth, so don't touch! Any attempt to connect earthed test gear - such as another 'scope - will result in a large "bang" unless you have an isolation transformer.

Using a lamp limiter is usually a bad idea with switching power supplies because of their negative resistance input characteristic and operating them at low input voltages can result in exceeding the current rating of the power switch and operating it at a high duty cycle.  This points directly to a major cause of failure; if the input capacitor wears out, then the low average voltage at the input causes the power switch to fail which usually causes other damage.

Oddly enough this failure mode probably does not apply to the early 2215 power supply which uses a TRIAC based preregulator.

Also, be aware that faults in power supplies usually "cascade". It is doubtful that there is one single failed component - there will be a chain of destruction that will often result in new components instantly being destroyed at power-up (the lamp-limiter will help to an extent here).

Tektronix was pretty good about including crowbar circuits in their power supplies to prevent catastrophic failures from progressing to the low voltage side.  The later 2215 power supplies includes this but the earlier one rely on a shunt zener diode.  In either case the crowbar or shunt zener diode should be checked.

In short, switched-mode power supplies had a bad reputation among service engineers at the time this 'scope was released, and for good reason. Today, they are more reliable thanks to better semiconductors and better design, and failures tend to be more about bad capacitors. From what I've read on this thread, it seems that you are a relative beginner to this, and I would suggest that this is not a good first project for you (I mean that completely respectfully ).

As I mentioned above, age related failures with the switching power supplies in the 22xx series also tend to be because of capacitors wearing out.

I have to admit that there are a fair number of mysterious failures in these power supplies but I have never gotten my hands on a bad one to figure out exactly what goes wrong with them.  I suspect the saturable inverter becomes unbalanced allowing the transformer to saturate causing excessive input current and failure of the inverter's power transistors.

I agree that this is not a repair for a beginner because circuit operation is complex and because of the safety issues when working on the line side of an off-line switching power supply.

Again, the 2215 really is nothing special.

The 2213 and 2215 have other problems as well compared to the later 22xx series oscilloscopes.  They were the very first ones and quickly replaced with the 2213A and 2215A which were much more reliable.

[Mark Hennessy]

Using a lamp limiter is usually a bad idea with switching power supplies because of their negative resistance input characteristic and operating them at low input voltages can result in exceeding the current rating of the power switch and operating it at a high duty cycle.  This points directly to a major cause of failure; if the input capacitor wears out, then the low average voltage at the input causes the power switch to fail which usually causes other damage.

Oddly enough this failure mode probably does not apply to the early 2215 power supply which uses a TRIAC based preregulator.

A well-designed SMPSU will have protection against an input voltage that is too low. Designers picked up on this fairly early on, as brown-outs are obviously a serious issue in some parts of the world.

But honestly I can't say if that applies to the 2215 as I haven't studied the circuit yet.

Also, be aware that faults in power supplies usually "cascade". It is doubtful that there is one single failed component - there will be a chain of destruction that will often result in new components instantly being destroyed at power-up (the lamp-limiter will help to an extent here).

Tektronix was pretty good about including crowbar circuits in their power supplies to prevent catastrophic failures from progressing to the low voltage side.  The later 2215 power supplies includes this but the earlier one rely on a shunt zener diode.  In either case the crowbar or shunt zener diode should be checked.

For clarity, I didn't mean this. Perhaps "cascade" wasn't the best word exactly; perhaps I should have provided more explanation.

I was referring to the PSU on its own. Often, if the switching transistor fails, it can take out many other components near to it, such as the control IC, the current sense resistor, perhaps many other things.

What happens here is that the service guy spots a shorted transistor, changes it without making other checks, and the new transistors fails short in the same way. So next time, one of the other faulty components is spotted, but now, two new components fail again at the first power-up attempt. Repeat until madness sets in

It's fair to say that switched mode power supplies rarely fail in an over-voltage condition. Most include a crowbar by default, of course, but generally failures stay within the PSU in most cases. Happily...

Away from that minor clarification, I happily agree with everything you say. Yes, capacitors are the usual enemy. High-value resistors in the start-up circuit can be another common problem. Both these are easy to diagnose and fix, but blow-ups are a different matter, and generally require a much deeper level of understanding and diagnosis skills. Sadly, I know that at least one of my 2215s has had a blow-up, so I'm not looking forward to tackling that - and I say that as someone with many years of experience with these things 

All the best,

Mark

[David Hess]

Using a lamp limiter is usually a bad idea with switching power supplies because of their negative resistance input characteristic and operating them at low input voltages can result in exceeding the current rating of the power switch and operating it at a high duty cycle.  This points directly to a major cause of failure; if the input capacitor wears out, then the low average voltage at the input causes the power switch to fail which usually causes other damage.

Oddly enough this failure mode probably does not apply to the early 2215 power supply which uses a TRIAC based preregulator.

A well-designed SMPSU will have protection against an input voltage that is too low. Designers picked up on this fairly early on, as brown-outs are obviously a serious issue in some parts of the world.

But honestly I can't say if that applies to the 2215 as I haven't studied the circuit yet.

Unfortunately studying the circuit in this case is not very informative since we still do now know which power supply was used in MXstarter's 2215; the two designs are very different up to the inverter.  The early one has simple brute force current limiting.  The later one has switch current limiting but it is not cycle by cycle and may not protect the switch from extended periods of low input voltage operation.  I do know that it works to protect the switching regulator from output shorts with the power supply dropping into tick mode.

Also, be aware that faults in power supplies usually "cascade". It is doubtful that there is one single failed component - there will be a chain of destruction that will often result in new components instantly being destroyed at power-up (the lamp-limiter will help to an extent here).

Tektronix was pretty good about including crowbar circuits in their power supplies to prevent catastrophic failures from progressing to the low voltage side.  The later 2215 power supplies includes this but the earlier one rely on a shunt zener diode.  In either case the crowbar or shunt zener diode should be checked.

For clarity, I didn't mean this. Perhaps "cascade" wasn't the best word exactly; perhaps I should have provided more explanation.

I was referring to the PSU on its own. Often, if the switching transistor fails, it can take out many other components near to it, such as the control IC, the current sense resistor, perhaps many other things.

What happens here is that the service guy spots a shorted transistor, changes it without making other checks, and the new transistors fails short in the same way. So next time, one of the other faulty components is spotted, but now, two new components fail again at the first power-up attempt. Repeat until madness sets in

This has certainly happened over on the TekScopes@yahoogroups.com email list.  Switching transistor failure usually results in a gate to drain short which takes out the gate driver circuit and the switching regulator control IC.  Luckily these failures usually result in a situation where changing just the switching transistor results in no operation and the blown up parts in the driver circuit are very visible.  I do not remember if the sense resistor in the source circuit usually dies or not.  The output diode may fail but it is not clear if the diode failing causes the power switch to fail or the reverse.

One of the poor aspects of the design is that the switching transistor was remote mounted with a three pin socket.  A later change notice removed the unreliable socket and soldered directly to the transistor leads.  In both cases however, I wonder if excessive peak gate to source voltage is causing poor reliability.  If I was rebuilding one of these, I would measure the Vgs waveform very carefully and probably add a snubber and protection at the gate lead and dress the leads for minimum inductance.

Away from that minor clarification, I happily agree with everything you say. Yes, capacitors are the usual enemy. High-value resistors in the start-up circuit can be another common problem. Both these are easy to diagnose and fix, but blow-ups are a different matter, and generally require a much deeper level of understanding and diagnosis skills. Sadly, I know that at least one of my 2215s has had a blow-up, so I'm not looking forward to tackling that - and I say that as someone with many years of experience with these things 

Sometimes the 22xx boards get so damaged that they are not worth repairing but I think that has a lot to do with a surplus of main boards which can be used as part mules for other 22xx oscilloscopes.  Someone I know who repairs these for money send me a stack of main boards to scavenge from.

There also seems to be one or more mysterious failure modes associated with the switching preregulator but when they have occurred over at TekScopes@yahoogroups.com, it has not been real clear if there was something really wrong because the people doing the diagnosis have always been inexperienced.  The experienced people fix these things without problems without asking for help on the list so we do not hear their stories.

Someday I hope to get my hands on one of the bad 22xx power supplies and figure out what is really going on.  All four of my 22xx oscilloscopes work great so I am loath to take them apart although at some point I need to refurbish them.

My incomplete list of possible causes of the mysterious failures which only applies to the later 2215 power supply includes:

• Improper derating of the power switch and/or diode.
• High peak Vgs.
• Saturation of the inverter transformer do to imbalance.
• Something weird happening to the buck inductor.  Maybe it overheats changing its characteristics

[Mark Hennessy]

Hi David,

Thanks for those additional thoughts. Thinking about all of this prompted me to dig out the manual (I have an original paper version that covers both versions), and I've been reminding myself of the differences. I've also popped up into the attic to take a quick look at both of mine, and they are both of the later variety.

I acquired the first one about 5 or 6 years back. It's in good overall condition, but unfortunately a previous engineer had stripped apart the PSU some years previously, and a lot of the mechanical parts are missing. At this point, I was ready to give up, but I put it in the attic and forgot about it. And about 2 years later, someone kindly gave me a second 2215, which is much tattier, but hopefully contains all the bits I need. I know nothing about its working state - I didn't have time to examine it (not even take the outer cover off) before I hid it away with the first in the attic. But I'm hopeful it'll yield the missing mechanical parts at the very least - and for all I know, it might even contain a working PSU. Although I doubt I'd be that lucky

This conversation has pricked my conscience, and I've moved this job further up my to-do list. If I learn anything interesting, I will happily report back. Perhaps it'll become an article on my website eventually. And I'll check out the Yahoo group you mention - I don't think I'm member of that one.

Back to the OP. Yes, I agree that we are in the dark. The manual doesn't appear to give the serial number when the change happened, but does state that later units should have an "OPT. 48" sticker on the rear - though none of my do! If the OP could provide the serial number of their unit, I could at least compare it to mine to see if it's similar or higher.

All the best,

Mark

[MXstarter]

WOW,
Sooo many replies.
Pictures coming tonight !!
It's the exam period here in University of Melbourne, I probably won't have enough time to take the shell off, will certainly post the series number.
I certainly love electronics, don't know why I chose mechatronics in the beginning and now I am stuck with it  . Dynamics of machine's exam coming in a few days, god I hate kinematics  .

It's a pity that this scope died, It looks soo lean and new when I got it.

Thanks again for your awesome replies.
 

[David Hess]

Back to the OP. Yes, I agree that we are in the dark. The manual doesn't appear to give the serial number when the change happened, but does state that later units should have an "OPT. 48" sticker on the rear - though none of my do! If the OP could provide the serial number of their unit, I could at least compare it to mine to see if it's similar or higher.

The scan of the manual that I have includes the change notice for the power supply and a bunch of rules for figuring out which units had it:

You instrument contains in the power supply either the Current Limit board (A19) or the Preregulator board (A18).  Instruments with a serial number B022000 or above contain the Preregulator board.  Some instruments below that serial number were built containing the Preregulator board.  To determine if your is one of these, look to see if there is an Option 48 sticker attached to the rear of the instrument.  If there is not an Option 48 sticker attached and the serial number is below B022000, your instrument contains the Current Limit board.

All instruments manufactured in Europe contain the Preregulator board (A18).  These instruments have serial numbers 200000 and up.

[Mark Hennessy]

You have a different version of the manual to me. Mine simply says:

Your instrument contains in the power supply either the Current Limit board (A19) or the Preregulator board (A18). To determine if your is one of these, look at the back of the instrument. If there is an "OPT. 48" sticker on the rear panel, the instrument contains the Preregulator board; otherwise it contains the Current Limit board.

No mention of serial numbers or Europe. How interesting

[David Hess]

There were other change notices for the 2213 and 2215 and I think one included adding a winding to the transformer to lower the heater voltage on the CRT to extend its life because as designed it was too high.  I am not the expert on the 2213 and 2215 change notices though.

[Mark Hennessy]

Totally irrelevant, but I had to LOL when I saw that David's post count has reached exactly 2215 

[SeanB]

[miguelvp]

You should have taken the screenshot when the count was actually 2215, you were off by almost 12 hours

[Mark Hennessy]

I did consider posting a screenshot, but decided that I'd probably consumed enough of Dave's bandwidth already 

[David Hess]

I did not see the post in time to grab a screenshot but I like the one taken better because the 2232 was the greatest of the 22xx series oscilloscopes.  It is the middle one in this photo: