Tektronix 2445A

[torquil]

Hi!

I just obtained a Tektronix 2445A. Apart from the display being erratic, it has the following problem: I only see the traces from Ch2/Ch3/Ch4 on the display.

Well, immediately after startup, the Ch1 trace can be seen briefly for a second or two as it moves downwards and disappears towards the south on the screen.

I will check PSU voltages and ripple when I get the chance, but perhaps someone here has an idea about the dissapearing trace problem?

Another question, for how long can I have it running while checking such things as PSU voltages before something overheats?

Best regards,
Torquil Sørensen

[Gixy]

Hi!
Hum, I think that first of all you need to recap the PSU board. There are lots of threads here explaining how to, even with the BOM needed. The best method is to replace each cap one at a time, as there is an error in the service manual with two references. Once the case is open and the scope running, you should provide some airflow to cool the main board. You should also check and replace the battery (see also the threads upon this, as you must maintain voltage with another battery during the operation).

[Electro Detective]

May I suggest BEFORE you get medieval with the soldering iron,
run some signal/s first and check and work ALL the switches and knobs, toggles etc

and see if that improve things, or changes things, or makes things worse.. and go from there.

I've rescued a few Teks from dry and unused switch banks,
and saved money on capacitors, solder, electricity,
and head banging  after replacing caps, and no difference 

That's my 50mA,

good luck! 


   

[torquil]

Thanks guys, I will jiggle and turn the knobs a bit before running amok with the soldering iron. At least the Ch1 trace exists. I'm hoping the hybrids are intact... I am planning on doing any type of necessary preventative maintainance on it when I get the time. So PSU rails will be analysed as my first order of business when time permits.

One thing I do know is that the vertical position knob seems to be working as it should. I acticated the "Exercise 01" diagnostic to see the encoder/knob values in real time, and the Ch1 position knob value that is displayed  seems nice and stable. So the trace is probably not dissapearing south because the position knob (encoder?) is acting up.

[MarkL]

Your first step to check the power supply voltages and ripple is the right approach.  If they are in spec I would strongly advise not doing a recap lest you compound your existing problem(s) with new ones.  I would not get out the soldering iron until there is a reason to suspect a component is bad.

The theory section of the service manual has a lot of good information on how the position controls work.  Most of the knobs are actually pots that are read by the processor.

It's good that you checked the EXER 01.  Ch2 operates exactly the same as Ch1, so you can use Ch2 as a comparison in your troubleshooting throughout.  Do you get the same range from both Ch1 and Ch2 as shown by the values in EXER 01?

Assuming you do, the processor sends these values to a DAC with sample/hold to set an analog voltage for each trace offset position (among many other analog outputs).

One good place to start is to compare the position offset voltages from the S/H for Ch1 and Ch2.  Take a look at U2430 pin 7 and pin 14 pin 1, which are the Ch1 and Ch2 position outputs, while moving the position knobs.

Also, you do not need to worry about external cooling with the cover off.  That's true on some Tek scopes, but not this one.

EDIT: Fixed wrong pin on U2430; should be pin 1 for Ch2 pos.

[torquil]

Hi!

Late status update to this thread: The PSU voltages were all perfect. Everything works fine except Ch1 trace is not visible. I used it a couple of minutes, and then two RIFAs blew along with a 68 Ohm resistor. So I replaced those and cleaned up in the PSU, and now the scope is back in business, but still missing Ch1.

However, I did find out something about the missing Ch1 trace. As you probably know, the 2445A displays an amplitude value in the top left corner of the display. When using Ch1, the value was large and negative. On Ch2, the value doesn't change when the input attenuator is changed. This is as expected. However, on Ch1, the amplitude figure went larger and larger as the V/DIV value was increased. So I started to suspect the input attenuator.

It turns out that when using the 1MOhm DC input setting, the resistance to ground is larger than 10 MOhm (i.e. O.L. on my DMM). So something is definitely wrong there. On DC 50 Ohm input, the resistance to ground is approx 50 Ohm. Note that there is no trace displayed for Ch1, even at the DC 50 Ohm input setting.

But when I opened up the scope and put in a jumper from the input connection on U200 (which is on the working Ch2) to the input connetion on U100 on Ch1, and specified the same V/DIV settings for both channels, I did get a a Ch1 trace which was a copy of Ch2, when using the built-in calibrator signal. I could control the vertical position of the two traces independently using the vertical position knobs of Ch1 and Ch2.

So this means that there is no problem with the famous hybrids, but with the input. I guess either with the relays, the chip that controls the relays, or perhaps the pins that connect the input attenuatur for Ch1 to the main board?

[mawyatt]

One of my 2465 had all sorts of problems with channel 1, this was traced to the input hybrid. I carefully disassembled it and cleaned the open relay contact areas on the ceramic substrate. Used a soft pencil eraser and then cleaned the residue. This solved the problem. Ended up doing the same to channel 2 also.

I suspect that the surface of the relays had some oxidation and possible residue (maybe tobacco smoke).

The hybrids are repairable if you are very careful, and good with this kind of repair.

Good luck with your scope.

Best,

Edit: Whoops it's a 2465 (changed), not 2445

[Tom45]

Those input relays on my 2445 tend to be flaky too. Lots of switching of the input mode (AC, Gnd, DC, 50 ohms) usually improves things for awhile. My 2445 is on the shelf now and I make do with a 2465 and 7904 instead.

[MarkL]

That's very strange the V/div was large and negative.  A bad input attenuator would not explain that.

The processor is in control of reading the V/div knob, and then setting the attenuator relays appropriately and also displaying the V/div in the upper left.  The fact that it's displaying a non-sensical value for V/div says to me that it's confused and is probably not setting the input attenuator relays correctly.

I would take another look at the Ch1 V/div knob encoding via EXER 01 to make sure it's sane.  Watch the sequencing for clockwise and counter-clockwise and see if they match Ch2.

EDIT:  Hold on there, the V/div is in the *bottom* left corner, not the top.  My mistake.  The value in the top left is the trigger level.  Preceding that is the trigger source.  "A" is the A trigger, "B" is B trigger, followed by the trigger source channel, 1 through 4.

It would make sense the trigger level display changes with the Ch1 V/div setting if the trigger source was, for example, A1.  I assume that's what you're seeing.

If the trigger is set to "AUTO LVL", the "large negative number" may be because the output of the attenuator is pinned to a negative voltage.  You could confirm by looking at the attenuator output (test point 6 on the schematic).

Sorry about the earlier mistake.

[torquil]

EDIT:  Hold on there, the V/div is in the *bottom* left corner, not the top.  My mistake.  The value in the top left is the trigger level.  Preceding that is the trigger source.  "A" is the A trigger, "B" is B trigger, followed by the trigger source channel, 1 through 4.

It would make sense the trigger level display changes with the Ch1 V/div setting if the trigger source was, for example, A1.  I assume that's what you're seeing.

If the trigger is set to "AUTO LVL", the "large negative number" may be because the output of the attenuator is pinned to a negative voltage.  You could confirm by looking at the attenuator output (test point 6 on the schematic).

Sorry about the earlier mistake.

I'm new to the 2445A, so I didn't know the "A1"-value was the trigger level. It is possible that I was using the AUTO LVL which you mentioned. I just assumed (without checking the manual) that it had something to do with amplitude of the input signal. I will check the attenuator output as you are suggesting. Thanks!

[torquil]

Well, it turns out that there was a diagonal hairline crack across the brittle subtrate inside the channel one input attenuator. The input to the channel one input amplifier was not even referenced to ground through the 1MOhm resistor near the output of the input attenuator. Replacement part has been ordered, and in the mean time I have grounded the input to the channel one amplifier, and I'm getting a nice baseline trace.

[torquil]

After receiving a new input module, Tek part no no. 119-2342-01 from QService on Ebay, channel 1 is now up and running:

[AJ3G]

Had a similar problem with the 2440, and Qservice in Greece also saved the day..  2400 are really marvel scopes of their era. I still have my 2440 and use it often.