Tektronix 2440 U844C missing in Schematic?

[HansMoleman]

Hi, this is my first post in this forum 

I'm trying to repair a broken Tek 2440, there is no output on the screen.
Following the service manual for troubleshooting the dead start (pdf page 306), voltage rails seem fine and clocks are working.
As described in the System uP Kernel test (page 307), I changed the jumper positions and checked if the nRESET signal on U844 pin 8 (input of NAND Gate) is high, which it was not.
Now I want to understand the power up and reset circuitry to debug it.
The System processor schematic (page 431) indicates the location in the schematic at 5A unfortunately, I can not find this Gate nor any information about its input signals.
I would be happy if somebody could give me a hint about where to lock.

Here is the link to the Service Manual: http://w140.com/tekwiki/images/b/b0/070-6603-00_complete.pdf

Thanks in advance,

Mario

[pbarton]

You are correct. U844C can’t be found at location 5A on the System Processor schematic. It appears to me that the text on page 307 is a different revision to the schematic on page 431. Probably a revision error in the manual.
However, the schematic is probably still valid for 2440 Oscilloscope Serial Number B014064 & below.

[HansMoleman]

Just my luck... everything else in the manual seems to be correct

[rf+tech]

Mario,

I would skip the diagnostics procedure and go straight to U942, a TL7705 supply voltage supervisor (datasheet below). Its purpose is to hold the microprocessor /RESET input low, until the +5 Volt rail has risen above a threshold of +4.55 ± .05 Vdc, and further delayed by CT on pin 3.

If U942 /RESET pin 5 is stuck low, check for +5 Volts on SENSE INPUT pin 7.

/RESIN pin 2 is pulled up to +5 Volts by R940 and R941. Check that U254E pin 10 is not pulling /RESIN low – or that a shorting jumper has been placed on the “Short for test” pins.

VREF pin 1 should be 2.53 ± .05 Vdc.
CT pin 3 should be well above VREF pin 1 voltage.

RF+ Tech

[HansMoleman]

Hi rf+ tech, thanks for your help.
All mentioned signal levels seem to be ok, nRESET on U640 is high and the waveforms 1 - 3 (page 428) are good, I think I'll have to ignore U844 and proceed the troubleshoot. 

[rf+tech]

Mario,

As a courtesy to others, please post the actual voltages measured. The next individual to encounter a TL7705 reset problem can then benefit from hard data.

Experienced eyes can sometimes pick out valuable clues in such data, that to the lesser experienced seems to be okay.

My recommendation would be to remove U942, power up the scope and measure the voltage present on pin 5.

TL7705 (and similar) have been the source of stuck low reset lines on several occasions, in my experience.

RF+ Tech

[HansMoleman]

You are right, that's the least thing I can give back to the community, I will do this as soon as I have the time.

As the reset circuitry seems to work as intended I think desoldering U942 is not necessary. I only suspected the reset circuitry to be damaged because the manual suggested so if there is a low on the nRESET line.
To my understanding pin 8 of U844 is not directly related to the nRESET signal and the description in the manual might be wrong.
Thanks again for your help!

[rf+tech]

Mario,

Yes, I would agree about the manual discrepancy and what was thought to be a stuck low on the /RESET line is incorrect.

Please confirm uP U640 /RESET pin 37 is *not* being held low.

Has battery BT800 voltage been measured? Are there any signs of corrosion present on component leads near BT800?

RF+ Tech

[HansMoleman]

All reset signals, except U844 pin 8, are looking fine to me.
Here is the measured data:

U942
Pin 1 Vref: 2.54V
Pin 2 RESIN: 5.07V
Pin 3 CT: 3.14V
Pin 7 Sense: 5.07V

U254 Pin 10: 5.07V

Image1:
ch1 U640 Pin 2 (Waveform 1 in Manual)
ch2 U640 Pin 37 (Waveform 2 in Manual)

Image2:
ch1 U640 Pin 2 (Waveform 1 in Manual)
ch2 U942 Pin 2 (Waveform 3 in Manual)

Image3:
ch1 U942 Pin 2 (Waveform 3 in Manual)
ch2 U844 Pin 8

The Battery Voltage is 3.70V and there is no sign of leakage.

Voltage levels for the NAND Gate are 1.65V at both inputs and 0.21V at the output (pin 8 ).

This is still strange, the manual clearly states to measure the nRESET signal of waveform 2  at U844 pin 8 (page 430).
Maybe someone with the same processor board (670-9746-17) can confirm that U844 Pin 8 is not related to the nRESET Signal.

[rf+tech]

Mario,

Good data points, thank you. The microprocessor is not being inhibited, based on your waveforms.

U844 appears on schematics 1 (two gates) and 20 (one gate). The one unused gate out of four gates, is output pin 8. Given this discrepancy between the troubleshooting procedure and the schematic, let’s put this issue aside.

Before delving any deeper into the processor control, there are some simple sanity tests that are easy to perform and can reveal what parts of this scope may be working. This can be especially helpful with a scope that is “blind.”

A quick check to assess the vertical deflection system can be made by connecting a battery powered DMM across the vertical deflection amplifier output. Note that this is a differential measurement, not common mode relative to the chassis.

R593 and R596 provide a convenient access point. The voltage difference observed should vary with the vertical position control, zero Volts when the control is nearly centered at 50% of rotation, and should swing close to equally positive and negative, in response to the vertical centering control.

Horizontal deflection can be verified in a similar manner, across R594 and R595. The differential voltage should respond with the Horizontal Centering control.

To facilitate these tests, set the Horizontal timebase for X-Y mode. Ch1 and Ch2 inputs grounded.

If these tests show the vertical and horizontal amplifiers are working, then CRT blanking should be checked. Something may be holding the beam in cut-off.

Once the easy sanity checks are confirmed, then return to the (more difficult) processor diagnostics.

RF+ Tech

[HansMoleman]

RF+ Tech, thank you for your input, you are very helpful.
I think before I do the measurements I'll have to do some reading to understand the systems I'm checking.

Btw, should I open another thread for the upcoming troubleshoot?

[rf+tech]

Mario,

It is best to keep a single thread, with all information in one place.

To aid understanding of the tests outlined, both vertical and horizontal amplifiers are differential or push-pull. This push-pull voltage difference is applied to the corresponding deflection plates, which push and pull the electron beam to form a trace on the screen, when everything is working normally.

When nothing can be seen on a scope CRT, a DMM connected as described can be used to determine if the corresponding deflection amplifier is working. Since a DMM cannot respond quickly to fast signals, one can substitute the position controls for an input signal. Manually moving the position controls through their range will produce a response on a DMM, exactly as the electron beam is seen to shift up-down and left-right. When the beam is centered vertically, the DMM will read very close to zero Volts. Same with the horizontal axis. This is the fastest method to qualify deflection amplifier operation. Another scope could also be used with two probes, but it's more cumbersome and fiddly than a DMM.

In the event of a failure in either deflection amplifier, the electron beam can end up being forced far off screen and nothing is seen. When this happens, a DMM will show a high differential voltage that never passes through zero Volts, when the position control is rotated.

At the end of each horizontal sweep, the beam is turned off, blanked, for the duration of time required to return the beam to the left side of the CRT. A failure in the blanking circuit can prevent a trace from appearing.

RF+ Tech

[HansMoleman]

RF+ Tech,

It took me a while to realize that I was looking in the wrong schematics.
The resistors you mentioned are also in the jitter correction ramp schematics which totally confused me

The output of the vertical and horizontal amp is not reacting to changes in the respective control knobs.
The differential measurements you suggested where took directly at power-up and after I heard a relais click.
The first values are relatively stable for each amplifier each time I powered up the scope, the second one varies very strongly.

Here is the data for the first measurement at the vertical amp (negative at R593, positive at R596):
at power-up: 0.42V
after relais click: 0.41V
the waveforms 137 and 138 are in the first image

The second one was:
at power-up: 0.41V
after relais click: -2.99V
waveforms are in the second image

The first measurement at the horizontal amp (negative at R595, positive at R594):
at power-up: -18.70V
after relais click: -20.66V
waveform 140 in the third image

The second one was:
at power-up: -18.71V
after relais click: -3.14V
waveform 140 in the fourth image

The third one was:
at power-up: -18.70V
after relais click: 11.15V
waveform 140 in the fifth image

The waveforms for the DACs do also not look healthy to me.
Image 8 shows the vertical DAC on channel 1 and the horizontal DAC on channel 2, the reference waveforms 123 (vertical) and 122 (horizontal) are in image 9.

At the end of the week, I might have time to dig deeper, for now, I would like to thank you again for your help and excellent explanations.

[rf+tech]

Mario,

Both Rigol scope waveforms show the timebase to be three orders of magnitude too fast. Waveforms 122, 123, 137, 138, 140 from the manual, are shown with 1 ms/div sweep.

Since the Vertical and Horizontal controls produce no effect, the signals to U902 on the front panel board are missing. U902 is a CD4051 8-channel analog multiplexer, under control of the Front Panel Processor U700. In turn, U700 is not receiving commands.

Referring to Self Diagnostics on page 6-28, these tests run when the scope is first powered on. As the tests are run, the Trigger LEDs flash to indicate tests are running. If a test fails, the Trigger LEDs will all be held on momentarily, indicating a failure has occurred. Then individual Trigger LEDs are turned on to produce a binary code that indicates the first failed test number. Page 6-29 explains how to decode the Trigger LED status.

The Trigger LEDs are driven by the System Data Bus and can reveal whether the kernel is operational. When time permits, please post the results.

Thanks for your words of appreciation, Mario. I’m glad to share my expertise with others, and appreciate a good challenge. My apologies for lack of clarity and the initial confusion of the whereabouts of those resistors.
 
RF+ Tech