TDS 380 Attenuator Resistor Replacement Specs


Richard Peterson
 

Hi Folks. New member, first post . . . I recently picked up a couple of new to me Tektronix scopes to replace my Hitachi V212. The 2236 is up an running beautifully after some cleaning, but the TDS380 has a (hopefully) small issue: No input on CH1. I've traced the problem down to what looks like a burned 50 ohm thick film resistor on the input of the attenuator board between the BNC and one of the capacitors on the board. (not quite sure how to post photos here yet but hopefully this link works: https://groups.io/g/TekScopes/photo/263498/3215960?p=Created,,,20,2,0,0)

I'm thinking I can simply solder a 50 ohm resistor between the BNC connector and the SMD capacitor that the original resistor links (the larger capacitor at the end of the first relay in the photo) . Any thoughts about doing this and what power rating the resistor should be?

Thanks.

Rich


Siggi
 

Hey Rich,

Matt D'Asaro fixed at least one attenuator in a TDS-420 that was similarly
bad. The article is here:
https://www.antiqueradios.com/forums/viewtopic.php?t=216792. The images
don't load for me, but you can still view them at these links:
http://dasarodesigns.com/~medasaro/TDS420/broken.jpg
http://dasarodesigns.com/~medasaro/TDS420/fixed.jpg

Good luck,
Siggi

On Fri, Apr 30, 2021 at 8:55 AM Richard Peterson <saaber1@comcast.net>
wrote:

Hi Folks. New member, first post . . . I recently picked up a couple of
new to me Tektronix scopes to replace my Hitachi V212. The 2236 is up an
running beautifully after some cleaning, but the TDS380 has a (hopefully)
small issue: No input on CH1. I've traced the problem down to what looks
like a burned 50 ohm thick film resistor on the input of the attenuator
board between the BNC and one of the capacitors on the board. (not quite
sure how to post photos here yet but hopefully this link works:
https://groups.io/g/TekScopes/photo/263498/3215960?p=Created,,,20,2,0,0)

I'm thinking I can simply solder a 50 ohm resistor between the BNC
connector and the SMD capacitor that the original resistor links (the
larger capacitor at the end of the first relay in the photo) . Any
thoughts about doing this and what power rating the resistor should be?

Thanks.

Rich







Richard Peterson
 

Thanks Siggi. I found that article and it inspired me to dig into my repair. Since Mark used an 0603 SMD resistor I assume the power requirement is pretty low but i thought i'd ask others. I can probably squeeze a 1206 in there, but it would be much easier scrape away the remains of the old resistor and solder in a 1/4 watt resistor (or two in series). I was mostly concerned about the power rating of the original part, but given the size, I guess it's pretty low. I'll try the metal film resistors, matched to the value of the resistor in CH2 and see what happens. Hopefully nothing on the underside of the board is also blown. I'll post my results . . .


Siggi
 

Good luck. I seem to remember someone sketching the schematic of these
attenuators, where this resistor was in-line with the signal. Presumably
it's for dampening and/or a fusible cutout, and so the power rating
shouldn't be a problem. I don't think the 380 or the 420 have a 50Ohm
coupling option?

On Fri, Apr 30, 2021 at 3:29 PM Richard Peterson <saaber1@comcast.net>
wrote:

Thanks Siggi. I found that article and it inspired me to dig into my
repair. Since Mark used an 0603 SMD resistor I assume the power
requirement is pretty low but i thought i'd ask others. I can probably
squeeze a 1206 in there, but it would be much easier scrape away the
remains of the old resistor and solder in a 1/4 watt resistor (or two in
series). I was mostly concerned about the power rating of the original
part, but given the size, I guess it's pretty low. I'll try the metal film
resistors, matched to the value of the resistor in CH2 and see what
happens. Hopefully nothing on the underside of the board is also blown.
I'll post my results . . .






Richard Peterson
 

Quick update - I installed a pair of 1/4 watt resistors to achieve 49.5 ohm to match channel 2 (https://groups.io/g/TekScopes/photo/263498/3217080?p=Created,,,20,2,0,0). I now have a good signal through the first relay. Still no CH1 display, so I'll have to do some signal tracing through the attenuator board and beyond to see where the signal dies.

A note to anyone working on the TDS3xx series. The ceramic attenuator board is not easy to solder. My first attempts to attach the resistors to the trace betwee the two SMD capacitors in the photo was an epic fail using a T18-BR02 fine point bent tip . . . I ended up using my hot air station which worked better. Hopefully the next issue I find is beyond the attenuator board so work will be easier.


Richard Peterson
 

A little progress. I now have a signal through the CH1 relays on the attenuator board and the voltage readings on all the pins match between CH1 and CH2 except:
- pin 15 and 16 (GND/OS and OFFSET). CH1 shows -2VDC and CH2 shows 0VDC.
- pin 28 (NEG_OUT). CH1 shows 0.25 VDC and CH2 shows 0.03 VDC.

Still no signal display for CH1 except a negative offset flat line which makes sense given the -2VDC reading on the offset pin of the attenuator. DC offset on the display for CH1 shows 0V . . . .

I'll spend some time checking components in the offset circuit next, but I've read more than a few posts where problems with the TDS3xx scopes are frequently tied back to the SRAM IC - AS7C164–12JCTR.

Your feedback would be appreciated.


Ozan
 

Hi Richard,
I don't have a TDS380 but I have debug experience on TDS520, TDS520A, and TDS7104. Looking at schematics briefly TDS380 looks similar.

- pin 15 and 16 (GND/OS and OFFSET). CH1 shows -2VDC and CH2 shows 0VDC.
- pin 28 (NEG_OUT). CH1 shows 0.25 VDC and CH2 shows 0.03 VDC.
Do you see a voltage of opposite polarity (i.e. -0.25V) at POS_OUT of Ch1?


Still no signal display for CH1 except a negative offset flat line which makes
sense given the -2VDC reading on the offset pin of the attenuator. DC offset
on the display for CH1 shows 0V . . . .
If you add offset from front panel (I don't know if TDS380 has that menu) or move vertical control does the -2VDC offset change?

The offset is sum of two parts: one is the offset set by front panel input, the other part is a calibration constant (actually two, DC balance and hybrid offset). On a TDS520/520A/TD7104 SPC (signal path compensation) menu self calibrates these constants. You may want to run SPC (procedure is in the service manual) to check if Ch1 recovers.

If Ch1 is still bad after SPC you can try swapping the hybrids between Ch1 and Ch2 to see if problem moves with hybrid or stays at Ch1. Since hybrid and Daculator U202 is in a calibration loop it is difficult to tell if the problem is with the hybrid of Daculator path. Swapping channels, especially when one is known good, is the quickest path to find the faulty block.

I'll spend some time checking components in the offset circuit next, but I've
read more than a few posts where problems with the TDS3xx scopes are
frequently tied back to the SRAM IC - AS7C164–12JCTR.
It is unlikely SRAM is the root cause of the problem you are seeing.

Ozan


Richard Peterson
 

Thanks for the feedback Ozan. Here are the errors I get:

https://groups.io/g/TekScopes/photo/263498/3219273?p=Created,,,20,2,0,0
https://groups.io/g/TekScopes/photo/263498/3219271?p=Created,,,20,2,0,0
https://groups.io/g/TekScopes/photo/263498/3219272?p=Created,,,20,2,0,0


- It does appear that the memory is not the issue. Swapping hybrids would be a very difficult undertaking from what I've read in another post by Dave Wilson . . . he tried to de-solder one to work on it using better equipment than i have and finally gave up.

I'll take some measurement to see if the -2VDC changes with offset adjustments. There is not a corresponding opposite DC voltage on the POS_OUT pin.
I'll also try the SPC and will spend some more time signal tracing both channels to look for differences that could help narrow the search.

Rich


Ozan
 

Thanks for the feedback Ozan. Here are the errors I get:

https://groups.io/g/TekScopes/photo/263498/3219273?p=Created,,,20,2,0,0
https://groups.io/g/TekScopes/photo/263498/3219271?p=Created,,,20,2,0,0
https://groups.io/g/TekScopes/photo/263498/3219272?p=Created,,,20,2,0,0

- It does appear that the memory is not the issue. Swapping hybrids would be
a very difficult undertaking from what I've read in another post by Dave
Wilson . . . he tried to de-solder one to work on it using better equipment
than i have and finally gave up.
I see from the pictures the hybrids are soldered. I agree, it is very difficult to solder/desolder those substrates. On TDS7104 they are connected by compression. Still replacing relays was very tricky.


I'll take some measurement to see if the -2VDC changes with offset
adjustments. There is not a corresponding opposite DC voltage on the POS_OUT
pin.
I'll also try the SPC and will spend some more time signal tracing both
channels to look for differences that could help narrow the search.
I assume you are using the technical reference guide (070-9436-04). On page 238 it shows hybrid outputs should be 0V common mode and 6.25mV/div. You should measure similar amplitude but opposite polarity at POS and NEG outputs. However, 2V offset may be so large that the balance is not there.

Figuring out if hybrid is bad or if the calibration loop is deciding the wrong DAC voltage for offset will help debug. You can try grounding pins 15&16 to see if the outputs (POS_OUT, NEG_OUT) move towards zero, at least differentially. R238 is large (1k) so you can short those pins without disconnecting R238, or you could remove R238 to be sure.

Is my understanding, Ch2 is good and only problem is with Ch1, correct?

Ozan


Richard Peterson
 

I have the technical reference (P256 for the TDS380) . . . I'll try grounding pins 15 & 16 to see what that does.

You are correct - CH2 is good, CH1 has the issue.


Richard Peterson
 

Grounding CH1 pins 15 & 16 had no effect on the DC outputs of the attenuator. it remained at 0.25 VDC and didn't shift the output waveform as a result. I looked at the waveform of the attenuator output pins with the DC offset filtered out (AC coupling). They are linked here:

https://groups.io/g/TekScopes/photo/263498/3219481?p=Created,,,20,2,0,0

Clearly there is a difference, besides the DC offset, between the good channel and the bad channel but I'm not sure what to make of it. I guess the next step will be to figure out why they are different and whether that difference is caused by something inside the attenuator or outside the attenuator . . .


Ozan
 

On Mon, May 3, 2021 at 06:00 PM, Richard Peterson wrote:


Grounding CH1 pins 15 & 16 had no effect on the DC outputs of the attenuator.
it remained at 0.25 VDC and didn't shift the output waveform as a result. I
looked at the waveform of the attenuator output pins with the DC offset
filtered out (AC coupling). They are linked here:

https://groups.io/g/TekScopes/photo/263498/3219481?p=Created,,,20,2,0,0

Clearly there is a difference, besides the DC offset, between the good channel
and the bad channel but I'm not sure what to make of it. I guess the next
step will be to figure out why they are different and whether that difference
is caused by something inside the attenuator or outside the attenuator . . .
TDS380 attenuator is on page 256 of SM but it doesn’t show the expected attenuator output voltage. Page 238 for 340A/360 has a note "0V common mode, 6.25mV/div/side" to show the expected output. It is likely TDS380 output is similar, in that case output of 0.25V is very large.

To get a rough idea about what is inside the hybrid you can review TDS530B hybrid on page 233 of:
https://download.tek.com/manual/070971003.pdf
TDS380 is missing the 50-ohm path, input attenuator and calibration paths are simpler. Looks like AC relay, one calibration relay, and one attenuator relay is implemented. Shift register is also inside the hybrid.

With zero offset and no-input the output should be ~ 0V differentially. I hope whatever caused the damage to the resistor didn’t damage the amplifier.

The signal you are seeing could be because of ground loops. If you haven't seen the Jim Williams app note before:
https://www.analog.com/media/en/technical-documentation/application-notes/an47fa.pdf
starting on page 18, especially Fig 37 for different ground connections. There is even a note for probe J.

Ozan


Richard Peterson
 

Ozan (and Siggi),

Thanks for your guidance. The TDS530D reference was exactly what I was looking for to figure this out. Unfortunately, what I discovered is that one of the OpAmps in the hybrid IC is shorted from its -Vsupply to its negative input allowing current/voltage to backfeed out of what should be the offset input from U202. Here's a sketch/photo of the issue to help anyone else who might run across something similar in the future: https://groups.io/g/TekScopes/photo/263498/3219781?p=Created,,,20,2,0,0

The IC is almost certainly unobtainium and would be nearly impossible to solder anyway, so I've buttoned the scope back up and will just enjoy my single channel TDS380 until i find a parts doner with a replacement attenuator. Still $20 well spent since I only use one channel for most things and have the 2236 if I need additional channels . . . On to the TM503 with installed DC501, DM501a & FG503. I'm getting hooked on Tektronix gear.


Richard Peterson
 

I wonder what would happen if i cut the input trace to the shorted OpAmp (the one carrying the -6.8 V out to U202) . . . . hmmmm. May have to open it back up. Nothing to lose, I suppose.


Ozan
 

Hi Rich,
Looks like magic smoke escaped out of that amplifier. Most likely input pair is damaged and there is a short between input and bias network (likely a tail current source).

Ozan


Ozan (and Siggi),

Thanks for your guidance. The TDS530D reference was exactly what I was
looking for to figure this out. Unfortunately, what I discovered is that one
of the OpAmps in the hybrid IC is shorted from its -Vsupply to its negative
input allowing current/voltage to backfeed out of what should be the offset
input from U202. Here's a sketch/photo of the issue to help anyone else who
might run across something similar in the future:
https://groups.io/g/TekScopes/photo/263498/3219781?p=Created,,,20,2,0,0

The IC is almost certainly unobtainium and would be nearly impossible to
solder anyway, so I've buttoned the scope back up and will just enjoy my
single channel TDS380 until i find a parts doner with a replacement
attenuator. Still $20 well spent since I only use one channel for most
things and have the 2236 if I need additional channels . . . On to the TM503
with installed DC501, DM501a & FG503. I'm getting hooked on Tektronix gear.


Siggi
 

On Tue, May 4, 2021 at 12:46 PM Richard Peterson <saaber1@comcast.net>
wrote:

Unfortunately, what I discovered is that one of the OpAmps in the hybrid
IC is shorted from its -Vsupply to its negative input allowing
current/voltage to backfeed out of what should be the offset input from
U202. Here's a sketch/photo of the issue to help anyone else who might run
across something similar in the future:
https://groups.io/g/TekScopes/photo/263498/3219781?p=Created,,,20,2,0,0

That's unfortunate.
Excellent sleuthing though!


Richard Peterson
 

Adding my pinout diagram for the H2565A hybrid attenuator for anyone who might be able to use it.

https://groups.io/g/TekScopes/photo/263498/3222929?p=Created,,,20,2,0,0

Found a parts TDS380 . . . hopefully the attenuators are intact.


Ozan
 

Adding my pinout diagram for the H2565A hybrid attenuator for anyone who might
be able to use it.

https://groups.io/g/TekScopes/photo/263498/3222929?p=Created,,,20,2,0,0
Hi Rich,
This is useful. Your drawing shows pin 42 (GND/CAL) going to relay coils. Pin 46 (SIG_EN) should go to the coil and pin42 should be going to one of the relay contacts. K2 relay enables calibration path (input disconnected and GND/CAL signal is fed to input) if GND/CAL goes to K2 coil.

Ozan


Richard Peterson
 

Good catch Ozan. I've uploaded a revised photo.

https://groups.io/g/TekScopes/photo/263498/3223503?p=Created,,,20,2,0,0


Richard Peterson