Tekronix 2430A CH1 is broken

[lordvader88]

Thanks to PB80 I learned my DSO is broken, so skip most this post, or I'll trim it down tommorrow

CH2 works as expected, I put a signal on there, and if I turn VERTICAL POSITION knob up or down, it will move the signal to the bottom or to the middle or to the top as expected, and the cursor moves accordingly. And the amplitude is the same the whole time.

CH1 however, if the cursor is centered, the signal gets completely flattened to zero amplitude, so its just a perfectly straight zero line all the way across, perfectly centered.

As I move the cursor up, the top peaks of the signal start showing, then more of the signal arises from that center position zero line

Then most of the signal will have arisen from the zero line and only the negative peaks are still flat inside the centered zero line.

And eventually the normal shaped/full amplitude signal is fully shown and its a division or so above the center position, with barely any space left on screen above it.

The horizontal shape is always normal and acts as expected, I can slide the whole works side to side like normal.

If I keep going up the vertical shape starts distorting more and more as it moves above the screen.

And the opposite happens when moved down from a centered cursor.



So what setting is that Surely its just some setting I can't find, but I've been trying everything I can fiddle with. I hope thats not the only way CH1 is meant to display stuff.

[lordvader88]

I got it used off ebay 1 1/2 years ago and never used it much, and I'm only getting used to it now. For all I know CH1 is sort of broke, and was like that when I got it, I don' remember, but at least it can display a properly shaped wave form



Its exactly like CH1 is put though a bandpass filter for the voltage amplitude.

And its so exactly like it thats why I think its just a setting

[tautech]

I got it used off ebay 1 1/2 years ago and never used it much, and I'm only getting used to it now. For all I know CH1 is sort of broke, and was like that when I got it, I don' remember, but at least it can display a properly shaped wave form



Its exactly like CH1 is put though a bandpass filter for the voltage amplitude.

And its so exactly like it thats why I think its just a setting

I'm not so sure, ch 1 is usually the one that gets the most use and abuse. I think you'll find ch 1 is damaged.
Thinking a FET in the input stages but I'm not familiar with this model.

Service manual:
http://elektrotanya.com/tektronix_2430a.pdf/download.html

Wait for "processing" to change to "get manual".

Tek's repair gude is useful too:
http://www.sphere.bc.ca/test/tek-parts/troubleshooting-scopes.pdf

Sphere is good for Tek and HP parts too.

[lordvader88]

Thanks for the trouble shooting manual, I have the actual service manual book here

When it does its self-diagnostic, it has a few things, I never looked into them much yet, time to turn it on again and see if anything stands out as I learn more.

[tautech]

Thanks for the trouble shooting manual, I have the actual service manual book here

When it does its self-diagnostic, it has a few things, I never looked into them much yet, time to turn it on again and see if anything stands out as I learn more.

REALLY, REALLY study the manual, waveforms and look for DC voltages listed in the schematics. Note for DC values to be valid the controls must be set in a particular way, it's all documented and to narrow down an area I think I'd check the DC measurement points first.

[lordvader88]

yup ch1 problem, this is the list of fails in self test. I looked at stuff on the list once in a while before, but for the 1 year I barely used the machine at all. So I just figured it was calibration stuff really, like parts wearing out and going out of spec.

7211 CCD/gain/short-pipe/CH1-1
7212                                 /ch1-3

7221 ccd/gain/FISO-slow/ch1-1
7222                              /ch1-3

7231 ccd/gain/fiso-fast/ch1-1
7232                            /ch1-3

7420 ccd/PD-offset/ch1-3

8121 PA/offset/norm-FISO/ch1
8141 PA/offset/Env-FISO-slow/ch1

8210 PA/POS-gain/ch1

8411 PA/gain/50MV/ch1
8421              /20
8431              /10
8441              /5
8451            /2

8511 PA/Inv-gain/50MV/ch1
and the same for 20, 10, 5, 2

8610 PA/Var-max/ch1

9211 trigs/gain/a-trig/ch1
9221                 /b-trig


well I paid CAD $200 for it, so thats fine for 1 and part of a channel, and it heats my room and it -10C outside

[lordvader88]

Just checking some stuff with both ch1,2 hooked to the same sine wave, 1 or 2V peak to peak, 1kHz or something.

And both are in phase just fine, and can be set to over lap once I move the CH1 vertical position far enough away from center.

As far as the + * and the function settings area that displays stuff like volts, time, V@T, 1/T

when ch1 is positioned vertically at center, and the signal voltage amplitude disappears to zero (cause the machine is broke) , the machine really thinks it is zero.

So it adds zero to ch2, and multiplies by zero. And as I move it away from center, so that the signal starts reappearing slowly, like from a bandpass filter, it add and multiplies like its so. And eventually the full signal is back and with the proper scale, and it + * correctly again.

And if I move it to far it it thinks there's more CH1 voltage then is really there, and + * as if thats whats happening. And then it goes off scale.


Likewise the function options stuff, when it shows upper and lower dotted lines 12v apart, when I center Ch1 and it goes to zero, it will think CH2 it 6 volts above it..............I don't have video but you'd see it.


So the DSO at that stage internally really thinks there is no voltage on CH1 when its vertically centered.

[Bud]

I recall it should be possible and easy to measure the analog signal at the CCD input, so see if you get a proper signal with the offset at the CCD input lines. If my memory serves me, it is two pairs of differential lines for each channel. They can be measured with a single ended scope or a two-channel scope in a differential manner.
But since you said you are getting totally flat line at zero offset that may be a problem between the CCD and the ADC. Under normal conditions you should see some noise on the trace because the CCD produces some noise. Check for how much noise you see on the working Channel with zero signal/zero offset, compare with the non-working channel to get some clues.
 

[siggi]

I recall it should be possible and easy to measure the analog signal at the CCD input, so see if you get a proper signal with the offset at the CCD input lines. If my memory serves me, it is two pairs of differential lines for each channel. They can be measured with a single ended scope or a two-channel scope in a differential manner.
But since you said you are getting totally flat line at zero offset that may be a problem between the CCD and the ADC. Under normal conditions you should see some noise on the trace because the CCD produces some noise. Check for how much noise you see on the working Channel with zero signal/zero offset, compare with the non-working channel to get some clues.

Since the triggers are also failing self tests, I'd bet on the preamps.
Have you tried reseating the hybrids? They tend to develop bad contacts due to oxidation. It's an easy thing to try...

Also note that you MUST provide airflow over the heatsinks on the mainboard if you operate this scope with the case off four more than a moment.


Sent from my iPad using Tapatalk

[Bud]

I would not be concerned much about self-tests other than the CCD ones. If the response to self-tests is measured through the CCD, many may fail and be missleading before the CCD and ADC are fixed. I had bunch of them failed and the actual problem was with the CCD clock, once I fixed that all other self tests passed.

[lordvader88]

I recall it should be possible and easy to measure the analog signal at the CCD input, so see if you get a proper signal with the offset at the CCD input lines. If my memory serves me, it is two pairs of differential lines for each channel. They can be measured with a single ended scope or a two-channel scope in a differential manner.
But since you said you are getting totally flat line at zero offset that may be a problem between the CCD and the ADC. Under normal conditions you should see some noise on the trace because the CCD produces some noise. Check for how much noise you see on the working Channel with zero signal/zero offset, compare with the non-working channel to get some clues.

I'm not going to risk turning it on with the cover off, I've read about how quick some stuff overheats. (I do have fans tho)

but I just don't know near to go probing around with my DMM


I have the case off tho, but I'm guessing its not some resistor of cap that fried and would be noticable

[tautech]

Does this model Tek have a fan ?

While it's not the same model there's some good advice in the first page of this thread:
https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/

Where to start ?
Well even though it's not directly related to your fault check the PSU first ALWAYS.
SM will list the voltages, test points and ripple spec that you MUST ensure the PSU complies with.
With just a DMM you can check the voltages and the ripple for which use an AC mV setting.

Just checking these will gain you some familiarity with the insides of the scope, where things are and some confidence to go further.
There is every possibility that you'll be able to scope the busted channel with the good Ch 2 by feeding the probe Cal output into Ch 1. Check the SM for expected waveforms at the various test points.

Others will have diagnosed busted inputs too and hopefully they'll chime in with some more guidance.

[lordvader88]

if it is that CCD IC 165-2235-00 U100 U200 they actually are on ebay for sale for $22


so what might I actually test ?

[lordvader88]

Yeah the fan I just mean, as u probably know, how much stuff heats up without the case on and proper air flow, but I have lots of CPU and case fans around

Everything on the bottom board looks good, the s\nuts on the CCD heatsinks look clean and new


I'm removing the 2 top boards, and will try probing the low voltage PS board

well probably in 3-4 hours when I get back from work, but at least I'll have daylight to help then too

[lordvader88]

The lowV PSB is buried deep inside, I pretty much have to take the whole thing apart. Will it power for testing at that point ? This is only possible with that army manual I found, but I certainly never read all the maintenance section

This guy had a smaller older Tektronix but no manual, he repaired PSU

[tautech]

Do the study and take your time. We can't hold your hand while you fix it as it's on your bench not ours.
Focus on one step at a time but do be assured we're keeping an eye on how things go so just scream if you need us.

Take your time, you can learn so much and have the pride in being able to use something YOU fixed.

[lordvader88]

Road block. I don't have a 1/16" allen key to take off 3 front panel knobs, and 1.5mm is too small.

So for now I can't take off the main board to access the bottom of the lowV PS board, or the bottom of the main board.

The CCDs are on the main board, what might I check on the topside of that board ? I have the schematic,

I'm going to check a local store for the allen key, but its a real long shot in this small city, even though its the capital


I see pins marked TP so test point I'm sure. I have to find in the book where it lists all that,

[Bud]

I recall it should be possible and easy to measure the analog signal at the CCD input, so see if you get a proper signal with the offset at the CCD input lines. If my memory serves me, it is two pairs of differential lines for each channel. They can be measured with a single ended scope or a two-channel scope in a differential manner.
But since you said you are getting totally flat line at zero offset that may be a problem between the CCD and the ADC. Under normal conditions you should see some noise on the trace because the CCD produces some noise. Check for how much noise you see on the working Channel with zero signal/zero offset, compare with the non-working channel to get some clues.

I'm not going to risk turning it on with the cover off, I've read about how quick some stuff overheats.

Good luck then. What do you think, how do i know what i wrote above? A 5v pc fan blowing across the CCD board will do the job.

[lordvader88]

You're right I'm going for it, yeah screw I'll have to wait till Tuesday for hex key, so I'll put it back together and start probing the main board test points.

I only have this scope, a DDM and an AMM, thats what I have to work with.

Exactly how safe is it to probe itself with CH2 The manual lists waveforms at test points.

But another problem is they presume I have and use external inputs signals calibration tests and stuff, I don't have that.

The manual will say stuff like this

Failure of tests in 7100 or 7200
The CCD output stage is a probable area for failure if a SELF CAL fails any of the 7100 or 7200 tests. Check these tests to see which channel did not pass, then perform the following steps.
CCD Output Troubleshooting Procedure (fig. FO-22):
NOTE
Channel 1 components are reference (Channel 2 components are in parentheses).
1. Input the OS-291/G calibrator signal to the channel that is not operating properly. if neither is working, start with CH 1. CH 1 components will be referenced, with the CH 2 circuit numbers given in parentheses. Set the bad channel to 100 mV/div, DC coupled, with 50 termination off. Adjust the screen waveform so the ground dot on the scope under test is 2 divisions below center screen if possible. Set the input coupling of the other channel to ground. Turn the A SEC/DIV to 5 ?s.
6

So IDK maybe most the waveforms pictures are of very specific stuff like that

But anyways this is fun

[lordvader88]

Well I'm back from the hospital......joking (adults watch "Trailer Park Boys - Conky takes Bubbles to the Hospital" for some good Canadian laughs) , but I'll write what I got before the wind/rain storm knocks out the power. I'm using a cheap Vici VC99, its 6000count, measures fairly accurately in reviews for a cheap DMM, it just lacks protection, the point is its pretty accurate for VDC anyways.

schematic pics and stuff at the bottom

I've powered it up with just the HV and maybe the front panel connected to the lowV PS. And I checked the marked voltage points most are fine enough, except a +5V pre-regulator feedback point. It measured 7.16-7.19Vdc over the course of 20 minutes. The VAC was 4-6mV.

I wrote down all the voltages, most were within a few mV to a few within 0.2-0.3V, the 2 marked specifically +10V and -5V REF were 10.03, -5.018

In the attached picture I wrote down all the voltages that are SUPPOSED to be there, not what I measured.

Attaching my COM to the rear backplate shielding lowered all voltages by 0.1-0.2V, and raised the VAC to 4-6mV compared to the sheilding/case that runs along by the edge of the board the length of the CRT. On that pretty much all VAC was 2mV

In the end I just went with the backplate to move the DMM 5-6" away from it all.


So that +5VD that measures 7.19V comes from a transformer winding, through some CR354, a 3 pin thing with Schottky diode symbols (u guys will know, a current rectifier ? much to learn yet), then an LC pi-filter, a fuse, then the 39.9k R, (which measures right but I'm charging a cap too and it went to +40k before I stopped) and into the INVerting pin6 of an LM358N dual opamp.

Here's the manual
http://kripton2035.free.fr/Resources/TEK2430A_serv.pdf

Page 352 is +5V pre-reg feedback and more
347 is PCB layout of lowV PS
349 shows the transformer and where that winding goes through the CR and LC pi-filter

That column of transistors with voltages, I measured all them, and the # on the PCB was always written across from the middle pin, but was not always meant for the middle pin (or somethings really wrong). SOme of them varied a bit.

Q479 at the top marked 8V, the upper pin was 5.76, middle 11.87, lower 8.17. But later I had 8.52 on the 5.76upper pin, (I'll double checked but I better check again that it wasn't the lower pin that changed) Most of those were 0.1-0.3V of what was marked


Wow is adobe reader DC ever annoying, can't bookmark pages, and screenshots are distorted, so I used win8 snipping tool

[lordvader88]

So thats the +5VD pre-reg feedback, and where that opamp goes to IDK yet, but the +5V rail itself was fine, it was like 5.01 with 5mVac.

Oh so +5VD is V DIGITAL, so where's it go next

[tautech]

So thats the +5VD pre-reg feedback, and where that opamp goes to IDK yet, but the +5V rail itself was fine, it was like 5.01 with 5mVac.

Oh so +5VD is V DIGITAL, so where's it go next

Good, it seems all the rails meet the spec in table 6-4 on P168 of the SM you linked, great BTW that we're all on the same page. 
The section CORRECTIVE MAINTENANCE P169 describes how to access any section of the HW/circuitry and tools needed.
In section 6-36 P194 read: Analog Section Troubleshooting:
Table 6-6 P199 too.

8700 ATTENUATORS 6-85  P244

Signal flow block diagram P291

Ch1 Attenuators and Preamp P321


Still studying..........

[nctnico]

This isn't going anywhere. Apply some signal to input one and follow it using channel two by going from the input BNC towards the ADC circuit. At some point the signal will be OK before a component and distorted after. The second channel is working so any shared circuitry like the power supply, display section, etc is OK. This leaves a relatively small part to worry about.

[tautech]

Ch1 Attenuators and Preamp P321
Schematic on P321 lists a # of voltages in boxes that are applicable only when the scope is set as per Initial Setup in 4-3 P119

BEFORE PERFORMING THE REMAINDER OF THIS PROCEDURE, DO THE “INITIAL SETUP” AT THE BEGINNING OF THE PROCEDURE STEPS.

Without further test gear other than a DMM these voltages can be checked against the P321 schematic in an attempt to find the faulty stage.
If you have access to a Sig Gen you could also check the attenuators give the correct output as described in 6-86 P244.

If using Ch2 to read waveforms ensure the grounding lead is not used, just use the probe point only taking care not to let the ground ring near the tip contact anything. Use the ground ring insulating sleeve that comes with probes or place some tape over it.

[lordvader88]

Yeah I'm going backwards 1st, this will help (but soon I really want to fire up the main board and check around those CCD stuff before that ebay sale ends). But I still can't get at the bottom of this board until Tuesday when I get the Hex key.

Ok so from 348-352, FO-31, 32, here's most of the break down on U189, the LM358N dual opamp running the +5V prereg feedback show.

Details in the attached picture, with a couple things I'll add after work in 3hr.

Also I had the 1st accident, I shorted pin8 to 7 so 23.43 to -1mV, it made R566 glow yellow and smoke for 1/2sec, just checked its 10R2, so it survived just fine, its marked 10R. It goes to pin3 of U579, a collector or something, it still measures 23V, and on the other side was a filter cap to gnd and it seems fine, so it was just argument between friends. Left a blaclk spot on the R.

The only other damage or mark of any type or even dirt part I've seen on this whole DSO was an R that had a little chip off of it.

I'll throw up the same for the +5Vreg too

I forgot a couple things in the picture, the voltage at the inductor was about right, but I can't access anything closer to the transformer from the top. That 23.5V on pin2, whats with that ? is it 18V+ some 5V ?=23 or something else. Something similar happens on U870 and U470, there outputs, I'll put a pic up after work


THANKS FOR THE HELP

[tggzzz]

This isn't going anywhere. Apply some signal to input one and follow it using channel two by going from the input BNC towards the ADC circuit. At some point the signal will be OK before a component and distorted after. The second channel is working so any shared circuitry like the power supply, display section, etc is OK. This leaves a relatively small part to worry about.

And if channel 2 is suspect, use a bench power supply as an input, set to DC coupling 1Mohm input, and trace the signal flow with a voltmeter!

[cheeseit]

The only other damage or mark of any type or even dirt part I've seen on this whole DSO was an R that had a little chip off of it.

Did you check that resistor and surrounding circuitry, preferably out of circuit? I've had (THT) resistors that looked fine but had a tiny mark that rubbed off easily, or an almost invisible tiny crack, that measured open or short with one lead lifted. Take a closer look to at least rule it out.

[lordvader88]

I don't have a proper PSU yet, thats on the short list tho.

The slightly chipped R is on the main I think, it measured ok, IDK. I'm gonna write down a few more voltages on the PSU now and then try and hook the main board back up

[george.b]

Have you checked the ASICs?
I got a 2430A with a faulty CH2. CH1 was okay. I swapped around the ASICs from CH1 to CH2 until I found the faulty one - in my case, it was the peak detector ASIC. Got one from eBay and, other than some faulty tantalum caps, that was all

(Trailer Park Boys rules, by the way!)

[tggzzz]

I don't have a proper PSU yet, thats on the short list tho.

In that case use an improper PSU (e.g. a cell) and a 10kohm potentiometer!

Think about what that test is trying to achieve, and how. You are looking for gross problems, not 1% problems!

[DimitriP]

I don't have a proper PSU yet, thats on the short list tho.

In that case use an improper PSU (e.g. a cell) and a 10kohm potentiometer!

Think about what that test is trying to achieve, and how. You are looking for gross problems, not 1% problems!

  On the "improper" PSU

[lordvader88]

I've been sick all week and haven't touched it since the 31st, and I never got the allen key yet to take off the buttons, hopefully next week. I'm hoping the extra 2.2V on the 5V digital rail is not on the feedback side with the 16DIP chip or all the stuff around it, thats way more complicated than if its a cap or diode off the transformer side. But can't check without allen key.

But tomorrow I hope to hook up the main board and poke around it, still never did that.