Date
1 - 20 of 59
Tektronix 2465A noise on readout
|
Gregor
Hello everyone!
I recently purchased a Tek 2465A. It's unfortunately not working completely, and I kindly ask you to help me repair it. I was able to resolve the 1st issue which it had, but I need help with the 2nd one. The 1st issue was that I was unable to get the trace and readout focused. I started troubleshooting at the power supply and I found that someone had already replaced the capacitors with new ones. But unfortunately, most of them head the wrong capacitance values, some were higher and some were lower than the should be according to the service manual. And they were also no low ESR or hight temperature rated caps. So what I did was replacing them with low ESR capacitors which head the right capacitance values. And now the focus is working fine. The 2nd issue is that the trace and especially the readouts have some noise on top of them. To describe the issue better, I made a video which you can watch here: https://www.dropbox.com/s/kdnso5fd07u264f/IMG_0658.mp4?dl=0 I already measured the voltages according to the service manual, and they are perfectly fine (much better than the allowed tolerance). I also checked the noise of the voltages with another oscilloscope and it's also looking good. The only thing I found was that the Analog +5V (J119-12) is a bit over its specified tolerance, I took two pictures of it: https://www.dropbox.com/s/0go88yc4sanc20r/IMG_0663.jpg?dl=0 https://www.dropbox.com/s/5kh5rn7gtekiafl/IMG_0660.jpg?dl=0 So what could cause this issue? Keep in mind since the caps in the power supply were replaced with wrong values, there could be still some other wrong components in the Oscilloscope. Here are some photos of the inside: https://www.dropbox.com/s/1kyq7ac0lrvmm9b/IMG_0664.jpg?dl=0 https://www.dropbox.com/s/pmywdes0hqx08pl/IMG_0665.jpg?dl=0 Best regards
|
|
|
|
John Gord
Does the noise persist with the case on?
toggle quoted messageShow quoted text
Check that the various shield and chassis grounding screws are in place and secure. There is a shield over the power supply area that is sometimes omitted after someone works on the scope. That may or may not make a difference. --John Gord
On Thu, Oct 8, 2020 at 11:39 AM, MrClassictronix wrote:
|
|
|
|
Gregor
Hello John,
Thanks for your reply! The noise persists with the case on. I have checked the shields and chassis grounding and I found nothing suspicious. Here is a photo of the power supply, just in case I missed something: https://www.dropbox.com/s/6fiq7ssh0g67g8z/IMG_0670.jpg?dl=0 https://www.dropbox.com/s/boiewt1jljxlyim/IMG_0671.jpg?dl=0 Is there something else that could cause such a problem? Best regards Gregor
|
|
|
|
Jean-Paul
I have seen this before, years ago on 2465B,
gnd the vert inputs and look at the traces, it affects only readout or also traces ? possible a switching power supply problème Check out the lv psu on the test header on mainboard Look at HH ripple at every LV test point with X 1 probe, AC coupling and short grd connection My Best guess is wrong cap or oscillating pwm in lvps switcher or regulators. good luck! just the ramblings of an old retired EE Jon
|
|
|
|
Chuck Harris
If you are seeing what I think you are seeing, the noise on the
toggle quoted messageShow quoted text
trace and readouts is there by design. The 2465 family uses time multiplexing of the CRT beam so that it can display 4 channels, and the readout/cursor lines all on a single gun CRT. What I believe you are seeing, and perceiving as noise is missing bits of readout, when the beam is busy displaying the trace, and missing bits of the trace when the beam is occupied drawing the readout onto the screen. If this is not what you are seeing, keep talking, I have many other ideas. -Chuck Harris MrClassictronix via groups.io wrote:
Hello everyone!
|
|
|
|
Bill E
Chuck, possible, but only under specific conditions and usually only if multiple chans are enabled. I'm actually amazed at how well the scope manages, there's a (by the standards of the time) complex algorithm to prioritize how the character display is done. I'll argue that the 2465A/B was the last and greatest of the analog scopes. But hey, that's just me. Speaking blasphemy, I love my HP 54542 digital scope maybe even more. 8 gs/sec over 4 500 Mhz chans? Spectacular triggering? What's not to love. I know one of the designers, says the triggering was to counter Tek's (valid) criticism of lame HP triggering.
|
|
|
|
Bill E
Do the 'vertical jitter' calibration from the service manual, free online. That's my guess.
|
|
|
|
John Gord
Gregor,
toggle quoted messageShow quoted text
It looks like the noise on the trace might be bursts at about a 22kHz rate, so perhaps related to the HV oscillator frequency. There is a 100uF filter capacitor on the -15V UNREG feed to the HV oscillator. If that were bad it might allow a lot of noise to run about. You could try paralleling it with a known good cap if there is room. Also check the general grounding integrity in that area. --John Gord
On Thu, Oct 8, 2020 at 12:34 PM, Gregor wrote:
|
|
|
|
John Gord
Oops, I meant 18us rate, probably HV oscillator.
toggle quoted messageShow quoted text
--John Gord
On Thu, Oct 8, 2020 at 02:39 PM, John Gord wrote:
|
|
|
|
Jean-Paul
Hard to see the videos, better for still photos
The-15 and other LV caps are critical and known to fail. The readout multiplexing will change with sweep rate, there's a calibration adjustment As the 2465A are decades old, consider a complete power supply recap. Finally the 2465B 2467B are newer and superior in every aspect, to 2465 or 2465A, consider an upgrade in the future Enjoy, Jon
|
|
|
|
Chuck Harris
Looking at the video, it is showing noise on the HV.
toggle quoted messageShow quoted text
I would take the cover off of the CRT's Anode supply, and replace the -15V unregulated's 100uf filter cap, C1972. It is the only defense against the current surges caused by the EHT supply's switching rate. When C1972 goes bad, the HV becomes modulated with the switching rate, and the display swims like that. Also, you should see an increased amount of EHT ripple on the -15V regulated supply... though if you were only looking for 60/120Hz ripple, you might not have noticed. -Chuck Harris MrClassictronix via groups.io wrote:
Hello everyone!
|
|
|
|
Gregor
Thanks for all your comments and suggestions! I am very grateful for that.
So let me answer all of them. It affects the readout and the traces. To me, it seems that maybe the readout is more affected. It does not change when I ground the vertical inputs. I already checked the test header and everything looks fine (DC with DMM and AC with another oscilloscope). Except maybe the noise on the Analog +5V (J119-12) is a bit over its specified tolerance, I took two pictures of it: https://www.dropbox.com/s/0go88yc4sanc20r/IMG_0663.jpg?dl=0 https://www.dropbox.com/s/5kh5rn7gtekiafl/IMG_0660.jpg?dl=0 But it's almost within its specified tolerance, so I don't think it is causing the problem. I don't think that what I am seeing is normal. Firstly it does not depend on how many traces are on. Secondly, It's way to prominent for being normal (It's even more visible in real than on the video). These were pretty good scopes, I don't think that these scopes head that when they were new. And lastly, I have looked up some other pictures and videos of this oscilloscope and they all look fine compared to mine. I already tried the vertical jitter calibration. Unfortunately, it doesn't help. If I turn it, it starts to add a low-frequency jitter and does not affect my problem. I already recapped the whole lv power supply. I soldered another 100uF cap in parallel to C1972 in the HV power supply. Unfortunately, this didn't improve it, absolutely no change. This is the Voltage measured on C1972 (with extra 100uF in parallel): https://www.dropbox.com/s/58odweyjg4selrz/IMG_0682.jpg?dl=0 https://www.dropbox.com/s/x8ffsk4pm24ipln/IMG_0681.jpg?dl=0 https://www.dropbox.com/s/ohauhj15oydotyz/IMG_0685.jpg?dl=0 This is the Voltage measured on C1972 (no extra 100uF in parallel): https://www.dropbox.com/s/xdd7ptgzdodff9h/IMG_0689.jpg?dl=0 https://www.dropbox.com/s/x14lsqxitht7rdy/IMG_0688.jpg?dl=0 https://www.dropbox.com/s/ey7hzj4epmkvq3s/IMG_0687.jpg?dl=0 And here are some extra photos of the problem: https://www.dropbox.com/s/jp3u59l2xf5kuw5/IMG_0674.jpg?dl=0 https://www.dropbox.com/s/qqlnm4qkh4pnc0a/IMG_0675.jpg?dl=0 https://www.dropbox.com/s/vnegjbcy6igo26y/IMG_0678.jpg?dl=0 Best regards Gregor
|
|
|
|
Chuck Harris
Analog scopes, and CRT's in general, are vector displays.
toggle quoted messageShow quoted text
What you are seeing as noise is the beam retracing the display, but not over the original path, but rather with slight variations due to noise in each pass. First, we need to visually determine if the trace noise is intensity, horizontal, or vertical, or some combination of each. In your recent snapshot pictures, The noise is little bursts that are only in the vertical direction. Since the bursts happen everywhere on the screen, that means that it is at some point that is during, or after the multiplexing of the input channels and display into a single trace. That would put the problem most likely in the Vertical Channel Switch Hybrid... which is a pretty common failure item. Do the bursts still occur when the readout is turned off? -Chuck Harris Gregor via groups.io wrote:
Thanks for all your comments and suggestions! I am very grateful for that.
|
|
|
|
Gregor
The bursts are still there when the readout is off.
-Gregor
|
|
|
|
Bob Wilson
My 2445 does this also. Maybe not as much as yours but close. The trace itself it very sharp.
Bob
|
|
|
|
Gregor
So I conclude that if I want to get it fully working, I need to buy the hybrid. According to the service manual, it's U400 with Tektronix part number 155-0236-00.
Is there something else that I should try before ordering? What's the best place to order such parts? I found a few on eBay, but is it possible to get them cheaper somewhere. -Gregor
|
|
|
|
Chuck Harris
Hi Gregor,
toggle quoted messageShow quoted text
If you look at the schematics, you will find that in general the output of one hybrid passes through some passives (or less) into the input of another hybrid. The only tricky stuff is the power connections... which can act as a bus for noise to go from one hybrid function to another. The only suggestions I can give is to use some deoxit on the hybrid's terminals, and resocket it. Sometimes that is all that is necessary. Be careful, the hybrids are easy to mis-socket, and break. Hold them down with one hand, and apply the nuts on the corners, finger tight, with the other. One of those plastic nut starters makes this easy. The channel switch has a couple of transistor arrays on its Channel 5 input. The arrays make up for a switching function that was left out of the hybrid ... probably they ran out of room... or forgot. It is more unreliable than the hybrid. Typically, problems in the transistor arrays result in noise on the CH1 and CH2 traces, but not the CH3 and CH4 traces. The noise goes away when the readout is turned off. This is because of the complex display algorithm for the display doesn't ever run the display during CH3 and CH4, only CH1 and CH2. I haven't spent any time analyzing what a failure in each of the transistors would do. It is possible that your failure could be from a transistor failure that I haven't witnessed yet. -Chuck Harris Gregor via groups.io wrote:
So I conclude that if I want to get it fully working, I need to buy the hybrid. According to the service manual, it's U400 with Tektronix part number 155-0236-00.
|
|
|
|
Gregor
While cleaning the contacts of U400 I did another test. I turned on the oscilloscope without U400 installed. And it turned out that even if U400 isn't installed the noise ist still there.
Here you can see photos of the screen without U400 installed (one with the readout off and one with the readout on): https://www.dropbox.com/s/b1zm0h2qvfkmtve/IMG_0707.jpg?dl=0 https://www.dropbox.com/s/duj6ruaexga16u9/IMG_0708.jpg?dl=0 So that means that the noise must come from somewhere else, and not from U400. -Gregor
|
|
|
|
Chuck Harris
Indeed. That also eliminates anything connected to an
toggle quoted messageShow quoted text
input on U400 as well, and points back to the power supply. Do you have another scope to use? -Chuck Harris Gregor via groups.io wrote:
While cleaning the contacts of U400 I did another test. I turned on the oscilloscope without U400 installed. And it turned out that even if U400 isn't installed the noise ist still there.
|
|
|
|
Gregor
Yes, I do have another oscilloscope. As mentioned in my first post (#172137), I checked the power supply noise with my other oscilloscope, and I found that the Analog +5V (J119-12) is a bit over its specified tolerance.
toggle quoted messageShow quoted text
Here are some Photos of the Analog +5V (measured with my Tek 2230): https://www.dropbox.com/s/0go88yc4sanc20r/IMG_0663.jpg?dl=0 https://www.dropbox.com/s/5kh5rn7gtekiafl/IMG_0660.jpg?dl=0 -Gregor
On Sat, Oct 10, 2020 at 03:02 PM, Chuck Harris wrote:
|
|
|