|
Hello fellas! I have been reading/researching a ton over the past couple months while trying to get my new to me 2213A in proper working order.
I bought one from the auction site and powered it on for less than a minute, seemed alright. Powered it down because I seen that its highly recommended to replace the power supply area electrolytics & safety caps before doing the initial LVPS tests.
I have been documenting this on eevblog forum and have had some great help there from David Hess, thought I may post here too just to document my issue for others and maybe get some insight to give poor Mr. Hess a break from me, hah!
In hopes of making this more concise and readable, I will section things up as such:
- briefly explaining the anomaly & including a video link
- listing the components I replaced
- listing the 2 instances of my novice induced damage/mistakes I encountered. (I dont believe these are related to the anomaly though)
----------------
Anomaly:
After replacing parts I powered up with scope settings set to "Baseline Trace", traces displays and everything looks good. After 2-3 minutes the trace starts shrinking, blinking and slowly fades out. The power led/trigger led also fade in and out during this time. After 10 or so seconds the power led finally goes dark and the machine displays nothing. Toggling power on/off doesnt help. After unplugging and waiting an hour or so and then trying to power on the unit works but will repeat the same process again in 2-3min.
Video of the anomaly, feel free to skip to 2:19 which is when it starts to occur:
?
Under the suggestion of Mr. Hess; I checked temperatures of a few components in the power supply area right after the anomaly occured. All areas seem normal (24-27C). The "highest" temp was 29-30C on T944. Mr. Hess advised that these temperatures seemed alright and if a transformer was the cause I would expect to see very high readings.
After much more reading around I haven't found any good leads in the right direction yet. Will keep this updated!
-----------------------
Parts replaced:
C900: KEMET?R46 Series?FILM 0.15UF 10% 560VDC RAD
C902, C903, C904: KEMET Y2 Rating Film 2200pf 1kvdc
C906: United Chemicon KXJ series?100?F?450V
C925: Chemicon KY?100?F 25V?
C940: Chemicon KYB?1200?F 63V?
C942, C943: Nichicon UCS?10uf 160v
C944: Kemet R82 Series?1?F Film Capacitor 40V 63V Polyester, Polyethylene Terephthalate (PET), Metallized
C954: Chemicon kxj?39?F 160V?
C960-C970:? Chemicon KYB?1000UF 20% 25V?
R892, 894, 888, 889, 890, 891: Vishay HVR37 510K 1/2W 1%
Schaffner AC inlet/IEC: Qualtek?858-03/007
------------------------------------
Accidents/Errors/Damage:
C944, Positive lead;? Lifted pad on bottom of board. The top side of the board still has pad intact. Used multimeter on bottom side of the board and the C944 positive lead shares continuity with other components connected to the ground, I believe this means that the lifted bottom pad isn't a big deal that requires repair.
C956: I made a mistake due to the capacitor list I was using being incorrect. Initially installed a Nichicon UPW?47?F 63V cap. I caught this mistake and replaced it with the stock 270uf +100-10% 40V cap which still tested good for ESR & Capacitance. Ordered & waiting on new cap (nichicon upw 330uf 50v). Not sure if powering on with this cap caused any damage but after removing the 47uf cap it still tested good, replacing the stock 270uf cap did not solve the scope anomaly.
----------------------------
Thanks for any insight, I will still be researching meanwhile and will update if I find anything out.
Cheers,
Gary R.
Treasure Coast, Florida
|
Nope, no break for Dave Hess; He's on TekScopes, too!Jim Ford?Sent from my Verizon, Samsung Galaxy smartphone
toggle quoted message
Show quoted text
-------- Original message --------From: guy232 <gry.russel@...> Date: 3/22/19 11:55 AM (GMT-08:00) To: [email protected] Subject: [TekScopes] 2213A Novice, unsolved anomaly encountered, documented & video Hello fellas! I have been reading/researching a ton over the past couple months while trying to get my new to me 2213A in proper working order. I bought one from the auction site and powered it on for less than a minute, seemed alright. Powered it down because I seen that its highly recommended to replace the power supply area electrolytics & safety caps before doing the initial LVPS tests.I have been documenting this on eevblog forum and have had some great help there from David Hess, thought I may post here too just to document my issue for others and maybe get some insight to give poor Mr. Hess a break from me, hah!In hopes of making this more concise and readable, I will section things up as such:- briefly explaining the anomaly & including a video link- listing the components I replaced- listing the 2 instances of my novice induced damage/mistakes I encountered. (I dont believe these are related to the anomaly though)----------------Anomaly: After replacing parts I powered up with scope settings set to "Baseline Trace", traces displays and everything looks good. After 2-3 minutes the trace starts shrinking, blinking and slowly fades out. The power led/trigger led also fade in and out during this time. After 10 or so seconds the power led finally goes dark and the machine displays nothing. Toggling power on/off doesnt help. After unplugging and waiting an hour or so and then trying to power on the unit works but will repeat the same process again in 2-3min.Video of the anomaly, feel free to skip to 2:19 which is when it starts to occur:? the suggestion of Mr. Hess; I checked temperatures of a few components in the power supply area right after the anomaly occured. All areas seem normal (24-27C). The "highest" temp was 29-30C on T944. Mr. Hess advised that these temperatures seemed alright and if a transformer was the cause I would expect to see very high readings.After much more reading around I haven't found any good leads in the right direction yet. Will keep this updated!-----------------------Parts replaced:C900: KEMET?R46 Series?FILM 0.15UF 10% 560VDC RADC902, C903, C904: KEMET Y2 Rating Film 2200pf 1kvdcC906: United Chemicon KXJ series?100?F?450VC925: Chemicon KY?100?F 25V?C940: Chemicon KYB?1200?F 63V?C942, C943: Nichicon UCS?10uf 160vC944: Kemet R82 Series?1?F Film Capacitor 40V 63V Polyester, Polyethylene Terephthalate (PET), MetallizedC954: Chemicon kxj?39?F 160V?C960-C970:? Chemicon KYB?1000UF 20% 25V?R892, 894, 888, 889, 890, 891: Vishay HVR37 510K 1/2W 1%Schaffner AC inlet/IEC: Qualtek?858-03/007------------------------------------Accidents/Errors/Damage:C944, Positive lead;? Lifted pad on bottom of board. The top side of the board still has pad intact. Used multimeter on bottom side of the board and the C944 positive lead shares continuity with other components connected to the ground, I believe this means that the lifted bottom pad isn't a big deal that requires repair.C956: I made a mistake due to the capacitor list I was using being incorrect. Initially installed a Nichicon UPW?47?F 63V cap. I caught this mistake and replaced it with the stock 270uf +100-10% 40V cap which still tested good for ESR & Capacitance. Ordered & waiting on new cap (nichicon upw 330uf 50v). Not sure if powering on with this cap caused any damage but after removing the 47uf cap it still tested good, replacing the stock 270uf cap did not solve the scope anomaly.----------------------------Thanks for any insight, I will still be researching meanwhile and will update if I find anything out.Cheers,Gary R.Treasure Coast, Florida
|
Looking at your Youtube, something is obviously failing with heat so freeze spray would be helpful (upturned canned air is fine). My initial thought was U760 as the timebase is collapsing but I've not seen one fail such that the power led would flicker. Unplug the timebase, see if it still fails. (You are looking for the power LED to flicker, nice symptom, unusual!). Surprised the scope isn't tripping completely. I'd suggest monitoring the power rails to see if its just one rail failing or all rails. Complicated with switching supplies because all will drop on overload as the switcher detects a fault but not impossible to ascertain which rail is overloading. I would start by disconnecting the feed to the multiplier (so nothing on screen) and just observing the power led (or monitor the supplies) for any failings. You can also remove power to the attenuator board, an unlikely culprit but I have had the opamps go short. You could just pull all the power supply links and monitor the rails but without a load the fault might not show. What is the Y amp doing when it fails? Amplitude still good? Your replacement parts look ok. Hopefully you have checked the obvious, I've received these with loose screws jammed in unlikely places, solder bridges due to careless repairs. JC
|
My first scope was a 2213A. I got it from the auction site. Stated all was working 100%. After powering on for several minutes it would lose its trace and go crazy. The previous owner had removed the board and there are little tabs that hold the board in place. One side of the board was not fit into the tab and the board was bent. after removing the board and making sure the board was in its proper place it has worked fine since and that was 7 years ago.
Brendan
|
I hope the 2213A owner follows JC's test plan, or some variation of it.
The 2213A needs a good logical plan to find this most unusual failure.
It's great to hear that David Hess has been helping him on this project over on the EEV Blog!
There are plenty of 22xx parts available once this mystery is solved.
tom jobe...
toggle quoted message
Show quoted text
On 3/22/2019 4:45 PM, chipbee40 via Groups.Io wrote: Looking at your Youtube, something is obviously failing with heat so freeze spray would be helpful (upturned canned air is fine). My initial thought was U760 as the timebase is collapsing but I've not seen one fail such that the power led would flicker. Unplug the timebase, see if it still fails. (You are looking for the power LED to flicker, nice symptom, unusual!). Surprised the scope isn't tripping completely. I'd suggest monitoring the power rails to see if its just one rail failing or all rails. Complicated with switching supplies because all will drop on overload as the switcher detects a fault but not impossible to ascertain which rail is overloading. I would start by disconnecting the feed to the multiplier (so nothing on screen) and just observing the power led (or monitor the supplies) for any failings. You can also remove power to the attenuator board, an unlikely culprit but I have had the opamps go short. You could just pull all the power supply links and monitor the rails but without a load the fault might not show. What is the Y amp doing when it fails? Amplitude still good? Your replacement parts look ok. Hopefully you have checked the obvious, I've received these with loose screws jammed in unlikely places, solder bridges due to careless repairs. JC
|
Hi Gary,
It might be helpful to drive an input from the 'probe adjust' output to be able to observe the state of the vertical when the horizontal collapses. Just a solid wire from probe adjust to the centre of the BNC will suffice. This would also expose the state of the trigger circuits as the collapse happens.
Cheers, Brian.
|
Sorry for the delay gentlemen, have finally started feeling better. Have been reading the theory of operation some and editing pics/vid meanwhile.
I have done a few things following your suggestions and appreciated all of the info.
In an attempt to be brief, I didn't mention this discovered issue in my original post:
A) H+ wire connector wasn't connected to CRT pin, carefully reattached with hemostat. (I caught this & reattached this prior to powering on at all, before shooting first the video etc)
-------
Things that I have done based on suggestions:
1) I searched everywhere for U760 so that I could monitor it's temperature. It's listed in the 2213a SM part list & theory of operation. But I couldn't find it on the instrument nor the diagram. I found it on the diagrams for the 2235 .pdf I have but on the 2213A it doesn't appear to be present on timing board?
2) Monitored LVPS, all solid and in spec. When the anomaly occured there was some fluctuation. I need to re-test the fluctuations though, I feel I may have made poor probe contact, unsure of accuracy.
3) I noticed that the neon bulb fired inside atleast once. I must re-test with shield off because I couldn't tell if it was DS858+DS856 (excessive grid voltage) or DS870 (excess heater/cathode voltage)
4) The square wave appears to compress horizontal during the anomaly. The LED still dimmed, also noted the Z-axis (intensity, yes?) was fluctuating alot.?
After approx 5 min total of operation the instrument finally went dark like last time. However, within 20s-1min it came back to life. At this time I touched side of chasis noticed the Q9070 heatsink area was rather warm. Blasted the side chasis with cold. The Z-axis instantly stopped fluctuating and returned to proper intensity level.
I waited, after 3-5min the tops of the square trace would get a bit of jitter, then intensity fluctuations, etc. I hit it again with cold and it went back to normal for atleast 5-10 min at which point I powered down.
----------------
I opened to inspect. I would assume that the reaction to spray may indiciate that Q9070 is potentially the cause of the issue. The molex doesn't appear discolored. Compared to pics of other 22xx, I don't see any excess white transfer paste on my Q9070.
Aside from that, I noticed a possible "mod". A pin was soldered near the front face PCB, hard to see under the attenuator/timing boards but it seemed to be a resistor and a couple other components attached to this. I know that some of the boards were updated/official mods were released. My attenutator board has an "01" sticker appended to it. Perhaps that explains the soldered pin?
Here is a video of my above mentioned test. I really tried to edit it down and caption it, only 58s long:
Here are a few captioned pics touching on things mentioned above:
I'm going to re-read your above posts again and see if I missed any instruction/advice. will update.
Thanks again for your time and help, fellas! Learning alot, hope I'm not typing too much, just trying to document this issue especially since the LED flicker isn't too common?
|
Since you have a good understanding of how to make the fault show, and not show, how about checking some voltages while this is happening and see if it will lead you to the problem and a better understanding of how the power supply functions?
A key voltage that is about in the 'middle of the scope' that you might start with would be from TP940 to TP950, and it should read around 43 volts.
If you flip the scope over, those test points (TP940 and TP950) are under that squarish plastic cover that protects you from some of the higher voltages.
One way to connect to those test points is to solder small loops of wire to the two test points so you can 'clip' onto them. Neither test point is at chassis ground, and I think TP940 is the floating ground for everything to the left of the main transformer (on the schematic) back to where the AC first gets rectified.
This voltage powers the primary side of the main transformer, and if it has problems the whole scope has problems.
If for example, your Q9070 was flakey in some way, it would affect your 43 volt reading and you would see the 43 volts change as the display changed.
Checking the voltages on the secondary side of the transformer is a waste of time until the 43 volts is stable.
Do double check whatever I have to say as I'm fairly old and stupid.
tom jobe...
toggle quoted message
Show quoted text
On 3/31/2019 6:56 AM, guy232 wrote: Sorry for the delay gentlemen, have finally started feeling better. Have been reading the theory of operation some and editing pics/vid meanwhile.
I have done a few things following your suggestions and appreciated all of the info.
In an attempt to be brief, I didn't mention this discovered issue in my original post:
A) H+ wire connector wasn't connected to CRT pin, carefully reattached with hemostat. (I caught this & reattached this prior to powering on at all, before shooting first the video etc)
-------
Things that I have done based on suggestions:
1) I searched everywhere for U760 so that I could monitor it's temperature. It's listed in the 2213a SM part list & theory of operation. But I couldn't find it on the instrument nor the diagram. I found it on the diagrams for the 2235 .pdf I have but on the 2213A it doesn't appear to be present on timing board?
2) Monitored LVPS, all solid and in spec. When the anomaly occured there was some fluctuation. I need to re-test the fluctuations though, I feel I may have made poor probe contact, unsure of accuracy.
3) I noticed that the neon bulb fired inside atleast once. I must re-test with shield off because I couldn't tell if it was DS858+DS856 (excessive grid voltage) or DS870 (excess heater/cathode voltage)
4) The square wave appears to compress horizontal during the anomaly. The LED still dimmed, also noted the Z-axis (intensity, yes?) was fluctuating alot.
After approx 5 min total of operation the instrument finally went dark like last time. However, within 20s-1min it came back to life. At this time I touched side of chasis noticed the Q9070 heatsink area was rather warm. Blasted the side chasis with cold. The Z-axis instantly stopped fluctuating and returned to proper intensity level.
I waited, after 3-5min the tops of the square trace would get a bit of jitter, then intensity fluctuations, etc. I hit it again with cold and it went back to normal for atleast 5-10 min at which point I powered down.
----------------
I opened to inspect. I would assume that the reaction to spray may indiciate that Q9070 is potentially the cause of the issue. The molex doesn't appear discolored. Compared to pics of other 22xx, I don't see any excess white transfer paste on my Q9070.
Aside from that, I noticed a possible "mod". A pin was soldered near the front face PCB, hard to see under the attenuator/timing boards but it seemed to be a resistor and a couple other components attached to this. I know that some of the boards were updated/official mods were released. My attenutator board has an "01" sticker appended to it. Perhaps that explains the soldered pin?
Here is a video of my above mentioned test. I really tried to edit it down and caption it, only 58s long:
Here are a few captioned pics touching on things mentioned above:
I'm going to re-read your above posts again and see if I missed any instruction/advice. will update.
Thanks again for your time and help, fellas! Learning alot, hope I'm not typing too much, just trying to document this issue especially since the LED flicker isn't too common?
|
Yes, sorry about the U760, forgot the timebase is different in the 2213A (was thinking 2235).
I read your comment on Eevblog, you say you disconnected the CRT anode connection, I was talking about disconnecting the feed from T948 to the multiplier. This would eliminate the multiplier as a problem and also removes any HV arcing associated with it. Insulate the wire from T948 or at least keep it clear of anything (and yourself). Interesting fault, freeze spray on the case? Odd. Screws all tight on those heatsinks?
toggle quoted message
Show quoted text
On Sun, Mar 31, 2019 at 06:56 AM, guy232 wrote: 1) I searched everywhere for U760 so that I could monitor it's temperature.
It's listed in the 2213a SM part list & theory of operation. But I couldn't
find it on the instrument nor the diagram. I found it on the diagrams for the
2235 .pdf I have but on the 2213A it doesn't appear to be present on timing
board?
|
Apologies for the delay in updating, was a bit under the weather.
Months ago I initially used a 5x mag lens to investigate the solder joints in the PS areas. Didn't see any issues.
Last week when I got back to working on this, I remembered I had a cheap $20 usb microscope laying around.
Discovered very tiny cracks on all 3 pads on both Q946 and Q947. C907 had crack on the component side of the board (the bottom was not cracked)
hosted pics of cracks:
IPA scrubbed, dried, applied some amtech flux and reflowed the joints.
Also swapped out C956, I was using the original mallory cap due to ordering the incorrect cap the first time. Used a new Nichicon UHW 330uF 50V that I had on hand. I seen that C956 is listed as +10-100% and believe this is a workable low impedance choice. If not, i'd appreciate any insight on the matter.
Powered up, the shrinking H and V issue
and extreme intensity change fail states no longer occurs. The only fault that I notice is that the traces just barely pulse brightness (as if you were just barely turning the intensity knob by 1-2 degrees back and forth) after 3-5min, they remains steady for until powered down.
Going to microscope a bit more and use a long wooden dowel to gently tap around while the unit is running as David Hess mentioned on my project thread on eevblog.
P.S.: Just adding a link to the project thread on eevblog in case it helps anyone in the future etc.
|
Thank you for sharing your experiences!
toggle quoted message
Show quoted text
On Mon, Aug 26, 2019 at 12:38 PM guy232 <gry.russel@...> wrote: Apologies for the delay in updating, was a bit under the weather.
Months ago I initially used a 5x mag lens to investigate the solder joints
in the PS areas. Didn't see any issues.
Last week when I got back to working on this, I remembered I had a cheap
$20 usb microscope laying around.
Discovered very tiny cracks on all 3 pads on both Q946 and Q947. C907 had
crack on the component side of the board (the bottom was not cracked)
hosted pics of cracks:
IPA scrubbed, dried, applied some amtech flux and reflowed the joints.
Also swapped out C956, I was using the original mallory cap due to
ordering the incorrect cap the first time. Used a new Nichicon UHW 330uF
50V that I had on hand. I seen that C956 is listed as +10-100% and believe
this is a workable low impedance choice. If not, i'd appreciate any insight
on the matter.
Powered up, the shrinking H and V issue
and extreme intensity change fail states no longer occurs. The only fault
that I notice is that the traces just barely pulse brightness (as if you
were just barely turning the intensity knob by 1-2 degrees back and forth)
after 3-5min, they remains steady for until powered down.
Going to microscope a bit more and use a long wooden dowel to gently tap
around while the unit is running as David Hess mentioned on my project
thread on eevblog.
P.S.: Just adding a link to the project thread on eevblog in case it helps
anyone in the future etc.
|
guy232
Brian Cockburn