Author Topic: (Solved) Old CRT circuit repair  (Read 19845 times)

0 Members and 1 Guest are viewing this topic.

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
(Solved) Old CRT circuit repair
« on: June 26, 2014, 05:53:26 am »
New to the forum, and while I have electronics knowledge, I'm still a beginner in the big scheme of things - so I thought I'd post my thread here, rather than Technical, but if I got it wrong just let me know.

As a hobby I maintain/collect computers from the 1980's. Generally if a PSU or CRT fails - I replace it - however in the future I'm going to need to be able to repair these properly, and doing that will take some learning piece by piece. I have 4 CRTs to begin with and I haven't bought myself an oscilloscope yet - but it's on my 'to get' list.

So the first project I want to take up is a monochrome display from 1987 - and I was hoping you guys could either discuss the circuit that's at fault so I can understand how it operates better, and/or point me towards which components are most likely to be failing to cause the issue.

So here is what I get:


What I think is happening is the horizontal deflection yoke isn't getting enough voltage to pull the beam to the left, so it starts drawing near the center, gets a little bit right, then oscillates backwards. So any lines which contain more than 5 characters, start overwriting each other. "C:>_" should be hard left, not centered, and the content above that is text laid on top of text - until you just get a mess.

Vertical is perfect.

This is the circuit that is responsible for feeding the horizontal deflection yoke:


The arrow is pointing to the trigger input - Hsync - which pulses at 18.4KHz. The line below the arrow is +12V.
The horizontal deflection yoke can be seen up the top - labelled H-DY.

My understanding (I could be wrong, I'm new to this) is that when Hsync comes in, it should apply full voltage to the H-DY - pulling the beam to the left, then gradually reduce moving it towards the center, then start applying an inverse voltage moving it right, then the next Hsync should come in, and it should repeat for the next line.

(I'm also aware of the dangers working inside these, I'm very careful, take my time, and make sure components are discharged before manhandling things)

Any information or discussion on this circuit appreciated! Apologies if this isn't the place for this, just I'm quite interested and looking for people with more knowledge to clue me in.
« Last Edit: July 23, 2014, 11:36:23 pm by SpidersWeb »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Old CRT circuit repair
« Reply #1 on: June 26, 2014, 07:33:03 am »
Unusual, because failing horizontal drive generally means HV is going to crap as well.  Which should cause the picture to dim and bloom (deflection sensitivity is way higher).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Sparc

  • Contributor
  • Posts: 16
Re: Old CRT circuit repair
« Reply #2 on: June 26, 2014, 07:40:10 am »
Its good to see someone looking after the old gear.  I've only repaired a small number over the years, but none recently.  CRTs are almost gone now.  For some background theory about CRT operations and repair, see "Sam's REPAIR FAQ".  Its a free document on web.  Lots of good info there for repairing CRT TV and computer monitors, and many other things.

As best I can see from the circuit:  When Q503 is OFF, current flows into the H-DY and charges up C505.  When Q503 is ON, current flows back out of the H-DY and C505 also discharges through the coil.  Generally, the horizontal output transistor acts as a switch, either ON or OFF.

For this monitor, yea, it looks like the H deflection coil isn't getting enough current to deflect the beam.  First thing I'd check is for burned/discolored components.  Check for bad solder joints with a magnifier.  They can especially develop in the high current, high temperature areas of circuit boards.   Check the power supply for correct voltage.  Electrolytic caps can go bad.

Check the components that drive the base of Q503,BU189.  If Q503 doesn't get enough drive current it won't pull enough current through the coil and could overheat and fail.  Check R507, C503 and everything backwards from it.

Looks like BU189 is a darlington transistor.  You might have to pull it from circuit and test it independently.  Maybe  check other components at the other end of H-DY, C504, C505, etc.
 

Offline PA0PBZ

  • Super Contributor
  • ***
  • Posts: 5121
  • Country: nl
Re: Old CRT circuit repair
« Reply #3 on: June 26, 2014, 07:50:23 am »
Since the HV seems more or less ok you should start by looking at the components that are just there for the horizontal deflection, like the deflection coil itself and the return circuit like C505 and the components around there.
Keyboard error: Press F1 to continue.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: Old CRT circuit repair
« Reply #4 on: June 26, 2014, 07:53:08 am »
It looks like you are getting EHT,so it would seem that the horizontal output is working to some extent.
Two possibilities come to mind:-

(1) C505 may be faulty.so that it places a high impedance between the junction of L503 & C504,& ground,reducing the picture width.
(Note that this cap needs to be a low ESR type)

(2) Even though you have some EHT,the voltage on Q503 collector may not be correct.

Normally,at start up,that voltage will be about +12v or so,but as soon as the Horizontal output transformer is fed with signal,this voltage will rise to,typically,around +120v due to the action of D502,C506,which make up what we called in the vacuum tube days a "Boost HT" circuit.

If either of those components are faulty,you will lose this voltage.

I doubt that (2) is the case,though,as you still have full vertical scans----vert output also uses "Boost HT".
 

Online Andy Watson

  • Super Contributor
  • ***
  • Posts: 2074
Re: Old CRT circuit repair
« Reply #5 on: June 26, 2014, 09:13:51 am »
Another vote for looking closely at C505. This cap is usually working under very stressed conditions - greatly accelerating its demise. They dry-out and become high impedance. This was a common fault leading to the classic shrinking screen on the old Apple Macs.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Old CRT circuit repair
« Reply #6 on: June 26, 2014, 10:02:47 am »
Looking at the schematic, I don't see why it needs to be nonpolar anyway.  It should have full DC boost on it.  Are NP lytics more robust (higher ripple / AC voltage ratings)?

An open somewhere in that path (also check continuity on L501-L503) would easily explain the symptom: in fact, I'd pretty reasonably guess the top end of L501 is still connected, because the switching edges are making it bounce -- it's still scanning a little bit, but very weakly, and only for a short period, pretty much going nowhere otherwise.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Old CRT circuit repair
« Reply #7 on: June 26, 2014, 10:28:34 am »
Looking at the schematic, I don't see why it needs to be nonpolar anyway.  It should have full DC boost on it.  Are NP lytics more robust (higher ripple / AC voltage ratings)?

An open somewhere in that path (also check continuity on L501-L503) would easily explain the symptom: in fact, I'd pretty reasonably guess the top end of L501 is still connected, because the switching edges are making it bounce -- it's still scanning a little bit, but very weakly, and only for a short period, pretty much going nowhere otherwise.

Tim

Yes I agree, I would say L501 or L503 (or soldering of) is open circuit. the reason it is scanning at all is the circuit made by R506,C504 & C505.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: Old CRT circuit repair
« Reply #8 on: June 26, 2014, 07:38:03 pm »
Looking at the schematic, I don't see why it needs to be nonpolar anyway.  It should have full DC boost on it.  Are NP lytics more robust (higher ripple / AC voltage ratings)?

An open somewhere in that path (also check continuity on L501-L503) would easily explain the symptom: in fact, I'd pretty reasonably guess the top end of L501 is still connected, because the switching edges are making it bounce -- it's still scanning a little bit, but very weakly, and only for a short period, pretty much going nowhere otherwise.

Tim

It needs to be a non polarised capacitor, as it is part of a resonant circuit and will swing to about the supply voltage below ground to drivve the coil to deflect to the left. These caps do die with age, so check them and the supply caps as well, and look for bulging ones. With these as well look for smaller electrolytics that dry out with age, and high value resistors that go high in value or open circuit.
 

Offline RazSlack

  • Contributor
  • Posts: 25
  • Country: 00
Re: Old CRT circuit repair
« Reply #9 on: June 26, 2014, 07:53:10 pm »
I've dug up my old "picture errors" CRT repair handbook. Here's what's in it:

First off, a precaution: Don't keep the screen running like that for a long time, it might damage the CRT, causing a permanent burn-in.

The book doesn't say any other useful stuff, so this is all my contribution I can give. Check the horizontal amp and the cabling, all it says. Bah.

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #10 on: June 26, 2014, 08:05:27 pm »
Thanks guys, this is exactly the kind of discussion I was looking for. I think I'll have to spend a lot more time here.

I'll start by checking the continuity of the inductors/circuit in that section. If I don't find any issues, I'll start replacing the 5xx electrolytic caps one by one. C505 is an easy starter. I was unable to see any visibly damaged components or dark/burned areas.

I also need to confirm that this circuit isn't modified from the diagram, logically it seems to match up with the diagram but there was a couple of open unlabelled spots on the PCB so I'll double check that.

I may not get back to it until tomorrow, but here are a few pictures in the mean time.

Neck PCB


Missing inductor spot, C505, L504 etc

NB: Black soot on resistor in last picture - it was in a factory controlling a welding machine for many years, it's just soot that wipes off, resistor isn't burned.

Most of the circuit is here, Q503 attached with a heatsink, Q502 nearby.


Bottom of main PCB:


« Last Edit: June 26, 2014, 08:11:14 pm by SpidersWeb »
 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #11 on: June 27, 2014, 11:59:49 am »
Looking at the schematic, I don't see why it needs to be nonpolar anyway.  It should have full DC boost on it.  Are NP lytics more robust (higher ripple / AC voltage ratings)?

An open somewhere in that path (also check continuity on L501-L503) would easily explain the symptom: in fact, I'd pretty reasonably guess the top end of L501 is still connected, because the switching edges are making it bounce -- it's still scanning a little bit, but very weakly, and only for a short period, pretty much going nowhere otherwise.

Tim

Yes I agree, I would say L501 or L503 (or soldering of) is open circuit. the reason it is scanning at all is the circuit made by R506,C504 & C505.

So, if one connector to L501 was disconnected,  that would still allow for a bit of scanning?

I'm double checking because the other coils checked out OK (continuity), and then I checked L501.
L501 actually does the vertical and horizontal, it's just clipped off the diagram I posted, it has 4 wires going in - one to each end of a coil.

Two of the connections show 0 ohms and gradually increase to 7.9 ohms.
The other two appear open circuit :/
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Old CRT circuit repair
« Reply #12 on: June 27, 2014, 12:07:03 pm »
Sorry my mistake. My post should have read L502 or L503......

If L501 was open then you should get a thin vertical line.

Although in saying that. I just thought, what would happen if L501 had a shorted turn?
« Last Edit: June 27, 2014, 12:09:08 pm by What_NZ »
 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #13 on: June 27, 2014, 12:23:21 pm »
Ahh well L502 and L503 check out okay (and L504).
L503 did give 1 ohm, the others 0.

And now I'm getting continuity readings on both of the coils in L501 - so possibly a bit of a red herring - I did carefully adjust one of the coils leads but the solder joins looks good and I can't be sure I just didn't make some mistake when measuring. But L501 is OK.

Edit: to check for continuity, I'm just using DC resistance, I haven't worked with inductors before so wanted to check that was correct.
« Last Edit: June 27, 2014, 12:25:03 pm by SpidersWeb »
 

Offline What_NZ

  • Regular Contributor
  • *
  • Posts: 150
  • Country: nz
Re: Old CRT circuit repair
« Reply #14 on: June 27, 2014, 12:30:56 pm »
Resistance is futile (Star Trek Borg joke) when measuring inductors for a shorted turn.

You need a shorted turns tester or the like. An inductance meter may not show a problem.
 

Offline Mr Smiley

  • Frequent Contributor
  • **
  • Posts: 324
  • Country: gb
Re: Old CRT circuit repair
« Reply #15 on: June 27, 2014, 06:00:02 pm »
Have a gander here,

CRT monitors-Part 1of 4


Randy should help you identify the problem  :-+

 :)
There is enough on this planet to sustain mans needs. There will never be enough on this planet to sustain mans greed.
 

Offline zaoka

  • Frequent Contributor
  • **
  • Posts: 376
  • Country: us
Re: Old CRT circuit repair
« Reply #16 on: June 27, 2014, 07:18:31 pm »
You should replace C501, C502 and C503 without testing if you dont have ESR meter. Test Q501 and Q502, D502, C505, C506 and C508. Test transistors and diode with analog multimeter (1K range).

C505 should be tested for both; capacity and ESR, you can use two 22uF 25V, connecte minuses together and use plus sides to solder in the board.

Also if power supply is low you will get the same symptom, L504 is supplying that. Make sure it works correctly. It uses either 10.8V or 12V. Also make sure that filtering for this voltage is OK.

L503 may be short, however, I dont this thats the problem, it usually change width only.

Also post full schematic if you have one.

Most critical parts are C505, C502, C503 and power supply.
« Last Edit: June 27, 2014, 07:26:51 pm by zaoka »
 

Offline sunnyhighway

  • Frequent Contributor
  • **
  • Posts: 276
  • Country: nl
Re: Old CRT circuit repair
« Reply #17 on: June 30, 2014, 09:48:55 am »


Could be nothing, but R414 seems to be chipped a bit.
 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #18 on: July 02, 2014, 05:53:42 am »
Thanks to everyone who replied, sorry for the slow updates, I still haven't got the caps to go in yet.

Mr Smiley - great link! I haven't finished watching the video but I'll make more time

zaoka - I'll be blindly replacing all the electrolytic caps in the 5xx section and see if that helps. C505 I'm thinking may have already been replaced. I went back to my old notes from before I got it, and the previous owner took it to a repair agent who said it was out of his knowledge range and was unable to get it running again - I think he may have replaced that cap. It's the only bright green one, higher voltage (100V) than on  the diagram, and identical to the NP model sold by our local JayCar.

D502 checked out ok, need to test, Q502 and Q501 still.

Sunny - I'll check it

I've attached the full diagram for anyone who is curious.


 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #19 on: July 02, 2014, 05:57:01 am »
Also I've assumed that inductors in this circuit should be near 0 ohms.
I get < 1 ohm from between the top of C505 to P503 (the plug for the yoke).
« Last Edit: July 02, 2014, 05:59:05 am by SpidersWeb »
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7563
  • Country: au
Re: Old CRT circuit repair
« Reply #20 on: July 02, 2014, 06:46:16 am »
Thanks to everyone who replied, sorry for the slow updates, I still haven't got the caps to go in yet.

Mr Smiley - great link! I haven't finished watching the video but I'll make more time

zaoka - I'll be blindly replacing all the electrolytic caps in the 5xx section and see if that helps. C505 I'm thinking may have already been replaced. I went back to my old notes from before I got it, and the previous owner took it to a repair agent who said it was out of his knowledge range and was unable to get it running again - I think he may have replaced that cap. It's the only bright green one, higher voltage (100V) than on  the diagram, and identical to the NP model sold by our local JayCar.

D502 checked out ok, need to test, Q502 and Q501 still.

Sunny - I'll check it

I've attached the full diagram for anyone who is curious.

Be cautious with generic NP caps.
I had one which was "obviously dead" in an Electrohome Monitor.
Replaced it with a new (much smaller) NP,marvelling at how technology had advanced since the original had been made.

The new one lasted 5 minutes  then let the "magic smoke" out----the "big" one was low ESR,& the new small one wasn't.

Mega-Oops!! :-[
1 minute's closer perusal would have shown me that the replacement would be inadequate.

As it happened,I had to make a "Christmas tree" of poly Greencaps  to replace it,as that capacitance, voltage rating ,& ESR value.was not locally available in an NP.
Two polarised electros of suitable rating connected in "reverse series" was even bigger than the poly stack! ;D
 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #21 on: July 02, 2014, 08:10:02 pm »
I'll check the labeling on it to make sure it has a low ESR rating (and check that it is actually new and not a 1988 factory substitute). This guy likely wont see power again until I've either found the faulty component and/or replaced the caps in that section and put it back inside it's case. This is also because I accidently pulled a +12 line when I moved the board and need to solder it back on :P

Last night I went through the diodes in that section and Q501-Q503 - all appeared fine (using a diode test while in circuit at least - showing 0.45 to 0.68V with correct polarity). Was going through the resistors and then got interrupted by dinner - but each one tested so far shows it's inline with the diagram. I'll post the list and recorded values when I finish (actually recording them this time).

If I don't see a specific issue, and caps don't help, I'll put it to one side until I get an oscilloscope running. I did grab a 10Mhz dual channel from a recycling center with the plug cut off - but no idea if it works yet - always planned to get one of those affordable Rigol models - and some probes with clips because I'm not a fan of poking around in a 25-30 year old CRT while it's on.

Others I need to look at:
- Amstrad PC-MD - brand new in box from 1987 - first power up was last week and it's got multiple issues (probably just needs a recap) plus I need to hack it slightly to make it useful (it has a full PC PSU inside and that's more of a problem than a help).
- Sprite RGB display - sync was crazy when powered up, I thought it was the source but then pow - thinking cap went bad then it took out the switching transistor - upside here is finding the fault should be easy - it'll be clearly marked!
- Taxan RGB - works fine for 5 minutes, then powers off, tapping it gets you a bonus 30 seconds.

 

Offline zaoka

  • Frequent Contributor
  • **
  • Posts: 376
  • Country: us
Re: Old CRT circuit repair
« Reply #22 on: July 03, 2014, 01:59:57 am »
Did you measure 12V on L504?

Also make sure you are getting proper input signal, horizontal should be arount 15kHz and vertical around 50-60hz. Without correct input signal this thing wont work.
« Last Edit: July 03, 2014, 02:03:34 am by zaoka »
 

Offline SpidersWebTopic starter

  • Supporter
  • ****
  • Posts: 62
  • Country: nz
Re: Old CRT circuit repair
« Reply #23 on: July 03, 2014, 05:05:45 am »
Not yet, I need to resolder the +12 line as I broke it off (genius) when removing the PCB and I was avoiding power ups until I'd put in the new caps - partly because I'm comfortable it's safely discharged at the moment. (Apologies for being a bit slow).

But I will check it when I get it powered up. I wasn't too suspicious of voltage levels because the other functions (which run off the same supply) seem to be 100%.
Edit: had to educate myself a little bit on inductors - now I get it it a bit more - will certainly test

I've tested the source with another MDA display (IBM 5151). Spec is a little bit higher the a normal old style CRT ~18.432Khz horizontal (720x350@50).

Edit2: Quick Update, unsoldered C505 - 1uF! That's at least a sign it's stuffed, others could be similar. It is identical to the one sold at JayCar though - MDL 100V 10uF NP.
« Last Edit: July 03, 2014, 07:53:49 am by SpidersWeb »
 

Offline Paul Moir

  • Frequent Contributor
  • **
  • Posts: 926
  • Country: ca
Re: Old CRT circuit repair
« Reply #24 on: July 04, 2014, 07:11:08 am »
Any chance it's been replaced before?  I was looking at the glue around it a few days ago which seemed to suggest it was.  That is to say it almost looks like the factory glue outlined a different diameter capacitor.  The soldering looked factory so I didn't mention it, but the fact that there's an identical one at Jaycar gives pause.

 
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf