Meguro MO-1252A 25MHz Analog Oscilloscope - nothing on CRT

• #25 Reply
  Posted by huelter on 14 May, 2018 21:02
• Good call, Q1003 was shorted (again)!
  
  I think it is because I used 2sb546 (20W) instead of 2sb861 (30W), all other specs match. Now with only a 8.2K ohm load on the 160V rail and 100ohm loads on the 12V rails, all rails are within spec!
  
  I will now carefully insert one module each time and monitor the supply voltages.
• #26 Reply
  Posted by huelter on 15 May, 2018 03:47
• OK, tested again with everything connected.
  
  There is a huge voltage drop on the 5V rail as soon as I connect CN2003 to the second board, becoming 2.75V. Connecting the rest of the cables/modules did not help.
  
  One thing to note is that all the other rails remained stable and within specifications, even with the added loads. Removing the loads also did not change the outcome.
  
  Now, what am I searching for? Looking every 5V component on the second board for shorts/opens?
• #27 Reply
  Posted by floobydust on 15 May, 2018 05:26
• The 5V regulator could be weak, or there is an overload on that rail. I would expect R1026 to smoke and Q1004 to get hot.
  
  Shorted Q307/Q308 vert. amp could backfeed and overvoltage the 5V rail at R319.
  
  The 5V rail is mostly for the digital IC's so one may have shorted, and would warm up.
  
  You could take a bench PSU (with current limit set to low 100-200mA) and inject only 5V power to the main board and see who heats up. Or measure the voltage drop across some of the 10R resistors on the 5V rail to see how much current a part draws. Like IC202, IC201 with R254, R253.
  IC202 74LS00 vert alt/chop
  IC201 74LS00 vert alt/chop
  
  IC401 3/4 74LS00 horiz sweep
  IC601 74LS122 horiz sweep
  IC602 TC4053 mux for chop
  IC401 74LS00 triggering
  IC4011 TC4011 CAL pcb square wave osc.
  
  I'd finger test these IC's and see if they are hot. Otherwise you'd have to lift a Vcc pin or guess and remove an IC. I'd put in a socket.
• #28 Reply
  Posted by huelter on 15 May, 2018 15:25
• IC602 was dropping the 5V rail. Replaced it by a hcf4053be, now rails are:
  
  160V OK
  5V not OK, measured 5.70V
  -12V OK
  12V OK
  
  Everything is connected again, no additional loads. CRT displays nothing, the filament is red and supposedly working...
  
  Another odd thing that I noticed is R635, it is way off. Gonna replace all these 1% resistors and capacitors on the region to see if it gets better.
• #29 Reply
  Posted by floobydust on 15 May, 2018 20:03
• Now the 5V rail is high? The 'LS TTL parts don't like over 7V and the 4000 CMOS to 15V. I'm not sure how to look for the reason, but it might explain why R1021 exists if the +12V or +160V can lift up the 5V rail. Or the regulator is not working. Voltage drop across R1026 will tell you the total current drain on 5V.
  
  If there's no sign of a trace, first I would look for the HV DC-DC signs of life. If Q2001/IC2001 are oscillating. Voltages on IC2001 would help.
  Next, I would measuring the CRT voltages (not HV) deflection plates etc.
  
  Is R635 off value in-circuit or out? Seems weird for resistors to fail there.
• #30 Reply
  Posted by huelter on 16 May, 2018 01:20
• R635 was low off-circuit below 20%. Replaced some caps as well but the voltages remained the same.
  
  I socketed all ICs on the second board (IC401, IC601, IC602, IC603, IC2001), then powered the board with no ICs. the 5V rail remains at 5.7V, the same as with all the ICs.
  
  The voltages on Q1004 when the second board is disconnected: base 5.8V, collector 14V, emitter 5.1V; voltage drop on R1026: 6.37V
  The voltages on Q1004 when everything is connected: base -9.15V, collector 19.7V, emitter 5.6V; voltage drop on R1026: 0V
  
  If I understand correctly, something is shorted in the second board, and it is not one of the ICs, which completely shuts off the 5V rail?
• #31 Reply
  Posted by floobydust on 16 May, 2018 21:50
• Because load R1021 is on the output of the +5V regulator, the current from other rails returning into the +5V rail normally doesn't bump up the 5V.
  But if you have no R1021, then there would be enough current to push the rail to 5.7V
  
  The only place I can see lots of current possible backfeeding is from the vertical and horizontal deflection amp output stages.
  
  Horiz. deflection amp at Q809/Q810 5V on the emitters and 160V on Q811/Q813, or if Q807 was on hard.
  Vert. deflection amp at Q307/Q308 5V on the emitters.
  
  So if there is still a problem in the deflection amps, they could push current back into the 5V rail.
  
  Which is the second board? At CN?
  Q1004 emitter voltage seems high, assuming R2008 is OK at 1R. But it would not affect the 5V issue.
• #32 Reply
  Posted by huelter on 17 May, 2018 12:48
• It is hard to say what that second board has, because the schematic doesn't follow. I think it has the vertical and horizontal deflection, 2KV high voltage circuitry, timebase selection, triggering selection.
  
  I checked every transistor on the second board. I found out that Q807, Q808, Q809 and Q3005 are busted. Just like you predicted, on Q807 there is backfeeding!
  
  So now, the hardest transistor to replace is Q807 2sc1907, because its spec has 1.1GHz... Is that really necessary? Q808 had a A838, with only 300MHz and easier to replace (nevermind the schematic, it is saying this PNP is also another 2sc1907, clearly wrong).
  
  Anyway, I swapped Q807 for a H945, Q808 for a BC558 (swapped base and collector legs), Q809 for another D669A and Q3005 for a C2371. The good news is that all rails are within spec now, and everything is connected! The bad news is that there's still nothing on the CRT.
• #33 Reply
  Posted by floobydust on 17 May, 2018 20:47
• 2SC1907 NPN UHF transistor 30V fT 1,100MHz popular with Asian scopes. Substitutes might be: SS9018 or 2N5770 but both have EBC pinout. Both at Digikey.
  
  Next would be seeing if the HV DC-DC is running, and then measuring the CRT voltages (X+,X-, Y+, Y-) and CN2006 astigmatism.
  The cathode and control grid, screen grid are around -2kV so no DMM there, although you can make a cheap HV probe voltage divider with 1/4W or 1/2W 10MEG resistors, if you include the DMM's input resistance for the ratio. Or a neon lamp in series with some 10MEG resistors as a poor mans indicator. It's ok to push 1kV through a 1/4W resistor for a few seconds.
  
  Q3006, Q3005 voltages could be checked, to see if the blanking is not stuck on or intensity shut off.
• #34 Reply
  Posted by huelter on 17 May, 2018 21:25
• Thanks, will look into replacing Q807 into a proper one.
  
  Now, There are several diodes that are busted as well!
  
  D433, 1S1588, high speed
  D819, 1SS86, schottky UHF
  D3024, BAV21, Vrrm 250V
  D3013, 1SS83, Vrrm 300V
  
  Those are all old diodes, do you have any recommendations for replacements?
  
  Edit: gonna replace them for:
  
  2SC1907 -> same (found one!)
  1S1588 -> 1N4148 (fast switching)
  1SS86 -> BAT83 (low capacitance schottky)
  BAV21 -> same
  1SS83 -> 1N4005 (general purpose)
  
  Edit2: replaced everything that was faulty. Here are some measurements:
  
  Q3005: E=57.7, C=156.6, B=58.6 (active)
  Q3006: E=148, C=156.6, B=15.4 (inactive)
  Q3007: D=2.3, G=0, S=0.8 (active)
  Q3008: E=2.2, C=0, B=2 (still inactive?)
  VR3002 = 420 ohms
  VR3031= 2560 ohms
  D637, IC side = 5V
  D637, unblank signal = 0.4V
  
  I find really odd what is happening on Q3008. The Intensity adjustment VR3002 is set at 420 ohms, setting a collector voltage on Q3008 of about 2.7V. But instead I see 0V. Then there is VR3031, which would set an emitter voltage of about 2.7V, and instead it is 2.2V. I checked the region, found no shorts to ground. Have no idea what is going on.
• #35 Reply
  Posted by floobydust on 19 May, 2018 07:15
• I'm not sure the best way to see if the beam is cutoff (blanked) all the time- if you can't measure the cathode and control grid voltages.
  You could try short the cathode and control grid, turning the CRT full on, which should be easy to see.
  On CN2005 jumper pin 3,4 (=CRT pins 3, 2) with a 100k resistor and turn on power and see if it lights up. Note the connector has -2kV so I would first power off the scope and discharge the CRT by grounding the cathode through a 100k resistor for a few seconds. Usually small CRT's don't bite if you power off and wait a few minutes.
  
  If you still have no display, then the HV is not there, or the trace is aimed at outer space, from a problem on X,Y deflection plate voltages.
  
  The beam blanking circuit is tough to work on, as the output floats at -2kV yet 180V-rated transistors are doing the switching.
  I'd have to be really awake to explain it.
  
  

  ---

  
• #36 Reply
  Posted by huelter on 22 May, 2018 03:17
• Jumper 100K ohm on CN2005 pins 3,4: nothing on screen.
  Measuring -2KV with a /6 voltage divider: 0V.
  
  Testing D2010 and D2011 for operation: blown one with a 50ma current. Damn these 28V~40V forward drops!
  
  Any current replacements for these GHV06SSN diodes? 7KV peak Vrrm. I was thinking about putting some 12KV Vrrm 2A diodes used in microwaves.
  
  Are those neon lamps NE3001 and NE3002 for overvoltage protection, or for high voltage indication?
  
  Edit: replaced IC2001 and the diode. Now we finally have a dot on the screen! 
  
  Now, the dot is shifted to the right. There might be a problem with the horizontal deflection circuit, right? Adjusting VR413 and the horizontal position knob doesn't move the dot.
• #37 Reply
  Posted by huelter on 23 May, 2018 18:15
• Replaced IC603. Now the dot is closer to the center of the screen, but not quite. Moving the horizontal position knob makes the dot move in the reverse direction, quite odd (clockwise moves it to the left). VR413 has not enough room for adjustment, so I cannot center it.
  
  I noticed that the vertical knob has no effect. The scope is in XY mode, but if I switch to another timescale, the dot remains unmoved. Don't know what to look at now...
  
  ICs that were replaced so far:
  IC401
  IC601
  IC602
  IC603
  
  No more ICs to be replaced on the second board, and the dot is still there on XY mode, but there is no more dot on any other timescale. I guess the sweep is faulty?
  
  Edit: using the signal generator from the calibration output, I was able to see the signal at CH1 OUTPUT with a second oscilloscope. Following the input, the signal becomes really weak at Q7 emitter, but at Q5 collector it is fine.
  
  Edit2: found the culprits. Q201 and Q202 are busted; any good 2sa711 replacements? I'm thinking about 2n3906, cheap and easy to find, but a little bit worse on the specs.
  http://pdf.icpdf.com/PdfOld/icpdf_datasheet_1/2SA71_datasheet_10346/171793/2SA71_datasheet.pdf
• #38 Reply
  Posted by floobydust on 27 May, 2018 19:43
• I would troubleshoot the H or V deflection systems separately in X-Y mode and get the V-amp working first, for both channels.
  
  Vertical deflection, you should be able to get a vertical line if you are injecting a signal like 1kHz probe cal or AC hum. A sine wave makes more of a line, a square wave makes top and bottom dots with not much in the middle.
  
  Replacing the transistors, the scope's alignment is probably out so the trace (dot) is not centered. There are trimpots as coarse adjustments to the front panel knobs. Like VR48/VR153 V-POS; with front panel knob centered, set VR48/VR153 for a centered dot. Or adjust it to keep the CRT V+ and V- voltages the same. You replaced many transistors in the diff amps so the balance trimpots may need to be adjusted, like VR147, VR148 too. Don't let the dot get really bright or you can burn the CRT phosphor.
  
  It's a bit confusing because there is one beam with one set of (vertical) deflection plates. For a two-channel display there are two vert amplifiers but one vertical deflection amp (driving the CRT) that is multiplexed by the alt/chop oscillator. i.e. in ALT mode it switches between Ch. 1, blanking, Ch.2, blanking and repeat again, with IC202.
  
  2SA711 (die alone is NEC NE71111) is PNP RF TO-18 fT=1GHz 40V 0.3W hFE=120 and run at higher collector currents here for speed.
  Equivalent parts are https://vetco.net/products/nte106-pnp-transistor-si-rf-if-amp?taxon_id=1872 and harder to find metal can RF transistors. I think a TO-92 part might run a bit warm but you could try it to get things going.
  
  The horizontal oscillator is more difficult to troubleshoot. You have to know if the H-deflection amp, sweep oscillator, or the trigger sections are working.
  I could not follow CN602/CN404 (Y-input to H section) on the schematics.
  
  I suggest getting the vertical sections working, as that can help troubleshoot the H sections.
• #39 Reply
  Posted by huelter on 30 May, 2018 01:36
• OK, using the scope on XY mode and on GND coupling, I'm able to see two dots on the screen now.
  
  Trying to center the dots vertically first, I adjusted VR48/VR148, but there is not enough room to reach the center of the screen. Tried replacing R49 with a lower value resistor, but now the front panel knob has even less room.
  
  A strange thing that I noticed is that adjusting the Y position panel knob makes the Y+ voltage vary wildly from 40V to 140V, but the Y- voltage stays the same at 150V. This happens on both channels. Maybe there's some leakage on the Y+ rail?
• #40 Reply
  Posted by floobydust on 30 May, 2018 19:32
• I think there's too much offset (difference) between the Y amps. The DC balance trimpots can help, but not fix a problem.
  The Y- can be high or the Y+ can be low, to make things off center (down). I think the Y- at 150V is high if there's only 160V available?
  
  The deflection amplifiers are identical (i.e same circuit for X+, X-, and same circuit for Y+, Y- (both channels up to alt/chop mux).
  So you can take comparison readings when the beam supposed to be right in the center.
  
  I would go back to the input stage and compare readings, to see if the imbalance starts there or is downstream.
  IC102 driving Q105 and Q106,IC2 driving Q5 and Q6, should be about equal output voltages. This applies for each stage, going to the CRT.
  A lazy transistor or bad resistor might be to blame.