Adrian: I used to design video drivers for CRTs back in the 1970s. The chief problem one faces is the slowness of response of the output transistor. In order to minimize parasitic collector to base capacitance and eliminate its multiplication by what is called the *Miller effect*, a technique is used, The output of the transistor collector is to the tube and has an inductor as well as resistor to rail to sharpen the response, but the base is grounded, and the input goes to the EMITTER.. This is called a *common base*, or *cascaded* configuration. We used it to multiply voltage, but not current. In essence two transistors are used to make one, with overall lower parasitic capacitance and higher voltage capabilities than either alone.. The old vacuum tube design of this was called a 'cascode' . I think it might have been from 'cathode to anode' or maybe 'cascaded cathode'. If you haven't seen this arrangement before, you will get puzzling voltages. There will be no signal on the base, and very little voltage change on the emitter either. Because what is changing is the CURRENT. I can't describe this well in words. But googling common base or cascade schematic should bring up some examples that make it all clear. It is a technique that one only finds when transistors are being pushed to the limit of what they will do. I met it first in driving a head up display CRT for fighter aircraft in the 1970s. The video signal needed to go to 75MhHz I think, and that was really hard with the transistors of the day. Whereas a typical TV video signal is only 5-8MHz. That we could do easily. 75MHz was the bleeding edge and I never quite made it, but my boss took what I had done and added some emitter followers to it and made the spec. Just.
Egad, man. You've brought back some memories here of working on the SPA-4 Plan Position Indicator. It had the "cascoded" arrangement you describe here. (I haven't touched one of these old beasties since I left "A" School.)
Cascodes also made great voltage amplification stages in audio amplifiers, for much the same reason. Later on transistors made for CRT video amplifiers were perfect for the job (high voltage, small current, fast, low Cob). All sadly obsolete now.
Well put! It's nice to hear from the electronics veterans. I don't remember learning about Miller capacitance as an EE student... although those years are a bit blurry :D
@@thirstyCactus Its alarming how little electronics I did learn from my degree. For sure we did the quantum theory of transistors and the existence of parasitic capacitance, but to understand pragmatic ways of dealing with it - the DIY magazines like Wireless World were better. As you say those years (1967-1978) were extremely blurry. If you can remember it all you weren't there, as they say!
@@BrickTamlandOfficial Assuming a CRT could have feelings you'd think the prospect of getting fixed would make it happy. Adrian rarely threatens his machines with the e-waste pile.
I've fixed a bunch of tube guitar amps, and I can tell you that many many times I've unplugged the tubes, plugged them back in again and the amp comes to life. Those pins get oxidized.
We use these things called pre concentrators that concentrate organics in air samples via flowing the air through a medium that's sitting in liquid nitrogen. As you can imagine, a lot of condensation and straight up icing forms around the traps which ideally drain but it still gets moisture around the electronics. Our weekly preventative maintenance involved pulling the cards out and reinserting them to remove any oxidation
I am convinced there was some sort of cold solder joint or something on the neck board. These types of PCBs are notorious for that. Back in the day when I troubleshot the issue to a particular circuit, I would reflow all of the components in that circuit.
Solder joints, dirty pots, oxidized tube pins, bad crimps on connector pins, another thing I've seen is condensation in/under/around the CRT socket or in the screen grid circuit, that's all it takes to shunt the screen voltage to ground and kill the picture.
I gotta say it's amazing the amount of effort you went to to show the audience what you are doing. The editing of this video could not have been easy. Great video Adrian.
13:00 When you touched the neckboard it moved. "Glue did nothing so let's pull that off". If the Neckboard was loose from be jostled, maybe one of the pins wasn't making a good connection or was oxidized?
It's been well over 25-years, since I had my analog classes at our local community college, but I recall our professor mentioning the variations in analog circuit design, & how they could accomplish a desired result, with several different approaches. For me, I liken this to alternate fingerings on a musical instrument, to get the same note or chord. I think this experience in your video also demonstrates the challenges, for anyone working on old electronics, without the service documentation (& sometimes replacement parts). In the absence of that documentation, reverse engineering a circuit prior to troubleshooting is probably your next best option. As always, thanks for your efforts & time spent making these videos. I enjoy your content, perspective, & approach to all things electronic. :)
i was a tv repairman in the 90thies. self healing problems are always comming back. i would knock with the back of a screwdriver on different parts of the board and wiggle the cables to search for broken joints/connections. i recommend to go from the end of the signalpath to the start. i would not disconnect the neckboard. i used a mirror so see the screen while working on the back, because sometimes unexpected stuff happens or i had to press a button and wanted to be sure what im doing. Addendum: Looking at the screen while measuring could also help identify thermal problems.
Sometimes for the mysterious process of self-healing to be instigated, machines just need to know they’re in a safe space and that they’re with someone who understands and cares about them. Adrian seems to be excellent at facilitating this🙏🤷♂️🕯️
It is more common than you expect: repair services have this problem all the time, sometimes just the movement of the device to the repair shop and change in temperature fix the problem temporally making it difficult to spot. A classical percussive treatment and a few days turned on can help but not always. Sometimes is electric, some capacitors self heal after being used for instance. It is particularly common in old semiconductors.
The problem is, if you don't know why the problem went away, it WILL come back again. And you STILL won't know how to fix it. I absolutely hate intermittent faults for this reason. "Yay! You fixed it!" "No, i didn't, it just started working. It isn't fixed, this problem will come back. But until it does, i cannot continue to troubleshoot."
Ee mit freundlichen Grüssen ihr Team vom Bahnhof abholen kannst du kannst mich gerne mit ihnen einen schönen Abend noch die Rechnung für die Info für dich okay wäre es möglich dass sie sich mit der Bitte die mit freundlichen Grüßen
Remember open-collector inverters sink current vs. sourcing it. So the outputs have to be externally pulled up, and when the TTL input is high, it pulls them low. Q402 and Q401 are mostly for contrast control. Bias on Q401 is set by the contrast knob, and that turns Q402 partially on but only when the video inverter is sinking current. When the V input is high, the cathode voltage is pulled down through Q402 and the inverter, turning the beam on. When the I input is high, the I inverter sinks current and grounds the emitter of Q401, which turns Q402 fully on and grounds the cathode for maximum brightness (but only if the V input is also high)
There is a bad solder joint next to Q401, follow the trace going right from the base pin, and to the pad that is left of the R411 silk screen, and you can see two pins that appear to touch, connecting the Collector and Base together (probably just perspective) but the pad connecting the trace to the collector looks like it's a cruddy cold solder joint. Maybe working in that area to remove and reinstall Q401 reflowed it?
Good catch! Indeed, the connection from the collector of Q401 to R411 looks like a cold joint (on the resistor, not on the transistor). As the trace between them is very short, un/resoldering Q401 might have reflowed that enough to make it work. A lot of the solder joints look bad in the picture but this one in particular looks like it did not wet the resistor pin correctly (32:39).
@@EdwinSteiner it may also have not reflowed it enough and it could fail again in the future. Maybe he could unsolder that leg of the resistor and see if it recreates the original fault?
There is a possibility that q401 was installed wrong at the factory. He removed it and tested it and reinstalled it. So he could of put it back in the correct way.
The base of the video driver transistor has a capacitor to ground. This would conduct any video to ground so the the base voltage of the transistor is a constant DC level. If that capacitor was shorted or leaking, it would have pulled the base DC toward ground and stopping the transistor from passing video to the CRT cathode. I would suspect the capacitor as the monitor was sitting unused for such a long time. Thanks for the great videos.
Agree on that, "L" should be an inductor coil. But also the poking test should have made it fail again (unless the inductor was not poked hard enough)..
I love the CRT repair videos and thanks making me realize that I was confused about transistors, I thought the emitter emitted the amplified signal! LOL.
I or someone else may have mentioned this in a comment on the last video, but this brand still exists and is a manufacturer of flat panel and curved displays.
Could have been oxide on the CRT pins (or the socket). Built up over years. Once you removed the neck base, you scraped the oxide off and then it worked. Maybe use a little deoxit on them.
Adrian, it's possible that the transistor was not soldered correctly. Removing it and then resoldering it seemed to solve the problem. Reflowing the solder points for each component might also help.
The voltage on the base of the cathode drive transistor is probably a supply rail that also powers the collector of the small transistor. The small transistor is an amplifier for the DC signal of the contrast pot. So the small transistor isn't even driving the base of the big one.
6:29 It’s actually the grids like G2 (screen) that has a positive charge that attracts the negatively charged electrons released from the cathodes. This is making the electrons shoot forward and through the screen grids and then finally hitting the fosfor on the inside of the tube. It’s not the HV second anode that attracts the electrons first. The HV second anode coating inside the tube takes care of the electrons right after they hit the fosfor to avoid them from further bouncing around inside the tube, creating a cloudy image otherwise.
Btw, never run a CRT device with the neck board disconnected, it could easily destroy the tube. There will be some leakage current from the HV anode to the cathode/grid which has nowhere to go, and can/will eventually spark over to somewhere, possibly causing a micro-leak in the glass if it doesnt hit something electronic. Monochrome sets with low-ish voltages often dont arc but especially later, large colour tubes do. If you have to leave the neck board off for some reason, wrap wire around all pins and ground to the CRT ground.
I'm not so good at analogue electronics either. A transistor is easy enough to understand, but it gets confusing when it's the PNP variant rather than NPN, or when it's used in "unusual" ways. Anyway, a thought came to me. Some people have mentioned that capacitors stop functioning properly when they haven't been subjected to voltage/use for a long time. It has something to do with the oxide layer inside. I noticed there's a capacitor in the circuit at the base of that transistor. Maybe tinkering with the circuits, making measurements and subjecting components to heat made this capacitor repair itself. I have no idea, but it was a thought. Congrats either way, on making the thing work again.
On the IBM monitor only a single inverter is used for the intensity and video signals, but on this one it appears they have connected three of the inverters in parallel for each of the signals. Perhaps the designer abhored leaving bits of that IC unused, or they wanted greater drive. In any case, that's an open collector inverter so you can parallel the gates. By pulling the neck board off and replacing it you cleaned a connection to the CRT socket and Viola! it worked?
Its possible the signal was being shunted to ground somewhere. Zieners can have some odd failure modes. But its very possible the a small solder blob made it from the wave solder. Moving board around board knock it off. Zieners are resistors until they reach their breakdown voltage. Sometimes they are faulty and breakdown sonner. This drops voltage lower than what they were designed to control.
Did you check out the functioning of the 7406? This is not a 'standard' Inverter TTL logic chip like the 7407. The output is Open Collector (Buffer/Driver) so you won't just get a logic voltage on its output, it needs a resistor or something it can drive current into. (Up to 30V)
i thought that earlier Adrian knew that was an open collector IC (yep 17:05), but at 45:07 he is puzzled by the changing voltage at the output of the chip.
I have a Dell crt monitor from the early 2000s where the brightness and the color go in and out. If I tap on the top of it sometimes it will go back to the correct brightness. Sometimes I get a blue tint and sometimes a pink tint. I suspect a loose connection. This video makes me think of it.
For all of us lovers of retrocomputing, I'd recommend watching the film/documentary named "The Rubber-keyed Wonder". It's about Sir Clive Sinclair and the ZX.-Spectrum, but not only. You too Adrian! 😉. A not very well known home computer in the US. It's a piece of history I'm sure, you'll all enjoy, specially those who had one.
I think I have read somewhere that sometimes heating a failing transistor can "revive" it. Maybe the act of unsoldering and resoldering it has "fixed" it.
Not done with the video yet, but this is fascinating. New-Old-Stock can 1.) just be an artifact that was ALWAYS defective, sometimes new products needed to be RMA-ed and never were! 2.) something aged poorly wherever this was stored in whatever conditions, causing a problem. 3.) it might have been New-Old-Stock, but that doesn't mean it wasn't kicked around, damaged in the years through many moves through different storage places, perhaps very unceremoniously once it was obvious it was going to be un-sold merchandise.
Adrian you have encountered similar errors on Apple monitors where you made it a routine to re-solder several components on the PCB. The phenomenon is called a "cold solder joint" which can be very difficult to see. Usually you can see a thin dark ring around the component leg in the middle of the solder joint. I learned in technical school that you were not allowed to cut the component leg after it was soldered because the shock led to the phenomenon of "cold solder joint". If the company has had their components in storage for a long time, the component legs become oxidized, causing you to get a nice looking solder but a cold solder joint with a bad connection. The flux in the solder has not been able to break down the oxide layer because it is only active in the short time the tin is hot enough. This is why new old stock is unreliable, it requires more finesse to solder correctly. When I watch your videos, words of wisdom from the past teachers pop into my head and I am glad that I had some wise people around me when I was learning the correct way to do things. I am impressed by your ability to analyze your way to a solution and make things work. Please keep up the good work, it is both entertaining and educational for many viewers including myself.
With lack of pucture issues, probe the neck of the tube first and work backwards in the circuit. I’ve fixed more issues than i care to remember by just reflowing a section. If you’d taken the swivel base off you would have enough wire to flip the main chassis up - these old monitors were made to be field serviced unlike modern rubbish which they just want you to throw away.😊
Finding a fault that isn't there is tricky! Good video. Perhaps a mirror on the back of the bench might help? Maybe a shaving mirror - if you have one?
If I had to foster a guess I'd say that some of the capacitors had lost their forming layer. Over the process of "fixing" the monitor the capacitors reformed allowing normal function. Beyond that I didn't see anything that might have fixed the problem.
Perhaps the magic self healing you seem to encounter often in your repair videos is just simply the caps reforming. Old, but still good caps may need to reform to start working properly. The same for old, good, but tired caps. This process takes a little time. Some caps may take longer then others. So one part of a circuit may be working but not another. At least for a while. All the power on's, power off's you do helps this reforming of the caps. This is probably why this happens. You should look into this phenomenon.
Potentially either a dry joint you accidentally fixed, or *perhaps* you "reformed" one or more capacitors by simply switching the unit on and off repeatedly.
That’s what I was thinking as well, I‘ve seen this happen (with used monitors, but that have seen long periods of being stored without being powered on)
@@peterhanraets Especially since it had never been turned on before. Still pretty dry probably so probably still a good idea to change the ones that are under stress.
I hate when crap fixes itself😂😂 and your analysis is correct, there was no signal out of the collector so... I would have probably look for MDA monitor schematics on Google that matches the input stage directly to the 74ls06 it was? So you have a closer to real schematic to work on, I love your repair videos!, Thank you
Ah yes, the monitor fixed itself, sorry about that. I it must have been me sending requests to the universe to fix all the old monitors so you'd be able to work on more actual retro computer stuff instead of fixing another monitor. Too bad the charm started working a tad bit too late. Must be the delay between the video comeing up on youtube and you already working on the next. 😂 Thanks for all your great videos. Please keep on doing what you like and don't listen to haters like me whining. 😊
Sometimes just jolting a device or parts around will reconnect something that had degraded over time. Taking it apart and putting it back together probably reconnected something.
Who loves CRTs more than Adrian? I just finished vacuuming out and cleaning and successfully testing a 12" toshiba crt in a recent video. After all the hard work it actually just powered up and worked. However after all my work it still stunk of animal piss which I knew it would. Still it powered up one last time.
I found myself yelling to my monitor "that's a common base amplifier." Single sided PC boards are notorious for having cracked solder joints that are not always visible.
Not quite. An emitter follower is where the signal goes into the base and the output is on the emitter, with the collector tied to a fixed voltage. This is also called a common collector circuit, because the collector is common to the input and output. When the signal goes into the emitter, the base is tied to a fixed voltage, and the output is on the collector. This is known as a common base amplifier, and is mainly used for low-impedance, high bandwidth circuits. The circuit with the two transistors in the schematic is a cascode circuit. The first transistor is a common emitter stage, from which the output on the collector feeds into the emitter of the second stage. Since the base of the second transistor is at a fixed voltage, the emitter is the same, and this doesn't vary with the signal. The point of this is that with the base at a fixed voltage, the voltage swings on the collector doesn't couple capacitively to the base, and conversly the collector of the first transistor doesn't have a swing, so there's nothing to feed back into the base. This results in a circuit with high bandwidth and useful voltage and current amplification.
That looks like something added by an external program. The lines are over the entire screen. Microsoft Powertools has a feature that can do it (mouse crosshairs)
stuff fixes itself, miracle! :) again some contact errors. fixed by replugin. you did remove the driver back cap and put it back later. too. you did bonk it, and it fixed it. by magic. unknown bits.
Now I'm wondering how color crts work. I always thought there was only one electron gun, that turned off and on when it got to the different subpixles. But now, I see theres three electron guns, one for each color. Now I'm very confused. The deflection yoke is fed by a clean saw tooth? Theres no steps?
Could one of the contrast knob connector pins have been at high resistance, and pulled the contrast (wiper) all the way one way (+ or 0V ref, whichever didn't have high resistance)? With the plug totally disconnected the wiper would float to maybe a reasonable value?
My guess is that the problem was the solder joint on R411 that looks fishy. When soldering the transistor it reflowed the R411 joint to a point that it works. R411 needs a proper resoldering.
When you get the win of fixing something, but don't know why it works again, that's the most irritating of repairs, it's basically the "Unplug it and plug it back in" fix, very annoying, but at least it's up and running now... :D
Might I recommend you create title cards with your disclaimers on them for display as needed? According to the manual for the Zoyi ZT-703S, it's scope input is good for 150 v at X1, 300 v at X10.
putting my bets on a tin whisker. by handling the PCB the tin whisker got remoded by accident during probing. Alternatively a capacitor which took really long to reformat itself.
Not likely. The solder used both for soldering the boards and for factory-tinning the component leads would have been leaded back then, and it’s not really susceptible to whisker growth.
26:40 probably to increase rise time, or if they only needed 2 drivers, use them all in parallel to prevent them from interfering/floating/oscillating etc. I think I read an article in Elektor way back that some tin can transistors (usually germanium) can form conductive filaments with age, used or new/unpowered. These can self heal when used. It could be one of those cases and might explain why at first it didn't work while the filaments were dissolving and started to work. It just might've been enough to leave the monitor running for a couple of hours and it would pop back to life. The monitor is unused so unless there's oxidation due to high humidity storage conditions, cracked PCB, dry joints etc issues that occur during use are unlikely. Not sure if the same can happen with electrolytic capacitors might be able to reform when not in use for a long time. My first suspect was the tin can transistor when I saw in in the footage. But it's hard to fault find/pin point if it self heals something else was going on. That's life I guess, of electronics repair.
Dang, these are the worst jobs because you know it will fail again.. some day. I don't think I've ever had a situation where something fixed itself. Typically, the problem comes back after I put the final screw back in the enclosure, or try to demo the thing to someone else :D
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
Adrian: I used to design video drivers for CRTs back in the 1970s. The chief problem one faces is the slowness of response of the output transistor. In order to minimize parasitic collector to base capacitance and eliminate its multiplication by what is called the *Miller effect*, a technique is used, The output of the transistor collector is to the tube and has an inductor as well as resistor to rail to sharpen the response, but the base is grounded, and the input goes to the EMITTER..
This is called a *common base*, or *cascaded* configuration. We used it to multiply voltage, but not current. In essence two transistors are used to make one, with overall lower parasitic capacitance and higher voltage capabilities than either alone..
The old vacuum tube design of this was called a 'cascode' . I think it might have been from 'cathode to anode' or maybe 'cascaded cathode'.
If you haven't seen this arrangement before, you will get puzzling voltages. There will be no signal on the base, and very little voltage change on the emitter either. Because what is changing is the CURRENT.
I can't describe this well in words. But googling common base or cascade schematic should bring up some examples that make it all clear. It is a technique that one only finds when transistors are being pushed to the limit of what they will do. I met it first in driving a head up display CRT for fighter aircraft in the 1970s. The video signal needed to go to 75MhHz I think, and that was really hard with the transistors of the day. Whereas a typical TV video signal is only 5-8MHz. That we could do easily. 75MHz was the bleeding edge and I never quite made it, but my boss took what I had done and added some emitter followers to it and made the spec. Just.
Egad, man. You've brought back some memories here of working on the SPA-4 Plan Position Indicator. It had the "cascoded" arrangement you describe here. (I haven't touched one of these old beasties since I left "A" School.)
Cascodes also made great voltage amplification stages in audio amplifiers, for much the same reason. Later on transistors made for CRT video amplifiers were perfect for the job (high voltage, small current, fast, low Cob). All sadly obsolete now.
Well put! It's nice to hear from the electronics veterans. I don't remember learning about Miller capacitance as an EE student... although those years are a bit blurry :D
@@thirstyCactus Its alarming how little electronics I did learn from my degree. For sure we did the quantum theory of transistors and the existence of parasitic capacitance, but to understand pragmatic ways of dealing with it - the DIY magazines like Wireless World were better.
As you say those years (1967-1978) were extremely blurry. If you can remember it all you weren't there, as they say!
Ok, Adrian, here's your book title: The Art of Getting Electronics to Fix Themselves
Guess he has the IT aura. I have that too. Basically people have issues and when you get there, it works now.
The monitor got so scared when it realized it was going to adrian's digital basement that it just fixed itself.
@@BrickTamlandOfficial Assuming a CRT could have feelings you'd think the prospect of getting fixed would make it happy. Adrian rarely threatens his machines with the e-waste pile.
@@moonrock41 the soldering gun hurts the most
Even though you didn't find the problem (yet), I still found it very informative and entertaining :) Love these kind of videos!
Showing the picture of the board as you are referencing points is pretty nice.
I've fixed a bunch of tube guitar amps, and I can tell you that many many times I've unplugged the tubes, plugged them back in again and the amp comes to life. Those pins get oxidized.
@@messmer777 knee high to "have you tried turning it off and back on again?"
We use these things called pre concentrators that concentrate organics in air samples via flowing the air through a medium that's sitting in liquid nitrogen. As you can imagine, a lot of condensation and straight up icing forms around the traps which ideally drain but it still gets moisture around the electronics.
Our weekly preventative maintenance involved pulling the cards out and reinserting them to remove any oxidation
I am convinced there was some sort of cold solder joint or something on the neck board. These types of PCBs are notorious for that. Back in the day when I troubleshot the issue to a particular circuit, I would reflow all of the components in that circuit.
Solder joints, dirty pots, oxidized tube pins, bad crimps on connector pins, another thing I've seen is condensation in/under/around the CRT socket or in the screen grid circuit, that's all it takes to shunt the screen voltage to ground and kill the picture.
I think this is what physicists refer to as 'spooky action at a distance'. Never knew engineers experienced this too!
Only an engineer would say "Unfortunately, it works now."
I gotta say it's amazing the amount of effort you went to to show the audience what you are doing. The editing of this video could not have been easy. Great video Adrian.
I don't know how you do it, but you make electronics repairs so interesting. Thank you so much for the video.
13:00 When you touched the neckboard it moved. "Glue did nothing so let's pull that off". If the Neckboard was loose from be jostled, maybe one of the pins wasn't making a good connection or was oxidized?
I really do love how you spend the time working the stuff out, I feel like I'm right there next to you. ❤
It's been well over 25-years, since I had my analog classes at our local community college, but I recall our professor mentioning the variations in analog circuit design, & how they could accomplish a desired result, with several different approaches. For me, I liken this to alternate fingerings on a musical instrument, to get the same note or chord.
I think this experience in your video also demonstrates the challenges, for anyone working on old electronics, without the service documentation (& sometimes replacement parts). In the absence of that documentation, reverse engineering a circuit prior to troubleshooting is probably your next best option.
As always, thanks for your efforts & time spent making these videos. I enjoy your content, perspective, & approach to all things electronic. :)
i was a tv repairman in the 90thies. self healing problems are always comming back. i would knock with the back of a screwdriver on different parts of the board and wiggle the cables to search for broken joints/connections.
i recommend to go from the end of the signalpath to the start. i would not disconnect the neckboard.
i used a mirror so see the screen while working on the back, because sometimes unexpected stuff happens or i had to press a button and wanted to be sure what im doing.
Addendum: Looking at the screen while measuring could also help identify thermal problems.
My dad was a TV repairman as well. He often would bring a TV from the customer to his workshop and just the small car ride would make it work again.
You are the King of Self-Healing devices. I don't like not knowing what a fault really was either.
Sometimes for the mysterious process of self-healing to be instigated, machines just need to know they’re in a safe space and that they’re with someone who understands and cares about them. Adrian seems to be excellent at facilitating this🙏🤷♂️🕯️
It is more common than you expect: repair services have this problem all the time, sometimes just the movement of the device to the repair shop and change in temperature fix the problem temporally making it difficult to spot. A classical percussive treatment and a few days turned on can help but not always. Sometimes is electric, some capacitors self heal after being used for instance. It is particularly common in old semiconductors.
The problem is, if you don't know why the problem went away, it WILL come back again. And you STILL won't know how to fix it. I absolutely hate intermittent faults for this reason.
"Yay! You fixed it!"
"No, i didn't, it just started working. It isn't fixed, this problem will come back. But until it does, i cannot continue to troubleshoot."
@@jeromethiel4323 My boss have a saying about software: problems that go away by themselves, comeback with company.
Ee mit freundlichen Grüssen ihr Team vom Bahnhof abholen kannst du kannst mich gerne mit ihnen einen schönen Abend noch die Rechnung für die Info für dich okay wäre es möglich dass sie sich mit der Bitte die mit freundlichen Grüßen
32:51 The Base of Q401 seems to have a dry solder joint. The actual Base pin seems to be completely lacking solder.
The leg of R411 attached to the collector looks pretty iffy too.
I dunno, Q401 base looks like a normal joint to me
What I don't like is all the corrosion around R410 and R414
@@jwhite5008 Do you think that is corrosion or dry flux?
Remember open-collector inverters sink current vs. sourcing it. So the outputs have to be externally pulled up, and when the TTL input is high, it pulls them low.
Q402 and Q401 are mostly for contrast control. Bias on Q401 is set by the contrast knob, and that turns Q402 partially on but only when the video inverter is sinking current. When the V input is high, the cathode voltage is pulled down through Q402 and the inverter, turning the beam on. When the I input is high, the I inverter sinks current and grounds the emitter of Q401, which turns Q402 fully on and grounds the cathode for maximum brightness (but only if the V input is also high)
30:58 Maybe a trick of the light, but check out the solder joint at right side of R411 - it looks cracked to me. That trace connects to Q402-base.
There is a bad solder joint next to Q401, follow the trace going right from the base pin, and to the pad that is left of the R411 silk screen, and you can see two pins that appear to touch, connecting the Collector and Base together (probably just perspective) but the pad connecting the trace to the collector looks like it's a cruddy cold solder joint. Maybe working in that area to remove and reinstall Q401 reflowed it?
Good catch! Indeed, the connection from the collector of Q401 to R411 looks like a cold joint (on the resistor, not on the transistor). As the trace between them is very short, un/resoldering Q401 might have reflowed that enough to make it work. A lot of the solder joints look bad in the picture but this one in particular looks like it did not wet the resistor pin correctly (32:39).
@@EdwinSteiner it may also have not reflowed it enough and it could fail again in the future. Maybe he could unsolder that leg of the resistor and see if it recreates the original fault?
There is a possibility that q401 was installed wrong at the factory. He removed it and tested it and reinstalled it. So he could of put it back in the correct way.
Yes. If you disconnect R411, the original symptom will return.
@@Doom2pro yes, I would worry about that, too
I had repairs like this and the fact that I didn't know why something fixed itself always drove me crazy ^^
Perhaps when desoldering and resoldering, you reflowed a bad joint?
The soldering on R411 and R408 looks kinda fishy.
Looks like those sleeping capacitors reformed after so many years sitting in a box.
The base of the video driver transistor has a capacitor to ground. This would conduct any video to ground so the the base voltage of the transistor is a constant DC level. If that capacitor was shorted or leaking, it would have pulled the base DC toward ground and stopping the transistor from passing video to the CRT cathode.
I would suspect the capacitor as the monitor was sitting unused for such a long time.
Thanks for the great videos.
Thought so too. Maybe cap(s) needed some time to reform.
its amazing how many things fix themselves Adrian LOL
32:19 - Some of those solder connections look a bit dodgy. Note the dark rings around the leads. Look at the upper lead of L401.
Agree on that, "L" should be an inductor coil. But also the poking test should have made it fail again (unless the inductor was not poked hard enough)..
Not really a no-name. I have 6 Scepter LCD Monitors purchased over the last couple of years. They really are very nicely made.
A lesser known or value name, yes. But they did earn market share. Definitely NOT no-name.
I love the CRT repair videos and thanks making me realize that I was confused about transistors, I thought the emitter emitted the amplified signal! LOL.
39:58 this is where you fixed it. Confirmed the pot is fully working. 420 magic bro!
Those solder joints all look like a reflow wouldn't hurt. 🤭
its hard to fix something that don't stay broke but you nailed it
I or someone else may have mentioned this in a comment on the last video, but this brand still exists and is a manufacturer of flat panel and curved displays.
R411 solder joints look pretty suspect. I bet the monitor was a warranty return from a dealer that tried it before they shipped it or bundled it.
Could have been oxide on the CRT pins (or the socket). Built up over years. Once you removed the neck base, you scraped the oxide off and then it worked. Maybe use a little deoxit on them.
Never mind, it’s great to see content on CRT repairs
Well, glad it works again 👍👏
Those amber monitors were the best really when it comes to monochrome with a black background.
Adrian, it's possible that the transistor was not soldered correctly. Removing it and then resoldering it seemed to solve the problem. Reflowing the solder points for each component might also help.
The voltage on the base of the cathode drive transistor is probably a supply rail that also powers the collector of the small transistor. The small transistor is an amplifier for the DC signal of the contrast pot. So the small transistor isn't even driving the base of the big one.
6:29 It’s actually the grids like G2 (screen) that has a positive charge that attracts the negatively charged electrons released from the cathodes. This is making the electrons shoot forward and through the screen grids and then finally hitting the fosfor on the inside of the tube. It’s not the HV second anode that attracts the electrons first. The HV second anode coating inside the tube takes care of the electrons right after they hit the fosfor to avoid them from further bouncing around inside the tube, creating a cloudy image otherwise.
@21:00 my guess is C204, it's always a crapacitor.
Btw, never run a CRT device with the neck board disconnected, it could easily destroy the tube. There will be some leakage current from the HV anode to the cathode/grid which has nowhere to go, and can/will eventually spark over to somewhere, possibly causing a micro-leak in the glass if it doesnt hit something electronic. Monochrome sets with low-ish voltages often dont arc but especially later, large colour tubes do. If you have to leave the neck board off for some reason, wrap wire around all pins and ground to the CRT ground.
I'm not so good at analogue electronics either. A transistor is easy enough to understand, but it gets confusing when it's the PNP variant rather than NPN, or when it's used in "unusual" ways. Anyway, a thought came to me. Some people have mentioned that capacitors stop functioning properly when they haven't been subjected to voltage/use for a long time. It has something to do with the oxide layer inside. I noticed there's a capacitor in the circuit at the base of that transistor. Maybe tinkering with the circuits, making measurements and subjecting components to heat made this capacitor repair itself. I have no idea, but it was a thought. Congrats either way, on making the thing work again.
On the IBM monitor only a single inverter is used for the intensity and video signals, but on this one it appears they have connected three of the inverters in parallel for each of the signals. Perhaps the designer abhored leaving bits of that IC unused, or they wanted greater drive. In any case, that's an open collector inverter so you can parallel the gates.
By pulling the neck board off and replacing it you cleaned a connection to the CRT socket and Viola! it worked?
I can't be the only one who saw POCUS in both spots on the block diagram where it shold have said FOCUS...
Its possible the signal was being shunted to ground somewhere. Zieners can have some odd failure modes. But its very possible the a small solder blob made it from the wave solder. Moving board around board knock it off. Zieners are resistors until they reach their breakdown voltage. Sometimes they are faulty and breakdown sonner. This drops voltage lower than what they were designed to control.
I may have zeiner's backwards. I sometimes do that. But either way they pull to a certain voltage and stop. Sometimes they don't stop.
Did you check out the functioning of the 7406? This is not a 'standard' Inverter TTL logic chip like the 7407. The output is Open Collector (Buffer/Driver) so you won't just get a logic voltage on its output, it needs a resistor or something it can drive current into. (Up to 30V)
i thought that earlier Adrian knew that was an open collector IC (yep 17:05), but at 45:07 he is puzzled by the changing voltage at the output of the chip.
I have a Dell crt monitor from the early 2000s where the brightness and the color go in and out. If I tap on the top of it sometimes it will go back to the correct brightness. Sometimes I get a blue tint and sometimes a pink tint. I suspect a loose connection. This video makes me think of it.
For all of us lovers of retrocomputing, I'd recommend watching the film/documentary named "The Rubber-keyed Wonder". It's about Sir Clive Sinclair and the ZX.-Spectrum, but not only. You too Adrian! 😉. A not very well known home computer in the US. It's a piece of history I'm sure, you'll all enjoy, specially those who had one.
Adrian is a literal miracle worker 😁
I think I have read somewhere that sometimes heating a failing transistor can "revive" it. Maybe the act of unsoldering and resoldering it has "fixed" it.
Near Q402, the pins immediately to the left of "E" and "c" look to have no solder at all. very likely causing at least part of the problem
[58:27] "Unfortunately, the monitor - is - working now..." LOL
Not done with the video yet, but this is fascinating. New-Old-Stock can 1.) just be an artifact that was ALWAYS defective, sometimes new products needed to be RMA-ed and never were! 2.) something aged poorly wherever this was stored in whatever conditions, causing a problem. 3.) it might have been New-Old-Stock, but that doesn't mean it wasn't kicked around, damaged in the years through many moves through different storage places, perhaps very unceremoniously once it was obvious it was going to be un-sold merchandise.
Adrian you have encountered similar errors on Apple monitors where you made it a routine to re-solder several components on the PCB. The phenomenon is called a "cold solder joint" which can be very difficult to see. Usually you can see a thin dark ring around the component leg in the middle of the solder joint. I learned in technical school that you were not allowed to cut the component leg after it was soldered because the shock led to the phenomenon of "cold solder joint". If the company has had their components in storage for a long time, the component legs become oxidized, causing you to get a nice looking solder but a cold solder joint with a bad connection. The flux in the solder has not been able to break down the oxide layer because it is only active in the short time the tin is hot enough. This is why new old stock is unreliable, it requires more finesse to solder correctly.
When I watch your videos, words of wisdom from the past teachers pop into my head and I am glad that I had some wise people around me when I was learning the correct way to do things. I am impressed by your ability to analyze your way to a solution and make things work. Please keep up the good work, it is both entertaining and educational for many viewers including myself.
With lack of pucture issues, probe the neck of the tube first and work backwards in the circuit. I’ve fixed more issues than i care to remember by just reflowing a section. If you’d taken the swivel base off you would have enough wire to flip the main chassis up - these old monitors were made to be field serviced unlike modern rubbish which they just want you to throw away.😊
Finding a fault that isn't there is tricky! Good video. Perhaps a mirror on the back of the bench might help? Maybe a shaving mirror - if you have one?
I think is was the transistor had a cold solder joint and removing and putting it in fixed it.
If I had to foster a guess I'd say that some of the capacitors had lost their forming layer. Over the process of "fixing" the monitor the capacitors reformed allowing normal function. Beyond that I didn't see anything that might have fixed the problem.
Perhaps the magic self healing you seem to encounter often in your repair videos is just simply the caps reforming. Old, but still good caps may need to reform to start working properly. The same for old, good, but tired caps. This process takes a little time. Some caps may take longer then others. So one part of a circuit may be working but not another. At least for a while. All the power on's, power off's you do helps this reforming of the caps. This is probably why this happens. You should look into this phenomenon.
Potentially either a dry joint you accidentally fixed, or *perhaps* you "reformed" one or more capacitors by simply switching the unit on and off repeatedly.
Could it be that some capacitors re formed while the monitor was plugged in for a longer period of time.
That’s what I was thinking as well, I‘ve seen this happen (with used monitors, but that have seen long periods of being stored without being powered on)
@@peterhanraets Especially since it had never been turned on before. Still pretty dry probably so probably still a good idea to change the ones that are under stress.
I have had a flyback clamp become rusted from that bad glue before, rusted right through it. Found it rattling around inside the chassis.
I hate when crap fixes itself😂😂 and your analysis is correct, there was no signal out of the collector so...
I would have probably look for MDA monitor schematics on Google that matches the input stage directly to the 74ls06 it was? So you have a closer to real schematic to work on, I love your repair videos!, Thank you
Sceptre still exists and sometimes they use other names like Ocosmo. One of my favourite cheap monitors I have is an Ocosmo branded one.
A slight possibility is that some tiny conductive piece was in the slot of one of the spark gap capacitors and shook loose.
My theory is that it was a cap that whent high impedance, and running it for a bit reconditioned it, which fixed it.
Ah yes, the monitor fixed itself, sorry about that. I it must have been me sending requests to the universe to fix all the old monitors so you'd be able to work on more actual retro computer stuff instead of fixing another monitor. Too bad the charm started working a tad bit too late. Must be the delay between the video comeing up on youtube and you already working on the next. 😂
Thanks for all your great videos. Please keep on doing what you like and don't listen to haters like me whining. 😊
Am i the only one who thinks it's funny that they used RGB colored wires to drive the neck-board on a monochrome monitor?
Prob same production line parts.
yeah, made me chucle too
@CybAtSteam back in those days they would use all white wires in a pinch. Do love the irony though. Good catch.
Most people accidentally break stuff, you accidentally repair stuff..🤔😁
This happend to me as well, didnt start but after being connected to power for awhile it turned right on
I suspect after sitting in a box for 20+ years that the monitor needed to be turned on and left alone for a while so that the capacitors could reform.
Sometimes just jolting a device or parts around will reconnect something that had degraded over time. Taking it apart and putting it back together probably reconnected something.
Just introducing heat can sometimes do wonders.
Who loves CRTs more than Adrian? I just finished vacuuming out and cleaning and successfully testing a 12" toshiba crt in a recent video. After all the hard work it actually just powered up and worked. However after all my work it still stunk of animal piss which I knew it would. Still it powered up one last time.
Coffee grounds?
Ozone generator?
I found myself yelling to my monitor "that's a common base amplifier."
Single sided PC boards are notorious for having cracked solder joints that are not always visible.
I think you still have a dry solder joint somewhere you were probing around and it will come back after some time.
Try wiggling the nearby components.
Pocus Focus? 😁
That type of transistor circuit where the signal is fed into the emitter is called an emitter follower.
Not quite. An emitter follower is where the signal goes into the base and the output is on the emitter, with the collector tied to a fixed voltage. This is also called a common collector circuit, because the collector is common to the input and output.
When the signal goes into the emitter, the base is tied to a fixed voltage, and the output is on the collector. This is known as a common base amplifier, and is mainly used for low-impedance, high bandwidth circuits.
The circuit with the two transistors in the schematic is a cascode circuit. The first transistor is a common emitter stage, from which the output on the collector feeds into the emitter of the second stage. Since the base of the second transistor is at a fixed voltage, the emitter is the same, and this doesn't vary with the signal. The point of this is that with the base at a fixed voltage, the voltage swings on the collector doesn't couple capacitively to the base, and conversly the collector of the first transistor doesn't have a swing, so there's nothing to feed back into the base. This results in a circuit with high bandwidth and useful voltage and current amplification.
@@RuneBivrin yes - Q402 is common base.
not a cracked solder joint on Q401, but maybe a cold solder joint?
You may have had bad connection to the Cathode on the tube and the incorrect trigger point on the scope when initially testing Q402.
Wonder if a capacitor reformed? How did you get that cool x/y line display in Google Photos at @40:05 ?!
That looks like something added by an external program. The lines are over the entire screen. Microsoft Powertools has a feature that can do it (mouse crosshairs)
My thought was a dry solder joint somewhere.
@@K-o-Ryeah, there was some dirty solder joints in that area. Could have been hiding a bad joint.
What about a cathode-filament short in the CRT that cleared up after heating it up and cooling down a few times?
stuff fixes itself, miracle! :) again some contact errors. fixed by replugin. you did remove the driver back cap and put it back later. too. you did bonk it, and it fixed it. by magic. unknown bits.
Now I'm wondering how color crts work. I always thought there was only one electron gun, that turned off and on when it got to the different subpixles. But now, I see theres three electron guns, one for each color.
Now I'm very confused.
The deflection yoke is fed by a clean saw tooth? Theres no steps?
I think a couple of others have mentioned it, but R408 and R411 solder joints don't look right. R411 in particular looks like it hasn't been soldered.
Again, something loose such as those connectors? Well, we all learning from trial and error, sometimes things just heal itself without reason.
There's a reason. We just don't know what it is.
Could one of the contrast knob connector pins have been at high resistance, and pulled the contrast (wiper) all the way one way (+ or 0V ref, whichever didn't have high resistance)? With the plug totally disconnected the wiper would float to maybe a reasonable value?
My my my... Adrian, 2 things: 1. you have a laying on hands capability, and 2., your oscilloscope probes and soldering iron are magic wands!
Capacitor reforming after all those years?
Corrosion on the variable capacitor and wiggling it fixed it?
My guess is that the problem was the solder joint on R411 that looks fishy. When soldering the transistor it reflowed the R411 joint to a point that it works. R411 needs a proper resoldering.
When you get the win of fixing something, but don't know why it works again, that's the most irritating of repairs, it's basically the "Unplug it and plug it back in" fix, very annoying, but at least it's up and running now... :D
Could the metal shield you removed have been touching something?
I’d love to see you work with a circuit emulator.
Downloading PDF files before opening them makes Firefox render them without delays instead of streaming them over the internet :)
Might I recommend you create title cards with your disclaimers on them for display as needed?
According to the manual for the Zoyi ZT-703S, it's scope input is good for 150 v at X1, 300 v at X10.
putting my bets on a tin whisker. by handling the PCB the tin whisker got remoded by accident during probing. Alternatively a capacitor which took really long to reformat itself.
Not likely. The solder used both for soldering the boards and for factory-tinning the component leads would have been leaded back then, and it’s not really susceptible to whisker growth.
26:40 probably to increase rise time, or if they only needed 2 drivers, use them all in parallel to prevent them from interfering/floating/oscillating etc.
I think I read an article in Elektor way back that some tin can transistors (usually germanium) can form conductive filaments with age, used or new/unpowered. These can self heal when used. It could be one of those cases and might explain why at first it didn't work while the filaments were dissolving and started to work. It just might've been enough to leave the monitor running for a couple of hours and it would pop back to life. The monitor is unused so unless there's oxidation due to high humidity storage conditions, cracked PCB, dry joints etc issues that occur during use are unlikely.
Not sure if the same can happen with electrolytic capacitors might be able to reform when not in use for a long time.
My first suspect was the tin can transistor when I saw in in the footage. But it's hard to fault find/pin point if it self heals something else was going on. That's life I guess, of electronics repair.
Well, germanium transistors were long gone by the time this monitor was made.
Maybe its a case of old hardware being giving care.
imo the solder joints at r411 look cracked right at the leads.
Dang, these are the worst jobs because you know it will fail again.. some day. I don't think I've ever had a situation where something fixed itself. Typically, the problem comes back after I put the final screw back in the enclosure, or try to demo the thing to someone else :D