Need Help with Finding an Amdek V220A Flyback (FMC-1245NL) Replacement

[Hugo Holden]

Ok that is a turn off spot, a different problem. This is happening because after you adjusted the brightness and sub brightness controls, the CRT gun is not cut off as well as it could be at turn off, so for now turn the brightenss down , to help avoid that turn off spot.

The reason that you cannot see anything on the screen with it operating and a video signal applied is that the cathode voltage of the CRT is far too high, cutting off the beam current. We need to find out why that is happening. We obviously don't have any data as to the correct voltages in the video amplifier.

Can you check the voltages with the meter on the Base, Collector and Emitters of the two transistors on the neck board and report those. The fault may be there but it could also be on the main board.

 

[andy]

We know for sure it's not a bad flyback. The problem is most likely in the area of the board labeled "video", or on the CRT socket board. I would start with a very careful examination for bad soldering, then do an in-circuit test of all the transistors in the video circuit (looking for open, or shorted junctions). You could do an in-circuit check of all the resistors in the video circuit. You're looking for major discrepancies, not exact values (eg. a 1k resistor that measures 80k is obviously open). The fact that it started intermittently makes me suspect a failing resistor, or possibly a bad transistor.

If you don't find anything there, I would start from the video input, and trace the signal though the circuit with your oscilloscope. It should be fairly obvious where the signal stops. Set your scope to AC coupling since some of the video stages may have a significant DC offset compared to the level of the video signal, and use a 10X probe.

As a general rule, you always use a 10X probe unless you have a very good reason not to (eg. looking at very low level signals). 10X probes load down the circuit under test a lot less, provide better frequency response, and provide some over voltage protection for the scope.

 

[awewsomegamer]

Ok that is a turn off spot, a different problem. This is happening because after you adjusted the brightness and sub brightness controls, the CRT gun is not cut off as well as it could be at turn off, so for now turn the brightenss down , to help avoid that turn off spot.

The reason that you cannot see anything on the screen with it operating and a video signal applied is that the cathode voltage of the CRT is far too high, cutting off the beam current. We need to find out why that is happening. We obviously don't have any data as to the correct voltages in the video amplifier.

Can you check the voltages with the meter on the Base, Collector and Emitters of the two transistors on the neck board and report those. The fault may be there but it could also be on the main board.

Adjusting VR501 and VR502 to lowest brightness setting

Brightness controls have been set to a minimum, no difference in the intensity or behavior of the spot. It still appears only after turning the monitor off and when the monitor has been left on for about a minute.

A broken picture did come back after leaving the tube on for about 2 minutes. This did not repeat in subsequent testing. The picture was across the full screen, and had the error from H-HOLD being set incorrectly in it, all pixels were the same color, contrast was set to a maximum and brightness controls were at a minimum.

I managed to capture this on video:
View attachment OneOffPicture.mp4

When the monitor did the one off picture, there is the changes in brightness, which was the original issue. That may be coming from VR501 as it showed similar behavior on the scope in post #38.
All pixels being the same color also means that contrast is not being taken into account, despite it being at a maximum; though, this could also be because of the brightness controls being set to their minimum.

Measuring BCE of Q105 and Q106 with the meter

Q105.B: 8.96 VDC
Q105.C: 69.8-70 VDC
Q105.E: 8.44 VDC
Q106.B: 1 VDC
Q106.C: 8.43 VDC
Q106.E: 384.6 mVDC

During measurement, I accidentally shorted the base and collector of Q105, it produced the following on screen:

View attachment ShortQ105.BC.mp4

So Q105 is likely bad.

 

[awewsomegamer]

I have reworked the schematic for the neckboard once more:

 

[awewsomegamer]

I have reworked the schematic for the neckboard once more:
View attachment 1320068

I have also put this into a simulator, and this is what the behavior looks like:

Red is the current through R8 (R124 in real life).
Green is the voltage on Q106's base.

 

[Hugo Holden]

THe reason the screen lit up with Q105's C & B shorted, it simply dropped the CRT's cathode voltage.

It could well be that Q105 is not conducting any significant collector current, along with the bias current (which are orders of magnitude lower) one would expect a higher voltage drop across R122 is low at 0.384 V, suggesting a current of 0.384/68 =5.6mA. That will only drop a voltage of 6.7V across the 1.2k load resistor, so its not surprising the cathode voltage is too high and the picture blacked out. Though from a dynamic perspective it could also mean that the drive at the base of Q106 is inadequate.

Can you test the junctions of both transistors on the meter ?

 

[awewsomegamer]

THe reason the screen lit up with Q105's C & B shorted, it simply dropped the CRT's cathode voltage.

It could well be that Q105 is not conducting any significant collector current, along with the bias current (which are orders of magnitude lower) one would expect a higher voltage drop across R122 is low at 0.384 V, suggesting a current of 0.384/68 =5.6mA. That will only drop a voltage of 6.7V across the 1.2k load resistor, so its not surprising the cathode voltage is too high and the picture blacked out. Though from a dynamic perspective it could also mean that the drive at the base of Q106 is inadequate.

Can you test the junctions of both transistors on the meter ?

All tests were done in circuit.

Q105 B->C: 0.573V
Q105 B->E: 0.611V
Q105 C->E: 0.699V
All other arrangements: 0L

Q106 B->C: 0.625V
Q106 B->E: 0.645V
Q106 E->C: 0.676V
All other arrangements: 0L

I observed the monitor some more, and it seems that when both the user brightness control and VR501 are turned all the way clockwise, the picture reappears more often as it did about a year ago.
Its behavior is exactly the same (the dimming and occasional disappearance) while running.
The turn off spot also behaves differently for no apparent reason. Sometimes it is very bright and stays on the screen for a long period of time, other times it is very short and not very bright, and rarely does it not happen.
I should mention that this turn off spot was not present for the past two years.

During this observation, I was also probing the base of Q106, which is the output of the video circuitry. It had some strangeness, sometimes the amplitude jumped up and down, but this strangeness was independent of a picture being displayed.

I'll have a look at some resistors on the main board.

 

[Hugo Holden]

I have also put this into a simulator, and this is what the behavior looks like:
View attachment 1320075
Red is the current through R8 (R124 in real life).
Green is the voltage on Q106's base.

I just noticed in that sim, the 100R resistor on the collector should not connect to ground directly, you could connect about a 10pF capacitor in series with it to simulate the cathode capacitance of the CRT or just leave it out.

 

[Hugo Holden]

All tests were done in circuit.

Q105 B->C: 0.573V
Q105 B->E: 0.611V
Q105 C->E: 0.699V
All other arrangements: 0L

Q106 B->C: 0.625V
Q106 B->E: 0.645V
Q106 E->C: 0.676V
All other arrangements: 0L

I observed the monitor some more, and it seems that when both the user brightness control and VR501 are turned all the way clockwise, the picture reappears more often as it did about a year ago.
Its behavior is exactly the same (the dimming and occasional disappearance) while running.
The turn off spot also behaves differently for no apparent reason. Sometimes it is very bright and stays on the screen for a long period of time, other times it is very short and not very bright, and rarely does it not happen.
I should mention that this turn off spot was not present for the past two years.

During this observation, I was also probing the base of Q106, which is the output of the video circuitry. It had some strangeness, sometimes the amplitude jumped up and down, but this strangeness was independent of a picture being displayed.

I'll have a look at some resistors on the main board.

Unfortunately those in circuit tests are not too helpful.

The B-E and B-C junction tests look ok, not the E-C though, but really the transistors should be tested out of circuit.

I would recommend using solder wick and removing the solder from the pads from two terminals on each transistor to disconnect the leads from the pad, it works well in single sided pcb's without plated through holes, you can leave the other one terminal soldered on both.

Then redo the junction tests. You have to use the meter in both polarities for each test.

If the transistors are likely ok, what you should find is that with each B-E and each B-C junction, they should only conduct on one direction (polarity) of the meter only, and give a drop in the region of 0.6 to 0.7V and read open on the opposite polarity. And when you check from the C-E it should not conduct in either direction. Tell me what you find.

 

[awewsomegamer]

Unfortunately those in circuit tests are not too helpful.

The B-E and B-C junction tests look ok, not the E-C though, but really the transistors should be tested out of circuit.

I would recommend using solder wick and removing the solder from the pads from two terminals on each transistor to disconnect the leads from the pad, it works well in single sided pcb's without plated through holes, you can leave the other one terminal soldered on both.

Then redo the junction tests. You have to use the meter in both polarities for each test.

If the transistors are likely ok, what you should find is that with each B-E and each B-C junction, they should only conduct on one direction (polarity) of the meter only, and give a drop in the region of 0.6 to 0.7V and read open on the opposite polarity. And when you check from the C-E it should not conduct in either direction. Tell me what you find.

I have de-soldered the collector and emitter of both transistors. The measurements are now:

Q105 B->C: 0.577V
Q105 B->E: 0.607V
Q105 C->E: 0.650V

C->B, E->B, E->C are all 0L.

Q106 B->C: 0.623V
Q106 B->E: 0.643V
Q106 C->E: 0L

C->B, E->B, E->C are all 0L.

I did measure these values twice before posting the message, the values were consistent (off only by a few mV).

Q105 C->E is very odd.

 

[awewsomegamer]

I just noticed in that sim, the 100R resistor on the collector should not connect to ground directly, you could connect about a 10pF capacitor in series with it to simulate the cathode capacitance of the CRT or just leave it out.

I have added this capacitor for completeness. The current through it is the opposite of the current through R8.

V1 has also been changed to be the measured B+ voltage.

Here the green trace is the current through the capacitor and the blue one is the video input.


Adding a resistor after the capacitor allows me to measure this voltage in blue (now the video input's voltage is in green):

 

[nunoalex]

Ok. decent photos posted, that is the way to do it, well done.

I have reverse engineered and studied the CRT neck board. It is a fairly classic arrangement with a Cascode video output stage driving the CRT's cathode. It will be fairly straightforward with this data to find your problem. I'm actually pretty sure I know where it probably is already. I'm sure @andy will see it too.

However, it took about 1 hour to do this, I will have to write again tomorrow to conclude the information and explain what to do.

Of course if I had this in front of me on the desk, I could have done it in 10 minutes, it is much more difficult with photos (even when they are good)

Thank you for helping diagnose this monitor Hugo

 

[Hugo Holden]

I have de-soldered the collector and emitter of both transistors. The measurements are now:

Q105 B->C: 0.577V
Q105 B->E: 0.607V
Q105 C->E: 0.650V

C->B, E->B, E->C are all 0L.

Q106 B->C: 0.623V
Q106 B->E: 0.643V
Q106 C->E: 0L

C->B, E->B, E->C are all 0L.

I did measure these values twice before posting the message, the values were consistent (off only by a few mV).

Q105 C->E is very odd.

I think Q105 might be defective though one aspect doesn't make a lot of sense in that collector-emitter conduction would tend to lower the collector & cathode voltage. I have had some difficulty finding its data sheet. Have you been able to find its data sheet ?

I think it is worth replacing it. You might have difficulty finding the original part, if you like you could replace it initially for a test with the MPSA42 or MPSA43 that is much easier to get. You just have to check the pinout before you fit it. So try that and we will see what happens.

 

[awewsomegamer]

I think Q105 might be defective though one aspect doesn't make a lot of sense in that collector-emitter conduction would tend to lower the collector & cathode voltage. I have had some difficulty finding its data sheet. Have you been able to find its data sheet ?

I think it is worth replacing it. You might have difficulty finding the original part, if you like you could replace it initially for a test with the MPSA42 or MPSA43 that is much easier to get. You just have to check the pinout before you fit it. So try that and we will see what happens.

I haven't found a datasheet for it either. However, there is a website (alltransistors.com) that lists some information about a 2SC2228 NPN BJT:

https://alltransistors.com/transistor.php?transistor=12905#pack

I have also found this listing for it, and the specs listed seem to match the ones stated by alltransistors.com:

2SC2228 Transistor Silicon NPN - Buy from Littlediode

This listing too states the same specification, and in the image the transistor is marked "C2228 D-4F" which is nearly identical to the "C2228 D- 4H" on Q105:

2SC2228 TRANSISTOR | Videotronics E-Store

videotronics.com.cy

The turn off spot is quite concerning, but so far it hasn't left any damage.
I will also follow through with Andy's request to look at the resistors on the main board to see if any of them are horrifically out of spec.

I am also working on revising my schematics so that they are hopefully easier to understand.

 

[Hugo Holden]

The ones on the videotronics site look like geniune NOS parts.

As far as I can tell the 2SC2228 is similar to a BF422,which they keep at Dalbani,

BF422 Original New Toshiba TO-92 Transistor | eBay

Let us fix the problem for you. We will do our best to get back to you as soon as possible. At Dalbani we like to differentiate between new, previously installed parts (used), and parts made by 3rd party manufacturers.

www.ebay.com

If those data sheets are correct, the 2SC2228 has a high range current gain for a single BJT.

With that collector to emitter reading you got, it made me initially wonder if it might have been a Darlington device or had an inbuilt C-E diode but going against the latter was the direction it conducted in, but I cannot convince myself this is the case, it appears it is a single BJT, so it should not have conducted in one direction between the C & E termnals with a floating base, unless it was damaged and has some base to collector leakage.

We already know that when you accidentally shorted this transistor's C-B the screen lit up, which is expected, because the crt's cathode voltage dropped. There are really only a couple of possibilities, one is this transistor is defective, or Q106 (which is likely ok on testing) is not getting a high enough voltage drive at its base from the main board. So if you could replace it with either a BF422 or an MPSA 42/43, we could rule it in or out and get on trying to find another cause.

With the spot thing, keep the manual brightness as low as you can, turn it up while powered and testing, but turn it right down before de-powering and try not to rapidly power cycle the VDU in short intervals, might help a little for now.

 

[Hugo Holden]

Also another excellent video output transistor that would work very well in the application, and is more rugged with 1W dissiaption is the 2SC3526, it was specifically designed for the application. click were it says pardon..:

Pardon Our Interruption...

 

[awewsomegamer]

I have revised my schematic once more, this should be hopefully more accurate and easier to read.
The rails are ordered by voltage (descending) and inputs come in from the left and outputs go out to the right.
There are some exceptions to the I/O ordering in the POWER and UNORDERED schematics:

Below are some close ups of each circuit:









The datasheet for IC401 can be found here: https://www.alldatasheet.com/datasheet-pdf/pdf/1718895/SAMSUNG/KA2130A.html.

The G label on the neckboard is also not a direct connection to ground, rather it is a connection to a terminal of L103 which is then goes to ground.
It looks like a thick diode:


I am not sure about the title "G1 GENERATOR", but the circuit looks like it would produce the square wave signal I observed with the scope on G1.

 

[Hugo Holden]

Your schematics are definitely improving, but there are still some issues.

In general, all schematics (if they are going to help some repair technician at least) are drawn from left to right, in the same way that we read a sentence, but in this case it represents the flow of signal through the circuit.

So in this case (mostly) transistors are drawn in with the base connections on the left side and the collector & emitter connections on the right, from the point of view of the observer's right and left hand, if they were looking at it on a piece of paper or on a VDU. So if you reach out with your R hand it would be on the side of the diagram where the transistor's collectors and emitter are.

( just clarifying that as there are a lot of right and left mix ups in life, for example if a person is standing in front of you and you reach out with your R hand to their shoulder on the same side, you are reaching out to their left shoulder)

There are exceptions of course where there are two transistors wired as a differential pair, or a transistor in a grounded base configuration. And those transistors get flipped around sometimes to make the schematic easier to read.

Consider the flow and processing of a signal and make it from left to right on the diagram, and your schematics will improve.

 

[awewsomegamer]

We know for sure it's not a bad flyback. The problem is most likely in the area of the board labeled "video", or on the CRT socket board. I would start with a very careful examination for bad soldering, then do an in-circuit test of all the transistors in the video circuit (looking for open, or shorted junctions). You could do an in-circuit check of all the resistors in the video circuit. You're looking for major discrepancies, not exact values (eg. a 1k resistor that measures 80k is obviously open). The fact that it started intermittently makes me suspect a failing resistor, or possibly a bad transistor.

If you don't find anything there, I would start from the video input, and trace the signal though the circuit with your oscilloscope. It should be fairly obvious where the signal stops. Set your scope to AC coupling since some of the video stages may have a significant DC offset compared to the level of the video signal, and use a 10X probe.

As a general rule, you always use a 10X probe unless you have a very good reason not to (eg. looking at very low level signals). 10X probes load down the circuit under test a lot less, provide better frequency response, and provide some over voltage protection for the scope.

I have measured some resistors on the board, most look OK, however there are a few questionable ones.

All measurements were done in circuit and repeated twice:

Resistor	Measurement #1	Measurement #2	Stated Resistance
R101	??? (meter couldn't get a reading)	74-78, unstable	75
R102	4.58k	4.58k	15k
R103	48.7	~100, unstable	100
R108	29.58k	29.58k	220k
R117	772	~770, unstable	3k

I double checked the resistor code as well for this table. There are some discrepancies between the schematic and some of the values I just read (R101 in the schematic is 7.5k instead of 75), so I will have to go over all resistors again and make sure I read them right.

The solder joints seem fine.

I also took Q105 and Q106 fully out of circuit and measured their junctions again:

Q105 B->C 580mV
Q105 B->E 611mV
Q105 C->E 0L
Q105 C->B 0L
Q105 E->B 0L
Q105 E->C 0L

Q106 B->C 624mV
Q106 B->E 644mV
Q106 C->E 0L
Q106 C->B 0L
Q106 E->B 0L
Q106 E->C 0L

I also put the two transistors into a component tester, and both were recognized as NPN BJTs, Q105 had an hFE of around 60 and Q106 had one of around 110.

 

[Hugo Holden]

In circuit tests of resistors are meaningless, you have to lift one leg. The test only has any meaning in circuit, if the resistance reads significantly higher than the resistors stated value. In circuit other current pathways shunt the resistor so you should expect lower readings.

Your first lot of transistor measurements must have been an error for Q105, now on the tests it looks ok.

We cannot help you if your test data is defective and unreliable, it just leads everyone up the garden path.