A600 Caps lock led stuck on at boot solution

[FOL]

Originally Posted by PeteJ

thanks, I should perhaps test the resistors in the video circuit next with the multermeter to see if one or more has failed. I will have a look at the 600's circuit diagram and try to understand what I'm looking at

No I've not replaced the NE555, I have replaced the failed capacitor there now that was causing the reset loop. When I googled the NE555 I read that it was a timer so I thought with the reset loop fault being caused by the bad cap there that this was used just for the systems startup / reset timing?

I would put it back on. Your trying to find an issue but you have parts removed.

[PeteJ]

Originally Posted by FOL

I would put it back on. Your trying to find an issue but you have parts removed.

Hi, I've not removed the NE555, it was just the bad cap there that I removed originally but that's now been replaced.

[FOL]

Originally Posted by PeteJ

Hi, I've not removed the NE555, it was just the bad cap there that I removed originally but that's now been replaced.

Sorry, think I was misunderstood. When I said replaced, what I ment is, have you reattached it to board. Replaced can be taken two ways in a certain context.

Spose its a good thing its effecting all outputs. Gives you a place to focus on. You could check U31, as your missing green.
Maybe swap U31 and U32 around see if issue switches to another colour.

[PeteJ]

Originally Posted by FOL

Spose its a good thing its effecting all outputs. Gives you a place to focus on. You could check U31, as your missing green.
Maybe swap U31 and U32 around see if issue switches to another colour.

thanks, that's exactly where I was up to this morning after reading the circuit diagram. I could see for the Green out from Gayle, the resistor R231B (1k) which I'll check with the multimeter and then the resistor in the analogue output further on from that circuit for the green R234B (75). I suspect as you suggest however that whichever one of the PC74HCT244T chips that includes green (U31 or U32) is faulty. It's a good idea to swap them around to confirm that, thanks for that suggestion, I'll try that next. I've not removed a surface mounted chip before, should I use flux and desolder braid?

[PeteJ]

I've looked again at the circuit diagram and the green is with U31. U32 has the blue with pixelsw and burst. Can anyone tell me what the function of pixelsw and burst is? I'm wondering if that has anything to do with the ghosting? The green in u31 is the same pins as the pixelsw and burst of u32. Surely both cannot have failed? Perhaps I should replace both? Talking of which, I've tried to look at where to buy the PC74HCT244T, I can see lots of different 74HCT244 logic gates available but not one with the same code that's printed on the chip. Looking at a photo of another rev 1.5 board I can see what looks like Motorola chips in position at U31 and U32. The scan is too low res to read it however. Are there many suitable pin compatible replacements that I should be looking for? The circuit diagram lists it simply as 74HCT244.

[PeteJ]

I can't seem to find any information about the (PC)74HCT244T or a datasheet. The 74HCT244D is readily available to buy from various sources. I don't know what the difference of the two are though and whether the 74HCT244D will work in the Amiga. I'll see if I can find a hi res photo online of other A600 motherboards and see what is installed on them.

[FOL]

Originally Posted by PeteJ

I can't seem to find any information about the (PC)74HCT244T or a datasheet. The 74HCT244D is readily available to buy from various sources. I don't know what the difference of the two are though and whether the 74HCT244D will work in the Amiga. I'll see if I can find a hi res photo online of other A600 motherboards and see what is installed on them.

Swap them around. It will help rule out if that is faulty.

[Jope]

The crucial part here is 74HCT244 .. if that matches and the chip is in the same package, then it will work.

[jimbob]

I would be suspecting the circuit of Q212 before U31. Definitely check it out before attempting to remove SMD chips.

I'm guessing you don't have a scope but with just a multimeter you can make a good stab at diagnosing a faulty transistor. With the board powered off measure the resistance and continuity/diode reading of each transistor terminal to ground and to the other terminals.

Compare these results with the same measurements of the equivalant circuit for red and blue ie. Q211 and Q213. If any measurement is a long way off from its red/blue partner you probably have localised the fault to this transistor circuit. In particular, the diode measurement between terminals should be in the ballpark of the known working circuits, pay attention to meter lead polarity when making this measurement. It could still be the transistor itself or one the passive components around it.

Then you have to find which one, you could for example check that R212 is not open or short circuit, or you could bridge C212 with a leaded capacitor, (any value ~0.1uF-1 uF will do), to see if the green flashes up, that will tell you that C212 is open circuit, etc.

(Sorry if this doesn't make sense to you, bit tricky to write detailed instructions as I don't have any of my amiga or electronics stuff with me just now)

You can also compare the voltage on each terminal of Q212 with Q211/Q213 when the board is powered on. Do this with a steady white or black screen if possible.

In case you don't have them, best available A600 schematics can be found here -

http://www.amigawiki.org/doku.php?id...ice:schematics

[PeteJ]

Originally Posted by Jope

The crucial part here is 74HCT244 .. if that matches and the chip is in the same package, then it will work.

Thanks mate I can order the 74HCT244's that I've seen then if the other tests prove to be ok.

Originally Posted by jimbob

I would be suspecting the circuit of Q212 before U31. Definitely check it out before attempting to remove SMD chips.

I'm guessing you don't have a scope but with just a multimeter you can make a good stab at diagnosing a faulty transistor. With the board powered off measure the resistance and continuity/diode reading of each transistor terminal to ground and to the other terminals.

Compare these results with the same measurements of the equivalant circuit for red and blue ie. Q211 and Q213. If any measurement is a long way off from its red/blue partner you probably have localised the fault to this transistor circuit. In particular, the diode measurement between terminals should be in the ballpark of the known working circuits, pay attention to meter lead polarity when making this measurement. It could still be the transistor itself or one the passive components around it.

Then you have to find which one, you could for example check that R212 is not open or short circuit, or you could bridge C212 with a leaded capacitor, (any value ~0.1uF-1 uF will do), to see if the green flashes up, that will tell you that C212 is open circuit, etc.

(Sorry if this doesn't make sense to you, bit tricky to write detailed instructions as I don't have any of my amiga or electronics stuff with me just now)

You can also compare the voltage on each terminal of Q212 with Q211/Q213 when the board is powered on. Do this with a steady white or black screen if possible.

In case you don't have them, best available A600 schematics can be found here -

http://www.amigawiki.org/doku.php?id...ice:schematics

Thanks so much, really appreciate your time and info with the reply. I don't have a scope so it will just be tests with the multimeter. I will definately go through all of those tests before attempting to replace the 74HCT244.

[PeteJ]

Originally Posted by jimbob

I would be suspecting the circuit of Q212 before U31. Definitely check it out before attempting to remove SMD chips.

I'm guessing you don't have a scope but with just a multimeter you can make a good stab at diagnosing a faulty transistor. With the board powered off measure the resistance and continuity/diode reading of each transistor terminal to ground and to the other terminals.

Compare these results with the same measurements of the equivalant circuit for red and blue ie. Q211 and Q213. If any measurement is a long way off from its red/blue partner you probably have localised the fault to this transistor circuit. In particular, the diode measurement between terminals should be in the ballpark of the known working circuits, pay attention to meter lead polarity when making this measurement. It could still be the transistor itself or one the passive components around it.

Then you have to find which one, you could for example check that R212 is not open or short circuit, or you could bridge C212 with a leaded capacitor, (any value ~0.1uF-1 uF will do), to see if the green flashes up, that will tell you that C212 is open circuit, etc.

After giving myself a quick crash course on how to use a multimeter I've run through the tests that you've suggested and you were on the money with a fault at Q212 it seems. I tested the red and blue Q211/Q213 and they both gave the same readings : touching either the positive or negative probe on the 3rd pin, they display with the other probe 0.32 at left pin and 0.18 at the right pin. Q212 (green) however displays a reading off the scale for the left pin and 0.50 for the right pin. I tested the resistor at R212 which shows a correct reading of exactly 151. I did also test the other resistors in the green circuit which gave correct readings R231B (1K) displays 977 and R234B (75) displays 75.

Now that I've already recapped the board and tested the resistors in the circuit, I guess that narrows it down to the transistor at Q212 so I should replace it. Can somebody please tell me which smd transistor to buy? There are no markings on it and the circuit diagram as far as I can see doesn't detail that. I'm unsure of the rating or package type to order.

Thanks.

[FOL]

Originally Posted by PeteJ

After giving myself a quick crash course on how to use a multimeter I've run through the tests that you've suggested and you were on the money with a fault at Q212 it seems. I tested the red and blue Q211/Q213 and they both gave the same readings : touching either the positive or negative probe on the 3rd pin, they display with the other probe 0.32 at left pin and 0.18 at the right pin. Q212 (green) however displays a reading off the scale for the left pin and 0.50 for the right pin. I tested the resistor at R212 which shows a correct reading of exactly 151. I did also test the other resistors in the green circuit which gave correct readings R231B (1K) displays 977 and R234B (75) displays 75.

Now that I've already recapped the board and tested the resistors in the circuit, I guess that narrows it down to the transistor at Q212 so I should replace it. Can somebody please tell me which smd transistor to buy? There are no markings on it and the circuit diagram as far as I can see doesn't detail that. I'm unsure of the rating or package type to order.

Thanks.

Look with good light,they have numbers on them.

[PeteJ]

Originally Posted by FOL

Look with good light,they have numbers on them.

oh, yes my mistake, you are right it does have numbers printed on them. They are very faint as well as minute It looks like 1AMx printed on them (the x has a line under it). I can see 1AM transistors for sale like "SOT-23 SMD NPN Transistor 1AM 40V 0.1A 100mA" Would they be the correct type?

[PeteJ]

Does anyone know if the transistor I'm looking for is an SOT-23 package? My board is rev 1.5 but I've looked at some other A600 motherboard photos and the rev 2b board has resistors in those positions with what looks like just 1A written on them. Does that mean DC Collector Current is 1A? Maybe the Fairchild Semiconductor BCW66G is suitable?

[jimbob]

I think SOT-23 is right.

Not sure about the fairchild part. Schematic says it is 2N3904 transistor so if it is equivalent to that then it should be good. I think I used these before -

http://uk.rs-online.com/web/p/bipola...392D3033303326

[FOL]

Originally Posted by PeteJ

Does anyone know if the transistor I'm looking for is an SOT-23 package? My board is rev 1.5 but I've looked at some other A600 motherboard photos and the rev 2b board has resistors in those positions with what looks like just 1A written on them. Does that mean DC Collector Current is 1A? Maybe the Fairchild Semiconductor BCW66G is suitable?

Your link is fine, that will work.
Id go with RS as its 8p as apposed to my link 12p.

MMBT3904 SOT-23

That should work, just checked the manual and that looks fine.
The 1A is a code, 1Amp would be extreme.

DataSheet for original

Datasheet for alternative

[PeteJ]

Great thanks guys, I have an account with Farnell so I'll order them right now.

[PeteJ]

I got the transistor this morning, with minimum ordering I had to buy one from ebay. I removed the old transistor with flux and desolder braid carefully but I then went over the pads with braid again to clean them up before rubbing with ipa and one of the pads lifted. It seemed to be still connected to it's trace so I moved the pad back into position with tweezers then tinned them all and soldered the new transistor in place. Booted the A600 and no change, the display was exactly the same as before. I checked the new transistor on the board with the multimeter (I should have also checked it before installing it but I forgot) at 20k I get the same wild readings as the old one with pin1 to pin3 off the scale and pin2 to pin 3 at .49? so now I don't know whether there is another component at fault or if I've damaged the pad of pin 1

[PeteJ]

I'm thinking perhaps it was U31 after all at fault. It's in the circuit via Denise and maybe why the readings of Q212 are wrong. I'll buy a replacement 74HCT244 and see how that goes I guess. I'd rather not swap u31/u32 to diagnose it to avoid the chance of me damaging u32 or the pcb. This last task caught me out, I'm always really careful with soldering. It's the first time I've ever lifted a pad or trace.

[jimbob]

You can try to verify U31 is good before replacing it by using the amiga video output as a kind of reverse oscilloscope. Sorry I can't confirm this will work but I think it should, though it may depend on your multimeter.

Use deluxepaint to make the whole screen pure white. It won't look white due to the fault but just check the colour you use has full RGB values. While you're at it, slide the value for G up and down to see if there is any visible change, that will verify that G is completely missing and not just certain bits of G data. Then measure the DC voltage at pins 3,5,7 & 9 of U31. (if it reads 0, make the screen black)

Then change half the screen to black, (or white) and measure again.

This should half the time that the 4 bits of green data are switched on. Hopefully the video frequency will be high enough that the multimeter will average the switching signal and give you roughly half the voltage. If you can't get a steady reading, try making the screen alternate black and white stripes.

Again you can compare this test with the RB channels. Pins 12,14,16 & 18 of U31 for red and 12,14,16 &18 of U32 for blue. You can also try experimenting with various greys to try to see what is going on with the G data.