For some reason I'm not getting any colour output from my II+ clone. The monitor (composite input) is attached to the video output on the back of the computer. Other than that there are no other issues with the video (it displays fine, but only in black/white), and everything else appears to work on the computer.
So I looked up the Applebox diagram (which illustrates what each IC does) and I've swapped the following ICs, alas without any difference:
74LS02 (B14)
74LS02 (B13) color ref
74LS11 (B12) color burst
74LS51 (C13) color decoding
74LS32 (C14) video blanking
I intended to swap F14 (74LS259) as well, but didn't have another one to try out. I also haven't replaced transistor Q3 (J14) close to the video level trimmer, but I suppose since video in general works it's fine.
Where else could the problem lie?
Before getting my II+ in fully working condition I had a problem with the keyboard not responding (see this thread). It turned out that a circuit trace had corroded away, but of course was very easy to fix once I tracked it down (thanks to people in this forum!). Could the reason be likewise here?
[quote=Fuji]
For some reason I'm not getting any colour output from my II+ clone. The monitor (composite input) is attached to the video output on the back of the computer. Other than that there are no other issues with the video (it displays fine, but only in black/white), and everything else appears to work on the computer.
[/quote]
Possibly silly question: You are displaying this on an NTSC screen? If not you're not going to get the colour without an additional card. Apple II colour comes from some nasty NTSC signal hacking and is black and white on PAL region systems without a PAL card.
John
Not a silly question at all. I wasn't aware of the monitor adhering to PAL or NTSC standards, thinking that was only an issue with TVs. So there's Composite PAL and Composite NTSC?
In any case the colour monitor (a Japanese NEC import, so it's definitely NTSC then) is the same monitor I've always used with this II+, and has always displayed colour from the II+.
That probably also explains why I neither got a color display on the monitor while attaching a (PAL) video camera composite output to it, nor when I attached the II+ rear output to our (PAL) LCD TV's composite input and yet again got a black/white picture! I wanted to see if the malfunction was within the monitor or II+, but assume from this that I would only get a black/white picture regardless because of the PAL/NTSC mismatch, so not much to gain from those tests.
While searching for a more specific answer I came across a very similar issue in this forum: no color for Apple II Plus. I gathered that the II+ clock has something to do with this and suddenly realized that my II+ (a clone actually) has a capacitor-trimmer mounted near the A1 area. It's not present in the original II+ schematics so I don't know what it's for and why, but could it have to do with fine-adjusting the main clock frequency? I do have a digital multimeter with a frequency counter mode: where do I point the probes to check the clock frequency while trying to adjust the trimmer to see if it makes any difference?
First try this monitor with another NTSC composite source to verify it is still capable of displaying NTSC color picture. It may have a manual switch somewhere to select the color standard...It may have color saturation control too...
Yes, it (the monitor) works in colour!
I checked it with a multi-region DVD-player, inserted a U.S. (NTSC) DVD and luckily it had a video output jack in addition to HDMI. Glorious colour!
OK, so that eliminates one source.
On the II+ I believe I've replaced most of the relevant ICs except the 74LS259 (F14). Could this chip be the cause of such issues? It appears hard to find stores who sell it, but I'll keep looking.
If I were you I wouldn't have wasted time with changing ICs especially led by this very doubtful German amateur source with the applebox diagram. At first glance this sketch lies about the data buffers near the CPU. The bidirectional 8T28 buffers cannot be replaced by 74367 unidirectional buffers. I would have tried to adjust the frequency of the 14.318 MHz crystal and its generator by adding small capacitance to it or by replacing the crystal itself...Nowadays this is still the most widespread and cheap quartz crystal used in every PC...
Some moderator has very unwisely switched off the editing of the messages in this forum but I needed to add this...Check also the color killer transistor operation with a multimeter, note that it should allow color only in graphic modes, so use at least GR from BASIC to enable your computer to output color subcarrier.
I wasn't aware of the diagram/site being doubtful, but being an amateur myself it seemed as a helpful start.
[quote=Casablanca]
I would have tried to adjust the frequency of the 14.318 MHz crystal and its generator by adding small capacitance to it or by replacing the crystal itself...Nowadays this is still the most widespread and cheap quartz crystal used in every PC...
[/quote]
So apparently the trimmer capacitor in my II+ clone is there for what you're suggesting above. I measured (freq. setting from a digital multimeter) from each pin of the crystal while adjusting, but have only reached around 14.292 MHz. It's very difficult to adjust, so maybe I should use a smaller value of the trimmer (mine is 220 pF). Is it very critical? Regardless, there's no hint of colour.
The colour killer transistor, is that the 2SC945 close to the "color trim" trimmer?
[quote=Fuji]
I wasn't aware of the diagram/site being doubtful, but being an amateur myself it seemed as a helpful start.
[quote=Casablanca]
I would have tried to adjust the frequency of the 14.318 MHz crystal and its generator by adding small capacitance to it or by replacing the crystal itself...Nowadays this is still the most widespread and cheap quartz crystal used in every PC...
[/quote]
So apparently the trimmer capacitor in my II+ clone is there for what you're suggesting above. I measured (freq. setting from a digital multimeter) from each pin of the crystal while adjusting, but have only reached around 14.292 MHz. It's very difficult to adjust, so maybe I should use a smaller value of the trimmer (mine is 220 pF). Is it very critical? Regardless, there's no hint of colour.
[/quote]
Apparently I would have written to use the color trimmer capacitor but I didn't. You can't adjust the 14.318 frequency with that trim capacitor and since it is lower that required your only oprion is to replace the quartz crystal. The transistor is Q6 (2N3904).
I haven't seen your clone so I can't tell for sure which transistor it is but yes, usually this transistor is placed near the trimmer capacitor. The above reference was for original Apple ][+.
[quote=Casablanca]
Apparently I would have written to use the color trimmer capacitor but I didn't. You can't adjust the 14.318 frequency with that trim capacitor and since it is lower that required your only oprion is to replace the quartz crystal. The transistor is Q6 (2N3904).[/quote]
I was referring to a capacitor trimmer right next to the 14.318 MHz crystal (which isn't present in the original II+) but in my II+ clone. It was broken from storage and I don't know its value, but the motherboard is marked 50 pF. I'm currently using a 200 pF trimmer as I didn't have anything else, but plan to order a 50 pF trimmer.
Do you think I need to replace the crystal as well?
I also have a "color trim" trimmer, which is marked 50 pF as well. It's worked in the past to adjust the color hue as far as I can recall. Yes, there's a transistor to the left of it -it's marked 2SC945 on my motherboard and that's what's used as well. I'm guessing it's an equivalent of the 2N3904 originally used, but the pinout is different (my clone motherboard was made in Japan I've been told, and those are perhaps more common in the Asian world than in the U.S.).
Unfortunately I don't have a replacement 2SC945 in my part box (and ordering online takes a while), but I do have several 2N 3904 so I'll double check if they're equivalent to replace and see what happens.
You don't have to buy all parts to verify their operation! Use some thinking. If you measure the voltage between the collector of Q6 anf GND it should be low (appr. 0.3V) in text mode and high (appr. 1V) if any graphics/mixed mode is active. Just type GR and TEXT at BASIC prompt to switch the modes. By adding capacitance to the oscillator you can only lower the frequency. So if with unconnected trimmer the frequency of your oscillator is lower than 14.318 MHz (AND you are sure in your frequency measurement!) than you have to replace the crystal. They tend to age and their frequency drifts with decades espcially in those ancient generation of quartz crystals that was used in your A2 clone.
Thanks for the clear explanation.
The computer won't boot without the trimmer capacitor (the one next to the 14.318 MHz crystal), but I tried to measure its frequency regardless. The trimmer capacitor (removed) must be crucial as the frequency drifts as I measure. It's around 2.7 KHz. So for now I'll put the trimmer capacitor back (it's marked 50 pF on the motherboard), but I'm still not reaching 14.318 MHz. I've read something about problems measuring some things like this correctly without a buffer circuit, but I'm not sure what that means and if it applies here.
In any case a new crystal doesn't cost much and is probably worth doing with such an old computer.
OK, next is measuring the voltage between C (collector) of Q6 and GND:
TEXT: 0.0108V
GR: 0.0104V
So it looks like Q6 needs to be replaced. In my clone it's a 2SC945, but according to online sources it's an equivalent to the 2N3904 as you point out and I assume used in the original II+ (but a different pinout). Luckily I have a spare 2N3904 in my parts box so I'll give that a go.
OK, I've changed the transistor (Q6) but it made no difference. The voltage readings are the same and I've also adjusted the "color trim" trimmer. No sign of colour. And of course the monitor is still adjusted to show colour (as tested with the DVD player).
Should I start looking for broken PCB tracks? In addition to the obvious (areas around Q6 and the "color trim" trimmer) do you have suggestions as to where else the problem might originate from?
Check with a scope the presence/absence of 3.58 MHz at the collector of Q6 while in GR/TEXT.
Keep in mind these are only color bursts and are not constantly present there. They probably can be seen better at pin 12 of 'LS11 (B12).
You acted again illogically -- you replaced Q6 without first checking the voltage at its base first during GR/TEXT!
I'm probably going to make more illogical mistakes as I'm no engineer :-)
I just measured the Base of Q6:
TEXT: 0.67V
GR: 0.14V
Measuring pin 12 of B12 (74LS11) in both GR and TEXT mode I get around 15.7 KHz for both.
I've done some further research and came across the Apple II red book 30th year anniversary edition which among other things explains where to take frequency readings from.
On page 143 there's a diagram that (if my understanding is correct) shows 14.318 MHz can be checked on pin 8 of B2 (74S86) and by adjusting the capacitor trimmer near the crystal I was able to get exactly 14.318 MHz, so that's good news!
There's no 3.56 MHz reading on pin 12 of 74LS11 (B12), but looking back at that same diagram (page 143 of the red book) I can read 3.58 MHz at pin 3 of B1 (74LS175) as labelled "color burst", so at least the signal is generated.
I meant 3.58 MHz (not 3.56 as in my typo above).
Too bad it's not possible to edit postings here.
I'm baffled!
I can't figure this one out, and just about all the "quick fix" documents (including Apple's own repair document) mention the following ICs for a "no color, everything else works fine" problem:
B12 (74LS11)
B13 (74LS02)
C13 (74LS51)
I believe I swapped them all, but I'm not ruling out that my replacement chips could be bad. Unless I figure this out quickly I think I'm going to order a bunch of different ICs that I don't already have (or just one of) and try some swapping again.
So what I've done next is read through a bunch of repair/service documents and in "Sam's computer facts -Apple II, II+" there are handy schematics with voltage readings in various places. I checked all of the ones I could find related to this issue (page 36/37 from the above book) and many of them were off. Here's that page with my readings in red (different) and green (more or less the same):
video circuits Apple 2.jpg
Another thing I don't think I've mentioned before: measuring the frequency of C of Q6 I get the following when entering these modes:
TEXT: 15.7 KHz
GR: 31.4 KHz
I'm not sure how to use the oscilloscope and make sense of the measurements.
Can anyone make any sense of all these findings?
[quote=Fuji]
I'm baffled!
I can't figure this one out, and just about all the "quick fix" documents (including Apple's own repair document) mention the following ICs for a "no color, everything else works fine" problem:
B12 (74LS11)
B13 (74LS02)
C13 (74LS51)
I believe I swapped them all, but I'm not ruling out that my replacement chips could be bad. Unless I figure this out quickly I think I'm going to order a bunch of different ICs that I don't already have (or just one of) and try some swapping again.
So what I've done next is read through a bunch of repair/service documents and in "Sam's computer facts -Apple II, II+" there are handy schematics with voltage readings in various places. I checked all of the ones I could find related to this issue (page 36/37 from the above book) and many of them were off. Here's that page with my readings in red (different) and green (more or less the same):
video circuits Apple 2.jpg
Another thing I don't think I've mentioned before: measuring the frequency of C of Q6 I get the following when entering these modes:
TEXT: 15.7 KHz
GR: 31.4 KHz
I'm not sure how to use the oscilloscope and make sense of the measurements.
Can anyone make any sense of all these findings?
[/quote]
I've had two boards do this and in both cases I replaced the crystal with a new one and it was fine. I guess they float out of spec with age... You can find my thread on here from many years ago about it.
Yes, I read that thread earlier, but confirming that I have a 14.318 MHz signal I thought I was OK and the problems must be elsewhere.
Did you measure fluctuations before replacing the crystal? It appears stable here.
It's worrying that you had to try several different crystals before it worked. I've just ordered three of these (14.31818 MHz) from Futurlec and hope it'll solve the issue. I've also ordered replacement ICs for what has to do with the colour signal. I've also started closely studying the motherboard itself for broken traces/vias. You never know (it was the cause of my previous keyboard malfunctions).
Regarding the voltage readings in my previous post -what can the cause of these be? The crystal can't be doing this, can it?
Apparently, the crystal tolerance is extremely tight. I had some 14.318 crystals that did not work. The ones that did were "NYMPH 14.31818" in my case.
I originally tried to swap between 2 Apple II's and then color worked in neither, so it seems the soldering iron heat took the other one out of spec also.
I suppose I can't go wrong to replace the (almost 40 year old) crystal in any case.
My II+ clone has a trimmer capacitor alongside the crystal which I found I could adjust to get the exact frequency, but for all I know the crystal may still not "be right" even though my meter readings don't pick up on it. From the documentation I've seen on the original Apple II+ there's no trimmer, so that probably means a tighter tolerance on the crystal itself is needed.
Do you remember the frequency reading when you replaced it? How much off the exact 14.31818 MHz would allow for it to still work properly?
Does the colour burst signal depend on the computer's main clock frequency at the time it's booted, or could I get colour to appear if I slightly adjust that frequency trimmer while the computer's powered on, then try to run a graphic test program (i.e. color bars or a game), exit back to text mode, then repeat with a slightly different frequency setting?
Or do I need to reboot every time I change the frequency for it to work?
If I'm not mistaken, the 3.58 MHz "colour burst" signal (which enables colour output) is as the name implies; a short signal "burst", then disappears again. My understanding might be totally wrong though as I'm just a hobbyist with a soldering iron and some practical skills, but that's my gathering from these postings and from the various books and service manuals I've read.
So the replacement parts finally arrived and I've made a new attempt to get the lost colour output back.
I have now replaced (with brand new parts):
74LS11 (B12) color burst
74LS02 (B13) color ref
74LS02 (B14)
74LS51 (C13) color decoding
74LS259 (F14) video page switch
2SC945 (Q6)
14.31818 MHz crystal (A1)
No change compared to using the old components and still no colour (but everything else works as it should).
I'm suspecting that (like my previous issue with the keyboard not working) a circuit trace somewhere is broken, but I don't quite understand which signals to trace, where they are and how (especially detecting the "color burst" signal). Does that signal originate mainly from the B12 chip?
The color burst appears on every line of the video signal. So you don't have to reboot or switch video modes to see if the frequency is correct. You can adjust it while looking at a graphics display and when it is correct the color will appear.
I'm not familiar with that "anniversary edition," but it looks to be just a reformatted version of the original 1978 reference manual. I find the 1979 version to be significantly improved; you may want to use that instead or in conjunction with the other sources you're using.
(Also, the 1979 version has a listing for the Autostart ROM as well as the older ROM; unless your clone is very old you likely have the Autostart ROM.)
You mean by adjusting the (50pf) trimmer (next to the crystal) that I mentioned earlier in this thread?
I replaced that as well by the way, but I'm not able to hit exactly 14.31818 MHz (my digital multimeter doesn't show that many digits but displays "14.317 MHz"). I can only go lower (i.e. "14.316 MHz") when adjusting it. If I hold an uninsulated screwdriver to adjust the trimmer I can go above that, so if the new crystal is the cause of my colour problem (it's a little lower than specified), is there something I can do to increase its frequency?
I adjusted it while having a game loaded, but still just black & white.
I read somewhere that the 2.7 uH coil (looks like a resistor) near the game port has something to do with the color burst. Would it be an idea to change it? I did receive one of those along with my other spare parts, but didn't think they usually "wear out".
Are there other components or circuit traces worth looking into for all of this?
From your previous posts, it's not quite clear if you are generating the proper color burst signal. To make sure that the color burst is working correctly, you should first verify that the 3,58MHz signal is present at pin 1 of B12. If so, you would next need to connect your scope to pin 12 of B12 and check for the actual color burst. Or just check the composite output itself to see if the burst is present in the graphics modes. You can find plenty of references online on what the composite signal should look like. Or just post a pix of what you see on your scope.
As for the accuracy of the 14.31818 MHz clock, I believe you said that you replaced the 50pF trimmer with a 200pF one. That would definitely cause the frequency to be too low. If you don't have the correct one, you could perhaps swap with the HUE adjust, C3. The circuit should work fine without C3 albeit with non-standard colors.
Yes, I did have a 200 pF trimmer (that's the only value I could find in my parts box to replace the one which appeared bad -but in hindsight probably wasn't).
But having received lots of spare parts I replaced it a couple of days ago with a 50 pF trimmer (which is the value printed on the PCB).
I have checked for the 3.58 MHz signal and believe it was correct, but apparently my multimeter doesn't show enough digits to determine how accurate it is. Is the frequency very critical?
I'll try your suggestion with the scope and also look for online references to compare with what I see/should see.
Assuming you're measuring the frequency at a buffered output of the oscillator, it's probably close enough. I wouldn't worry about the last few digits; the meter probably isn't accurate enough to measure with that precision anyway. A look at the composite signal on the scope should be most telling.
I'm not sure if there's buffering involved or not. Using various Apple II repair manuals found online I've found out which pins on the motherboard output which frequency readings and they all looked fine to me. I can recheck this and which pins I used but IMHO it all looked fine to me, which was why I suspected the tolerance demands to be tighter than I first assumed.
Regarding the NTSC colour burst signal, the following is what I'm looking for to show up on the oscilloscope, right? (source: understanding the 1VPP composite video signal)
NTSC_colorburst.png
Silly newbie questions, but how should I set up my oscilloscope to measure this? I have a cheap, simple and small Jyetech DSO-150 (specifications here) with various modes, settings and configurations (AC/DC/GND coupling, Auto/normal/single trigger mode etc.)
I assume I should attach the scope to GND on the motherboard and the video ouput pin?
Ah. Sorry but this really isn't an oscilloscope. For one thing the bandwidth is way too low to view a video signal. So unless you can get your hands on a real scope, you'll have to make do with other troubleshooting techniques like you've been doing. BTW, if you do get a scope, note that Figure 2 above is actually inverted (i.e. viewed from a broadcast TRANSMITTER perspective). The actual composite video waveform that you'll see on the scope will have the sync pulses at the bottom.
I see. Well, you usually get what you pay for ;-)
Unfortunately I don't have access to anything better, so back to the old fashioned troubleshooting methods...
I've now done several things to troubleshoot:
- replaced Q3 (2SC945)
- replaced Q6 (2SC945)
- removed C3 (color tint adjustment)
- removed C2 (47 pF) capacitor next to C3 trimmer
- removed a decoupling capacitor next to the above
I removed the capacitors in case they had short-circuited as I've read can happen, but put them back again afterwards since it made no difference.
And with my digital multimeter I've checked several places as suggested earlier in the thread. The frequency and voltage readings on the C pin of Q6 doesn't make sense, indicating something is wrong earlier in the chain (I replaced Q6 just to be sure it's not the problem).
Next I loaded a game (to get colour output) and check for 3.58 MHz.
I found the book "Understanding the Apple II" (page 8-12 "video generation") useful for tracing the signal as shown below:
A2_video_schematic.png
No, 3.58 MHz is NOT present on pin 12 of B12 (74LS11), but I do find 3.58 MHz on pin 2 of B13 (74LS02), so that means the source is OK.
But why does the signal stop after B13? I've already replaced that IC so it should be OK.
UPDATE: problem solved !!!
I first found a broken trace which resulted in pin 2 of B14 (74LS02) not being connected to the rest of the circuitry, but reconnecting it made no difference.
then I found a broken trace between pin 1 of B12 (74LS11) and pin 1 of B13 (74LS02). I had to desolder and remove both IC sockets and replace them with new ones (they were badly oxydized anyway) and this made the colour come back!
Strangely there's still no 3.58 MHz reading at pin 12 of B12 (74LS11), but perhaps this was only supposed to be a short "burst" (i.e. color burst) and not something I can pick up with my digital multimeter set to read frequency rates, but only an oscilloscope?
So all looks well now, and even the "color tint" trimmer works as it should.
And yes -just as commented earlier, by adjusting the trimmer next to the crystal I could see the colour disappear and reappear again. I didn't have to be super-accurate when adjusting the trimmer, but my multimeter didn't pick up on the changes (it showed 3.579 MHz, sometimes jumping over to 3.580 MHz, but back again. It's just not accurate enough). A normal capacitor could probably be used, but I assume the trimmer capacitor is probably very useful when parts age and change their values.
Phew! That's a lot of work done, so now I can finally put the computer together again and actually use it!
Congrats on fixing the problem!
And yes, the output of B13 will be bursts of 3.58MHz. Wider than the actual color burst in the video signal but nevertheless not continuous. Thus the lower reading that you got when trying to measure it's frequency.
Thanks! That was a great relief, only to have another issue (rolling display, for others joining) introduced.
Good to know as I was looking for a continous 3.58 MHz signal with my multimeter. And luckily the new crystal worked (despite not being the Vishay/Dale type discussed in this and this thread among others). If case someone reading is wondering about that crystal, here are a couple of links:
Vishay/Dale XT9UNLANA14M31818 (14.31818 MHz crystal)
Mouser part no. 73-XT49U1431-S (obsolete) (PDF datasheet)