I've posted this question in the Apple II Enthusiat's group on Facebook, but I haven't had a ton of luck with the reponses - even from the experts. I'm hoping to get some solid answers before anyone pulls out the soldering iron.
I have an Apple IIc+. It works nearly 100%. The only real issue that it has is that the internal drive wants to continue to keep spinning when it is accessed. It won't stop spinning. It still reads and writes just fine. For example, I can be playing Ultima IV and have a perfectly normal game, starting, playing and saving, but during the entire time, the drive won't stop spinning.
Now, that's not entirely the case. If I boot Apple ][ Desktop from the 3.5" drive, again, it will continue spinning until the IIc+ seeks other drives.. That's right. If I run a command to "Search all drives" and I have an external anything connected (3.5", 5.25", wDrive, etc), the drive will stop spinning.
Yes - I've swapped out the drive. That didn't solve the issue.
I've heard that the likely culprit is a tempermental IWM. I've also This being a IIc+, Uncle Bernie's IWMless won't help me here. So I found a regular IIc on Facebook that I bought for $100 so I could put IWMless in it and harvest the IWM from it. But, it's an A1S4100 and I've been told to keep it pristine and not mess with it.
I've also heard that there are LS chips that could be the issue, but the ones I've been told about were IIc LS chips that I couldn't find on my IIc+ MB.
I don't have the skillset (or spare chips) to start swapping out chips. I know people who do have the skillset, but they aren' t exactly next door.
My big question here is - is it possible to test the functionality of the IWM (and perhaps the LS chips) to isolate the issue without frying my IIc+?
1st question: do you have the computer's schematic? Without it, all talk of troubleshooting is completely inutile.
2nd: does this spinning happen when using DOS?
The Apple 3.5" drive spins its motor when the drive is enabled (when it's /ENBL input is low). The IWM chip has two enable outputs (since the original design was to support only two drives). But the IWM outputs don't connect directly to the drive's inputs. There is a complicated circuit in between that includes the custom MIG chip and others. There's even a LM555 timer that connects to the internal drive's /ENBL. Why would there be a timer there?
The timer is triggered (starts) when the signal from the MIG goes active to enable the internal drive. After that signal goes inactive, it turns on the integrator made of the charging resistor R32 (220 KΩ) and storage capacitor C42 (4.7 µF). Until this charges up to the threshold voltage, the timer keeps the internal drive's motor turned on. But the timer will be reset by the computer enabling any external drive, and the internal drive immediately disabled.
So the first suspects given the symptoms are R32 and C42. These passive components are simple to test and cheap to replace compared to any IC (especially custom chips that cannot be bought like the MIG or IWM).
2nd Question: I don't know. If you mean any DOS, I mean, it happens with ProDOS, but IDK about DOS 3.3 (I don't think DOS 3.3 support 3.5" disks).
I have some Capacitor and resistor hunting to do. I took a lot of pics of my MB but I can't find those two so it's gonna be time to open the case again and look.
Thanks for some useful intel. I'll see what I can do with it.
IMG_5821.jpeg
IMG_5822.jpeg
IMG_5818.jpeg
Sorry for the large pics - didn't post well from the phone and when I got to my iMac, it was too late to edit.
First pictures is of the back side of the motherboard.This was the only place where I saw that there was any discolorment. The 2nd image shows where it is on the front side the MB. Its actually just beneath the stripe under the MC1489N chip. Does this chip have anything to do with disk drive activity?
The 3rd images shows the C42 capacitor and the R32 resistor (number got chopped off but its the one directly to the right of the IN4 diode.
Physically, they all look fine. I guess I'll have to break out the multimeter next.
If that blue drop tantalum cap is shorted (which is rather common), it won't ever charge to the 555's threshold voltage, so the enable signal to the internal drive will stay active instead of stopping after ~ 1000 msec.
The wet spot is suspicious, but I don't think it's related. The tantalum capacitors don't contain any liquid (aluminum electrolytics do). Just clean it with alcohol (or flux remover is OK) and cross your fingers.
The MC1489 is unrelated to the floppy drive. It's a serial receiver for the serial ports.
Tantalum capacitors often fail in a shorted state. Most of the time they explode if there's enough current passing through them. But often they just sit there...
I'd replace it just because it's an old tantalum capacitor. Along with R32 as robespierre suggested. Then report back.
If you're really stuck and it is the IWM, the same device exists in every Mac Plus. I've harvested IWM chips from those motherboards in the past since they often come up for sale on ebay for relatively low prices.
This is just uncleaned flux. It means someone had to unsolder/resolder whatever element is on the other side.
Can the blue tantalum be replaced with a yellow one with the same specs?
Tested the resistor. It SLOWLY climbed and landed at 187kOhm
Yes. It can sometimes even be replaced with an electrolytic, although in this case I'd stick with the tantalum, mostly because of their tighter tolerances (10% vs 20% for electrolytics).
The capacity in microfarads (4.7 uF) and voltage rating (16v) are what is important.
You can up-rate the voltage to a higher value if desired for extra reliability or parts availability, provided the physical size of the device fits the available space (careful with electrolytics - they're often physically larger)
Edited previous post but just saw you already responded. Tested R32 and unlike other resistors along that row, it SLOWLY climbed and stopped at 204 kOhm, then stopped at 190 kOhms. It keeps changing but never got to 220 kOhms.
Did you take it off the board to measure the resistance? if you're measuring it in circuit and there's a charged capacitor connected to it, it will taint the resistance reading. De-solder and lift one leg to do a proper measurement.
If you removed it from the board and it drifts like you describe then it's no good.
Well, I dont know how good my technique is but I pulled the whole MB so I could test those two components from the back. I could not confirm a short in the C42 capacitor and the R32 resistor climbed to 207KΩ- within 10% of 220.
sigh
An in-circuit test will always struggle to read the true value of a component, because what the instrument is really measuring is the equivalent circuit of the whole board. In this case, the current from the ohmmeter is not only passing through R32, but also through everything connected to it. That will typically make the meter read low.
By the late 1980s, the old "carbon composition" resistors were rarely used in new computers. Those consisted of graphite mixed into a fired ceramic, brick-like material. They were known for increasing their value over time. The replacement film resistors don't have that problem, so R32's value isn't really in doubt. What you want to look for is whether it is physically cracked, has corroded leads, or bad soldering to the board (poor solder fillet, or cracked solder).
The tantalum cap C42 might fail not by its value changing, but by an internal short. If you attach the leads to C42's pins and measure resistance, you will again be measuring the whole circuit, but at least if the resistance is below 10 Ω or so, you know there must be a short. The leaded tantalums are still available, and an aluminum electrolytic, or (more expensive) multilayer ceramic type could also be used. Neither R32 or C42's values are crucial: they just need to have a "good enough" product to get a long enough delay from their RC time constant. Resistance times capacitance has units of seconds, so 220 kΩ × 4.7 µF = 1.034 s. These components were chosen to get around 1 second delay time.
Why the 1 second delay: the disk access might stop and then quickly start again. It's better from a transfer speed and data integrity standpoint if the drive keeps spinning instead of stopping and starting.
If neither R32 or C42 appear to be bad, then the search needs to spread out to other components that affect /INTENB, like D2, the 555, or the 74LS11. An oscilloscope helps a lot because it tells you what the voltage is doing at each step.
There is no point replacing components which under test, show no signs of failure.
I don't agree that there has been any rework in this area. The datecode of the chip there matches the one next to it ('89) and the machine's production run, and the solder fillets look the same. They were wave soldered.
The reason I think it's suspicious is the moisture had to come from somewhere. There may be leaking aluminum electrolytic caps on the board or in the drive.
It is flux, you can tell by the brownish color. It is definitely not "moisture". I agree that nothing was replaced, since these are all vias under the chip. It is quite possible that someone was trying to desolder something else nearby and ended up misidentifying its location, melting the vias inlead. I have made this mistake myself on several occasions when flipping a large board. You can even tell that there is quite a lot of solder on the via sitting between the chip and the ceramic cap, that it's bulging on the front side. This is way more solder than a normal factory via.
I think you may be right, but I would like to smell & touch it, myself, before pronouncing ex cathedra.