A500 joystick direction stuck

[Amiga1992]

I was just testing my newly flashed Gotek and it seems like since the last time I booted it (maybe a year ago) my A500 (rev 5) isn't working as expected.
As far as I can explain, the only symptom seems to be that the UP direction is stuck on joy port 2.

I did nothing to the machine except connecting the Gotek. The same happens with no joysticks connected in any port.
Is this a short, a CIA or what?

Thanks in advance for any light you can shed on this. I never had this problem before!

[edit] found this LemonAmiga thread, seems related. Possible?
http://www.lemonamiga.com/forum/viewtopic.php?t=12219

Really late and tired now, will investigate more tomorrow

[edit2] stupid Tapatalk adds too many spaces, wtf

[Daedalus]

It could be a short - pin 1 of the port relates to the up direction so check that the pin isn't bent against the shield or something like that. Then you can have a look at the traces and solder joints related to pin 1 to make sure there aren't any obvious shorts. After that you're getting into failed components. The multiplexer IC U15 as found in that thread is a likely candidate alright, and fortunately is a very common and cheap chip, well worth popping out and replacing for a couple of dollars at most. After that, the EMI filters and resistor packs should be checked. But the CIAs can be discounted since they're only involved with the fire buttons.

[jotd]

isn't it possible to damage those by plugging/unplugging the joysticks when the amiga is running? (didn't prevent me from doing that, though )

[Daedalus]

Yep, in theory you can damage semiconductors in particular by connecting and disconnecting while powered up (and one of my A1200s even has a sticker to that effect on the keyboard). However, in the case of joysticks the risk is practically nil because no power is going anywhere. The real risk is attaching other devices that have their own power supply, e.g. printers, scanners, monitors, modems etc., which may have different grounding from the Amiga and so cause can cause a surge if the grounds aren't connected first. This is likely what has killed many a CIA in the past...

[Amiga1992]

Originally Posted by Daedalus

It could be a short - pin 1 of the port relates to the up direction so check that the pin isn't bent against the shield or something like that. Then you can have a look at the traces and solder joints related to pin 1 to make sure there aren't any obvious shorts. After that you're getting into failed components. The multiplexer IC U15 as found in that thread is a likely candidate alright, and fortunately is a very common and cheap chip, well worth popping out and replacing for a couple of dollars at most. After that, the EMI filters and resistor packs should be checked. But the CIAs can be discounted since they're only involved with the fire buttons.

Thanks for that Daedalus!
I checked the pins, at least externally, they are not shorted.
I will have a scope today with the multimeter and see what's up. I have not done anything to this computer, I barely use it, so I would not discard a component fail, because I can't even recall if I plugged a joystick on it ever until yesterday.

By the way, how can I know if a component is bad? I never really have done such analysis with a multimeter. What am I supposed to read?
I'm sorry, I know this is an absolute moron/newbie question, but I Never really have done this kinda work, but I'd love to be able to fix it. An A500 is more like an old car, no SMD components to deal with, and seems like right up my alley abilities-wise!

[Daedalus]

Well, if the signals going into the multiplexer are ok but the computer isn't reading it right, it's a fair indication that it's dead, though without an oscilloscope or logic probe you can't check the output to be sure. The signal going into the multiplexer should be +5V at rest, and 0V when the corresponding direction is activated on the stick. I'm not sure which port is which on the A500, but checking them both will show you what it should look like and let you see if one of them is behaving differently. Pin 2 is the up signal from the left port, pin 11 is the up from the right port, so they're the ones to check.

If they're fine, (i.e., both switching from 5V to 0V when the corresponding stick is pushed up), next to check is if left is working fine on the joystick. If left is working fine, and the up signal is reaching the multiplexer but going no further, it's pretty much the multiplexer that's dead, since from that point on, left and up share the same line.

If however the signal isn't reaching the multiplexer correctly, or is reading much less than 5V when no stick is attached, check the resistor bank RP401 by measuring the resistance between pin 2 and +5V (make sure the Amiga is turned off for measuring resistances and continuity!) +5V is available on pin 1 of the resistor bank and pin 16 of the multiplexer. It should be 4.7K - significantly more than that and you'll have false triggering of the up direction. You can also check the EMI filter - pin 2 of the multiplexer should have a low resistance to pin 1 of the left joystick port (pin 11 can be checked to pin 1 of the right joystick port for comparison).

Edit: The A500 isn't a bad machine for troubleshooting - most parts are easy to prod and check without getting too fine or hiding on the underside of the board. You'll learn a lot with a multimeter and the schematics in front of you Just take your time and be careful of shorting nearby pins with the probes.

[Amiga1992]

You're a star, I appreciate this a lot!
Will test and report

[Amiga1992]

Originally Posted by Daedalus

The signal going into the multiplexer should be +5V at rest, and 0V when the corresponding direction is activated on the stick.

Well had a go with the multimeter and I have to say that I don't understand, from the schematics, which pin is which (schematics don't mark the chip nub anywhere so I have no idea how I am looking at it, and schematics don't correlate to location on motherboard!! how am I supposed to read this?), so I probed them all ,and while some do give 5V readings, a lot don't.

I wrote down what I had:



I measured with no joystick or mouse plugged to any port.

Some readings changed a few times, which I thought was weird (from 0 to some value) but maybe it was me losing contact from gnd

I don't know what to make of it now. From reading the LemonAmiga thread, should I be suspecting the resistor pack? I guess I can check after I understand which pin is which, these schematics make no sense to me in regards of direction

Thanks for your help!

[Daedalus]

Yes, the schematics tend not to bear a physical resemblance to the circuit board itself and are laid out to make them more readable rather than to represent the physical pinouts. I should've mentioned that the chip pin numbers start where the notch is in the chip's case and go anti-clockwise from 1 to 16 (or whatever number of pins there are. So in your case, pin 1 is reading 1.70, pins 2 & 3 are reading 5, pin 4 is reading 2.80 and so on, all the way around to pin 16 which is reading 5.

Ok, well that's something anyway - Pin 11 should be 5V - that corresponds to the up signal on the right joystick port. Was there a mouse plugged into either port at the time?

It's normal for the readings on pins 1, 4, 7, 9 and 12 to give funny, changing readings because they're rapidly changing signals - far too quick for the multimeter to make sense of so you can't get a proper reading there. But if nothing was connected to either port, it looks like the multiplexer itself is fine and the fault is elsewhere. Next things to check:

- Make sure the power is *off*
- Measure the resistance between pins 11 and 16. This should be around 4.7K. If it's really hi or infinite (usually shown by a "I" on the meter display, there's either a track break, a bad joint or a broken resistor pack, so use the meter to check the resistor pack itself (pin 6 of the resistor pack is the one to check from pin 1 of the same pack). If that's ok at 4.7K, use the continuity mode to check the traces from there to pin 11 on the multiplexer and to the port itself.

[Amiga1992]

Thanks for clarifying that! I caught the pinout of the IC and was confused, because the schematics make it seem like the numbers alternate between left and right rows



versus:




I will proceed to measure what you told me. Nothing was plugged when I tested.
How to recognize where in the resistor pack pin numbers start?

[Jope]

The diagrams in the schematics "never" match the physical location of the pins on the chips. The lines are positioned so that the schematic stays legible.

Good luck!

[Daedalus]

Yeah, if the schematics kept to the same layout as the physical parts, there's be so much more crossing of signals, splitting of buses etc. that it would make them far more difficult to follow. In this case the schematic has the inputs coming in on one side and the outputs going out on the other side - and the power connections are in a different section entirely!

The resistor pack will usually have a dot or some other mark at one end, and the PCB itself will have a matching dot or notch. In fact, almost all multi-leg components have pin 1 marked in some way, be it a dot, notch or something else that is unique to that end or corner. Generally, the PCB will have a corresponding mark too for when it's being assembled. Sometimes the exact marking varies, but there'll be something. It looks like the 500 uses a box printed around pin 1 separating it from the rest of the pins in the row...

[Amiga1992]

Originally Posted by Daedalus

Yeah, if the schematics kept to the same layout as the physical parts, there's be so much more crossing of signals, splitting of buses etc. that it would make them far more difficult to follow.

Oh, I totally get that.

BUT, why, when you get to the IC itself, not stick to the standard pin numbering? That gets on my nerves. You are connecting lines to it anyway, at least keep the numbers, their positions and the nub mark. It ain't hard!

It looks like the 500 uses a box printed around pin 1 separating it from the rest of the pins in the row...

I think I saw the box. Thanks for that and for withstanding all my noob questions!

Originally Posted by Jope

Good luck!

Thanks buddy, I'll need it! And I can always bother YOU With a message, haha :P

[xArtx]

Take a look at Budgie and Gayle in the A1200 schema.
There are two of each chip! but they couldn't have done it neatly any other way.

You could connect a 10K resistor between the stuck
direction pin and +5V to eliminate the possibility of
the failure of one resistor in the resistor pack.

[Amiga1992]

Hi guys! finally found a second to do this:

Originally Posted by Daedalus

- Measure the resistance between pins 11 and 16. This should be around 4.7K. If it's really hi or infinite (usually shown by a "I" on the meter display, there's either a track break, a bad joint or a broken resistor pack, so use the meter to check the resistor pack itself (pin 6 of the resistor pack is the one to check from pin 1 of the same pack). If that's ok at 4.7K, use the continuity mode to check the traces from there to pin 11 on the multiplexer and to the port itself.

I set the multimeter at 20k and the result was 0.10 or something like that. Measuring between pin 1 and 6 of the resistor pack gave the same result,.. So I am guessing the resistor pack is broken? Or am I reading this wrong? It kind of gave the opposite result you said (leaning towards no resistance instead of infinite resistance)

[Daedalus]

Hmmmm, that is strange. It does indeed sound like you measured it correctly.... That being the case, you shouldn't have the problems you're having. Perhaps I've been confusing (or confused) about the ports. What do you get if you measure between pin 1 and pin 2 of the resistor pack? This corresponds to the up direction of the other port.

The power was definitely off when you measured it, yeah?

[Amiga1992]

Yeah the power was off indeed.
It measures 1.90 between 1 and 2
But now that I re-measured between 1 and 6, it gave me the far away 1 which I think means Infinite. I am confused.

[edit]: here's a video of me measuring the resistor pack and the IC. Now it gave 0.10 again!

[ Show youtube player ]

[edit 2] I realize I had a mouse connected to port 1 *facepalm*, I removed it and now measuring between pins 1 and 2 of the pack give 4.7k as expected. 1 and 6 still give 0.10~0.12

[Daedalus]

Ugh, I wish I had the machine in front of me to poke about with. A low reading like that doesn't prove that the resistor pack is at fault unfortunately - if it was, the direction wasn't stuck so there appears to be some other fault. Looks like it could indeed be the multiplexer that's failed after all.

[Amiga1992]

Changing it should be easy, i should just do it and see if it fixes it. Thanks for your help so far, I will report back!!

[edit] done, bought 5 pieces of SN74LS157N, will desolder the old one and wait for the new one to come. Should get sockets too so I don't need to deal with that again.

[edit2 ] figured it out, it's DIL16/PDIP-16 sockets

[Amiga1992]

Well that was a shit-show of a desoldering . Why was it SO HARD to remove this component? This should have been an easy job!

While trying to remove the thing, I broke C15
I can't read the tiny bitsy number on it so I have no idea what to replace it with.
This seems to be a key fucking component and all, listed on page 9 of the schematics: http://www.amigawiki.de/dnl/schematics/A500_R6.pdf

I don't know what it is doing but it's connected to the 68000 so I take it my Amiga is as dead as it could be for the time being.

So the schematics say 0.33 pico farrad, is there anything else I need to look for?
It looks exactly like this: www.ebay.com/itm/152072835341
Would other types work, like this one? www.ebay.com/itm/171995362926

[edit] I think the capacitor says:
A5E
104

Z

but I can't be sure.