MegaSTe hardware problems -help locate it please!

Troubles with your machine? Just want to speak about the latest improvements? This is the place!

Moderators: Mug UK, Zorro 2, Greenious, spiny, Moderator Team

User avatar
Fujiyama
Atari Super Hero
Atari Super Hero
Posts: 509
Joined: Thu Jul 12, 2007 8:21 am
Location: Norway

Re: MegaSTe hardware problems -help locate it please!

Postby Fujiyama » Wed Sep 13, 2017 6:37 pm

Finally a few free moments to work on the MSTE again :D

In preparation for recapping the PSU I've measured the various voltages and they all seem to be within spec. How "clean" and stable the voltage rails are is another question which I have no idea.
Next I've started to check the IC sockets and have so far removed and re-inserted all the DIL type ICs. The PLCC chips are another story as I need some more time for that, given that I don't have a dedicated PLCC extractor tool, but very carefully pry them out.

It may very well be that there is some contact problem(s) somewhere, but at the same time things seem to point towards the opto-isolator (U-302). I say this because I finally got the computer to boot, so I tried out Notator again, but as before no MIDI input. But then I realized (I might have mentioned this before) that the missing keyboard response (no click while the TOS 2.06 memory test is running), and no response when pressing the keys as well as missing mouse response had to do with having the MIDI IN cable connected from the synth.
Furthermore I attached my old MIDI input-selector box (which I also added a "soft THRU"/"hard THRU" switch to, utilizing the non-standard MIDI connection so I could play other synths without running a sequencer program) and the "MIDI activity" LED I added lights up constantly!
Now, this is a DX-7 I attached, which is know to output an "active sensing" message, but this should flash the LED, not light it constantly. To me, this indicates a short between the MIDI IN and OUT jacks. An estimated guess tells me the opto-isolator is broken.
I decided to try out my MIDI switchbox before soldering new MIDI cables for testing.

Another discovery: I mentioned that the MSTE schematic says that U-302 is a Sharp PC-900, and that the one actually used was another type. Well, I misread the markings. It actually says H11L3 (not HIL3 as previously quoted), and that IC is still available for purchase without going to eBay etc. So that's good news. So what do you think? Does it sound like a broken opto-isolator could be the reason for my problems? I could desolder it, put in a socket and try out a new H11L3.

Well, that's a little update. I'm not giving up quite yet :wink:
Mega STe | MonSTer with dual IDE-CF memory card adapter | STe | SM-144 |NEC Multisync 1990SXi | IDE doubler | ST_ESSC | RSVE | ICD Link II | Link '97 | HD floppy drive/AJAX | HD floppy module | Minolta PCMCIA card-drive | Realtime Clock module | Discovery cartridge | Unitor-2 | Export | Combiner | Steady Eye | Human Touch | Unicorn USB

siriushardware
Captain Atari
Captain Atari
Posts: 341
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: MegaSTe hardware problems -help locate it please!

Postby siriushardware » Fri Sep 15, 2017 5:18 pm

The default state for the output device in the optocoupler is OFF, so that means that with the power on to the machine and no MIDI IN activity you should measure a steady HIGH state (5V) on pin 4 of U302.

If you then connect a midi source and generate a blizzard of MIDI activity (rolling the pitch bend wheel back and forward is an excellent way to do that) you should see the voltage temporarily lowering. Using a logic probe or a scope would be a better way to observe this - then you would see a train of low pulses coming out of the opto output whenever there is a big burst of MIDI activity.

If you see a steady LOW (0V) on the opto output regardless of whether there is MIDI IN activity then it is possible that the opto output has failed and is pulling that point in the circuit low.

If you do find a steady low state on pin 4 of the opto, remove the opto from the PCB and make the measurement again.

If the voltage on the pad normally occupied by pin 4 of the opto has now risen to 5V, your opto is faulty.

If you've removed the opto but you still have a low (0V) state on the pad normally occupied by pin 4 of the opto then your problem is elsewhere - could be that the pullup resistor (R307, 4K7) is cracked / open circuit. The input of U301F would also normally be a suspect but I believe you've already changed U301. It is also possible that the input (pin 2) of the 68B50 ACIA has failed short-circuit / low resistance down to 0V, but let's not go there for the moment.

User avatar
Fujiyama
Atari Super Hero
Atari Super Hero
Posts: 509
Joined: Thu Jul 12, 2007 8:21 am
Location: Norway

Re: MegaSTe hardware problems -help locate it please!

Postby Fujiyama » Mon Sep 18, 2017 8:33 pm

That's useful info as always which I'll get back to, but I've made some new discoveries.
First, I replaced the H11L3 opto-coupler (U302) with an IC-socket, then put it back in, but guess what? It makes no difference if the opto-isolator is present or not! The LED of my MIDI input selector box lights constantly regardless, so I suppose there might be a short-circuit somewhere.

Just to make sure there wasn't something wrong with the box (which is probably 25 years or so old) I pulled out my trusty old STe (works fine except its floppy drive and doesn't have a MonSTer board installed yet, as planned, which has completely spoilt me with impatience when booting and starting up software!) and tried out what you suggested earlier: set the box to "MIDI THRU" mode, setting the keyboard synth to "local off" then play a few notes -yes! it does play the synth, so it works fine. By setting it to the standard "MIDI OUT" mode it wouldn't play anything as expected.

About the MIDI input selector box: I finally took the time to open it up, check the wiring and draw up a schematic which I'm enclosing here.
Its main function is (or should I say "was" as I since then replaced it with a Roland A-880 MIDI patchbay/merger) to allow sysex dumping from a variety of MIDI modules' OUTs without unplugging cables. The everyday setting is to have it selected to the master keyboard input, which in turn controls a variety of MIDI modules.
I also took advantage of the rather unique (and may I add; non-standard) MIDI OUT port so that by setting it to the (non-standard) "THRU" mode I could play around with my synths without having to run any MIDI sequencer (or "software THRU" type programs) first. As far as I remember the computer has to be powered on though, but that's no biggie.
By switching it over to the normal MIDI "OUT" mode I would have to run a MIDI sequencer or other software in order to play the MIDI modules attached to the OUT of the ST.
MIDI IN selector with Atari ST THRU.png

The way I attached MIDI cables for testing via the box was as follows:

Synth OUT to box IN#1
Box OUT to synth IN
Box "ST OUT" to Atari STe OUT
Box "ST IN" to Atari STe IN

The synth was set to "Local OFF" and the switch on the box set to "THRU mode".
The LED dimly flashed whenever I played something on the keyboard.
Good thing I have two Ataris for troubleshooting!

I've taken a closer look at the MSTE schematic around the MIDI IN section and something puzzles me....
mste_MIDI_section.png


For some reason I found pin 2 of U302 (the opto-isolator) to connect to GND! And that's with the IC removed from the socket!!!
So I traced the path from pin 5 of the MIDI IN connector, through L306 and there it was! I don't know exactly how inductors work or their purpose here, but unlike the schematic says, they both (one's connected to pin 4 of the MIDI IN connector, seemingly marked "L309", but according to the actual MSTE PCB silkscreen it should be "L305") have three pins, not just two!
The centre pin of L306 goes to GND -it's not a short circuit but an actual trace. I don't know what to make of that, but apparently it was designed that way.
Anyway, like I said the MIDI input selector box's LED lit up constantly when connected to the MSTE, even without the opto-isolator IC present, so I suspect there's a short circuit or malfunctioning component somewhere along the line. I don't know if this has anything to do with the sometimes working keyboard/mouse (and sometimes not, likewise the keyboard click sound) but it does point in the direction of a bad connection at the very least.

I've just received a missing connector for my oscilloscope, so when that's in place I'll check what you suggested and look closer at R307 as well.
You do not have the required permissions to view the files attached to this post.
Mega STe | MonSTer with dual IDE-CF memory card adapter | STe | SM-144 |NEC Multisync 1990SXi | IDE doubler | ST_ESSC | RSVE | ICD Link II | Link '97 | HD floppy drive/AJAX | HD floppy module | Minolta PCMCIA card-drive | Realtime Clock module | Discovery cartridge | Unitor-2 | Export | Combiner | Steady Eye | Human Touch | Unicorn USB

siriushardware
Captain Atari
Captain Atari
Posts: 341
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: MegaSTe hardware problems -help locate it please!

Postby siriushardware » Tue Sep 19, 2017 12:18 am

Just check one thing: With -nothing- connected to the MIDI IN port on the computer, do you still have your measured short from MIDI IN pin 5 to 0V on the MSTe? I just want to be sure that the short you are measuring is not actually down the connected MIDI cable, through the equipment at the other end and connected to 0V on the MSTe via the shared mains earth connection.

As you've noticed L305 /L306 are only supposed to have two actual connections. However, components with only two pins are prone to wobble back and forth so sometimes coils / inductors / transformers have extra legs which are not connected to any part of the coil winding. Their job is just to provide more solder anchor points to the PCB.

If the schematic is correct, there should not be a direct connection from any part of the inductor L306 to 0V. If you really do have a connection to 0V through the inductor, the inductor is faulty - it may be that there is a short between the coil winding and the third pin.

So here's the good news, L305 and L306 are non-essential from the point of view of MIDI functionality. The ST has lots of sharp-edged high frequency waveforms running around inside it, and if allowed to exit from the machine via any of the long cables connected to it, these signals could interfere with radio reception in the local area. The purpose of L305 and L306 is to filter out radio frequency interference and prevent it from travelling out through the MIDI IN port and being radiated by the MIDI cables.

For diagnostic purposes, you can remove L305 and L306 and replace each one with a wire link to see if that fixes your problem. Note that if you do, you should ignore the third pin which is soldered to the ground plane on the PCB. Your wire links should connect the other two pins together, so that one wire link connects MIDI IN pin 5 directly to opto pin 2, and the other wire link connects MIDI IN pin 4 directly to R308.

You don't need a scope to check R304, the pullup resistor. With the power off, measure the resistance across R304 in-circuit. It should be 4.7K, or something lower than 4.7K. If you get a reading higher than 4.7K, desolder one end of R304 or remove it altogether and measure the resistance of the resistor by itself.

czietz
Hardware Guru
Hardware Guru
Posts: 413
Joined: Tue May 24, 2016 6:47 pm

Re: MegaSTe hardware problems -help locate it please!

Postby czietz » Tue Sep 19, 2017 5:51 am

siriushardware wrote:As you've noticed L305 /L306 are only supposed to have two actual connections. However, components with only two pins are prone to wobble back and forth so sometimes coils / inductors / transformers have extra legs which are not connected to any part of the coil winding. Their job is just to provide more solder anchor points to the PCB.


No, actually the third pin of these components is not just for mechanical support. The schematic gives it away: These are not simple inductors, but T-type EMI filters, comparable to this one:
Image

I.e. even though they are marked Lxxx, they consist of two inductors and a capacitor like this:
Image

Hence, it is perfectly normal that the center pin is connected to ground.

User avatar
Fujiyama
Atari Super Hero
Atari Super Hero
Posts: 509
Joined: Thu Jul 12, 2007 8:21 am
Location: Norway

Re: MegaSTe hardware problems -help locate it please!

Postby Fujiyama » Tue Sep 19, 2017 11:40 am

What Czietz is saying makes sense (given the circuit trace from the middle-pin of the inductor which goes to the GND plane) but also makes it all more confusing.
I will do some more troubleshooting tonight and compare the results with that of the plain STe. Comparing the two schematics they appear to be identical when it comes to the MIDI ports.

And yes, Siriushardware; I checked with nothing connected to the MIDI ports, and with power off (actually I think had removed the motherboard from its case for checking the solder connections underneath for any shorts or cold joints). That's why I was so surprised to find that MIDI pin going to GND as nothing external (even the monitor, hard drive etc.) was connected.
It'll be interesting to see how it compares with the STe MIDI IN port.

UPDATE: I haven't had time to do any extensive tests, but using a multimeter set to continuity testing I found out that on the STe, none of the MIDI IN port pins are connected to GND and the MIDI OUT portonly had pin 2 connected to GND. Just like the schematic says it should.
Looks like I should have a closer look at the inductor in the MSTE and compare it with one of the other inductors.
Are they all the same? The schematics say nothing about their values.
Mega STe | MonSTer with dual IDE-CF memory card adapter | STe | SM-144 |NEC Multisync 1990SXi | IDE doubler | ST_ESSC | RSVE | ICD Link II | Link '97 | HD floppy drive/AJAX | HD floppy module | Minolta PCMCIA card-drive | Realtime Clock module | Discovery cartridge | Unitor-2 | Export | Combiner | Steady Eye | Human Touch | Unicorn USB

siriushardware
Captain Atari
Captain Atari
Posts: 341
Joined: Thu Aug 21, 2014 7:55 pm
Location: UK

Re: MegaSTe hardware problems -help locate it please!

Postby siriushardware » Tue Sep 19, 2017 5:41 pm

czietz wrote:The schematic gives it away: These are not simple inductors.


True, although the fact that the drawing represents them as normal inductors is misleading. But yes, I've only just noticed the words 'T Filters' near those components.

In the light of that clarification from Czietz, it is actually possible that the capacitor in the T-filter assembly has gone short circuit.

You can verify this by taking out the suspect coil/filter and measuring the resistance from each terminal of the component to the others. Your meter should see a dead short, more or less, between two of the pins. From those two pins to the third pin (the one which normally goes to zero volts on the PCB) there should appear to be infinite resistance, although you may see a brief flash of lower resistance when you first connect the meter. If all three pins appear to be constantly shorted to each other the device is faulty.

If that proves to be the case, what I said before still applies - these components are just interference filters and you can bypass a suspected faulty one by linking from one side to the other and leaving the third (0V connected) PCB hole unconnected. If that solves the problem, by all means try to source a replacement filter.

czietz
Hardware Guru
Hardware Guru
Posts: 413
Joined: Tue May 24, 2016 6:47 pm

Re: MegaSTe hardware problems -help locate it please!

Postby czietz » Tue Sep 19, 2017 8:13 pm

Fujiyama wrote:Are they all the same? The schematics say nothing about their values.


The STE (not MegaSTE) service manual mentions them as "noise filters" and gives several possible suppliers. Luckily, for the TDK part number ZJSR5101-102 a datasheet can still be found: http://pdf1.alldatasheet.com/datasheet- ... 1-102.html


Social Media

     

Return to “Hardware”

Who is online

Users browsing this forum: No registered users and 5 guests