MB-6582 CS troubleshooting

[highrider]

Once again I ask for your help. Hopefully this will be my last troubleshooting session.

I have successfully connected the CS and LCD. The MB6582 is functional through all 4 SID engines but there are a few quirks that don't let me seal this beast of synth inside the box just yet.

I'll Start with the LEDS

1) the whole vertical row 3 of the mod matrix is not working.- fixed

2)Waveform LEDs don work (although waveforms change when I press the respective button)- fixed

3)Control ENV MISC KNOB LEDS don't work (again this section is fully operational)

-fixed

4) LR SID section LEDS don't work while sids can be selected.- fixed

5) Sync MODE LED doesn't work ( just the LED)-was not a problem

6) SID ENGINE 2 LED doesn't work as well as the button i.e. it can't be selected even though I can play it through MIDI CH2 - Fixed

Buttons

1) The aforementioned SID Engine 2 button. I can change the engine via the menu encoder and buttons. - fixed

1)2) The horizontal row of the mod matrix. I presume that it's not working because any two buttons of this row pressed at the same time don't change the MODE.- fixed

Encoder

1)- Just the Depth Delay Encoder dowen't work at all.- fixed

One more thing that may help :

When I select LFO waveforms the LEDs in the LFO 123456 section light up but very dimly (one at a time with every push of the button in the wave selection).

I have noticed that through out the CS this happens quite often . The light is so DIM it can barely be distinguished from being fully of.

I have went through all the connections between the CS and Base and have found no shorts and no breakage in connection.

After tracing the tracks on the CS diagram I couldn't establish any pattern to determine the source of the described faults.

To the best of my knowledge the PIC's and SID's are in full working order. Patches change/notes play through all SID engines.

So far I think that I might have just damaged the non working LEDs apart from row 3 in the MOD matrix. I will do the LED test next and post the results.

Please reply if you have experienced similar problems or have any advice to share on how to carryout the troubleshooting.

Many thanks,

Vlad

Edited March 9, 2010 by highrider

[Filch]

If you haven't already, double check that your LCD board isn't shorting on the CS. This has been the cause of many weird and wild problems in the past.

[highrider]

If you haven't already, double check that your LCD board isn't shorting on the CS. This has been the cause of many weird and wild problems in the past.

Thank you for the advice Filch. I have removed the LCD from the CS. Apparently there were no shorts because all of the problems persist. Otherwise it works as it has before.

V!

[highrider]

I managed to fix two sections . There was a damaged track going from the diode to the LED in the CONTROL and in the SID ENGINE sections. Now the LEDS light up like they should!

Edited February 20, 2010 by highrider

[jbartee]

Encoder

- Just the Depth Delay Encoder dowen't work at all.

This might not be your problem, but depending on the encoder brand you're using, they can be quite twitchy. I had several of my alpha brand mouser encoders stop working after the pins got just *slightly* bent. Bending them back straight fixed the issue. I actually had to open up one or two and bend the contact wipers inside the encoders. They're very simple mechanically, so if the issue is with the encoder and not with the wiring, they're easy to fix.

[highrider]

This might not be your problem, but depending on the encoder brand you're using, they can be quite twitchy. I had several of my alpha brand mouser encoders stop working after the pins got just *slightly* bent. Bending them back straight fixed the issue. I actually had to open up one or two and bend the contact wipers inside the encoders. They're very simple mechanically, so if the issue is with the encoder and not with the wiring, they're easy to fix.

Thanks for the advice! Unfortunately this didn't solve my problem.

I opened the encoder (from Voti) made sure the wipers are straight and the right height. I even put the metal ring back to where it was (to make it detent again). After putting it back together it still doesn't change value.

Than I changed the depth value with the menu encoder, all the way to -127. Interestingly after trying to modify it with the encoder in question the value would rapidly go up all the way to 127 regardless of which direction I would turn the encoder. Also, If I keep turning the encoder anticlockwise the value changes only in very small increments (1-2) but immediately returns to 127. It's as if it's been pulled there.

I am confident it's not the wipers unless this is what you experienced with your encoders?

V!

Update :

As it turnes out with an encoder stuffed (i tried two different ones) the result is the same. Depth values go up to 127 and barely nudge from that value. On certain occasions the CS freezes on the envelope screen i.e. everything works but I can't navigate back to any other editing screen. Note : this doesn't happen when there is no encoder stuffed.

I'm shure there is a short or damaged track/pad somewhere thats causing this behaviour in the envelope section but following the tracks from where the encoder used to be I see nothing wrong with the PCB.

Could it be the 165N's?

Edited February 20, 2010 by highrider

[Wilba]

Encoders only going in one direction suggests that one pin is always shorted to ground. This could be at the encoder (one pin of the encoder is always connected to ground) or it could be in the tracks leading to the bottom, the cables or on the base PCB.

You should be able to remove the ICs in U16-U19 (to be sure) and with power off, test continuity between ground and the two "signal" pins of the encoder, while turning the encoder, you should get "beeps" from the multimeter from both pins. Both pins will be connected to the 5V rail via 10K resistors but this shouldn't affect continuity testing between the pin and ground, I think.

[highrider]

Encoders only going in one direction suggests that one pin is always shorted to ground. This could be at the encoder (one pin of the encoder is always connected to ground) or it could be in the tracks leading to the bottom, the cables or on the base PCB.

You should be able to remove the ICs in U16-U19 (to be sure) and with power off, test continuity between ground and the two "signal" pins of the encoder, while turning the encoder, you should get "beeps" from the multimeter from both pins. Both pins will be connected to the 5V rail via 10K resistors but this shouldn't affect continuity testing between the pin and ground, I think.

I have found a short between pins D0 and D1 of JD3. These pins are also shorted to ground. Resoldering the wires hasn't solved the problem. Taking the "DEPTH" encoder hasn't helped either. I triple checked the tracks and it looks like there isn't anything to cause a short starting from U17 all the way upto the encoder.

Also I have uploaded the LED test up. All the LEDS that didn't work still did not light up apart from the sync led. Does this one work only when there is a master MIDI device connected to the MB6582 and it's sending tempo info?

Thanks

[Wilba]

"Sync" LED just isn't used yet... I had planned for it to toggle between master and slave (i.e. LED on when it's slave to MIDI clock) but I never got around to actually coding it up (or suggesting to TK to do it).

Re depth encoder, can you check you're testing the right pins? The Env Depth encoder isn't connected to D0 and D1 of J3, it's actually D2 and D3. See pic.

Regarding shorting to ground... this CAN happen depending on the position of the encoder, i.e. those pins WILL short to ground as you turn the encoder, that's how it detects the turn, as one or both will be shorted to ground.

[highrider]

"Sync" LED just isn't used yet... I had planned for it to toggle between master and slave (i.e. LED on when it's slave to MIDI clock) but I never got around to actually coding it up (or suggesting to TK to do it).

Re depth encoder, can you check you're testing the right pins? The Env Depth encoder isn't connected to D0 and D1 of J3, it's actually D2 and D3. See pic.

Regarding shorting to ground... this CAN happen depending on the position of the encoder, i.e. those pins WILL short to ground as you turn the encoder, that's how it detects the turn, as one or both will be shorted to ground.

Wilba, thanks that picture really helped. The fault was in the IC in U18 after replacing it with a new one everything just worked perfectly including the MOD Matrix and all the LED's that wouldn't light up before. Now I have a question, is this normal? Does the 74HC165 actually make all of that work or have I fixed something without knowing it?

Finally, after putting the case together I've encountered two more problems. The EDIT button and LED don't work (I assume because of the LCD causing a short) and some random LED's are on at a very low brightness even when not selected. Has anybody had this problem before?

Thanks for your help,

Vlad

[Wilba]

Wilba, thanks that picture really helped. The fault was in the IC in U18 after replacing it with a new one everything just worked perfectly including the MOD Matrix and all the LED's that wouldn't light up before. Now I have a question, is this normal? Does the 74HC165 actually make all of that work or have I fixed something without knowing it?

Maybe, the LEDs don't depend on working 74HC165 ICs, but they do share two common tracks, perhaps there was some interference.

Finally, after putting the case together I've encountered two more problems. The EDIT button and LED don't work (I assume because of the LCD causing a short) and some random LED's are on at a very low brightness even when not selected. Has anybody had this problem before?

What kind of LEDs are you using? Ultrabright LEDs with low resistance resistors might result in some "ghosting"... LEDs being lit because that row was lit in the previous column of the LED matrix.

[highrider]

What kind of LEDs are you using? Ultrabright LEDs with low resistance resistors might result in some "ghosting"... LEDs being lit because that row was lit in the previous column of the LED matrix.

I'm using either 1000Mohm or 47ohm resistors (the bands are too close for me to distinguish where the first band is) with diffused Red LED's 250 mcd. I have attached a picture of the glowing LED's. This is how the CS appears on start up. I've changed all the 165n's and 595's. I'm confident it's not them.

[Wilba]

It's definitely not 1 gigaohm... more like 47 ohm which is way too low.

Why are you using resistors which you don't know the value, and which are much lower resistance than those specified in the parts list (220 ohm)?

In the build guide, I suggest using a higher resistance for brighter LEDs, and testing the LEDs with resistors first (if you're not using known resistors like SmashTV's red LEDs). I recommend removing the resistors and replacing with 1K resistors, or at the very least, 220 ohm resistors. After you desolder the resistors, you can run the LED matrix test app and test which resistor value you prefer (1K or 220 ohm) by touching them on the pads... only some LEDs will light, but you can judge from that if it's too dull or too bright.

[highrider]

Why are you using resistors which you don't know the value, and which are much lower resistance than those specified in the parts list (220 ohm)?

They were supplied with LED's. Stupid mistake.

[Wilba]

:rolleyes: :smile:

[highrider]

Ok the Edit mode LED and button work. Work as in mechanically work, the LED lights up the screen doesn't change though. What is it's purpose? I assume once it's depressed :sad: all the changes I make to the patch/ensemble will be stored once I save them, no? I'm slightly confused here because I have saved patches without pressing the edit button.

I will change the resistors over the weekend and report back, then I will put the the knobs and the MB-6582 will be finished. It has a pretty white case so photo's coming up :ahappy:

Thanks for your help guys!

[Wilba]

I can't find Edit or CC buttons described explicitly in the V2 manual.

They were in the original V1 specs:

http://ucapps.de/midibox_sid_csB.html

and then "reused" for Up/Down buttons for original V1 hardware.

Since I added dedicated Up/Down buttons to MB-6582, CC and Edit buttons now do what are referred to in the "Special Button Combinations" section of the User Manual:

http://ucapps.de/midibox_sid_manual_fp.html

• SHIFT + UP: CC mode on/off. In CC mode, a "knob" (rotary encoders within the OSC->Knob layer, or analog pots/joysticks) sends out value changes via MIDI Out, e.g. to your sequencer. The values can be recorded for MIDI automation.
  
• SHIFT + DOWN: EDIT mode on/off. In this mode, all parameter changes will be saved automatically into the currently selected patch whenever you change to a new patch. The LED flashes slowly if there is something to save in order to warn you, that the old patch will be overwritten.