[Snyderman]

So I've been going through and building the SammichFM with only a few small and silly hiccups but right now I'm at the step with voltage checks. I haven't plugged it in at all and am just doing continuity tests. The ground is fine throughout the base PCB and there's no shorts between the OPL 5V and ground but there's continuity on all of the audio 5V and ground (as far as I can tell). Where should I start looking? Are there a few specific places where there's a bad joint that's causing it to go all through the board?

Thanks for help in advance.

EDIT: I have looked, I just couldn't find anything egregious on the base PCB.

This post has been edited by Snyderman: 06 December 2011 - 08:08

[nILS]

"As far as I can tell" - How can you tell? What's the resistance? Does your meter have a "beep" mode, if so does it constantly beep or just for a short time?

[Snyderman]

nILS, on 06 December 2011 - 23:44, said:

"As far as I can tell" - How can you tell? What's the resistance? Does your meter have a "beep" mode, if so does it constantly beep or just for a short time?

My apologies, I'll be more detailed. The continuity tester at all points between ground and Audio 5V beep continuously, with a resistance of 0.050 exactly. It's not exactly 0 however as it is with ground (understandably). It's consistent at all points that I tested (using the diagram, between ground and all orange points). For reference, resistance between ground and OPL3 5V is roughly .408 (I assume that means 408 ohms) and between ground and Digital 5V is about .512.

EDIT: I realize now it's not technically a short, but resistance seems really low.

This post has been edited by Snyderman: 07 December 2011 - 02:24

[nILS]

If your resistance measure of .408 actually means 408Ohms (read: your meter is set to 1k) then 50Ohms between +5V and ground is in fact too low, making it look like you have a short somewhere.

Follow the audio +5V path and see if there's any shorts:

C5, C7, IC3 -> power jumper

power jumper -> IC51, C53, C54, IC52 (IC51/52 have pin 1 and 3 being +5V and pin 2 Gnd, make sure you have no shorts on the pins there)
power jumper -> IC54, C55, R60, R62
power jumper -> IC53, C52, R56, R58

Have some non-blurry, closeup pics?

[Snyderman]

I think I found my problem... I'm pretty sure all of my 10k and 220 ohm resistors have been switched. They look so similar, there's a good chance that's what I did wrong.

[Snyderman]

After fixing my stupidity, I have got the thing working and it sounds good. Now I'm noticing a host of relatively unrelated problems.

First I'm getting some inconsistencies on the LEDs. The top row seems fine, but the bottom row might be fried. The third and fifth from the left seem to light up just fine, but the other three seem to light up very dimly if at all. What is the bottom row supposed to look like? When do they light up?

Second, the controls on the board seem fine, MIDI seems to work okay since I can play it from a MIDI keyboard, however the patches won't upload. I upload them via SysEx on MIOS Studio and it CLAIMS to do it but afterwards I'm stuck with 128 gunshot patches for some reason. It lists all of the patches it supposedly sends via SysEx but once it gets to the last patch, all of the patches seem to be the same. Is it because I installed the banksticks before installing the SammichFM firmware? Is there a way to override that or manually have the banksticks format again? Do I need to remove them and re-insert them? I tried all of the programs listed and none of them seem to work. JSynthLib seems to just hang, but I haven't tried the newest version, and to be honest I'm not sure what to do with the contents of the JSynthLib Snapshot file anyway.

The last problem is a low hum in audio 1/2. The audio in 3/4 seems to be perfectly fine. Any clues would be appreciated.

This post has been edited by Snyderman: 13 December 2011 - 10:01

[Wilba]

The bottom row of LEDs show which operators are selected... so they should light up when pressing the button above, when that "layer" is also selected via the button between INS and OP... that one toggles between selecting instruments or selecting operators, and the current state is shown by the LED next to the label.

The LEDs are in a small matrix, so each row of LEDs have cathodes connected, and each pair of LEDs above and below a button (INS/OP, 1, 2, 3, 4) have anodes connected, and are then connected to a digital output of the 74HC595. So if one row isn't working, it is more likely to be a due to something wrong after the point that those LEDs' cathodes are conneced... such as the BC547 (which "current sinks" the row of LEDs), or the digital output of the 74HC595 which drives the BC547.

For example, you could have fried the BC547, or have a bad solder joint somewhere between the 74HC595 pin and the collector pin of the BC547.

Alternately, you could have some LEDs backwards, or fried individual LEDs, making them not work or be not bright. It's hard to say exactly what's wrong.

As for patches, try uploading the SysEx patches with another tool like MIDI-OX and use 750ms delay between messages.

[Snyderman]

Wilba, on 13 December 2011 - 12:31, said:

The bottom row of LEDs show which operators are selected... so they should light up when pressing the button above, when that "layer" is also selected via the button between INS and OP... that one toggles between selecting instruments or selecting operators, and the current state is shown by the LED next to the label.

The LEDs are in a small matrix, so each row of LEDs have cathodes connected, and each pair of LEDs above and below a button (INS/OP, 1, 2, 3, 4) have anodes connected, and are then connected to a digital output of the 74HC595. So if one row isn't working, it is more likely to be a due to something wrong after the point that those LEDs' cathodes are conneced... such as the BC547 (which "current sinks" the row of LEDs), or the digital output of the 74HC595 which drives the BC547.

For example, you could have fried the BC547, or have a bad solder joint somewhere between the 74HC595 pin and the collector pin of the BC547.

Alternately, you could have some LEDs backwards, or fried individual LEDs, making them not work or be not bright. It's hard to say exactly what's wrong.

As for patches, try uploading the SysEx patches with another tool like MIDI-OX and use 750ms delay between messages.

Given that it's just three LEDs (and not vertical pairs), I'll go back and check my solder joints or I might just have to replace the LEDs altogether which is fine since I was given spares.

I tried the patches business and the only conclusion I can come to is it's not recognizing the banksticks and it's only saving the very last patch received, and it's not currently capable of storing the patches at all (one default patch that just keeps getting overwritten). Any ideas for fixes? I have the banksticks in there, the orientation is correct.

[Wilba]

Perhaps there is an issue with the connections to the banksticks.
Check the connections there... perhaps there is something wrong with the solder joints on the PIC, or the pullup resistors.

[Snyderman]

It's been a while since I posted in this topic but I've been busy with other work, I finally got back to things and checked my soldering once again and found that a few joints needed fixing but it didn't fix my problem and even made it "apparently" worse (now there's more of an issue with the LCD light).

Once again, there has been no improvements in the lights or the banksticks. Wilba, if you could clarify what you mean by "pullup resistors". I've checked all of the contacts on the base board with the banksticks but I will check again.

Now, I've noticed that if I push in to the MIDI sockets too hard, sometimes the backlight of the LCD goes off so the text is dim and greenish. I haven't put it into the full case though, but it shouldn't really matter. I never had that issue when I was testing the sammichSID.

While I'm not sure about the hum issue, I think it has to do with the fact that I'm splitting a Stereo signal into two mono signals to go into my Mbox2. When I fiddle with the levels, I get just about the same amount of noise regardless of if I'm using Stereo Out 1/2 or 3/4. For example if the output is 1/2, one controls the level and the other just adds hum. If it's output 3/4, one brings up the level and the other just adds a high frequency hiss (same effect, at the end, more noise). Not sure what's causing it to be so noisy, but I'm also noticing I've got some issues with the decay of the signal going on indefinitely (though I think this has to do with my bankstick problem).

The frustrating thing is that the sound is actually really good (although a bit noisy). Everything seems to work other than a few really irritating issues. I should probably mention that when I screwed up with the resistors way back, I was forced to go to RadioShack and pick up some there instead. I hope that's not the problem. It was hard enough desoldering all of the incorrect resistors in the first place. However, I'm not sure that's it since the voltage tests were fine and the sound works great and the control surface seems to control pretty well.

Here's a video of the LED problem: https://docs.google....M5M2Yz&hl=en_US
For reference, the top row lights up when it is on the top row, but they are only very dimly lit when the bottom row is active (and 3 of 5 light up there anyway).

Here's the LCD backlight issue: https://docs.google....Tk3YjkwYjU3ZThj
It gets dim if too much pressure is applied (but it doesn't shut off).

This post has been edited by Snyderman: 11 January 2012 - 00:53

[Wilba]

Sorry, by "pullup resistors" I meant R2 and R12, which are on the two signal lines between the PIC and the 24LC512 ICs (IC5-IC19)

Regarding LCD: check the trimpots... I've had some go intermittent due to pressure, i.e. push them in too much with a screwdriver and it loses contact. So perhaps try pressing them to see if that's the cause.

Regarding hum: nearly all the time, hum is caused by the input power. Perhaps it's not enough power and the voltage regulator can't produce stable 5V... or the power supply is switchmode and noisy. Also you might get hum if you're feeding the output into a balanced audio input. I would advise checking whether you get the same hum when connecting into a totally different amplifier, different cables, etc. i.e. home stereo or computer line in. If it's still there, then consider the power supply as a cause... it could even be a faulty capacitor or voltage regulator.

Regarding LEDs: The LEDs that don't light up are either dead, backwards or have a bad solder joint. Perhaps you can confirm they're working by powering them directly with 5V with a 1K in series. Since the other LEDs seem to be working as expected, the problem isn't with the 74HC595, it appears to be outputting OK. Maybe you just got unlucky and burned a few while soldering. I can send you some replacements if you want, just email me.