preenfm Forum
PreenFM => preenfm2 and preenfm3 => Topic started by: horridus on September 14, 2015, 11:25:23 PM
-
Hello,
I put the first kit together last week. It appears to really "want" to work. It boots as expected, but no chime. Although there are a series of fast audio "pops" as the blue leds flash rapidly. There is no audio when attempting to play the instrument. At first I assumed that I have a cold solder joint on the board some place, but I have inspected them with a magnifying glass and a multimeter / continuity tester from the top of the pins or leads on a component to the edge of the solder pads on the other side of the board and get strong signal everywhere. The preenFM2 navigates as expected, USB drive is easy to access to load patches etc. Power seems clean on both USB connector as well as when I use an external power supply. I bought 3 kits, so I swapped out all of the chips in pairs and have gotten the exact same results each time. The Oled screen works great. I'm stumped. I ran continuity on all the pins of each socket to the edge of the solder pads as well, also clean signal. I double and triple checked the resistor values, polarity of the Electrolytic Caps and diodes, all seem ok.
Any ideas? Maybe I have missed a crucial step in the build? Pics are attached.
[attachment deleted by admin]
-
Another pic ... other side
[attachment deleted by admin]
-
I don't see any obvious error and everything looks nice (although a bit too much soldering ;) )
How do you test audio ?
Do you see a note blinking on the display when you send midi notes ?
Have you tried holding down "MENU" button and press one of the five left buttons ?
What happens then ?
-
Hi Xavier, thank you for your reply
I don't see any obvious error and everything looks nice (although a bit too much soldering ;) )
Yeah, I'm pretty rusty on the soldering ... will be less zealous on the next!
How do you test audio ?
I've tested by loading DX7 patches / instruments and then pressing the Inst button and one of the buttons on the left side. The note lights up, but still no sound.
Do you see a note blinking on the display when you send midi notes ?
Yes. The instrument appears to play, but still no audio. Mysterious!
-
The pop you hear is the empty audio signal around the op amp reaching 2.5v.
As you have both audio channels not working, it must be the connection between the preenF405 and the DACs.
I'll be back tonight with some aditional testing you could do.
Xavier
-
That would be great .... It is late morning here now, and I am awake and working. I can easily take a break and try your tests.
-
I attached 2 pictures.
1. First you can check all 5v, ground and 2.5v voltage. (2.5v only if the preenfm2 is powered 0v otherwise).
2. on the other picture, I colored the wires that the preenF405 uses to send audio samples to the DACs. Check the solder points and the connections (null resistor).
Xavier
-
Ok thank you, I'll take a look.
-
Hi Xavier,
I checked all of the power 5v, 3.3v, 2.5v .. and ground lines to each chip and the OLED etc ... Voltage and ground continuity appear everywhere they are supposed to. Although, the 2.5v lines were reading 4.02v. All others read perfectly. Nothing is hot or smells strange.
The DAC lines read clear continuity from the PreenF405 on both DACs ... Voltage and ground around the board appear good on both sides, pins to pins solder joints to solder joints.
-
I worry about the 4.02v.
That should really be VCC/2. So 2.5 if you get 5v on VCC or 2.4 if your USB power is 4.8.
Seems like there is something wrong in the audio path after the DACs.
Please, follow the following picture.
Remove your op amp, and try to connect directly with a metal wire the 2 blue points (or the 2 green points).
Do you hear something when you play midi notes while the note are connected ?
-
I removed the op amp and connected the 2 points while playing a note, still no audio heard.
Also, I triple checked the points that should 2.5v ... The 2 resistor points (68 ohm) read 4.02 ... The 2 DAC points read 3.85v each.
I tested this using a ground connection from the small 5v holes on the bottom left, also from the top of the 5v regulator. Hopefully this was a correct move for an accurate reading. I tried also using the top of the 3.3v regulator as well, which reads .7v on the DACs and .72v on the 68 ohm resistors.
-
Have you tried to swap the preenF405 with an other one one ?
-
I did ... and it gave the exact same results, which was somewhat a relief :>)
I just ordered all of the other chips from Mouser and they showed today ... I think I may change out the DACs and see if maybe there was just a bad run of DACs? The 2.5v on VCC is regulated internally on the DACs?
-
The 2.5v on VCC is regulated internally on the DACs?
No, this is the zero of the preenF405 audio output signal. It sends 2048 on the 4096 DAC range.
So another test, put back the opamp, remove the preenF405 and test the 2.5v points. You should get 0v.
-
Yes, zero as expected.
-
I just tried preenF405 #2 .. and tested the voltages ... 4.02v on the resistors ... 3.95v on the DACs
-
Last test of the day :
Power it up by pressing a button.
Check on the display you are in bootloader mod.
Then check one more time the 2.5v. Still 0v ?
-
I meant : "Power it up WHILE pressing a button."
-
Still reads zero .. have pleasant night Xavier
-
Seems like the DACs are receiving the signal from the preenF405 firmware when it's running.
But so i really don't understand why you don't hear sound while shortcuting directly to the audio jack while I suggested with the blue/green points above... :-\
In any case you can double check the area circled in the attached picture.
These are all points used by the audio path. Don't hesitate to heat the points again and remove a little soldering.
-
And you turned up the volume pot right ? ::)
-
I'll try the audio test with the op amp out again .... Then I'll clean up those points in the circle ... I have an iron just for that ... then I'll try the audio test without op amp again.
-
Yes, volume is definitely turned up ... the audible "pops" on boot go away when it's down.
-
Hi Xavier,
Ok, cleaned up and hit all of those audio section solder joints looking for shiny. Took away a lot of solder, checked for voltage (still 4.02v at the 68 ohm resistors / 3.85v on DACs) and removed the op amp and checked for audio (volume all the way up) ... precisely the same results.
Ric
-
Not easy...
I attached two pictures so that you can check any potential shorcut between junctions and all contuity.
I colorized the 5v in red, so as your audio path is higher than 2.5v i would first try to see if there is no 5V at some place where it shouldn't.
Then, check one by one every part of the circuit following the 2 pictures.
-
Do you have access to an oscilloscope ?
-
I have 2 oscilloscopes, but no probes! I can work on that though. They were out on loan and the probes didn't return.
After work today, I will carefully go through the tests you've outlined and check to see if voltages / values are correct. No worries about it being easy ... i'm patient and I don't mind checking :>)
Thank you for all of your help.
-
5v test complete. Most points show slightly under 5v .. More like 4.88v ... The op amp shows 4.88 on the 5v pin, and 4v - 4.02v on all but the pin which is ground (continuity is good on that pin back to the top of the 5v regulator)
I'll check the continuity next ... I took a coffee break for this test, so may be a while before results of the other test.
-
I ran the round of continuity tests in the green picture. All chime in strong. I checked a couple of times to make sure.
-
I would go back to the first idea that the DACs do not receive correctly the SPI data from the preenF405.
Audio_SPI.png picture in message #6.
Maybe if one of the signals green (Clock) or blue (sample data) does not reach the DAC, that could lead to not null but undetermined DAC output voltage.
Following the Audio_SPI.png picture, try to check and resolder all the concerned pins. DAC + preenF405.
Then see with your oscilloscope if the 5 color signals reach correctly on the DAC pins.
They're all fast signal.
The blue one is the fastest signal : several mhz.
-
Will do!
-
Ok, de-soldered, then carefully re-soldered all of the pins that match the preenF405 to the DACs on both sides (all 3 chip sockets) ... Used an oscilloscope to view the preenF405 spi_ck and on both DACs ... The clock signals are solid and appear the same on both sides. The 2.5v pins on the DACs still read 3.99v each ... resistors also read 4.02v still .. so no audio / no real change. I tested continuity after re-solder and the signal is clean on all pins preenF405 to both DACs. I used a different voltmeter this time too :>)
-
I just switched out both DACs with new chips from Mouser ... same results
-
If the digital audio signals from the preenFM to the DACs are clean and if the analog audio signal path after the DAC is OK too. There should be sound ;)
Unless your problem comes from the mixer/soundcard you use to hear the sound, but as you hear a "Pop" when switching on, it seems to be OK.
You're talking about the clock signal. It's one of the 5 i colorized in audio_spi.png picture.
I'll try to post this week a screenshot of each of the color showing what you should have.
-
Sounds good ... yes, the pops are perplexing. They don't happen every time though. Somewhat intermittent. They are in time / sync with the flashing LEDs on boot. Yes, the signal between preen and DACs are strong. I have the board layout to look at now. I'm going to carefully trace the full length of the audio signal path after DACs again. The 3.99v at the DACs 4.02v at the resistors seems like the best (only) clue so far. I tested midi via DIN again the other night too .. still works like a champ.
-
I just removed both DAC dip sockets, cleaned the pads nicely, and replaced the sockets with machined sockets. Also new DACs. Still no change. Same 3.85v on the DACs. Next I'm going to remove and replace every resistor and cap on the board. The only ones that look like they may be odd are the 115k resistors. I'll have to locate some though. My local electronics store didn't have any in stock. I have everything else here now.
-
Sorry, didn't have time to post SPI oscilloscope screenshots.
But if you post yours i can tell you if they are good.
You removed the DAC sockets :o Ouch. Impressive...
The 115K resistors (with the 1.8K ones) are only to add the 2 DACs output : with a 64 times ratio (115/1.8 = 64).
I don't think they can be responsible for the bad voltage.
Didn't you try to build the other preenfm2 you have ?
-
Here are the expected signals measured directly on the pins of the MCP4922 while the preenfm2 is turn on without any sound.
Colors correspond to Audio_SPI.png:
http://ixox.fr/forum/index.php?topic=63509.msg65602#msg65602
-
And the 2 last...
-
Hi Xavier,
Nope, waiting until all paths are exhausted in troubleshooting before moving on to another build. I have hope that this one will work eventually. The socket removal was a bit of work, but the sockets come off safely (no pads harmed) if one carefully crushes the plastic around each pin carefully. This saves hitting the pads with a lot of heat for no reason. I have a pretty decent desoldering iron here too. Anyhow, it came out clean, and I wanted to see if any of the pads were somehow connected by excess solder underneath the sockets on the top side of the board. This was not the case. They had (still have) plenty of clearance.
Thanks for the tip on the 115k's .. I'll let them be. I'll look at the oscilloscope today too and compare your pics to the signals here.
Thank you for the pictures too! Maybe by the end of this thread, there will be some good troubleshooting tips for some other novice / future builder of the preenFM2.
-
Handheld pics ... not using a desktop capture app :-[
Pins 3,4,5 on the DAC side
The last 2 pics you sent, which pins are they representative of?
-
...
-
...
-
The last 2 pics you sent, which pins are they representative of?
The file name of the screenshot is the color you see on the SPI schema.
Xavier
-
Ahh, I see .. thanks
-
...
-
...
-
Everything looks good. Digital sample data reach the DACs.
This is the Biggest mistery in the whole preenfm2 history :-\
You can use your oscilloscope to follow the audio path.
Clear the default Combo preset (MENU>SAVE>Default>clear). REBOOT. so that we know you have a "sounding" preset.
Set the voice number to 1 for the current instrument and 0 to the others to maximize the voltage output of the preset.
Play a repeating note from you sequencer, loop it, and be sure the instrument show up on the display.
Now put your oscilloscope on the output of the DAC.
Bottom left pin on the gerber1.png
http://ixox.fr/forum/index.php?topic=63509.msg65620#msg65620
If you see your analog waveform, Then try to continue the audio path. 1.8K => 3.3nF => 8.2k => opamp => 68 => 22uF => pot.
If the output is FLAT at the direct output of the DAC, try to unsolder the 1.8K resistor and try again the DAC output.
-
I'll give this a try :)
-
Here is some audio from an MFB Step64 sequencer playing the preenFM2 over midi DIN ... I traced the audio path with the oscilloscope all the way to the phone jacks and recorded this audio sample at the same time. Also, the leds flash on the preenFM2 as midi is received, I reset to the default instrument on the preenFM2 as instructed. The audible pops are in a different pattern (matching notes from the sequencer pattern), but similar to what the chime on boot sounds like in terms of final audio product. There are some "extra" artifacts that blast occasionally that are not native to Audacity with no signal input during a recording (tried this to make sure it was not a ghost in the machine saying hello :>) ... Sorry for the harsh noise! Please make sure and turn your audio output down before playing the files.
The signal as read by the oscilloscope appears all the way to the phone jacks, but it is obviously not the audio we are hoping for. The 3.8v at the DAC output and the 4.02v at the 68 ohm resistors are still the strongest indicators of the underlying problem I think.
-
audio from chime for comparison
-
OK So there is actually a sound. That's not a no audio problem ;)
Forget what i said about the 115K resistors.
Can you double check the resistor values 1.8K and 115K ?
-
I am quite dyslexic and color blind (short list), but I will definitely check again right now.
Here is some video of part of the test too: https://youtu.be/JXA-8qQG5RE
-
I read both just now ... the 115k positions read 103.8k ... and the 1.8k positions read 1.788k
-
Congrats...
You're the winner of the contest : the unfixable preenfm2 8)
Your prize : a new PCB and its components.
Send me an email to preenfm1@gmail.com and let me know if you already have some of the IC sockets, resistors, capacitors, diode in your stock...
Time to go to bed for me :)
-
One last question, have you tried to power it with a real 9V wall power supply ?
-
I did ... a great deal of the first rounds of tests were done with a known to be good 9v power supply. The audible pops have always shown up on boot or with midi coming in ...
Good night Xavier sleep well .. thank you again for your time and tutelage :(
-
You should send the unit back to xavier.
I bet he would be thrilled to actually see what happened.
-
Horridus, your new PCB is on it way.
Served is right, in the meantime if you can send me back your faulty preenfm2, i'd love to take a look.
Thanks,
Xavier
-
that can be done ... just returned from a seattle noise show ... i ended up playing as a last minute fill in ... great fun :)
-
ps ... middle of the night here ... time to rest while not driving 70
-
Xavier,
The preenFM2 PCB has been shipped and is on the way to you :)
-
Xavier, if you ever work out what happened after having a look for yourself, please post here! I've been following this thread the whole time and I really want to know what was up.
-
I found it. :)
It was a shortcut between 2 solder points.
The blue one (see pics for color) responsible for sending the sample to the DAC and the close junction responsible for lightning the LEDs.
That must have distored the SPI signal which became ununderstable for the DACs.
I removed the tiny wire with my soldering and the sound came...
I'll send your PCB back. You can keep the extra one i sent you ;)
-
Xavier, wow, that is really great! I'll send the board and parts back to you as well :>)
-
Wow, that is a tiny bridge!
-
Looks just like one of my grey beard hairs ;D
-
(clap)