Jump to content

Troubleshooting midiphy SEQ v4+


latigid on
 Share

Recommended Posts

14 minutes ago, Hawkeye said:

@synaptech - that's good news with the encoder! :cheers:

As written above, i'd recommend to thoroughly check every SMT IC pin on LeMEC L - i had more than once problems with solder connections i thought were ok. You can "reflow" them quite easily. Please check the top pins of IC2 on LeMEC L, can't really see it in the photo, but there might be some solder missing, or they might not "reach" the PCB - check from the side with a magnifying glass.

Good luck!
Peter

Thanks Peter, I did an inspection with a magnifying glass tonight and it looked ok, but I'll run over them all with the iron tomorrow to see if that clears things up. :cheers:

Link to comment
Share on other sites

I went through and reflowed all the IC and transistor/diode pins, double checked with a magnifying glass but still no luck with the buttons. I'm admittedly a bit weak when it comes to digital circuits, would it make sense to look at the IC that joins the serial data from LeMEC L and LeMEC R or is the problem likely before the LeMEC L data gets that far?

Is the button design similar to the original seqv4? I was looking in other threads to see if I could find someone else that's had similar problems.

Link to comment
Share on other sites

It might be interesting to solder down the superflux cathodes, then you could check if the problem is on the sink side or the DIN side.

IC2 is the sink side of the matrix; you can see the sink part of the matrix here on the left. MIOS pulses each output in turn and this allows the switches to conduct to 0V when pressed (also lighting LEDs when that column of LEDs is on). This registers as a switch press in MIOS by the 165 inputs going low.  

IC3 is the DIN part of the matrix. D0-D7 represent the switch columns and there are eight per board from left to right. D0-D7 connect to the anodes of diodes located above Matias switches and beside MEC switches.

lemecDIN.thumb.JPG.40140ee0cfdaf33c45f14

 

As you have a functional lemec board, you can compare voltages on the two. The only difference with the _R board is that there are three extra shift registers (74HC595) at the start of the chain. 

Because you have some activity with encoders, the issue lies with either the DIN or sink side of the matrix. Do all encoders work? If so the SC, RC2 and SI lines are all fine. Check that the correct resistor networks are installed (also the orientation), check that you have +5V on all aforementioned 4148 diode anodes, 

 

  • Like 1
Link to comment
Share on other sites

@latigid on thanks that definitely gives me a better understanding of how the button matrix is working. I'll try tacking down the cc pin of the superfluxes tonight. 

All of the encoders seem to be working fine, I can see the CC values increment and decrement correctly in MIOS Studio.

As much as troubleshooting can be frustrating it definitely helps teach me about circuit design! Appreciate the hand holding. :cheers:

Link to comment
Share on other sites

Definitely looks like something is up around IC2

Pins 15, 1-7 (QA-QH): 0v (pins 3-5 seem to occasional show .001v)

Pin 9 (QH*):.422v

Pin 10 (SCL): 4.8v

Pin 11 (SCK): 2.7v

Pin 12 (RCK): 4.8v

Pin 8 & 13 (GND): 0v

Pin 14 (SER): .46v

Pin 16: 4.8v

On the working LeMEC_R:

Pins 15, 1-7 (QA-QH): .586v,

Pin 9 (QH*): .493v

Pin 10 (SCL): 4.8v

Pin 11 (SCK): 2.7v

Pin 12 (RCK): 4.8v

Pin 8 & 13 (GND): 0v

Pin 14 (SER): .422v

Pin 16: 4.8v

I reflowed the resistors around IC2 just to make sure the top soldering wasn't bad and confirmed the resistor networks are in correctly they're the same part and orientation between both boards. Also checked all the 4148 diodes and the voltage seems good (4.8v everywhere).

Edited by synaptech
formatting/pin labeling
Link to comment
Share on other sites

Thanks for the voltages, so the issue seems to be on the sink side of the matrix. 

As you're troubleshooting the _L board and the _R board works fine, it would seem as though the data is getting though correctly.

Could you try the following commands in MIOS terminal:

set dout d0 1 %turns first output on
set dout d0 0 %turns first output off

set dout d7 1 %turns eighth output on

Here you manually set the pins, so this can rule out any errors from the software side. You can also try it with the .NGC loaded. When d0=1, the first four encoder switches should work when pressed. set d0=0, then d1=1 should allow the first four Matias switches.

If the component values/types, cabling and soldering are all sound, then I'm beginning to suspect a dud IC. Could have been overheated or subject to static for example.

 

  • Like 1
Link to comment
Share on other sites

12 minutes ago, latigid on said:

Thanks for the voltages, so the issue seems to be on the sink side of the matrix. 

As you're troubleshooting the _L board and the _R board works fine, it would seem as though the data is getting though correctly.

Could you try the following commands in MIOS terminal:


set dout d0 1 %turns first output on
set dout d0 0 %turns first output off

set dout d7 1 %turns eighth output on

Here you manually set the pins, so this can rule out any errors from the software side. You can also try it with the .NGC loaded. When d0=1, the first four encoder switches should work when pressed. set d0=0, then d1=1 should allow the first four Matias switches.

If the component values/types, cabling and soldering are all sound, then I'm beginning to suspect a dud IC. Could have been overheated or subject to static for example.

 

Yep I'm beginning to think the same! I ran the following (minor correction to use 0 instead of d0, MIOS didn't like the d):

[1240259.981] set dout 0 1 %turns first output on
[1240260.142] DOUT Pin 0 (SR#1.D7) set to 1

Tested the encoders again with no luck. Then reset to d0=0:

[1240259.981] set dout 0 1 %turns first output on
[1240260.142] DOUT Pin 0 (SR#1.D7) set to 1

Then enabled d7:

[1240259.981] set dout 0 1 %turns first output on
[1240260.142] DOUT Pin 0 (SR#1.D7) set to 1

Tested the matias switches with no luck.

I tried once immediately after booting up, and once after loading seq_l. Same results both times. I'll remove the IC and get another one ordered. Thanks again!

Link to comment
Share on other sites

Ah yep,

set dout x 0|1

is the correct syntax.

You could try the same on the second 595 chip, these would be dout 8--15.

You've copied the same terminal data each time, but I assume you turned the correct outputs on and off. Note that pressing the up arrow on the computer keyboard will go through a history of the terminal commands.

 

  • Like 1
Link to comment
Share on other sites

Whoops, forgot I have to right click to copy the terminal data instead of using Ctrl+C! Here's the actual input:

[1240259.981] set dout 0 1 %turns first output on
[1240260.142] DOUT Pin 0 (SR#1.D7) set to 1


[1240383.356] set dout 0 0 %turns first output off
[1240383.517] DOUT Pin 0 (SR#1.D7) set to 0


[1240401.956] set dout 7 1 %turns eigth output on
[1240402.117] DOUT Pin 7 (SR#1.D0) set to 1

Edited by synaptech
Link to comment
Share on other sites

Hah I'm an idiot and just realized % is the MIOS comment character! I was wondering what %turns was referencing. :decayed:

Testing the LeMEC_R - it seems to be booting up with the dout pins enabled and doesn't seem to react if I manually set the value to 0. Both buttons and encoders all fully function.

The core has NG loaded but behaves this way with and without the sd card loaded. Is that suspicious?

 

Edit: For the next person that does this - YES this is suspicious, doing the above test on the second board caused lights to turn on on the JA board. If you have the boards in the correct order this should not happen!

Edited by synaptech
Link to comment
Share on other sites

10 hours ago, latigid on said:

No worries, you provided good info (voltages and pictures) and had a decent attitude. If it works, we're very happy!

 

:D

Got all the MEC switches/LEDs installed and working correctly now!

Having a strange issue with the Matias switch LEDs - all the buttons and encoders on both boards are still working.

I initially had a few red LEDs that weren't working, I've replaced those with the spares and they're great now, red lights are all good on both boards. Blue lights working great on LeMEC_R.

On my LeMEC_L (I've got the right board this time :grin:) every 4th blue LED isn't appearing. Top row missing #4 and #8, bottom row missing #4 and #8. I did replace the #4 bottom row LED to get the red working again but the blue problem still happens.

Picture of the board: https://imgur.com/a/39nPcFc

Link to comment
Share on other sites

4 hours ago, latigid on said:

Looks like R16 is not soldered correctly?

You rock! :cheers:

Had a few more hurdles to get over (bad I2C cable) but it's alive!

Still have a couple things to finish before calling it complete - beat LED and I ran out of 10 pin connectors for the LineTX cable from LeMEC_R, but it's looking good!

Thank you Andy, Peter, Bruno and the rest of the midibox community for all your patience, wisdom, and troubleshooting help! Thank you TK and Adrian for your efforts on this project as well! Can't wait to learn the ins and outs! :cheers::grin::love:

its-alive.png

  • Like 2
Link to comment
Share on other sites

2 hours ago, Antichambre said:

Yeah!!! which number is it? 7 or 8 maybe?
cheers :cheers:

Number 8 perhaps? :D

Though I think I have to get the beat LED installed to get the official stamp! How much clearance should I leave to fit your LED diffuser? 

Thanks again everyone! If anyone has similar problems to the ones I faced I'll be happy to share my experience. :D

Edited by synaptech
Link to comment
Share on other sites

55 minutes ago, synaptech said:

Number 8 perhaps? :D

Though I think I have to get the beat LED installed to get the official stamp! How much clearance should I leave to fit your LED diffuser? 

Thanks again everyone! If anyone has similar problems to the ones I faced I'll be happy to share my experience. :D

Let 3 mm, it will be fine! :)

Link to comment
Share on other sites

Hi,

i have a problem on the LE-MEC LH after pushing on the left most encoder cap the leftmost row of buttons doesn't work any longer:

leftmost encoder click, leftmost upper and lower DIN and MEC don't work any longer.

leftmost encoder turning works and the LEDs are also ok.

i suspect something got shorted as i pushed on the cap. any hints at which components i should look at
(and if i can fix it without desoldering all the Matias switches ...)

Thanks!

Link to comment
Share on other sites

Hi,

On lemec_L, left side and both versions, right side, the encoder can touch the header pins. Best is to trim both sides (encoder pins are easier) first and also put a piece of kapton tape or similar between.

From the sounds of it, you followed the testing procedure and had all DINs working before soldering in the Matias?

Please try to isolate/insulate first and see if that fixes the issue. You can also check if there are longer through-hole components (especially diodes) that short out in beween.

If not, please upload a good photo of the rear of the assembly and I will try to help.

Link to comment
Share on other sites

yes i had everything working before i soldered the mathias, also before i installed the encoder caps.

but i probably shouldn't have tested DURING pushing on the encoder caps ...

i did cut the encoder legs but there seem to be a few possible other shorts, in particular between the resistor networks and the thru-board headers...

i guess i'm going to have to borrow my neighbours fancy desoldering station :-/

Link to comment
Share on other sites

Schade/dommage! 

You might be able to remove the Matias switches (if really necessary) with your soldering iron heating both pins at once and pulling it out with the other hand/use a vice. A simple vacuum pump can also work wonders.

If the LEDs still work, then the sink side of the matrix is okay. Not sure what could be fried just from pushing together.

Link to comment
Share on other sites

ok ... so i desoldered all the Matias (ripping off most of the pads on top ...), clipped everything on the Enc-Plate even more and also the resistor networks and soldered back in the Matias ... it works now :-)

(resoldering the Matias is not recommended - one of the more stupid mistakes i've made. it would probably be an idea to provide a kapton isolation sheet in the shop to prevent this?)

i loaded seq v4+ and started testing a bit. this went fine until i tried to go back to midibox ng where the flashing failed halfway and i got stuck in bootloader mode.

now when i powercycle the core is not detected any longer in a MIDI scan and does not show up in the USB device tree (i'm on macOS)...

the green led on the back of the Core still flashes slowly when i power the Core but i don't get the bootloader message on the screen.

maybe i'm stuck in some kind of low level boot mode? any suggestions?

thank you!

Link to comment
Share on other sites

At least on Windows you should power cycle the Core between updates. (EDIT: I mean close and restart MIOS, the MCU is automatically reset after updating.)

If you really can't get it to boot/be detected any more, JPA0 is there to help on the wCore PCB. If you stuff this jumper, the bootloader will hold during startup. Normally you can then flash in another .hex again.

Edited by latigid on
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

×
×
  • Create New...