Jump to content

Troubleshooting midiphy SEQ v4+


latigid on

Recommended Posts

Hello, everyone,
I am nearing the end of the assembly of my midibox but I admit to be stuck at the second test (32. LeMEC: Testing DIN Switch Function). With the Lemec card connected, the kernel does not boot at all (no activity leds etc...), I also noticed that IC5 becomes burning hot in few seconds. I proceeded by elimination by checking the welds / cables and testing without the top card without success. I cross my fingers to get some help, thanks in advance!

IMG_20190513_055947.jpg

IMG_20190513_055958.jpg

IMG_20190513_063732.jpg

Edited by otropom
Link to comment
Share on other sites

When you say the "kernel" doesn't boot, do you mean the waveshare MCU breakout board? If this is the case, a typical cause is a short between 0V/+5V.

Please check the orientation of all ICs. At least IC4 is in backwards! (The stripe on the IC should go towards pin 1).

  • Like 1
Link to comment
Share on other sites

Thanks for your fast answer latigid :) ,

yes it's the core I was talking about but you have the eye and after welding in the right direction ic2 and ic4 (on both pcb) everything was working and I was finally able to do the test.

In this case, the encoders work perfectly but their push action doesn't work, its the same for the mec switches and the matrix (i was on the correct cc/channel). Do you have any idea what I need to check ? pretty hard to debug when you are a noob i admit.

thanks by advance

Link to comment
Share on other sites

No worries.

It might be that IC2 was damaged by being reversed. Did you use new chips or the same?

I was troubleshooting with @niles the past weeks on a similar problem. The same advice applies, so check cables, resolder IC2 and IC3. Many solder joints look "cold" Any MIOS output with 

set debug on

?

Link to comment
Share on other sites

These are the same ICs, I unsoldered them with hot air and soldered them back with iron (my paste is poping !). I'll apply your advices in the evening and I'll give you a feedback:)

With Mios I have a normal feedback by turning all the knobs but nothing from pushing them, same with mec switches on both pcb.

 
Edited by otropom
Link to comment
Share on other sites

Okay, so I just checked the cables: no problem, obviously. I also unsoldered / soldered back ic2 and ic3 together without success. The problem is 100% identical on both lemec pcb.

Under MIOS, when I turn the knobs, I have a correct return of the values from both boards with the debug mode enabled, but no action is triggered when I press one of the buttons (alp push or lemec switches).

Out of curiosity, I tested the pcb lemec (the smallest) by completely removing IC3 and the problem is the same, maybe that can confirm that IC3 is dead ?

In your opinion, does this mean that certainty both IC3s are dead (because of the inversion?) and does the problem with the matrix also correspond to these ICs?

I am waiting for your answer and I will order some spare parts.

Have a good day and thanks :)

  • Like 1
Link to comment
Share on other sites

Just check first that the MB_NG app is flashed and you have loaded the correct seq_l or seq_r.ngc. Was testing of the JA board okay?

To simulate DOUT behaviour you can try 

set dout all 1

Or go through and count the dout pins (8 per chip) until you get to the correct IC -> replace all with the one you want to toggle. DOUT rows/cathode sink rows are given in previous pages.

The matrix is controlled by ICs 2 (cathodes) and 3 (DIN).

Having spare parts is always useful and could be used in future projects. I typically get 100 shift registers at a time.

  • Like 1
Link to comment
Share on other sites

Double checked but yes, MB_NG is flashed (MIOS detect MIDIbox NG on midi IN/OUT) and i tested both seq_l and seq_r (with seq_r, all ja board leds are lighting up, what the exact diffrences of r & l config files, simply for left/right version ?) .

Actually JA Board seems to works perfectly, i have correct feedback within MIOS console (push btns / alp push action / knob) and leds are working correctly when i test buttons.

With MEC boards its another story and only knobs turning are sending feeback to MIOS (tested with both seq_l / r).ngc) here the screenshot of what i get in the console. for info between each knob test i'm pushing them (no action triggered in console as you can see), same for mec switches and the matrix on CC16 (chn 1), not working . No succes with "set dout" function either unfortunally (dunno if its possible to test ic with a multimeter ?).

 

So for you , only IC2/3 (on both MEC boards) need to be changed or its better to replace all IC's ?

for now its only my second diy project (first one was a xoxbox few years ago), i didnt had ocassion to stock spare parts but for sure i'm going to order enough "in case of"  :)

 

Thanks

 

mios.jpg

Edited by otropom
Link to comment
Share on other sites

If _NG is flashed and JA is working it's a good sign. The _l and _r .NGC files specify the position of JA. You can use _r to test lemec_L. It's possible to test lemec_R as the first board in the chain, but you need to reconfigure the file.

The boards are chained JA-lemec_L-lemecR, correct? 

You should measure +5V on the 74HC165 input pins and 74HC595 outputs should go to +5V with set dout all 1. Check the preceding pages of this thread or the datasheets for pin assignments. You can follow the SRIO chain (especially the SO signal) to see how far the data gets.

Seeing as you have some DIN activity, consider replacing IC2/IC4, which were reversed and then heated up probably quite a lot during rework. 

 

Link to comment
Share on other sites

i was finishing all the test then.....I can thank the all mighty god of randomness, i switched one of the lemec board cable with the core one then everything was......working :rolleyes: i have pretty luck after unsoldering twice the IC's that i didnt kill anything finally . I was far too confident with my cables, will check them twice for sure now ! i'm out of cable headers but finally the beast should be alive very soon

Edited by otropom
Link to comment
Share on other sites

That might be a hard combination to tune. But yes, normally blue should have a lower-value resistor as it's typically dimmer. 

At least on the older SEQ firmware, the measure LED was not "XORed" i.e. beat and measure would trigger at the same time. That might be a way to get all three going at once for white-ish, though I can't remember if it's still part of the current firmware.

Link to comment
Share on other sites

Hi there

Everythig is done and before assembling the sequencer, I wanted to test all the functions (always with midibox ng and seq_l/r). R.A.S (good news) except for the screens that don't work. At the second boot I had the right screen (nothing on the left one) which displayed 88888...over its entire surface and at the next boot, nothing, the same for both screens.

I test this way (core->JA board->lemec1->lemec2 + both screens)

- I checked that J15S is set to 3.3V,
-Rework each core weld (in case...)
- Checked the voltages, I have the same results as on these images (posted by "SimonSays"):

http://midibox.org/forums/uploads/monthly_2019_05/IMG_0137.jpg.7a321f39f767fe44ff31e9f676c01c8a.jpg

http://midibox.org/forums/uploads/monthly_2019_05/IMG_0788.jpg.a5e22618fcb98c9437a10976bc2d058c.jpg

-tested the lcd command...

no idea what to do at this stage;(

thanks for any help !

IMG_20190518_202603.jpg

IMG_20190518_202623.jpg

IMG_20190518_202628.jpg

IMG_20190518_203758.jpg

Link to comment
Share on other sites

The soldering on the OLED headers could be better, try to rework. If the fillet looks like a teardrop, it might not be contacting the PTH properly. Try to get the heat into the pad and the solder will flow there.

Soldering could be better on IC2, pin 8. Recheck R33abcd.

Try swapping the OLED cabling to the other J15 port. Could be bad cables again, could also be that some pins on the cable short out components on the OLED, so isolate with even a scrap piece of plastic or card on the OLED PCB right under the connector.

Link to comment
Share on other sites

Just followed every of your recommandations. Checked R33 abcd : 3.2v on all
you will find attached also the values that i have on both display ports pins and on the oled itself.
after reworking on ic2 a saw a problem with the pin that was in front of C2 (solder did not stick) so in doubt i unsoldered IC2 and the pin was simply missing :/ . Do you think its the origin of oled problems ? also can i bridge directly the pin to C2 or i must bridge from somewhere else ? thanks !

 

 

IMG_20190519_050046.jpg

pad.jpg

values.jpg

Link to comment
Share on other sites

On 14/05/2019 at 2:29 PM, otropom said:

 

Well, after repairing IC2 pin 16, (i suppose my problem came from pin/pad misalignement) i installed New câbles for each screen and protect the bottom of both oled connectors with adhesive tape, it worked !  With one problem, however, as you will see on the picture. 

I tested each screen independently / on each port without any change and it appears that the problem probably comes from the core, any idea if it can still be IC2 or some other element?

(So close from the end!) Thanks 

IMG_20190519_190451.jpg

Edited by otropom
Link to comment
Share on other sites

Each display gives this output when only one is plugged in? Then the cables are okay, or at least have the same issue :).

At a guess, probably the register data is corrupted, so look again on IC2 for cold joints or shorts. Make sure the power is jumpered to 3v3 (it was before I know). 

Check that the 407v board is correctly inserted into the 2x 2x25 header pins. You may have bent the connectors by inserting/removing the 407v a lot. Reference the schematic here and check for continuity back to the top of the 407v header pins PD7, PD6 etc.

Check that the E (enable) lines of the displays are not connected together.

Link to comment
Share on other sites

:cheers: everything is running great now, you were right about IC2 cold joint. I just have the matrix that display "mirrored" text can it come from a backward cable or something like that. ?

in any case merci beaucoup for your kindness and your advices, i will finally be able to enjoy the midibox seq ! and it wasn't that easy for noob (second diy project) that was not comfortable with English  language :) learnt lot of things !

Link to comment
Share on other sites

5 hours ago, otropom said:

:cheers: everything is running great now, you were right about IC2 cold joint. I just have the matrix that display "mirrored" text can it come from a backward cable or something like that. ?

The activity matrix? Check the HW config file as you need:

# set to 1 or 2 to enable the relative track position display
# 0: TPD disabled
# 1: TPD enabled - columns are cathodes, rows are anodes
# 2: TPD enabled - columns are anodes, rows are cathodes
# 3: TPD for midiphy frontpanel enabled - columns are anodes, rows are cathodes, inversion
TPD_ENABLED 3

 

Quote

in any case merci beaucoup for your kindness and your advices, i will finally be able to enjoy the midibox seq ! and it wasn't that easy for noob (second diy project) that was not comfortable with English  language :) learnt lot of things !

De rien! You had a great attitude and provided the right info (pictures). It's really great work if it's just your second project! Félicitations!

Edited by latigid on
Link to comment
Share on other sites

hi all,

after testing i found an minor issue with my breakout-box. everything is working as intended but i get annoying double triggers on some modules. when i connect the same modules to my robokop sequencer or other trigger source everything is fine. could it be a grounding issue?

the +/-12V for the AOUT NG is coming from my eurorack and the +5V for AOUT NG and the LineDriver Receiver from an additional psu. the external psu (5v) is in the middle of the ribbon cable between AOUT NG and the receiver as suggested by latigid on. i also enabled the 1ms trigger but it makes no difference.

please help

rbv2

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...
×
×
  • Create New...