Jump to content

Troubleshooting midiphy SEQ v4+


latigid on
 Share

Recommended Posts

Thank you again for the fast response. I tried to reflow some pins and get rid of short where they seemed to be possible.

Status Quo: Japcb + Les Mec LH work fine together but when Reconnected all the PCBs and loaded SEQ_L - with both les mecs  installed all the encoders still dont work (no turns and press on all of the encoders on both boards) - when i press switches on JApcb i get the following in console:

"[10389.729] load seq_l
[10390.157] [MBNG_LCD] no response from CLCD #1.1
[10390.157] [MBNG_LCD] no response from CLCD #2.1
[10390.157] [MBNG_FILE_C] Event Pool Number of Items: 348
[10390.157] [MBNG_FILE_C] Event Pool Allocation: 12250 of 65536 bytes (18%)
[10390.178] Patch 'seq_l' loaded from SD Card!
[10390.184] [MBNG_FILE_R] /seq_l.NGR (optional run script) not found
[10396.484] set debug on
[10396.655] Debug mode turned on
[10402.023] MBNG_MATRIX_NotifyToggle(2, 34, 0)
[10402.023] MBNG_DIN_NotifyToggle(3034, 0)
[10402.023] [EVENT] id=BUTTON:3034 hw_id=BUTTON:3034 bank=0 fwd_id=LED:3034 type=NoteOn value=0 label=
[10402.023] MBNG_DOUT_NotifyReceivedValue(3034, 127)
[10402.023] MBNG_MATRIX_DOUT_NotifyReceivedValue(3, 34, 127)
[10402.026] MBNG_MATRIX_NotifyToggle(2, 50, 0)
[10402.026] MBNG_DIN_NotifyToggle(3050, 0)

"

Can i assume that failure must be on the Lemec-board (RH)?

Additional question: is it possible to progress without the lemec (RH)? - install and test the OLED, MIDI8 etc.?

Edited by knochenfabrik
Link to comment
Share on other sites

12 minutes ago, knochenfabrik said:

Can i assume that failure must be on the Lemec-board (RH)?

Not sure.

Disconnect all PCBs, connect lemec_L directly to the Core J89 (without JA) and load seq_R.ngc. You can also connect lemec_R directly to the Core, but you need the following modifications:


################################################################################
# Left Panel
################################################################################

...


# 8x8 Button/LED matrix - we emulate "normal" LED/Button functions:
DOUT_MATRIX n=1  rows=8 inverted_sel=1 inverted_row=1  sr_dout_sel1=4  sr_dout_r1=5  led_emu_id_offset=1001
DIN_MATRIX  n=1  rows=8 inverted_sel=1 inverted_row=0  sr_dout_sel1=4  sr_din1=2     button_emu_id_offset=1001


 

(i.e. offset the DOUTs by three)

Link to comment
Share on other sites

OK when the LH pcb is connected to the core J89 it works fine and everything is registered so I modified the SEQ_R and tried it with the RH pcb but there is nothing - will have to measure continuity later - the pcb just seems to do nothing right now :-/.

I also tried to set on/off pins individually how it is mentioned in this thread a bit earlier but it didnt realy do anything - maybe i also should try it again.

 

Thank you very much...

Link to comment
Share on other sites

21 minutes ago, Dimduj said:

Hi all,

@knochenfabrik just my two cents: Did you check the last/first encoder pins are not touching the pins of the ribbon header .

You can check that looking at the side of your build?

Thanks for your input - i checked it but there seems to be no short anywhere - On the LEMEC RH PCB however i have only installed the header in the J89 slot and not the J89A because it is the last module in the chain (am I wrong there) so there isnt realy the possibility of such a short right now. So my chain right now is from JApcb J89-port to J89-port of LEMEC LH and from the J89A-port to J89 port of the LEMEC RH.

7777.jpg

Edited by knochenfabrik
Link to comment
Share on other sites

Mmmhhh cannot help further sorry :/

One last advice, Before soldering mathias switch take time to recover them with electrical tape or paint sides with a marker to avoid light bleed between them.

i was really annoying with light bleed between steps.

Caps are hard ( impossible ? ) to remove once placed and it took me lot of time to place tape to isolate light with caps installed.

Cheers

 

Link to comment
Share on other sites

16 minutes ago, Dimduj said:

Mmmhhh cannot help further sorry :/

One last advice, Before soldering mathias switch take time to recover them with electrical tape or paint sides with a marker to avoid light bleed between them.

i was really annoying with light bleed between steps.

Caps are hard ( impossible ? ) to remove once placed and it took me lot of time to place tape to isolate light with caps installed.

Cheers

 

thanks, nice info - will do it when i figured out whats wrong at the moment :(

Link to comment
Share on other sites

On 5/23/2019 at 1:21 PM, pat_00 said:

Ugh, I seem to have problems at every stage :(

So I have gotten to the Fluxtest and have issues on both boards.

On the LH board, no red led on SW14. Have tested the LED with multimeter and it's fine.

On the RH board, no green or blue on SW7 and SW15, same as above, LEDs are OK.

I've tried the usual, looking for obvious faults, diodes are all correct., soldering all looks OK. Any hints on where to look?

 

 

Thought I'd report back, the above issues are solved. They were soldering issues, some really stupid some harder to find.

I still have an issue I'd like to fix, the stop, rewind, fast forward and pause buttons are constantly lit up. They function correctly when pushed, but I'm assuming this is not normal?

Midiseqv4.jpg

Link to comment
Share on other sites

Hi Pat,

Good that the other issues are sorted.

For the multiple LEDs, load up _NG again and look at the DIN events (set debug on) when you push the buttons. Probably you get just one switch event per press/per depress. If the LED events are back to normal perhaps the HW config file is mis-configured?

I'm a bit surprised that not all of the bottom row is on. Maybe SW17 has its LED in backwards? In fact, check all LEDs for correct polarity.

Check whether pins of R1/R2 are shorted together. 

Check whether any of the cathodes are shorted together. For this, test for continuity between the top-left pads (as viewed from the front of the MEC switches.

As usual, pictures of the soldering would be useful for diagnostics.

 

 

Link to comment
Share on other sites

Thanks for the suggestions. I'll upload some photos later tonight. I took some yesterday, but the quality is too bad to be useful.

So far I checked:

r1 and r2 are not shorted,

All LEDs are the correct orientation with the anodes going to the holes marked with the circles. In testing then I found that the LEDs on SW14 to SW18 the LEDs passed voltage (approx 1.6v) in both directions, I'm guessing this isn't as it should be?

I also found my Multimeter does not light any of the LEDs in diode test mode except the red 'recording' one. Maybe it doesn't put out enough power for the clear ones?

The Cathodes are shorted in pairs to the switch below eg. SW9 is connected to SW14. I guess this is by design and not an issue?

I'll do the tests with NG firmware tonight as well, but when I originally tested all button functions were OK (no latching, registering on and off as separate events) but the problem LEDs were still stuck on.

 

 

Link to comment
Share on other sites

13 hours ago, pat_00 said:

Thanks for the suggestions. I'll upload some photos later tonight. I took some yesterday, but the quality is too bad to be useful.

So far I checked:

r1 and r2 are not shorted,

All LEDs are the correct orientation with the anodes going to the holes marked with the circles. In testing then I found that the LEDs on SW14 to SW18 the LEDs passed voltage (approx 1.6v) in both directions, I'm guessing this isn't as it should be?

Power off or on? The diodes should only work in one direction though.

 

Quote

I also found my Multimeter does not light any of the LEDs in diode test mode except the red 'recording' one. Maybe it doesn't put out enough power for the clear ones?

It should work with a little bit of current. Another option that Peter likes is to use a coin-cell battery. If the voltage is limited to 3V or so it should be safe. 

 

Quote

The Cathodes are shorted in pairs to the switch below eg. SW9 is connected to SW14. I guess this is by design and not an issue?

That's correct, the cathodes are connected across eight columns:
SW1-4, SW9,14, SW10,15, SW11,16, SW12,17, SW13,18, SW5-8 and the switch terminal on the encoder.

 

Quote

I'll do the tests with NG firmware tonight as well, but when I originally tested all button functions were OK (no latching, registering on and off as separate events) but the problem LEDs were still stuck on.

If the switches work, your issue is on the anode side. Check the soldering around IC2 and R1-6.

Reset with _NG loaded and don't load an NGC file. You can control the outputs manually with 

 

set dout x boolean

set dout all 0
set dout all 1

 

e.g. 

set dout 15 1
set dout 6 0

turns on the lower anode row and sinks current on SW9,14

 

set dout 14 1
set dout 5 0

turns on the second-to-bottom anode row and sinks current on SW10,15.

 

Measure the output pins (15, 1-7) on the 595s to check the voltages.

Might be a dodgy 595?

 

 

Link to comment
Share on other sites

@pat_00 It looks like the LED on switch 17 is dead. I had the same problem (strange) with the same effect! First it worked all perfectly well and after some testing the same LEDs as on your picture constantly lit.

Maybe the problem is the bending to fit into the switch and the colored LEDs are more sensible in this regard.

Link to comment
Share on other sites

38 minutes ago, CJ55 said:

@pat_00 It looks like the LED on switch 17 is dead. I had the same problem (strange) with the same effect! First it worked all perfectly well and after some testing the same LEDs as on your picture constantly lit.

Maybe the problem is the bending to fit into the switch and the colored LEDs are more sensible in this regard.

@pat_00@CJ55 I also have the same issue but when NG is loaded the LEDs light up just fine when the buttons are pressed (also the green one) - will anyway swap it out to make sure its not the fault of said LED

Link to comment
Share on other sites

On 6/16/2019 at 7:08 PM, latigid on said:

Not sure.

Disconnect all PCBs, connect lemec_L directly to the Core J89 (without JA) and load seq_R.ngc. You can also connect lemec_R directly to the Core, but you need the following modifications:



################################################################################
# Left Panel
################################################################################

...


# 8x8 Button/LED matrix - we emulate "normal" LED/Button functions:
DOUT_MATRIX n=1  rows=8 inverted_sel=1 inverted_row=1  sr_dout_sel1=4  sr_dout_r1=5  led_emu_id_offset=1001
DIN_MATRIX  n=1  rows=8 inverted_sel=1 inverted_row=0  sr_dout_sel1=4  sr_din1=2     button_emu_id_offset=1001


 

(i.e. offset the DOUTs by three)

Hello, i tried most of the things you told me so I need to report :-). When connecting the LH Lemec directly to the uCore all switches and encoders work and I got no strange debug-messages. When connecting the RH Lemec directly to the uCore i have no response of the PCB. I modified the seq_R.ngc as shown above and when loading it i get a bit of debug output (see picture 1 please).

When connecting the whole Chain  uCore - JA - Lemec LH - Lemec RH, i now get triggers and debug-messages all over the place and none of the encoders on both boards seem to work (hard to tell because of the many random messages). Also the LEDS on the JAboard dont work in this setup (without Lemec RH they work fine when buttons are pressed). I tried to reflow the ICs and checked for contuinity between the ICs. You explained that there should be contuinity from IC1 Pin 10 to IC2 Pin 9 and I got contutinity from IC1 Pin10 to IC3 Pin 9 -I refer to the labels on the PCB (which is also the case on Lemec LH) So I at least got contuinity but could you please explain me what comes before IC1 that could have the effect of making trouble for the whole PCB and chain respectively causing random trigger messages (see picture 2 please). :angelnot:

I'm very thankful for your help and since I'm pretty stuck now it always feels so nice to read your thoughts.

Greetings, Alex

debuglemecrh_to_core.png

debugrandommiditriggers.png

Link to comment
Share on other sites

8 hours ago, CJ55 said:

@pat_00 It looks like the LED on switch 17 is dead. I had the same problem (strange) with the same effect! First it worked all perfectly well and after some testing the same LEDs as on your picture constantly lit.

Maybe the problem is the bending to fit into the switch and the colored LEDs are more sensible in this regard.

 

7 hours ago, knochenfabrik said:

@pat_00@CJ55 I also have the same issue but when NG is loaded the LEDs light up just fine when the buttons are pressed (also the green one) - will anyway swap it out to make sure its not the fault of said LED

If there is a systematic error here, please provide all of the info possible, especially the voltage outputs of 595 registers, results of NG tests including manual control.

 

6 hours ago, knochenfabrik said:

Hello, i tried most of the things you told me so I need to report :-). When connecting the LH Lemec directly to the uCore all switches and encoders work and I got no strange debug-messages. When connecting the RH Lemec directly to the uCore i have no response of the PCB. I modified the seq_R.ngc as shown above and when loading it i get a bit of debug output (see picture 1 please).

Random trigger behaviour can indicate issues with the DIN shift registers. Check that all resistor networks are the correct type and properly installed with the polarity mark aligned with that on the silkscreen. Check soldering once again, check and swap ribbon cables.

 

6 hours ago, knochenfabrik said:

When connecting the whole Chain  uCore - JA - Lemec LH - Lemec RH, i now get triggers and debug-messages all over the place and none of the encoders on both boards seem to work (hard to tell because of the many random messages).

Did you reload _L.NGC and use the unmodified file?

 

6 hours ago, knochenfabrik said:

I tried to reflow the ICs and checked for contuinity between the ICs. You explained that there should be contuinity from IC1 Pin 10 to IC2 Pin 9 and I got contutinity from IC1 Pin10 to IC3 Pin 9 -I refer to the labels on the PCB (which is also the case on Lemec LH)

Correct, IC1/3/5 are the 74HC165s.

 

6 hours ago, knochenfabrik said:

So I at least got contuinity but could you please explain me what comes before IC1 that could have the effect of making trouble for the whole PCB and chain respectively causing random trigger messages (see picture 2 please). :angelnot:

Before IC is the J89 connector. The SI pin is pin five (middle pin closest to the notch).

 

6 hours ago, knochenfabrik said:

I'm very thankful for your help and since I'm pretty stuck now it always feels so nice to read your thoughts.

Greetings, Alex

Keep on trying and it should work someday :)

Link to comment
Share on other sites

13 hours ago, knochenfabrik said:

@pat_00@CJ55 I also have the same issue but when NG is loaded the LEDs light up just fine when the buttons are pressed (also the green one) - will anyway swap it out to make sure its not the fault of said LED

13 hours ago, CJ55 said:

@pat_00 It looks like the LED on switch 17 is dead. I had the same problem (strange) with the same effect! First it worked all perfectly well and after some testing the same LEDs as on your picture constantly lit.

Maybe the problem is the bending to fit into the switch and the colored LEDs are more sensible in this regard.

Bingo, switched out the LED and everything works like it should. Thanks for the tip. Almost there!

Link to comment
Share on other sites

17 hours ago, latigid on said:

 

If there is a systematic error here, please provide all of the info possible, especially the voltage outputs of 595 registers, results of NG tests including manual control.

 

Random trigger behaviour can indicate issues with the DIN shift registers. Check that all resistor networks are the correct type and properly installed with the polarity mark aligned with that on the silkscreen. Check soldering once again, check and swap ribbon cables.

 

Did you reload _L.NGC and use the unmodified file?

 

Correct, IC1/3/5 are the 74HC165s.

 

Before IC is the J89 connector. The SI pin is pin five (middle pin closest to the notch).

 

Keep on trying and it should work someday :

I modified _R.NGC for the testing with only Lemec RH so when I connected all Pcbs (ucore + ja +lemec LH + rh) I used _L. NGC which was unmodified. Should I modifie _L. NGC instead for the lemec rh test? 

Also, the resistor network next to IC1 seems to have troubles since I measure 8.3kOhm between first and second pin (shouldn't it be 10k?).

Another question. I found out that IC17 is not connected to VCC on Pin16 (no contuinit to the pin on the J89 connector). Is that on purpose? 

Thanks in advance... 

Link to comment
Share on other sites

11 hours ago, knochenfabrik said:

I modified _R.NGC for the testing with only Lemec RH so when I connected all Pcbs (ucore + ja +lemec LH + rh) I used _L. NGC which was unmodified. Should I modifie _L. NGC instead for the lemec rh test? 

Unmodified _L should be used when JA is connected on the left. Unmodified _R should be used when JA is on the right. Modified _R should be used when lemec_R is the first board in the chain. You can of course freely modify the .NGC files to suit the configuration.

 

Quote

Also, the resistor network next to IC1 seems to have troubles since I measure 8.3kOhm between first and second pin (shouldn't it be 10k?).

Pin 1 being the dot marker? The resistors should connect in parallel to the common pin. You should also measure about +5V on the 165 inputs and they shouldn't be shorted together. 

 

Quote

Another question. I found out that IC17 is not connected to VCC on Pin16 (no contuinit to the pin on the J89 connector). Is that on purpose? 

Thanks in advance... 

IC17 should be connected to power. There are two larger plated through-hole vias that you can consider using.

Link to comment
Share on other sites

Hi Guys,

I've finished my midibox seq V4+ few weeks ago but didn't time to post until now ;)

 

Thx to latigid on for his awesome support and TK for the amazing software. I’ve almost scratched the surface but it is above my expectations on every aspect !!

I really enjoy the build !!! and here is some feedback about my build:

About the video build form Hawkeye:

* Super awseome work, i cannot imagine the time it took. It is enough detailed, Hawkeye's voice is pleasant and enjoyed and that's great :p. I learned some tricks during the video so :+1:
* I admit I'm not a video guy and prefer to read. So maybe take this with a pinch of salt. I always tend to skip some minutes on video tutorial and you can miss some precious informations (5mm nuts/ 3mm nuts)
* It will be a great to have high res pictures of boards on important steps like checkpoints contro, MIOS tests, etc ...
* I would prefer smaller video and a playlist: this is easier to find a precise step of the build (granluarity of video head isn't easy on a 2hours long video)

About documentation:
It could be great to have a documentation of common pittfals:

  • Do not solder encoder before led matrix,
  • Do not mout header that wrap 745 to soon
  • 5mm 3mm nuts when connecting the board (they are in the same and If Pay attention to 
  • Do not solder sd card header without attaching the USB board first to be sure get enough clearance.

A screenshot of all the ribbon cable to do with length and pin count and precise header orientation would be really usefull.

I wish I had paint sides of mathias with black to avoid light beed on sides.
Now my mathias do not pop out and I'm stuck with light bleed. especially when you are not exactly in front of your midibox (I've made a video bellow eight first I put tape no bleeding button 8 to 16 no tape => bleeding).
This is espacially true with cusom key caps (Thx Antichambre I love them 
:love:)

I will upload pictures of my awesome SEQ a little later but had to resize them first....

Cheers

 

 

 

Dimitri

Link to comment
Share on other sites

On 6/20/2019 at 11:00 AM, latigid on said:

Unmodified _L should be used when JA is connected on the left. Unmodified _R should be used when JA is on the right. Modified _R should be used when lemec_R is the first board in the chain. You can of course freely modify the .NGC files to suit the configuration.

 

Pin 1 being the dot marker? The resistors should connect in parallel to the common pin. You should also measure about +5V on the 165 inputs and they shouldn't be shorted together. 

 

IC17 should be connected to power. There are two larger plated through-hole vias that you can consider using.

I actually resoldered IC1 and IC17 IC18 very carefully and after that the Error changed a little bit. The Random messages stayed but seem to appear more in a burst..i already tried to switch cables but it is still the same.

but if I now klick on for example SW1 of the Ja-boards MEC-switches SW2 lights up - if i click SW2 , SW3 lights up and so on. Before that none of the LEDs lit up.  Fixed the issue by resoldering IC 17, IC18, IC19,  RN4 and RN5. Ja-boards switches and Leds work fine now when LEMEC LH and RH are chained. Also the Switches on LH board seem to work now when RH is connected to it (which wasnt the case - RH board seemed to kill the LH board when connecting).

Got 4.89V on all ICs and all VCC pins seem to be connected together - only IC17 is not connected to the same VCC rail? i still measure ~5V from pin 8 to pin 16 on IC17 so i guess it still got power (i just cant find out where it comes from haha)

But I guess you dont mean the VCC pins with inputs? On the parallel inputs of IC1 on the failing board i only have 0.6 V on PIN 14  and 1V on Pin 13 for example - on the working LEMEC LH it seems to be that there are 4.9V on all parallel inputs. So maybe I*m coming closer to finding out whats wrong?

I fixed nearly everything, the only thing that is left are those stupid random messages that appear in lightspeed :((

They also appear if i only attach the RH-Lemec to the Ucore and dont load any app or NG file:

[22732.580] set debug on
[22734.666] Debug mode turned on
[22734.675] MBNG_DIN_NotifyToggle(129, 0)
[22734.675] No event assigned to BUTTON hw_id=129
[22734.675] MBNG_DIN_NotifyToggle(130, 0)
........ (counts to 256)
[22734.832] MBNG_DIN_NotifyToggle(256, 0)
[22734.832] No event assigned to BUTTON hw_id=256
[22734.844] MBNG_DIN_NotifyToggle(1, 1)
[22734.844] [EVENT] id=BUTTON:1 hw_id=BUTTON:1 bank=0 fwd_id=LED:1 type=NoteOn value=127 label=^std_btn
[22734.844] MBNG_DOUT_NotifyReceivedValue(1, 0)
[22734.844] MBNG_DIN_NotifyToggle(2, 1)
[22734.845] [EVENT] id=BUTTON:2 hw_id=BUTTON:2 bank=0 fwd_id=LED:2 type=NoteOn value=127 label=^std_btn
[22734.845] MBNG_DOUT_NotifyReceivedValue(2, 0)

............................. and so on. It repeats in a loop. First it comes in slow bursts and then suddenly it starts to go wild

Edited by knochenfabrik
failure changed a bit, fixed the issue with JA boards leds
Link to comment
Share on other sites

On 6/22/2019 at 1:26 PM, Dimduj said:

Hi Guys,

I've finished my midibox seq V4+ few weeks ago but didn't time to post until now ;)

 

Thx to latigid on for his awesome support and TK for the amazing software. I’ve almost scratched the surface but it is above my expectations on every aspect !!

Nicely done!

 

On 6/22/2019 at 1:26 PM, Dimduj said:

I really enjoy the build !!! and here is some feedback about my build:

About the video build form Hawkeye:

* Super awseome work, i cannot imagine the time it took. It is enough detailed, Hawkeye's voice is pleasant and enjoyed and that's great :p. I learned some tricks during the video so :+1:
* I admit I'm not a video guy and prefer to read. So maybe take this with a pinch of salt. I always tend to skip some minutes on video tutorial and you can miss some precious informations (5mm nuts/ 3mm nuts)

Good lesson! Always WTFV (watch the full video).

On 6/22/2019 at 1:26 PM, Dimduj said:

* It will be a great to have high res pictures of boards on important steps like checkpoints contro, MIOS tests, etc ...

I think the video is shot in 4k or better? So you could always pause and zoom? Can you zoom on YT?

 

On 6/22/2019 at 1:26 PM, Dimduj said:

* I would prefer smaller video and a playlist: this is easier to find a precise step of the build (granluarity of video head isn't easy on a 2hours long video)

There are arguments for/against. Ultimately it is up to Peter to decide. Note that there are bookmarks for all chapters.

 

On 6/22/2019 at 1:26 PM, Dimduj said:

About documentation:
It could be great to have a documentation of common pittfals:

Good advice, I can think about an FAQ thread/post to summarise the most popular errors.

 

On 6/22/2019 at 1:26 PM, Dimduj said:
  • Do not solder encoder before led matrix,

Low before high!

 

On 6/22/2019 at 1:26 PM, Dimduj said:
  • Do not mout header that wrap 745 to soon

?

 

On 6/22/2019 at 1:26 PM, Dimduj said:
  • 5mm 3mm nuts when connecting the board (they are in the same and If Pay attention to 
  • Do not solder sd card header without attaching the USB board first to be sure get enough clearance.

Good tips

 

On 6/22/2019 at 1:26 PM, Dimduj said:

A screenshot of all the ribbon cable to do with length and pin count and precise header orientation would be really usefull.

There's a list of cable lengths comprising once part of the BOM on the shop website.

 

On 6/22/2019 at 1:26 PM, Dimduj said:

I wish I had paint sides of mathias with black to avoid light beed on sides.
Now my mathias do not pop out and I'm stuck with light bleed. especially when you are not exactly in front of your midibox (I've made a video bellow eight first I put tape no bleeding button 8 to 16 no tape => bleeding).
This is espacially true with cusom key caps (Thx Antichambre I love them 
:love:)

A personal taste thing, but you can always make it your own :).

Link to comment
Share on other sites

9 hours ago, knochenfabrik said:

I actually resoldered IC1 and IC17 IC18 very carefully and after that the Error changed a little bit. The Random messages stayed but seem to appear more in a burst..i already tried to switch cables but it is still the same.

Your board has some "bodges" correct? Are you certain that they all make the right connections? Post some pics and I can check.

 

9 hours ago, knochenfabrik said:

but if I now klick on for example SW1 of the Ja-boards MEC-switches SW2 lights up - if i click SW2 , SW3 lights up and so on. Before that none of the LEDs lit up.  Fixed the issue by resoldering IC 17, IC18, IC19,  RN4 and RN5. Ja-boards switches and Leds work fine now when LEMEC LH and RH are chained. Also the Switches on LH board seem to work now when RH is connected to it (which wasnt the case - RH board seemed to kill the LH board when connecting).

Good that it's better. It suggests that the soldering on all points could be investigated.

 

9 hours ago, knochenfabrik said:

Got 4.89V on all ICs and all VCC pins seem to be connected together - only IC17 is not connected to the same VCC rail? i still measure ~5V from pin 8 to pin 16 on IC17 so i guess it still got power (i just cant find out where it comes from haha)

IC17 is certainly connected to the same power rail as all other ICs. Remember that CMOS ICs can be parasitically powered by data signals, so verify the connects with the power off and reconnect as necessary.

 

9 hours ago, knochenfabrik said:

But I guess you dont mean the VCC pins with inputs? On the parallel inputs of IC1 on the failing board i only have 0.6 V on PIN 14  and 1V on Pin 13 for example - on the working LEMEC LH it seems to be that there are 4.9V on all parallel inputs. So maybe I*m coming closer to finding out whats wrong?

Sorry, this refers to the parallel inputs (pins 11-14, 3-6). The input states are determined by the encoders and switches, so sometimes you might get 0V on an input that would change if you turned the encoder. The advice was more for checking IC3, which should always have +5V on the switch inputs unless one of a row is depressed. 

These state changes can be monitored in MIOS Studio with set debug on or monitoring DIN switch/encoder events.

 

9 hours ago, knochenfabrik said:

I fixed nearly everything, the only thing that is left are those stupid random messages that appear in lightspeed :((

They also appear if i only attach the RH-Lemec to the Ucore and dont load any app or NG file:

[22732.580] set debug on
[22734.666] Debug mode turned on
[22734.675] MBNG_DIN_NotifyToggle(129, 0)
[22734.675] No event assigned to BUTTON hw_id=129
[22734.675] MBNG_DIN_NotifyToggle(130, 0)
........ (counts to 256)
[22734.832] MBNG_DIN_NotifyToggle(256, 0)
[22734.832] No event assigned to BUTTON hw_id=256
[22734.844] MBNG_DIN_NotifyToggle(1, 1)
[22734.844] [EVENT] id=BUTTON:1 hw_id=BUTTON:1 bank=0 fwd_id=LED:1 type=NoteOn value=127 label=^std_btn
[22734.844] MBNG_DOUT_NotifyReceivedValue(1, 0)
[22734.844] MBNG_DIN_NotifyToggle(2, 1)
[22734.845] [EVENT] id=BUTTON:2 hw_id=BUTTON:2 bank=0 fwd_id=LED:2 type=NoteOn value=127 label=^std_btn
[22734.845] MBNG_DOUT_NotifyReceivedValue(2, 0)

............................. and so on. It repeats in a loop. First it comes in slow bursts and then suddenly it starts to go wild

As before, the common cause here is floating DIN inputs resulting from poorly soldered resistor networks (cold joints or incorrect polarity).

Another possibility is that pins of the 165 are shorted together. Imagine if the clock line was shorted to an input, like pin 2 to pin 3, then you might get a weird feedback loop, or if pin 1 was shorted to pin 2, which would shift in the inputs on every clock.

Did you swap the cables?

Link to comment
Share on other sites

@latigid on Hi, I did swap cables but it didnt change anything for the better. Suddenly also de Leds on the JA Pcb started to flash randomly O.o. So maybe the cables are the reason but with only the Lemec LH connected they work fine and dont do anything bad (will check them again).

Here are some pictures of the ICs. For IC 1 i had to scrape the protection away because i removed the pads (rings) on 3 pins of the resistor network (Pin 1 (dot) ,5 ,7)

Will check the parallel inputs on  IC3 now but its so strange that suddenly the JA board leds begin to do crazy stuff :) pretty confusing for a noob like me :D

Thanks for your help - will definitely have to tip you some beer when I have finished this beatuy - already learned a lot about this project...

debug2306.jpg

debug23062.jpg

Link to comment
Share on other sites

14 minutes ago, knochenfabrik said:

@latigid on Hi, I did swap cables but it didnt change anything for the better. Suddenly also de Leds on the JA Pcb started to flash randomly O.o. So maybe the cables are the reason but with only the Lemec LH connected they work fine and dont do anything bad (will check them again).

Here are some pictures of the ICs. For IC 1 i had to scrape the protection away because i removed the pads (rings) on 3 pins of the resistor network (Pin 1 (dot) ,5 ,7)

Solder bodges like this are not very reliable. Rather I would use a piece of insulated wire and really make sure that you're properly connected to the RN, also that the trace is not shorted to 0V (there is a ground plane everywhere) or to +5V (less likely).

You may have, for instance, shorted the +5V rail to 0V, causing nothing to work when plugging in that board. Not that the SIL header 0V connection sits close to the common +5V pin of the RN.

Floating 165 inputs resulting from no connection to a pullup resistor could explain your random DIN triggers.

Check that all pins are properly connected (view from the rear of the board). You should measure about 10k to the common pin on RN1.

lelemec_IC1.thumb.PNG.d11b3b47e8fffb7f32

 

 

Soldering on ICs could be much better. The blobs on IC2 indicate too much solder and potential cold joints. You can wick off the rest or even try with a solder pump (here be careful to remove any "splashes" of solid solder that come out). It might be the camera angle, but soldering on IC3 looks like some pins may be shorted or close to it.

5d0fe0eaa9dd0_lelemec_IC23.thumb.GIF.65d

 

 

 

14 minutes ago, knochenfabrik said:

Will check the parallel inputs on  IC3 now but its so strange that suddenly the JA board leds begin to do crazy stuff :) pretty confusing for a noob like me :D

You'll get it, don't worry :).

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...