SEQ V4 Wilba panel LED troubleshooting

[tabularasa]

Hello,

 

I soldered the LEDs on a SEQ V4 panel and some of them keep glowing dim when they should be off and one is always glowing very bright. I checked for shorts but couldn't find anything. Then I changed the 74HC595s but I have still the same effect. Debugging is really difficult since I can always measure a voltage between the pins of the LEDs.

 

The config file is is a copy from here: http://svnmios.midibox.org/filedetails.php?repname=svn.mios32&path=%2Ftrunk%2Fapps%2Fsequencers%2Fmidibox_seq_v4%2Fhwcfg%2Fwilba%2FMBSEQ_HW.V4

 

According to Wilba's schematic all of the missbehaving LEDs share a connection to Out14 but I could no find a shortcut between this line and other connections on the board.

http://midibox.org/forums/index.php?app=core&module=attach&section=attach&attach_id=8731

 

The Track and Track Group LEDs are less bright if I switch on the step LEDs.

 

The power source I'm using is a Meanwell RTP-60B with 100kHz switching frequency connected to J2 and on Core STM32. Might there probably an issue with the multiplexing frequency or do you have an idea what could be the root cause? I'm quiet lost and don't know how I can effectively troubleshoot this issue.

 

 

Martin

 

[ilmenator]

The power supply is plenty and the switching frequency will not affect the problem.

 

Double-check all soldering joints at the shift register ICs. I guess you have forgotten one, hence there is a floating voltage on some LEDs which causes your problem. A hi-res picture of the underside might help spotting the problem (but this board is large).

 

Edit: Oh, and also check orientation of the resistor networks (RNs)!

Edited September 17, 2013 by ilmenator

[tabularasa]

Thanks a lot for the hints.

 

Today I resoldered the 2 74HC595 today but there is no difference. The orientation of the resistor networks is correct andI think they are in line to read encoders and buttons which is working fine.

 

There is a resistance of 43 kOhm between the QA-QH outputs and the ground connector of the 595 shiftregister, is this normal or shall I look look for shorts?

 

I can also measure a resistance between the outputs of the shifregister 1 and 2 (Out vs DOUT) but these are serveral MOhm, so I don't think this is the root cause of the problem.

.

Is there a good methology to track down the root cause?

[blingy]

I saw similar behavior caused by a reversed LED...double check LED orientations as well

[tabularasa]

I feel so embarrassed. I checked the LED orientation at least 5 times and I always overlooked that the Step View LED has a different orientation than its neighbours in that row.

Thanks a lot for the help, now everything works :thumbsup:

Edited September 20, 2013 by tabularasa

[rosch]

phew, i'm glad it works. embarassment would have been on my side ;)

[tabularasa]

I really tried hard but could not find bad solder joints ;-)

The good thing is that I learned a lot about shift registers and how the panel works. I only killed two fully operable 74HC595s ;-)

Edited September 21, 2013 by tabularasa

[smashtv]

Hi guys,

 

I added a note about LED 31 orientation (twice + bold text) to the SEQ construction guide in the wiki.

 

Best regards

 

Tim

 

 

[tabularasa]

That's a good addition and can save others from troubleshooting. Thanks Tim

[niklasni1]

Just so it's clear to everyone: It IS possible to misorient other LEDs than L31, so check all of them. Resoldering LEDs with cut legs is no fun..  :smile:

Edited October 24, 2013 by niklasni1

[slo]

I added a note about LED 31 orientation (twice + bold text) to the SEQ construction guide in the wiki.

Can some post a link to this construction guide in the wiki, I can't seem to find it. I can find the wilba one, but it does not contain the reference to L31.

 

George

[niklasni1]

 http://www.midibox.org/dokuwiki/doku.php?id=wilba_mb_seq_construction_guide

[slo]

Thanks, but there is no reference to L31 on that page

[niklasni1]

It's written as LED31.

[slo]

I'm an idiot, thanks for your patience

 

George