Antichambre Posted November 15, 2017 Report Share Posted November 15, 2017 (edited) 9 hours ago, Hawkeye said: I've also looked through the thread to see high-res photos of front and backside of the core (without the stm32f4 module installed) and did not find them (but maybe i did not search good enough) - so, if you already have them, please upload them, if not, you'd need to take photos directly from above the frontside and the backside of the core (not from an angle), so we could look at it, sometimes, solder points can go bad, even if you think they look good and that you have checked them. I found it in previous page of topic, but there's too much flux residue on the back of your board to discern something. There's already some suspicious pads like those: Maybe this is just reflect, and it's difficult to see the relief from a top view picture... I suggest you to clean all pads with paste flux and iron, no need to add more solder, then clean the board(good flux remover) take a picture and share with us please. This is the paste flux I use: Best regards Bruno Edited November 15, 2017 by Antichambre Adding howto video Quote Link to comment Share on other sites More sharing options...
Antichambre Posted November 15, 2017 Report Share Posted November 15, 2017 You call it 3 times!!! maybe "Evil" is in your circuit!!! And we have to call an exorcist now! Quote Link to comment Share on other sites More sharing options...
EvilEvilEvil Posted November 16, 2017 Author Report Share Posted November 16, 2017 (edited) Thanks Bruno - for the vid and the suggestions. I am going to try one more LCD and then hit your steps. Yesterday I gave up. Today I'm willing to try some more Can I just use isopropyl alcohol as flux remover? Edited November 16, 2017 by EvilEvilEvil Quote Link to comment Share on other sites More sharing options...
Antichambre Posted November 16, 2017 Report Share Posted November 16, 2017 I know we are on a DIY platform;) But sometimes it is better to invest in tools and consumables than to damage one's own work. Isopropyl will do the work but there is a risk of not stopping there and attacking the varnish and all the plastic components. So go for it with delicacy. Personally I use a French brand that does not melt soldermask, silkscreen or plastic. I'm sure you can find the equivalent near you. Translated name is "low surface tension flux cleaner", exact composition is not specified. Best Bruno Quote Link to comment Share on other sites More sharing options...
Antichambre Posted November 16, 2017 Report Share Posted November 16, 2017 I found it, there's isopropyl inside but not pure List of ingredients isopropylalcohol Hydrocarbons, C6-C7, isoalkanes, cyclics, < 5% n-hexane ethanol; ethyl alcohol methylal 1-ethoxypropan-2-ol carbon dioxide 2-ethoxypropan-1-ol methanol Quote Link to comment Share on other sites More sharing options...
EvilEvilEvil Posted November 19, 2017 Author Report Share Posted November 19, 2017 Could the fact that I haven't yet installed the SD reader contribute to this issue? Does the SD reader need to be installed to see results on the display? Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted November 20, 2017 Report Share Posted November 20, 2017 16 hours ago, EvilEvilEvil said: Could the fact that I haven't yet installed the SD reader contribute to this issue? Does the SD reader need to be installed to see results on the display? Nope, should work. Can you go to the bootloader and set the LCD type again? It is explained here: http://www.ucapps.de/mios32_bootstrap_newbies.html Here is the relevant section for the SEQ: Example2: two HD44780 based 2x40 character LCDs are connected to your MIDIbox Enter: lcd_type CLCD lcd_num_x 2 lcd_num_y 1 lcd_width 40 lcd_height 2 store Many greets and good luck! Peter Quote Link to comment Share on other sites More sharing options...
EvilEvilEvil Posted November 20, 2017 Author Report Share Posted November 20, 2017 Thanks again Hawkeye. I'll try that. In the meantime, here's the voltages I'm measuring at J15A Quote Link to comment Share on other sites More sharing options...
EvilEvilEvil Posted November 20, 2017 Author Report Share Posted November 20, 2017 8 hours ago, Hawkeye said: Nope, should work. Can you go to the bootloader and set the LCD type again? It is explained here: http://www.ucapps.de/mios32_bootstrap_newbies.html Here is the relevant section for the SEQ: Example2: two HD44780 based 2x40 character LCDs are connected to your MIDIbox Enter: lcd_type CLCD lcd_num_x 2 lcd_num_y 1 lcd_width 40 lcd_height 2 store Many greets and good luck! Peter Quote Link to comment Share on other sites More sharing options...
latigid on Posted November 20, 2017 Report Share Posted November 20, 2017 MIOS Studio shows the SEQ application, you have to upload the Bootloader: http://www.ucapps.de/mios32/mios32_bootloader_v1_018.zip Where are you located? Perhaps you could send one of your LCDs to someone to test on known good hardware. Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted November 20, 2017 Report Share Posted November 20, 2017 Andy: probably he listed the MIOS dump to show, that the LCD type is set correctly (need to zoom in the second screen to enlarge). Regarding the J15 voltages: I've never measured all voltages on J15, but what puzzles me, is that the voltages of DB0-DB3 (the four data bus pins) are all "high". But maybe this is the default state in 4-bit mode, i am not sure about it, but it looks strange. It sure would be nice, if someone with a core lying around could confirm those measurements, i unfortunately just have no time right now... :-/ It is really a mystery! Many greets, Peter Quote Link to comment Share on other sites More sharing options...
slo Posted November 20, 2017 Report Share Posted November 20, 2017 I have taken the values from an NG flashed core that has a working LCD on J15A. Measurements taken starting at top pins (#1), key on left side 2 Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted November 20, 2017 Report Share Posted November 20, 2017 Thanks for the super fast measurements, slo! Awesome! This would confirm my doubts about all data lines being "high" (rows 4/5 in slo's table, which are all 5v on Evils measurements), this might be the problem. Evil, I'd suspect maybe an accidental "bridge" to 5v somewhere on the core or a faulty 595 (maybe inserted the wrong way around??!?). As you said, that you have a second core, can you install the seq firmware on that second core exactly as you did on the first one and measure again, and see if the voltages are different? Edit: also please check the correct value and alignment of the resistor array R33a/b/c/d (i think Bruno recommended to check this, first)! 2nd edit: yes, the resistor array could be the cause, and maybe you soldered in both in the wrong alignment on both cores? You'd own Bruno/Antichambre a beer, then! ;-) Many greets! Peter Quote Link to comment Share on other sites More sharing options...
latigid on Posted November 20, 2017 Report Share Posted November 20, 2017 Better than a mulimeter would be an oscilloscope to verify the pulses coming from the Core pins, 595 and display. The voltages between ~0-5 are a result of time multiplexing I think, possible if the update rate on the multimeter is faster. If the display never initialised properly, the 595 outputs may be stuck waiting for data. Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted November 20, 2017 Report Share Posted November 20, 2017 Just now, latigid on said: Better than a mulimeter would be an oscilloscope to verify the pulses coming from the Core pins, 595 and display. The voltages between ~0-5 are a result of time multiplexing I think, possible if the update rate on the multimeter is faster. If the display never initialised properly, the 595 outputs may be stuck waiting for data. Fully agreed with the oscilloscope, there are DIY kits for ~20$, that would be sufficient for the job, so this is no excuse anymore :-) Quote Link to comment Share on other sites More sharing options...
Antichambre Posted November 20, 2017 Report Share Posted November 20, 2017 (edited) 2 hours ago, Hawkeye said: Better than a mulimeter would be an oscilloscope 2 hours ago, Hawkeye said: Fully agreed with the oscilloscope, there are DIY kits for ~20$ Better than Multimeter and Oscilloscope is the Logic Analyser. It can be very cheap too, 9€ here! 24MHz is enough. You will see multiple waveform as concrete things on a timeline, very good to learn or dig in with a lot of fun... Edited November 20, 2017 by Antichambre 1 Quote Link to comment Share on other sites More sharing options...
latigid on Posted November 21, 2017 Report Share Posted November 21, 2017 Do your cables look like this: Notice both connectors face the same direction relative to the cable. On your one, one IDC is "up and the other "down" TK.'s has a bend on the left connector. I need a coffee, so not sure if this makes a difference electrically, but worth a try at this point. Quote Link to comment Share on other sites More sharing options...
Hawkeye Posted November 21, 2017 Report Share Posted November 21, 2017 me needs a coffee, too ;-) <detective_mode> Good catch! It makes a difference electrically, that's also what TK. saw and why he posted his cable pictures, methinks. The cables are ONLY valid, if the connector housing arrow(notch) on both ends points e.g. to the red wire, but to complicate things, it does not matter, if one connector is turned by 180° to make an "s-type" cable or not. The reason behind that is, that you can bend the cable by 180° around the plug and the upper wire will always stay the upper wire, that's why you can optionally mount the plug strain-relief or not (which bends the cable by the same 180°), the wiring will stay the same. And you are right, in Evil's first picture (your first quoted picture) above, the notch of the cable plugged into the display clearly was not to the left of the screen, and that's what TK. also saw, methinks. It must look exactly like TKs picture (your second quoted picture), so this might be the reason, why the display got burned after some time of flawless operation in the first place, when it was plugged in again. Evil: it is not sufficient to just "rearrange" cables, you might have burned all displays, that were plugged in that way. </detective mode> ;-) Many greets, Peter Quote Link to comment Share on other sites More sharing options...
Nirokesan Posted November 21, 2017 Report Share Posted November 21, 2017 As there is no shroud on the LCD's header, the orientation of the notch on that end really doesn't matter either. What matters is that the cable goes "straight/flat" to the LCD, no twist. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.