TK. Posted December 2, 2013 Report Posted December 2, 2013 Yes, it somehow related to the 74HC595. I haven't found out yet, what exactly could cause such an effect. Best Regards, Thorsten. Quote
novski Posted December 2, 2013 Report Posted December 2, 2013 (edited) cold it be because of the open pin 9? edit: no its QA (pin15) behaving strange... Edited December 2, 2013 by novski Quote
John E. Finster Posted December 10, 2013 Report Posted December 10, 2013 (edited) Hi, I´m sorry to barge in on the current discussion, I have a display related question but don´t want to open another topic for that. Are T6369c displays supported by MB_NG? I´ve noticed there are some drivers available in the mios modules section. I have some of those laying around and would want to use them for at least 2 future projects I´m planning out right now. Thanks my regards Edited December 10, 2013 by John E. Finster Quote
TK. Posted December 10, 2013 Report Posted December 10, 2013 The T6369c has a very poor performance and requires buffering in memory. Therefore it's not part of the universal driver (it's a dedicated driver), and probably compile will fail due to unavailable RAM resources. Best Regards, Thorsten. Quote
John E. Finster Posted December 10, 2013 Report Posted December 10, 2013 (edited) Too bad... Just another box of scrap for the junkyard :happy: . Thanks for the info. Are there other alternative graphical displays this size with at least 240px width? The old KS0108 with 240x64 pxs can not be found anymore so it seems. There are some KS0108 compatible with 240x128 pxs on ebay at the moment, would they be compatible? I almost bought some of them but after the other conversation about the KS0108 displays not being selected with the normal lcd_parameters I´m not so sure after all. my regards Edit: Sorry, just found out, that these 240x128s are also based on a t6369c controller, but are sold as ks0108 compatible. Oh well... Edited December 10, 2013 by John E. Finster Quote
TK. Posted December 10, 2013 Report Posted December 10, 2013 Do you have the datasheet of such a display? Best Regards, Thorsten. Quote
John E. Finster Posted December 10, 2013 Report Posted December 10, 2013 The ebay offer doesn´t state what display it is, just that it´s ks0108 compatible here. But I looked up the number visible on the back of the screen and found a matching datasheet here. On that a T6360C is listed as the controller. Quote
TK. Posted December 10, 2013 Report Posted December 10, 2013 No, this display is definitely not KS0108 compatible. E.g. it would require 8 CS inputs for 8 segments... such a higher number doesn't make much sense, therefore KS0108 isn't used for large display sizes anymore. The display interface looks more like a T6369c Best Regards, Thorsten. Quote
John E. Finster Posted December 10, 2013 Report Posted December 10, 2013 Well, I guess I have to keep lookin´ then... Thank you very much for your input. my regards Quote
TK. Posted December 13, 2013 Report Posted December 13, 2013 @KUI @novski There is new hope! While stress-testing the OLEDs today (more about the app in the next days...), I found a potential flaw in the LCD driver, and managed to fix it with a plausible method. Could you please check if this app is working at your side? -> http://www.ucapps.de/tmp/mios32_test_app_lcd_ssd1306_LPC1769__20131215.zip @KUI: if you would try this with a local build, just update your SVN repository Best Regards, Thorsten. Quote
novski Posted December 14, 2013 Report Posted December 14, 2013 Hi TK. My prototype is on the way to me. Cant't take long anymore... as soon i have it, i will test. best regards novski Quote
TK. Posted December 15, 2013 Report Posted December 15, 2013 Ok, I will be very interested on the results! Meanwhile another version which should be even more stable: -> http://www.ucapps.de/tmp/mios32_test_app_lcd_ssd1306_LPC1769__20131215.zip I tested the changes at my side with the MBCV V2 application, see this video :) Best Regards, Thorsten. Quote
KUI Posted December 15, 2013 Report Posted December 15, 2013 (edited) Hi TK, Finished my last show for 2013:) I have tested the changes and can report that there are some improvements and still some undesirable issues. The boot issue has disappeared. I was testing the display 16 corruption and thought to tell the MBNG that it had 24 displays attached. What should be display 16 appears to be display 24. If I type " lcd@(24.1.1) test " the text appears on display 16. Also now having trouble with dispalys 4 and 6 staying dark. If I disconnect the second display board completely, reboot, displays 1-8 are fine. If you would like pics please just say so. Also very impressive video with the synth:) Today: Continued testing. it looks like there is something amiss in the SR routing. With mbng configures with 24 displays. On the SmashTV dout module J3 O7 should be display 16. J3 O7 is definitely display 24. On J4 O7 does nothing. O0 is display 23, O1 is display 22, O2 display 21, O3 is display 20, O4 is display 19, O5 is display 18, O6 is display 17, O7 is nothing. When configured for 32 displays the pattern repeats, but with displays 32-25 with display 32 missing, it should appear on J5 O7. When configured for 40 displays the pattern repeats, but with displays 33-40 with display 40 missing, it should appear on J6 O7. Also, during all these tests J3 O7 continued to be Display 24. Hope this helps Keep it up, we'll find it:) KUI Edited December 16, 2013 by KUI Quote
novski Posted December 20, 2013 Report Posted December 20, 2013 (edited) maybe off topic: wasn't there once a guy who found a fail routed (or not connected) Net on the SmashTV LPC17 Core? what connection was it? I can't find that topic it anymore. I has happened to be January 2013 or sometime around that... Edited December 20, 2013 by novski Quote
TK. Posted December 20, 2013 Report Posted December 20, 2013 I can't remember this - please keep me updated if you find the posting! Best Regards, Thorsten. Quote
novski Posted December 20, 2013 Report Posted December 20, 2013 (edited) i don't know if the following can be and if, how it can be. I got my PCBs yesterday and soldered them together as fast as i cold. What i now noticed is that a different core works. I measured al connections and can't find a difference, but it seams somehow that the core MBHP_CORE_LPC17 V1.0 that i have here on my table does not work and my fresh made core works. I then looked at the Video again and saw that the core in the right upper edge has a red solder mask color on it. So it seams that me and KUI are working with a Green one and TK with a red. I realy can't tell what the difference between my two cores is but i can load all of the software in this post that is still linked and it works correct. (the links from apps older than mid Nov. are not available anymore...) I think the best i can do is pack my stuf in a pack and send it to TK. @TK do you have a Green MBHP_CORE_LPC17 V1.0 Core? best regards novski Edited December 20, 2013 by novski Quote
KUI Posted December 20, 2013 Report Posted December 20, 2013 Novski, Are you referring to the core pcb solder mask color? KUI Quote
TK. Posted December 20, 2013 Report Posted December 20, 2013 Very interesting! No, I don't have a green colored PCB Before sending me the hardware, first let's try to analyze this remotely. Could you please create high-resolution pictures from the top and bottom side of the LCD (J15) related circuitry? I need pictures from the "good" and from the "bad" PCB. Best Regards, Thorsten. Quote
KUI Posted December 20, 2013 Report Posted December 20, 2013 I have 3 core pcb's, All from SmashTV. Two are stuffed. I get the same behavior from both. KUI Quote
TK. Posted December 20, 2013 Report Posted December 20, 2013 The question was directed to Novski, but KUI: since you've an unstuffed module, pictures would be interesting as well! This would allow me to check the traces at the top (because I guess that on Novski's PCB they will be hidden by the 74HC595 and some sockets) Best Regards, Thorsten. Quote
novski Posted December 20, 2013 Report Posted December 20, 2013 I will make pictures as soon as i can. @KUI: thats perfect. May you build uf the third board? I wold sugest to just place the parts you need to run the oleds. Leave out everything bevor J2. (No rectifier, LM317 and Pol-Cap. Then power it with 5V straight to the LF33. You will need Ic2 and the resistornetwork as well as his cap. But leave out the BC337 and the potentiometers 10k fur brightness and ilumosity. Maybe tat it will work then... Best regards Novski Quote
novski Posted December 21, 2013 Report Posted December 21, 2013 i just made a discovery. As i said the self made pcb works, but as soon the cable length precedes more than 70cm also there i get corruption on OLEDs 9-16. whats quite strange is that the corruption on OLED 1-8 appears already after about 20-30cm of cable length and gets worse. I drawn a schematic to explane the cable lengt in detail: my Cable A is and always stays 8cm, B is the cable i vary. Cable B = 10cm - no problem every oled works fine Cable B = 30cm - oleds 1-8 corrupt, mostly n°3 & n°5 first. Followed by 6,7,8. one an two are OK as well as all oleds on second SR (oled9-16) so D0,D1 & DC seams to be very sensitiv on cable length. What i see is a total stable second SR oled line (oled9-16) . So why don't we try to test without oleds on j15 as a workaround? Wold it be possible to not initialize the IC2 as shift-register for oled 1-8 and initialize the shiftregister that now drives oled 9-16 as oled CS 1-8? best regards novski ps: i once scanned the smashTV PCB before i assembled it but just top side... i can't see anything bad or exceptional or so... (i changed the C because i wanted to test it and go sure its not a broken part that makes me stumble... ) Quote
TK. Posted December 21, 2013 Report Posted December 21, 2013 I've another idea (and will explain the reasons if my assumption is true): Could you please try this version: http://www.ucapps.de/tmp/mios32_test_app_lcd_ssd1306_LPC1769__20131221.zip Required hardware change: connect the D1 lines (SDA) to J15B:E2 instead of J15A:RW J15A:E1 will output SCLK (D0 lines) as before. Best Regards, Thorsten. Quote
KUI Posted December 21, 2013 Report Posted December 21, 2013 Hi TK, I will try the new test shortly. Cheers KUI Quote
novski Posted December 23, 2013 Report Posted December 23, 2013 OK, i tried it... it does not work on my side. And to me it seams a bit worse than bevor. bevor i had 5 oleds display something, now with mios32_test_app_lcd_ssd1306_LPC1769__20131221.zip i have 3 displays showing something. #3 is corupted #5 is working for 1 min than corrupted #7 is corrupted #8 sometimes there are some digits apear alone... most of the time its black. #9-16 work perfect. best regards novski Quote
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.