Jump to content

Troubleshooting midiphy SEQ v4+


latigid on
 Share

Recommended Posts

On 4/26/2019 at 5:42 PM, latigid on said:

Hello,

For testing les mecs, you'll want the  seq_r.ngc.

The easiest place to test for power is on the J14 three-pin connector on the east side of the board. 0V in the middle, +5V above (marked with the silkscreen). Always test that these pins aren't shorted together before applying power. The other pins I've already mentioned I think. The voltages will depend on the speed of your multimeter and would be best viewed with a scope.

For 595 chips: pins 16 and 10 = +5V; pin 13 = 0V (resistance may be 10k).
For 165 chips: pins 3-6, 10-14, 16 = +5V; pin 15 = 0V.

For J89 header, pins 1,2 = 0V, pins 3,4 = +5V, 

 

 

 

Thanks!

Link to comment
Share on other sites

So neeeeeeeearly there...

Remembered the 3.3v jumper on the core, put everything in the case (Adrian has done a great job here - built like a tank!), flashed Seq with L/H config and... the right OLED works, the left was gibberish. Rebooted and the left was blank.

I'll check all the connections, but I found an old thread talking about screen issues with low and/or noisy voltage. Is this still a problem? Could it be the 3.3 vs 5v at the board? Or a need for a cap?

Cheers!

Link to comment
Share on other sites

Correct R33D installed, cables and OLEDS both functioning correctly - whatever is plugged into J15A - left-hand screen - shows nothing. The seq is functional - the R/H screen swaps top show whichever half the encoders have been moved for (ie. if I moved encoders 1-8, it shows steps 1-8... if I move encoder 9+, it shows steps 9-16.

I've reflowed the J15A header - there was a bit of resistance on one pin - but it's made no difference...

Any more for any more?!

Link to comment
Share on other sites

I've looked at the schematic (I'm not good at reading them, so take analysis with a pinch of salt!) and checked voltages on the J15A and B headers... when grounding to bottom left pin as I look at the core board installed in the case, I get 5v on the bottom right pin, 3.3v on the top left and nothing significant on top right (very slight negative voltage). Exactly same on both headers.

I have a scope, but no digital analyser so I'm not sure I can sensibly read anything from the data pins but I can certainly give it a go.

I've got correct readings on all the Res to the left of J15B. The contrast pots are doing nothing but I assume that's the nature of the OLEDs?

Link to comment
Share on other sites

5 hours ago, latigid on said:

Check the IDC cable strands don't contact the rear of the OLED PCB. Swap cables to rule out bad cables. 

If you haven't done so already, place a bit of tape or even plastic or cardboard between the IDCs and the OLED PCB.

 

2 hours ago, SimonSays said:

if I moved encoders 1-8, it shows steps 1-8... if I move encoder 9+, it shows steps 9-16

Strange! It might be something to do with the E(nable) lines of the cable, so one display is updating with both displays' data.

The schematic is here:

http://ucapps.de/mbhp/mbhp_core_stm32f4.pdf

Note that the pins are "mirrored". If you check J15A/B, you'll see that all pins are common between the two, apart from the E lines. With the power off but everything (except the USB cable) plugged in, use a multimeter in diode/resistance/continuity mode to check that all signals are common between the corresponding pins of the OLED header, also that no adjacent pins are shorted within pin columns of a header. Especially check the enable lines are not shorted to adjacent pins or to each other! 

Pins are arranged (view from the top of the display)

02 04 06 08 10 12 14 16
01 03 05 07 09 11 13 15

so you need to test e.g. pin 09 for shorts between 08 and 10 etc.

 

From your previous posts, I think the LCD type is correctly configured (plus I believe the SEQ overrides the bootloader setting), but it could be worth uploading the bootloader app again:

http://ucapps.de/mios32/mios32_bootloader_v1_018.zip

MIOS terminal commands are:

set lcd_type 0x02
store

then restart MIOS Studio and upload the SEQ app again.

Edited by latigid on
Link to comment
Share on other sites

@SimonSays wild guess, but maybe bridged "E" lines? E.g. on the waveshare board headers? Could you upload hi-res photos of the front and backside of your wCore PCB? Have you pushed in the waveshare board all the way? The wCore with installed waveshare board must be able to stand on 20mm hex standoffs, if the assembly is higher, the waveshare board is not pushed in fully.

Many greets,
Peter

Link to comment
Share on other sites

Thank you @latigid on and @Hawkeye... given the behaviour with the one screen being shared I'm going to guess at bridged 'e' lines and/or a config issue (the former most likely). I'm at work today and at an event tonight so I won't get the chance to troubleshoot more until tomorrow evening... but I'll have received my extra shipment of 5x2 headers and plugs by then so I can clean-up those cables as well.

I'm pretty sure the Wavemaker is seated as I've done it twice and my palms are a map of the cut pins on both sides as it's a good friction fit!

@latigid on - just to check on theJ15A/B pins, mine run top to bottom as I sit in front of the Seq... is pin 01 top-left, or bottom-right?

Thanks!

Edited by SimonSays
Additions
Link to comment
Share on other sites

Mirroring is very confusing. As you have identified the +5V (unused backlight) and +3v3 (OLED power supply, J15_S is correctly jumpered), you can orient yourself this way. 

I don't think I was clear before: with everything plugged in (power off), check for continuity between the two OLED headers, i.e., left OLED pin #1 with right OLED pin #1 and so on. Also check the adjacent pins within one header for shorts, especially on the enable lines.

Link to comment
Share on other sites

Couldn't resist some probing!

Plastic inserted between OLED pins and headers for surety (no difference, though).

Reflashed bootloader and ran LCD commands... that put some gibberish (some correct characters) on LCD#1 - ie. the problem one. Reflashed Seq and got something on LCD #1 on first boot but blank thereafter. Even with gibberish on LCD #1, LCD #2 was still swapping to display the actively parameters, as before.

Continuity as expected on all mirrored pins EXCEPT e pins. No connection there. I got a limited continuity beep (single pip, 0.303 on screen of DMM) for the adjacent pins to the top-right and top-left of J15A and B (as picture them below) - I think this is actually top left and right on the schematic. I'm jiggered if I can see a short, though! So not sure how to interpret this.

Images attached - can't see any shorts/bridges under a loupe... I can reflow the board tomorrow, though.

IMG_0787.thumb.jpg.ff753980929a51da2bd8eIMG_0786.thumb.jpg.7004f0162d60b330f0d99IMG_0785.thumb.jpg.a361d02513e83f50c70b0IMG_0784.thumb.jpg.6782bb786dc2b96ddf2b9IMG_0783.thumb.jpg.2e3431cc00a76915f95eeIMG_0782.jpg.a8c67f086b104ecc76542b67fdd

Link to comment
Share on other sites

Next two things to try: check that all required lines are correctly pulled up to 3v3. These are RW (R33A), E2/E1 (R33B/C) and RS (R33D, 560R). All of these pins should be at 3v3 when powered up. Always take care not to short pins together on the OLED when probing!

If you don't have another cable/headers around, you could try to use the SD card one (J16E; the application should start without an SD card). You'll probably have to bend the metal parts holding the display in to remove a cable. 

It could also be worth tracing the signals back to the MCU pins.

 

Link to comment
Share on other sites

I'm getting some confusing (to me) readings here... images attached.

Measuring the noted resistors above, I'm getting -3.3v grounding on left/+ve on right, as per image.

On the headers, when I ground to Vs, I'm getting ~0.1 - 0.15v on both E, RW and RS pins.

Hmmm... have I measured this wrong? If not, how is the R/H OLED working?

I've already ruled out issues with the OLEDs/screens - either screen works perfectly when plugged as LCD#2 (R/H).

I've checked continuity between:

PC8/9 on the Wave board and the E pins in the headers (all good);

PC11 and the mirrored RW pins (all good);

 PA8 and the RS pins - no continuity.

So.... I reflowed R33D, got continuity between PA8 and RS thinking that I had the fix.... and no change :(

IMG_0137.jpg

IMG_0788.jpg

Edited by SimonSays
Additions
Link to comment
Share on other sites

I confess that I'm not 100% sure about how the driver works here. It could be that the data edges are positive-going, meaning the MCU pin drives low for most of the time. This you could probe with your scope.

As the displays work when plugged into the other J15, this would seemingly rule out problems with the OLED itself or cables. Meaning that it's further upstream.

I would trace the E lines back to PC8 and PC9 on the v407 breakout (the header pins are labeled), and also back to the MCU pins. It could be that you have a weird contact on the female header for example.

https://www.waveshare.com/core407v.htm
https://www.waveshare.com/wiki/Core407V
https://www.waveshare.com/w/upload/5/58/CorexxxV-Schematic.pdf

Core407V-size.jpg.bc8f6bc1949d28a01d5d31

 

STM32F407-pinout-datasheet.thumb.gif.a10

Link to comment
Share on other sites

Could you check the value/number written on R33D? It does look a bit like an initialisation error. Often this is to do with R33D, hence it is a lower resistance to make the pull-up "stiffer".

Reflowing the resistor shouldn't affect the connection between PA8 and the OLED. Hence it could be something like a cold solder joint or unintentional connection to 0V on R33D, or an intermittent contact on the female header.

 

Link to comment
Share on other sites

Following on from the above, reflowing R33D has made changes.

I now DO have 3.3v on the headers at the E pins and RW pins... but still nothing at the RS pins. No short between E pins.

I've traced continuity all the way from the MCU pins to the headers.

STILL nothing on LCD #1!

 

 

Link to comment
Share on other sites

@SimonSays R33D should be ok - please also check connectivity of all datalines - from the problematic J15 port to the waveshare board (through the headers) - some years ago, i had a display with similar garbage and one data line was disconnected.

You might also want to reflow the pins of the waveshare headers (on the wCore PCB) before measurements, just to try that out before you measure a lot of pins. If it still does not work after reflowing, measurement from J15 to the waveshare core MCU and then to the STM32F4 IC pins are the only way to check for continuity (quite some work!). Also check for "neighboring" unwanted shorts, as those could well cause the problem, too. I'd recommend to reflow every pin first, especially as reflowing on R33D seems to have helped!

Many greets,
Peter

Link to comment
Share on other sites

Oh dear - this is now going very much not according to plan.

Having just checked all data lines through the 595 and back to the MCU, all header pins, etc. I put it back together and it's now completely unresponsive. No green light for power, no nothing. Changed USB cables, still nothing.

Wave board is responsive to it's own USB input power.

Link to comment
Share on other sites

@SimonSays indeed, that should not happen! But: could you check the settings of the switches on the waveshare board? The power switch should be set to "5V in" - if it is wrong (e.g. accidentally changed it during the measurements), that could explain, why the core is unresponsive!

https://www.youtube.com/watch?v=QaN26uzUA1A&t=3362s

Many greets,
Peter

Link to comment
Share on other sites

The Waveshare board settings are as per... after another 2 hours I've localised the power issue to (I tihnk) the Waveshare board itself. I've replaced the SD ribbon connectors now, removed and resoldered the SD card header and the power is still intermittent (basically it works out of the case, and as soon as I put it in, it stops working again).

I've eliminated shorts and can now see that the power only connects if I put moderate pressure on the USB socket on the WS board itself. The board is connected as far as it can onto the core (the pins are too long for it to be flush header to header, but it definitely won't go further), there are no shorts against the bottom of the case. What can be causing the power to intermittently connect/break based on pushing on the WS USB socket?

While it's out of the case, I've managed to get both OLEDs working - basically, by reflowing the entire core.

It's driving me nuts today!

 

 

 

 

Link to comment
Share on other sites

@SimonSays can you measure the connectivity of the PWR 5V and GND pins of the waveshare board to the wCore board (the two bottom-right pins in Andy's picture above)?

If you have inserted and removed the waveshare board from the wCore PCB a lot of times, there might be bad conductivity between gold pinheaders and the header sockets on the wCore, if some headers are bent to the side (this "pinbending" can happen after forceful "uneven" removal of the waveshare board, when you lift one side quicker than the other).

If you measure "fluctuating connectivity" (i.e. the continuity beeper beeps when you push down on the USB socket, and does not otherwise), you could just solder top to bottom bodge wires to connect the pins respectively. I've seen something like that happen once, when the gold pin headers of the waveshare board were bent a little bit after too many removals/reattachments of the waveshare board.

It also requires quite some force to push it fully in, please measure if the standing height of the wCore/waveshare board assembly is just below 20mm (e.g. by using the provided four 20mm hexspacers), so it sits flat on your table, then it is ok.

Many greets and good luck!
Peter

Link to comment
Share on other sites

The mini USB shouldn't have much to do with the +5V power, and it could be a dodgy pin contact (the PWR 5V pin as Peter noted) that reconnects when you put pressure on the board.

A little note on connectivity tests: if you probe for connections between +5V/0V lines, you may get a short "beep" as there are capacitors that charge up temporarily. A short circuit on the +5V lines to 0V (which is quite nasty for your USB power supply and probably other components) is only diagnosed as such when you get a continuous beep or 0 ohm resistance.

 

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...