Author Topic: Midi controller not recognised (Device descriptor request failed). (Read 3584 times)

Califauna · « **on:** July 20, 2020, 03:30:13 pm »

Hi all. First post here.

I have an EKS XP10 midi controller with a Burr Brown DAC which isn't being recognised by Windows.

I've taken it apart and found the chips have had the codes scratched off them. Apparently manufacturers do this sometimes so their products can't be reversed engineered.

Any suggestions? I was thinking of reflowing any chip or solders that might be badly connected. I have a solder reflow station. Perhaps I could start with the chip that contains the device descriptor information?

Audiorepair · « **Reply #1 on:** July 20, 2020, 07:00:32 pm »

Isn't this far more likely to be a software/driver problem?

Califauna · « **Reply #2 on:** July 20, 2020, 10:34:09 pm »

Definitely something up with the controller. Forgot to mention I have two more of the same controllers. Only this one has this problem.

magic · « **Reply #3 on:** July 21, 2020, 05:39:22 am »

If it's not some stupid damage to the USB connector or its PCB traces, see if supply voltages are in spec and there is no ripple.

Califauna · « **Reply #4 on:** July 21, 2020, 12:35:03 pm »

OK. I have checked and confirm there is steady 5v across legs 1 and 4 of USB socket on the underside of the board. Also 3v across data pins 2 and 3.

Califauna · « **Reply #5 on:** July 21, 2020, 03:02:30 pm »

I have also checked the voltage between ground and various testing points on the board and they seem to be the same, except across the two silver contacts at the top left of the board . On the bad board, there are two components with 472 written on them, which look like green painted capacitors to me, soldered on the other side of the board (the side without the components on it). The voltage across them is 0.26 volts.

On the good boards, the components at that spot are 800 ohm resistors soldered on the component side of the board and have 0.5 volts across them.

magic · « **Reply #6 on:** July 22, 2020, 11:27:45 am »

Not sure what those components are but if they are placed at some analog connector, meh, unlikely to be the problem.

USB input seems to go straight to U2. I see no obvious soldering problems with that chip. See if there is continuity from the USB port to the chip. It seems there are two series resistors there so test them too.

As for power supplies, wrong voltage is one thing, ripple might be another. This would take a scope to test for, or just replace all electrolytics (there seems to be only like 3 different types on the whole board) or at least those powering U2.

fzabkar · « **Reply #7 on:** July 22, 2020, 06:31:46 pm »

U2 could be an FT232RL:

https://www.ftdichip.com/Support/Documents/DataSheets/ICs/DS_FT232R.pdf

Califauna · « **Reply #8 on:** July 23, 2020, 10:50:31 am »

Photo of chip.

magic · « **Reply #9 on:** July 23, 2020, 11:51:34 am »

Nice fingerprints

That's not FT232 because it's a standard SO28 package with 1.27mm terminal pitch, not SSOP28. No idea what it is.
Check if the 22Ω resistors are good, check if there is continuity from them to the USB port and a pair of pins on U2.

The 1k5 resistor almost certainly goes to 3.3V supply, by the way.

Califauna · « **Reply #10 on:** July 23, 2020, 01:05:01 pm »

The two resistors read 22.5 ohms.

There is continuity from the resistors and various legs of the chip to the data pins on the usb.

fzabkar · « **Reply #11 on:** July 23, 2020, 07:55:04 pm »

The logo is very familiar, but it's not coming to mind. :-(

Califauna · « **Reply #12 on:** July 23, 2020, 10:52:56 pm »

Yes there's clearly the bottom half of a circle in the bottom right hand corner where they haven't fully sanded off the print.

fzabkar · « **Reply #13 on:** July 24, 2020, 03:27:59 am »

The A, B and I (index?) terminals of the jog wheel appear to head towards U2. If U2 takes input from the wheel, and if it also connects to the USB port, then it is probably a custom programmed microcontroller rather than a USB bridge of some kind. One of the adjacent 8-pin ICs might be a serial flash memory (?), otherwise the uC would have internal flash. There appears to be a second uC (QFP) which receives input from the pushbuttons and switches.

You could use USBDeview to check whether any device appears on the USB port. Just be aware that USBDeview will report all USB devices which have ever been connected to the PC.

https://www.nirsoft.net/utils/usb_devices_view.html

BTW, I'm wondering where the RAM is, assuming that data is buffered somewhere??

Califauna · « **Reply #14 on:** July 24, 2020, 11:14:24 am »

Pins 2 and 5 (from left to right) from jogwheel have direct track connection with legs 10, 11, 15, 16, 25, and 28. 1 and 3 go down to two test holes by R66., which then go (EDIT) left and down to test holes between r61 and r75 via track on both sides.

The track changes to the other side, connecting two test holes, then comes back to the component side by r66.

Califauna · « **Reply #15 on:** July 24, 2020, 11:23:12 am »

USBdevview doesn't seem to provide anything beyond windows device manager. Its listed as unknown device with failed device descriptor and loading the BADDEVICE.Dev.NT driver.

In windows is it possible to manually associate a device with a driver and a device description, on the offchance that it could work even with this chip failing, if that is the cause?

mikerj · « **Reply #16 on:** July 24, 2020, 12:09:14 pm »

Clean the top of the ICs and drop a small amount of IPA or similar on top of them, often partially sanded markings can be made visible.

Califauna · « **Reply #17 on:** July 24, 2020, 02:20:31 pm »

Still not visible.

fzabkar · « **Reply #18 on:** July 24, 2020, 06:55:51 pm »

Quote from: Califauna on July 24, 2020, 11:14:24 am

Pins 2 and 5 (from left to right) from jogwheel have direct track connection with legs 10, 11, 15, 16, 25, and 28. 1 and 3 go down to two test holes by R66., which then come up to test holes between r61 and r75 via track on both sides.

If you look closely at the hub of the wheel, you will see "B5vAIG", with a "1" above the "G". If you haven't come to the same conclusion already, I would think that this identifies the pins as follows:

Califauna · « **Reply #19 on:** July 24, 2020, 08:16:58 pm »

Ok so I should have said:

Jogwheel pins are 1 to 5 right to left looking from the component side of the board. Pins 4 and 1 (5v and Ground) have direct track connection with legs 10, 11, 15, 16, 25, and 28 on the presumed UART chip U2. Pins 2 and 5 (Index and B) go down to two test holes by R66, which then go left and down to test holes between r61 and r75 (via test holes connecting to track on both sides).

Califauna · « **Reply #20 on:** July 24, 2020, 08:43:01 pm »

The DAC IC (Burr Brown 2702) is between C37 and L5.

https://pdf1.alldatasheet.pl/datasheet-pdf/view/56839/BURR-BROWN/PCM2702.html

I checked and 6 and 7 of the 2702 go to legs 21 and 22 on the presumed UART chip U2. From there I can't see. There is 1300 or so ohms between legs 6 and 7 and the USB port data pins.

fzabkar · « **Reply #21 on:** July 24, 2020, 09:25:21 pm »

Could the PCM2702 be one of several USB devices on the PCB, and could U2 be a USB hub???

Does the EKS appear as several devices in Device Manager, or maybe a composite device (I'm referring to a working device, not your faulty one)?

Can you identify the markings on any of the 8-pin chips? If we identify as many of the ICs as possible, perhaps we will have a better idea of the overall block diagram of the circuit. It would help to associate each of the input/output components (eg pushbuttons, jog wheel) with their respective ICs.

Where is the +3.3V supply generated?

Califauna · « **Reply #22 on:** July 24, 2020, 10:55:23 pm »

This is what Devview gives me:

==================================================
Device Name : Burr-Brown Japan PCM2702
Description : USB Audio Device
Device Type : Audio
Connected : Yes
Safe To Unplug : Yes
Disabled : No
USB Hub : No
Drive Letter :
Serial Number :
Created Date : 7/25/2020 12:50:51 AM
Last Plug/Unplug Date: 7/25/2020 12:50:51 AM
VendorID : 08bb
ProductID : 2702
Firmware Revision : 1.00
USB Class : 01
USB SubClass : 01
USB Protocol : 00
Hub / Port :
Computer Name : DESKTOP-42V9VIS
Vendor Name :
Product Name :
ParentId Prefix :
Service Name : usbaudio
Service Description: @wdma_usb.inf,%USBAudio.SvcDesc%;USB Audio Driver (WDM)
Driver Filename : usbaudio.sys
Device Class :
Device Mfg : (Generic USB Audio)
Friendly Name : Burr-Brown Japan PCM2702
Power : 0 mA
USB Version : 1.00
Driver Description: USB Audio Device
Driver Version : 10.0.18362.1
Driver InfSection : USBAudio
Driver InfPath : wdma_usb.inf
Instance ID : USB\VID_08BB&PID_2702&MI_00\9&2094a6b5&3&0000
Capabilities : SilentInstall, SurpriseRemovalOK
==================================================

==================================================
Device Name : Burr-Brown Japan PCM2702
Description : USB Composite Device
Device Type : Unknown
Connected : Yes
Safe To Unplug : Yes
Disabled : No
USB Hub : No
Drive Letter :
Serial Number :
Created Date : 7/25/2020 12:50:51 AM
Last Plug/Unplug Date: 7/25/2020 12:50:51 AM
VendorID : 08bb
ProductID : 2702
Firmware Revision : 1.00
USB Class : 00
USB SubClass : 00
USB Protocol : 00
Hub / Port :
Computer Name : DESKTOP-42V9VIS
Vendor Name :
Product Name :
ParentId Prefix : 9&2094a6b5&3
Service Name : usbccgp
Service Description: @usb.inf,%GenericParent.SvcDesc%;Microsoft USB Generic Parent Driver
Driver Filename : usbccgp.sys
Device Class :
Device Mfg : (Standard USB Host Controller)
Friendly Name :
Power : 0 mA
USB Version : 1.00
Driver Description: USB Composite Device
Driver Version : 10.0.18362.693
Driver InfSection : Composite.Dev.NT
Driver InfPath : usb.inf
Instance ID : USB\VID_08BB&PID_2702\8&889cc25&0&4
Capabilities : Removable, SurpriseRemovalOK
==================================================

Califauna · « **Reply #23 on:** July 24, 2020, 11:34:16 pm »

It installs the following in ' software devices' in Windows device manager:

EKS Midi Driver
SWD\MMDEVAPI\MIDII_E83C0101.P_0004

Califauna · « **Reply #24 on:** July 25, 2020, 12:16:03 am »

Quote from: fzabkar on July 24, 2020, 09:25:21 pm

Could the PCM2702 be one of several USB devices on the PCB, and could U2 be a USB hub???

Does the EKS appear as several devices in Device Manager, or maybe a composite device (I'm referring to a working device, not your faulty one)?

Can you identify the markings on any of the 8-pin chips? If we identify as many of the ICs as possible, perhaps we will have a better idea of the overall block diagram of the circuit. It would help to associate each of the input/output components (eg pushbuttons, jog wheel) with their respective ICs.

Where is the +3.3V supply generated?

Composite device by the looks of it?

Only one with anything partially visible is the ones attached.

Can't tell for sure where the 3.3 V is generated but the 5v supply goes down to U2 and legs 1 and 3 on that have 3.2V at them.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Midi controller not recognised (Device descriptor request failed). (Read 3584 times)

Share me