search

PaulStoffregen / USBHost_t36

USB Host Library for Teensy 3.6 and 4.0
167 stars 86 forks source link

CP2104 doesn't appear to rx/tx data in the serial mode #24

Closed ladyada closed 5 years ago

ladyada commented 5 years ago

tested two CP2104 boards/cables, they are detected but don't send or receive data. swapping in an FTDI cable works just fine. @KurtE dya remember if you tested CP210x? i could revert to an earlier commit if ya did :)

KurtE commented 5 years ago

I believe I tested a few boards, but some could be different. Which boards are not working? Maybe need to pickup one of them...

ladyada commented 5 years ago

hmm - thats odd - i tried a genuine CP2104 on a Metro Mini - https://www.adafruit.com/product/2590 and ran SerialTest from the latest master commit

KurtE commented 5 years ago

Hi @ladyada,

I believe I only tested with one Silex CP210 board, I need to go figure out if I can find it. Some of this code looks at VID:PID values to know which Serial board it is... Currently the only one setup for this is: 0x10c4:0xea60, which is in the library file serial.cpp.

You might check to see what the pid/vid value is for your board.

I might try to order one of yours or see if this is the one I tested with: Like does this one work: https://www.adafruit.com/product/3309

May need to order one of the ones you mentioned: https://www.adafruit.com/product/2590

Need to see if Amazon carries it, as I can no longer order from you as I don't have a mailbox at my physical address as I don't want the mail delivered a half mile from my house, so I use a PMB in town...

But again you might try checking to see what the product and vendor ids are and edit that serial.cpp file table around line 58 and add it and see if it makes it work.

Kurt

ladyada commented 5 years ago

Thanks i have a https://www.adafruit.com/product/3309 and tried it and it also doesn't work. enumerates but no data goes out when i type into the serial console.

The VID/PIDs all match and are correct - this is the TestSerial output:

*** Device USERIAL1 10c4:ea60 - connected ***
  manufacturer: Silicon Labs
  product: CP2104 USB to UART Bridge Controller
Starting with baud: 115200
Serial Available
plain data
plain data
plain data
plain data
plain data
plain data

(i added a println right before userial.write(ch); to verify)

You mention "Silex", but these chips are SiLabs - just to verify, do you know what chip you tried? maybe it was a knockoff or clone?

PaulStoffregen commented 5 years ago

I ordered the 3309 product.

KurtE commented 5 years ago

Earlier today I ordered the two I mentioned (both the breakout board and the metro board) from Amazon. I think the breakout board is scheduled to arrive Friday and the Metro on Monday.

The one currently have is some generic one I think I either ordered on Amazon or E-Bay.

Markings on back say D-SUN, USB TO TTL, I will try it out again and also again try it when the Adafruit ones come. As you mentioned, the one I have may not be an official one, or different model

The output from Serial test shows:

CP210X:  0x41, 0x11, 0, 0, 0 - reset port
*** Device USERIAL1 10c4:ea60 - connected ***
  manufacturer: Silicon Labs
  product: CP2102 USB to UART Bridge Controller
control callback (serial) F
control callback (serial) E
control callback (serial) C
control callback (serial) 8
control callback (serial) 0

I typed in some text, note, I have debug turned on to see what is going on:

Serial Available
txtimer
  TX data (21) 49 66 20 61 74 20 66 69 72 73 74 20 79 6F 75 20 64 6F 6E 74 0A 
rx token: 803F8100 transfer length: 64 len:1 - 49 0
rx: 49 
tx1:
Irx token: 3B8100 transfer length: 64 len:5 - 66 20
rx: 66 20 61 74 20 
f at rx token: 80348100 transfer length: 64 len:12 - 66 69
rx: 66 69 72 73 74 20 79 6F 75 20 64 6F 
first you dorx token: 3D8100 transfer length: 64 len:3 - 6E 74
rx: 6E 74 0A 
nt

With debug turned off, I have:


USB Host Testing - Serial
*** Device USERIAL1 10c4:ea60 - connected ***
  manufacturer: Silicon Labs
  product: CP2102 USB to UART Bridge Controller
Serial Available
I just typed in a line of text

So maybe 2102 is different than 2104 here. Note: for this test I have the RX jumper to the TX pin of the converter, such that everything that I type into the Arduino Monitor is sent out to the device which then sees the data and reads it back in and prints the data out.

ladyada commented 5 years ago

ooh yeah 2102 is probably different than the cp2014 - but they have the same VID/PID? weird!

KurtE commented 5 years ago

Also maybe I need to try some other tests, like to verify that setting the baud rate works. The main test I did was a simple loop back test, which if the baud rate setting does not work, may not be caught. But will do more testing once I receive some of your products with the 2104 chips. I may also dig around in my storage to see if I have any other boards that used their USB to Serial chips. I think some of the old Basic Micro boards did, although they are no longer produced and I may not still have their IDE installed anywhere…

On Dec 26, 2018, at 2:16 PM, ladyada notifications@github.com wrote:

ooh yeah 2102 is probably different than the cp2014 - but they have the same VID/PID? weird!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/PaulStoffregen/USBHost_t36/issues/24#issuecomment-450034126, or mute the thread https://github.com/notifications/unsubscribe-auth/ABfGQPkzeRSPMr6yusYMxAE0a3TOCBRGks5u8_VYgaJpZM4ZhSdc.

FrankBoesing commented 5 years ago

According to the datasheets: https://www.silabs.com/documents/public/data-sheets/CP2102-9.pdf and https://www.silabs.com/documents/public/data-sheets/cp2104.pdf

They have the same default VID/PID: 10C4h/EA60h If the OTP-ROM has not been customized.

KurtE commented 5 years ago

Thanks Frank,

Note: I just tried running the test again with one of my Saleae logic analyzers attached to the TX/RX pins, and verified that the default program was properly outputting at 115200. I then recompiled for 9600 and the Autobaud setting on logic Analyzer said that the new baud as 9606, so looks like that part is working. Again not sure what else I can check here until new hardware arrives.

ladyada commented 5 years ago

hihi, just wondering if anyone else was able to replicate the bug (at least) - or if its something amiss with my setup!

KurtE commented 5 years ago

Hi @ladyada (and @PaulStoffregen)

Yes - I am able to reproduce. So far I have only tried with the Arduino board... Which shows the problem.

I know that some of the init code is failing. I know that some of the chips had an issue where something like some register value gets reversed. I have seen some tests that the Linux driver does to see if a board has the issue or not. I did not put that level of support in yet and don't remember if the board I started off had it or not... My guess is one of these two boards has it and the other does not...

What I probably need to do is to run the setup code on a Linux box again and hopefully capture the USB startup sequence using my Saleae Logic Pro 8, and try to diagnose what packets and values are seen or not seen, ... I Know earlier I would save away some reports and use something like grep and editors to figure out the data... But has been awhile

Problem is I am having too much fun trying to help out on some other projects, so I have looked any deeper yet. Also don't have as much time due to 13 week old puppy

KurtE commented 5 years ago

Hi @ladyada and @PaulStoffregen -

I played around some and looked at the differences... Also looked through Linux and a Android driver and changed a little of the configuration code.

It appears to work for both Adafruit boards as well as the previous board I had.

Hopefully pull this one in to next beta? Also hopefully pull in other PR in this library to fix some keyboards that were broken by the gigabyte keyboard fix

ladyada commented 5 years ago

thanks - tested and works!