Detecting Pen Drive Insertion on LPC1768

Hello Tsuneo,

Thanks for your detail response.

My observation is as follow

>1) Host plug-in detection on Device mode
>The voltage at VBUS pin of B receptacle is supplied to a MCU VBUS pin over resistor divider.

The voltage at VBUS pin is 5V while plug-in the cable.

>2) Device plug-in detection on Host mode
>VBUS power should be supplied to the A receptacle.
>Plug-in of device is detected on the LPC1768 chip by D+/D- line voltage change.

There is no changes on D+/D- line while plug-in of usb stick.

Is there any other method to detect the plug-in of USB Stick?

> There is no changes on D+/D- line while plug-in of usb stick.

Huh?
How did you measure it?

In Host mode, D+/D- lines are pulled down by 15k Ohm resistors on the host side, respectively. When a device is plugged in, and when VBUS (5V) is supplied to the device, the device pulls up D+ (or D-) line by a 1k5 resistor. It causes voltage change at D+ (or D-) line.

It's the standard way for the host to detect device connection.

Tsuneo

Your Host circuit is almost similar to Olimex one.
Which option do you take, a) direct connection, or b) Connection over analog switches?

When you take b) over analog switches,
the Olimex circuits around the receptacles are applied as is. Just replace the jumper blocks with analog switches.

But when you take a) direct option,
the 15k pull-down resistors on D+/D- lines are switched by Nch-MOSFETs or NPN transistors in series, respectively. These transistors are enabled on the Host mode by a spare port of the MCU. Sorry, I've forgot to write on this modification in above post.

Tsuneo

The transistor switches for 15k pulldowns are fine.

> Although i am not quite sure if i can use the transistor part on top of the circuit to detect the device (pen drive) plug in.

For Host mode, you don't need any extra circuit for device plug-in detection. It is integrated in the host controller on the LPC176x chip.



For a) Direct connection option,
a sliding door in front of the USB receptacles is necessary, so that just single receptacle is exposed to users at a time. You may need to detect the door position, to switch USB mode between Host and Device automatically.

For b) Over analog switches option,
such a sliding door is not required. Both receptacles are always exposed to users.

Tsuneo

Hello Tsuneo,

I made the hardware change as you suggest for Direct connection option.

The code is as follows.

  if(IS_USB_CABLE_CONNECTED())
  {
    if(!(device_and_host_switch_flag == DEVICE))
    {
      LPC_USB->OTGStCtrl = 0;
      enumerated_flag = DEVICE;
      USB_Init();       // USB Initialization
      USB_Connect(TRUE);        // USB Connect
      while(!USB_Configuration);        // wait until USB is configured
      device_and_host_switch_flag = DEVICE;
    }
  }
  else if(is_usb_stick_connected())
  {
    if(!(device_and_host_switch_flag == HOST))
    {
      enumerated_flag = HOST;
      Host_Init();      /* Initialize the lpc17xx host
      device_and_host_switch_flag = HOST;
      /* Do data transfer */
      LPC_GPIO0->FIOSET  = BIT4 | BIT5; /*switch on trasistors connected to D+ and D- /*
      rc = Host_EnumDev(); /* Enumerate the device connected*/
      if (rc == OK)
      {
        /* Initialize the mass storage and scsi interfaces */
        rc = MS_Init( &blkSize, &numBlks, inquiryResult );
        if (rc == OK)
        {
          rc = FAT_Init(); /* Initialize the FAT16 file system */
          if (rc == OK)
          {
            Main_Copy();        //Copy MSREAD.TXT to MSWRITE.TXT
          }
        }
      }
      LPC_GPIO0->FIOCLR = BIT4 | BIT5;/*switch off trasistors connected to D+ and D-  */
    }
  }
  else
  {
    device_and_host_switch_flag = USB_NONE;
  }

The USB Device is enumerated fine for above code. But when I remove the USB cable and connect the USB stick host fails to enumerate.

If swap the LPC_GPIO0->FIOSET = BIT4 | BIT5; /* and
LPC_GPIO0->FIOCLR = BIT4 | BIT5; /* */ then HOST is enumerate fine but during USB_Init(); the code gets abort.

After detecting the USB stick, software should turn on the both trasistors before transferring data and after completion of data transfer software should turn off the both trasistors right?

-Prakash

In a) Direct connection method, you can't switch device/host role by plug in to the connector. The current role (host/device) is detected by a switch attached to a sliding door in front of the USB receptacles. When the door switch detects role change, firmware disable current role, and it calls stack initialization for each role, host or device. In addition, before the host initialization, the firmware enables D+/D- pull-downs. Also, device initialization is preceded by disabling D+/D- pull-downs.

For a) Direct connection method, a sliding door is necessary so that users don't plug in to both receptacles at a time.

Anyway, try host and device separately, first.

Tsuneo

On the host circuit of above post,
Is the D+/D- pull-down resistors (R8/R9) on the first figure are switched by NPN transistors, as you show in the second figure?

On the device circuit,
Replace NPN transistor (Q1) for the D+ pull-up with PNP.
On the NPN emitter-follower circuit, V_CE doesn't saturate, less than V_BE. You may take this circuit from some KEIL board, but it is wrong one.

Tsuneo

Hello Tsuneo

Is the D+/D- pull-down resistors (R8/R9) on the first figure are switched by NPN transistors, as you show in the second figure?

I am switching ON these two NPN transistors and trying to communicate with the pen drive.

On the device circuit,
Replace NPN transistor (Q1) for the D+ pull-up with PNP.
On the NPN emitter-follower circuit, V_CE doesn't saturate, less than V_BE. You may take this circuit from some KEIL board, but it is wrong one.

It can be seen that Q1 that is being used is already a PNP configuration.

As you rightly suggested, I have tested the HOST and the device code saparatly and everything seems to work fine.

It would be very helpful if you can post the relevent changes. probably a circuit would be very helpful.

-Prakash

>> On the device circuit,
>> Replace NPN transistor (Q1) for the D+ pull-up with PNP.

> It can be seen that Q1 that is being used is already a PNP configuration.

Aha surely, It's PNP.
But the emitter and the collector are swapped on the schematic.

Tsuneo

>Aha surely, It's PNP.
>But the emitter and the collector are swapped on the schematic.

The schematic I have send of USB device is working fine.

-Prakash