LPC4357 heats up at high frequency ?

[lpcware]

Content originally posted in LPCWare by Algorithm on Wed Aug 19 11:31:17 MST 2015
I'm using LPC4357 chip, I followed the Ramp up instructions on the user manual to raise the frequency to 200 MHz . When the frequency is raised , the chip gets warm but its working fine, I still can touch it as it is not really hot , I wonder if that is normal ?
//
Thats my Ramp up code

uint32_t status;
// Setting the CORE-M4 clock as the IRC (12 MHz)
status=CGU_EntityConnect(CGU_CLKSRC_IRC,CGU_BASE_M4);
if(status!=CGU_ERROR_SUCCESS) while(1);
//

// Set the XTAL freq
status=CGU_SetXTALOSC(12000000U);
if(status!=CGU_ERROR_SUCCESS) while(1);
//
// Enable the XTAL OSC
status=CGU_EnableEntity(CGU_CLKSRC_XTAL_OSC,ENABLE);
if(status!=CGU_ERROR_SUCCESS) while(1);
//
// Reconfiguring PLL1
// Setting the PLL1 clock as the XTAL_OSC
status=CGU_EntityConnect(CGU_CLKSRC_XTAL_OSC,CGU_CLKSRC_PLL1);
if(status!=CGU_ERROR_SUCCESS) while(1);
//
// Confiugre multipliers
LPC_CGU->PLL1_CTRL|=(1<<0); // Power down temporarily
LPC_CGU->PLL1_CTRL&=~(1<<6); // CCO is the feedback divider input clock
LPC_CGU->PLL1_CTRL&=~(1<<7); // Direct mode diabled
LPC_CGU->PLL1_CTRL&=~(3<<8); // PSEL=0
LPC_CGU->PLL1_CTRL&=~(3<<12); // Clear NSEL
LPC_CGU->PLL1_CTRL|=(2<<12);// NSEL= 3
LPC_CGU->PLL1_CTRL&=~(0xFF<<16); // Clear MSEL
LPC_CGU->PLL1_CTRL|=(49<<16);// MSEL=50
//
// Turn on PLL1
LPC_CGU->PLL1_CTRL&=~(1<<0);
//  wait for PLL1 to lock
while((LPC_CGU->PLL1_STAT&1) == 0x0);
//
// Connect PLL1 to CORE M4
status=CGU_EntityConnect(CGU_CLKSRC_PLL1,CGU_BASE_M4);
if(status!=CGU_ERROR_SUCCESS) while(1);
for(status=0; status<1000000; status++);
//
// enable direct mode
LPC_CGU->PLL1_CTRL|=(1<<7); // Direct mode enabled

[lpcware]

Content originally posted in LPCWare by mch0 on Sun Aug 23 14:21:30 MST 2015
Unfortunately it could(!) be normal.

I'm using several different members of the 43xx series in different configurations (range 4320@96MHz to 4370@200MHz) and at 200 MHz and some peripherals active they all get at least *quite* warm.
With the 4370@200MHz and both the built-in USB-HS  and the  HS-ADC turned on it got (and gets) so hot I thought I made some HW-blunder somewhere at first, like exceeding the supply voltage limit or some short circuits.
But when you add up the numbers given in the DS for power consumption it turns out that this amount of heat is to be expected :/

The good news so far is that none of them has died yet (air cooled & room temp) during "normal" operation.
However, you *can* mess up with the main PLL such that the chip freezes at obviously an out-of-range frequency and during these periods I was really scared about the temperature level reached.

I the lab, I could remove power to get the chip out of the freeze but if that condition persists any amount of time in the field I'd rather not guarantee for a survival.

Mike

[lpcware]

Content originally posted in LPCWare by Algorithm on Thu Aug 20 07:55:06 MST 2015
Yea bavarian , I know that already, but I used to deal with ST 32-bit chips and I could turn on a lot of peripherals, but the chip did not get hot (180 MHz operation) . So I was just wondering if this is normal for LPC chips or not .
Thx a lot

[lpcware]

Content originally posted in LPCWare by bavarian on Thu Aug 20 05:50:00 MST 2015
I simply assume that not the high frequency warms up the chip but the high current   8-)

The higher you clock the chip, the more power it sucks. The reason for this lies in the relatively large number of peripherals which are switched on after startup. A lot of gates are toggling with a high frequency. So to save power - and to cool down the chip - you need to switch off the blocks you don't need. "Switch off" means that you switch off the clock to these peripherals.

Regards,
NXP Support Team.