Hello,
I finally received my two webplatforms boards today from seed :) great.
Well,
I only connected one of them to my rooter, uploaded original Microchip webpages with Microchip MPFS generator (only webpages, not new firmware).
I launched a ping from my laptop on the board to make sure the server is working!
when I go to my desktop computer and launch a web browser I can access the demo pages and can toggle Led2.
fine. everything seems to work fine.
BUT when I touch the ENC28J60 chip, I find it hot!! the board is powered from a little PSU with only 4.5Volts output (arround 4.7Volt).
I measured the Vcc (after the regulator) and found it arround 3.28Volts.
The ENC28J60 chip becomes hot, not enought to warm wy fingers, but it becomes really hot.
Is that a problem on the board? or is it "normal" ?
Regards
This is normal with the ENC28J60. See the Hackaday "How To: webserver on a business card (Part 2)" blog for more details (http://http://hackaday.com/2008/09/25/web-server-on-a-business-card-part-2/)
Beware the ENC28J60 runs really hot ‘n hungry; it uses over 200ma even just sitting idle. Don’t expect to run it off of AA batteries for very long…
I can confirm this - the ENC28J60s on my webplatforms gets real warm (but not to the point of burning my fingertips) as well.
~250mA @ 3.3V --> 0.825W!
Other users have reported the same:
Markus Gritsch (http://http://dangerousprototypes.com/forum/index.php?topic=324.msg3384#msg3384):
I had it running for 4 hours at 50 MIPS overclocked, then I switched back to the nominal 40 MIPS, since the dsPIC gets a bit warm when running at 50 MIPS (but not as hot as the ENC28J60).
When searching the web you will find other boards and blogs where users ask/are worried about ENC28J60s getting warm/hot in completely different designs as well ... this seems normal, as long as it doesn't get to the point where you will burn your fingers.
One of my web platforms has been running non-stop on the network since May and the ENC28J60 temperature is currently around 38.1C (ambient air temp is 25.5C). It's certainly not hot to touch - just perceptibly warm. (Temps tested with tropical aquarium contact thermometer over the course of 5 minutes.)
It depends on the Ethernet transmit activity how warm/hot the ENC28J60 will get ... but other factors like SPI clock and LEDs driven by the ENC28J60 contribute to the power dissipation as well. If your webplatform just sits there and hardly sends any Ethernet data it will not get very warm.
There could be problems with the specific board (e.g. wrong VR - 5V instead of 3.3V) but that's rather unlikely. Get the full marking on the chip and see if the ENC28J60 silicon errata (http://http://ww1.microchip.com/downloads/en/DeviceDoc/80349c.pdf) (quite extensive for some revisions) lists anything relating to out of spec power dissipation ...
I will set up two webplatforms, running different applications (for high/low Ethernet transmit activity) and measure the tempreature with an IR thermometer/scanner ... results to follow.
Thank you for your answers.
I got the chip very hot when I make a permanent ping on the board (==permanent data exchange).
I'll also measure the temperature and let you know.
Here are the results from some simple tests (uIP stack running, idle/permanent ping - as fast as Windows XP will ping with the -t option). The webplatform board was placed flat on a table. I used an IR thermometer with and without a K-Type surface contact sensor. Room temperature was 27.5-28°C during all measurements:
1. webplatform idling - no ethernet cable connected (after 5 min) - measured with IR sensor only:
- ENC28J60: 32°C between pins 1/28; 41°C between pins 12+13/16+17
- dsPIC33FJ128GP204: 38°C center
- FT232RL: 36,5°C center
2a. webplatform connected and getting pinged (after 5 min - total run time 15 min) - measured with K-Type surface contact sensor
- ENC28J60: 33°C between pins 1/28; 38,5°C between pins 12+13/16+17
- dsPIC33FJ128GP204: 39°C center
- FT232RL: 34,5°C center
2b. webplatform connected and getting pinged (after 10 min - total run time 20 min) - measured with IR thermometer
- ENC28J60: 33,5°C between pins 1/28; 40°C between pins 12+13/16+17 (however, there was a source of 46.5°C between pins 14/15)
- dsPIC33FJ128GP204: 38°C center
- FT232RL: 36°C center
I expected higher temperatures on the ENC28J60 surface towards pins 14/15 because that's where I suspect the TP drivers to be placed on the die but I did not expect the maximum to be so close to the edge of the housing somewhere between pins 14 and 15 as the die is usually much smaller than the housing and placed in the center of the housing.
Hello,
in my case the ENC28J60 chip go up to about 46°C and remain at this temp when there is continuous eth trafic. The dsPIC stays at about 38°C.
Regards