Propeller "endurance" loading

Timothy D. Swieter

I have a new design I am working on with the propeller. I got my PCBs today and I don't have all the external components populated just yet. However I want to do some current testing and loading on the voltage regulators/PCB. Is there a program that exists that doesn't drive I/O but runs all eight cogs as much as they can to simulate a full processing load? Does this make sense?

I was thinking some simple program that is in SPIN or ASM that basically exercises the counter and such. I believe there was a program a while ago that Chip and Paul used for loading the Propeller while temperature extreme testing, but I would like a program that wouldn't toggle I/O Pins since many of them are connected to device that are populated (or will be).

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, uOLED-IOC, eProto for SunSPOT, BitScope
www.tdswieter.com

cgracey

Without swinging the I/Os around rapidly, you're not going to get the Propeller to draw much in excess of 80mA, or so. That program we made for heat testing toggled all pins at 160MHz and drew something like 600mA. It was a real burner. You could find that thread and change the program to only toggle a few pins.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Chip Gracey
Parallax, Inc.

Timothy D. Swieter

OK - that makes sense. I think I might be able to swing around some pins so I will do some searching for that thread.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, uOLED-IOC, eProto for SunSPOT, BitScope
www.tdswieter.com

Timothy D. Swieter

Here is the link to the thread Chip was referencing: http://forums.parallax.com/showthread.php?p=616057

I double checked a couple things and since my board isn't fully populated yet I could run the demo as is. I ran it for a couple minutes and I measured the temp of the Propeller to be around 50C. I didn't have a specific probe on the 3.3v or 5v DC circuit, but I could see my power supply current jump up. That is what I wanted to see and let it run for a while to see the design, regulators, etc handle it.

Thank you.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, uOLED-IOC, eProto for SunSPOT, BitScope
www.tdswieter.com

soshimo

This should probably be a new topic, but wow! 50C! I realize that's during endurance testing but even so, at that temperature the propeller isn't going to last long (what IC would?). What do you do about cooling and are there any solutions out there that illustrate cooling methods? The simplest that comes to mind is a dumb dc fan with no temperature control. From there I can envision adding a DS1820 and ESC to control the dc fan speed as the unit changes temperature (though I haven't worked out how to actually measure the temperature of the chip and not the ambient air - but that's an implementation detail). What kinds of things are you doing or thinking about for cooling (if you don't mind sharing)? TIA

Mike Green

soshimo,
The Propeller is rated for operation over the full military temperature range (-55C to +125C). 50C may be hot, but it's well within the operating range. The chip would be fine much hotter, but don't touch it.

soshimo

Mike
Noted - thanks for the info and my fingers thank you for the last part!

Carl Hayes

(1) I assume you measured case temperature, not die temperature. 50 degrees C is only 122 degrees F, warm but not painfully hot. Your fingers are not in any danger.

(2) Is that 125-degree-C specification an ambient temp, or is it a die temp? The case-to-die thermal resistance is likely to be in the range 50-70 degrees C per watt, though apparently not specified; at 600 ma and 3.3 volts it's dissipating 1.98 watts and the die (that is, the chip inside the case) will be about 100 degrees hotter than ambient. So if the 125 is die temp, that's pushing it, but if the 125 is ambient, we're coasting. Which is it?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

Phil Pilgrim (PhiPi)

Carl,

The datasheet specs "ambient temperature under bias" under "Absolute Maximum Ratings" and a T_a range for operating temperature — the "a" which I take to stand for "ambient". By also specifying the maximum power dissipation, this indirectly limits the die temperature with no heatsink. By adding a heatsink, one could probably take some liberties with the maximum power dissipation, since the die temperature would then be lower for a given ambient.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Just a few PropSTICK Kit bare PCBs left!

Timothy D. Swieter

My measurement was of course the case temperature and I used my IR Heat Temp Gun. I had to get close to the IC in order to ensure I was measuring and not the surrounding area. I suspect that the IC may have been hotter as I could not leave me finger on it for too long.

I don't plan on running the Propeller this hot, I was only testing the I/O and current draw of a fully loaded Propeller to make sure the voltage regulation and design was OK. I ran the test for about as long as it could before the Propeller reset, which I am not sure why it reset yet. For now I am satisfied though so I am moving on to the next task I have.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, uOLED-IOC, eProto for SunSPOT, BitScope
www.tdswieter.com