fastest safe operating clock - overclocking

IanM

I've done a search but can't find anything, so apologies if this question has already been asked. What is the fastest over-clocking speed (ie internal clock) the prop can get to without resorting to cooling etc. I've seen Chips comments on the conservative nature of the 80MHz clock but can it run at higher speeds safely enough for non-critical applications not exposed to temperature extremes?

Many thanks, Ian

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ian Mitchell
www.research.utas.edu.au

Paul Baker

Each Propeller is different on it's max speed. For room temperature a typical Propeller can achieve somewhere in the range of 96 to 110 MHz, but not all Propellers can run at 110. 96 is a good number because most if not all Propellers can run at this frequency and this is 6MHz + 16xPLL.

If you have 2 Propeller's you can use one to test the other's maximum speed by using the counter of one to generate the clock of the other.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

IanM

Excellent, thanks Paul.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ian Mitchell
www.research.utas.edu.au

cgracey

Propeller chips are production tested at room temperature to 104MHz (6.5MHz·* 16). So, every part passed a functional test at that frequency. Before the part ever gets hot due to speed, it will just cease to operate properly. However, if you can actively cool it to something like 0 degrees C, you could probably run it pretty reliably at over 128MHz.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Chip Gracey
Parallax, Inc.

IanM

yeah, i saw the DLP-PROP uses a 6MHz clock so i assumed that was quite safe

thanks Chip

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Ian Mitchell
www.research.utas.edu.au

Bill Henning

Chip,
What misbehaves first?
ie what is a good functional test to know that a chip is passing a certain frequency.

Chip Gracey (Parallax) said...
Propeller chips are production tested at room temperature to 104MHz (6.5MHz·* 16). So, every part passed a functional test at that frequency. Before the part ever gets hot due to speed, it will just cease to operate properly. However, if you can actively cool it to something like 0 degrees C, you could probably run it pretty reliably at over 128MHz.

Paul Baker

The core of each cog breaks down first. The PLL is capable of 160MHz and the hub is half rate clocked.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Bill Henning

Umm... I was hoping for some symptoms - ie shifts going wrong, adds coming up with the wrong result, etc.

The current prop does not have very complex instructions, personally I'd suspect that the cog/hub interaction would be the first to go; but I have no evidence for this, just a gut feeling.

Paul Baker (Parallax) said...
The core of each cog breaks down first. The PLL is capable of 160MHz and the hub is half rate clocked.

cgracey

Bill Henning said...

Umm... I was hoping for some symptoms - ie shifts going wrong, adds coming up with the wrong result, etc.

Bill,

The first thing to fail as speed increases is the hub memory access (~112MHz), particularly the hub SRAM read -- this is the critical path in the chip. The next thing to fail is the Z flag (~120MHz), then the C flag (~128 MHz).

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Chip Gracey
Parallax, Inc.

CJ

so Beau's LED light show would be a good test(lots of hub reads) when testing a speed?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Who says you have to have knowledge to use it?

I've killed a fly with my bare mind.