Me too -- I stay on the conservative side, 850 mHz tops and the ideal settings. I tried starving but crashed with video calls. I'm very tempted to go higher, but I don't really need it.

Hmm, I guess I will take the bait and jump into the discussion, given that I figured out the initial voltage calculations for this thing.

The stock kernel sets the core voltage at 1.35V @ 600MHz
Now, from what I remember from my good old engineering days, for a 65nm device, leakage current and active current (and hence the power dissipation) offers the biggest challenge

P_active depends on Vcore^2 f
P_leakage depends on Vcore^2

There would be other dissipation from things like memory etc. but we don't control that.

So for stock kernel @600MHz

P_active = Ax1093.5
P_leakage=Bx1.8225

Where A and B are constants

For comparison, for the overclocked XLV kernel at 810MHz

P_active=Ax1166
P_leakage=Bx1.44

So while active current is slightly higher, leakage current is significantly lower. Thus active power usage is 6.7% higher than stock but leakage current would be 21% lower

Using a rule of thumb that at that size of device, leakage is circa 30% of current drawn, which would be there of there abouts for a 65-90nm chip, you can see that the the slight increase in active power would be compensated by reduction in leakage power, ergo no extra heat till that frequency...and I think that is what Titan said in the first place.

By the way Bill, can you stop using LOL after every sentence? Kinda gives away your age.

Not for nothing, but if anyone here has been impolite, it's been you. You called the OP irresponsible without reading the thread and/or doing any investigation to see that he has been quite responsible in warning people of the risks involved. You called other members here "dicks" and worse for disagreeing with you, or making light of the situation. You flooded the thread with a bunch of non-referenced replies, and replied rudely to people. Then when people reported you for being rude, you called on the moderators and started a separate thread about moderation. To me, and many others, all of those actions were far more impolite than anything others have said here.

I've found Titan to be quite responsive to questions and suggestions about kernels, and overclocking in general. He's integrated several other mods from other people into his kernel, and has improved the way several people use and monitor their N900. I don't suspect that will change any time soon.

Yes, I also think it would be great to look at implementing a system to better cool the CPU. I also delight in seeing the creation of responsive scripts (or even a kernel-based module?) that can monitor core temperatures and auto-reduce the clocking when temperatures hit settable limits. All great ideas. But when you come in spewing those ideas with combativeness, using words like "irresponsible", and calling people names, you're not going to get people to pay attention to anything you say, yet alone engage you in polite conversation.

I want you to do something: Take a second (or a day, or two), think of a project you've worked on, one that had a slight component risk involved maybe. Then go back and re-read your posts as if someone else has posted them about your project. I bet you get infuriated quickly. Communication is not the strong suite of then engineering class (I say this being one myself). Look at this as a learning experience for yourself, as you above state you hope others would.

woody14619 your the kind of person that never gives up and leaving me no choice but to reply to you, i wanted this to drop and carry on in a good way but people like you just will not allow that.

"I want you to do something: Take a second (or a day, or two), think of a project you've worked on, one that had a slight component risk involved maybe."

This is actually boring me now to say but i have no choice so i will say as clear as possible... the project in this case are about15 N900's that i have opened up and inspected including my own, now i think that gives me enough experiance to know what i am talking about don't you think?(have you actually opened one up even).
Theory and actual practice do not always come together (sigh this is boring) so while you can work something out on paper does not mean that in reality it ends that way.
Every single N900 i have seen has the same problem of the usb port and i have managed to repair every one within minutes, that gave me the chance to look at the area around the cpu underside of the m/b.
I actually AGREE and support Titan in OC and the very reason i wanted to try and put something in to help him by making the device safer to run up heat.
As per usual people get the wrong end of the stick (you also) and don't understand where it is going.
Maybe after this post from me it can be laid to rest and maybe if you lot just get on with OC,ing and i will post some pics of something that can be done to make this situation safer in the long term (that was indeed always my aim).
@arbitrabbit how old you think i am?. lol.

PS my wish on this forum is everyone work as a team and help each other.

Woody i am sorry i forgot to add something... in my work i am usually the one to look at problems and not had the chance really to start many projects but i really am aware of risks and i do try to avoid them as i have had to deal with bad design for many years.
Please just trust me that all i want to do is help ok and no disrespect at all to you.

Ha ha ha.... I won't bring your age into it. I think your approach that assumed that there were no engineers on this forum and we didn't know what we are doing kinda pissed me off. But hopefully now you know that there is some science behind this madness.

I am sure that some people on this forum would almost certainly fry their phones, but Titan has given sufficient warning about that. We don't condone the use of frequencies higher than 850MHz for ideal/xlv for that reason, but if someone wants to do it, they do so at their own risk. Hell, look at some of the tags on this thread and that would give you the idea of what we have been talking about.

are you really that deluded?
i know you have not read the whole thread but at least read back thro your posts, respectful are they?
respectful is a word that cant really be applied to your earlier posts, condescending, argumentative, arrogant, cocky are words that i may use to describe your earlier posts....

IF you had read more of the thread you would know that all the arguments you are trying to put forward have pretty much been discussed in detail already anyhow!!

and i REALLY dont get your point!!!
so you are saying "dont overclock as it WILL damage your device" but let me do a mod that involves drilling holes and sticking bits of metal into your phone then overclocking will be ok!!

dude, wake up to the fact you are possibly taking hypocrisy to new levels..........

@geneven: is there an easier way (or code) for changing governors rather than open cpufrequi? im currently searching pages back but its taking me years...

Yes!

You'll have to look at the Kernel info because I forget the exact syntax, but you know there's a command to save things like my-settings to a file?

So what I did was use cpufrequi to set the governor to what I wanted, used the command to save the settings to a file, then used the procedure also mentioned in the Kernel tutorial to make those settings the default.

Oh, but I found my settings. just add

GOVERNOR=conservative as a separate line in your settings. I have it just below POWERSAVE_BIAS=25

So, something like that. If you forget how to do this, it's really worthwhile to go thru Titan's advanced Kernel tutorial, and it doesn't take long.

Uhm... I just got here and read this after all of it happened. If you look back, you'll note this is my first and only comment on this topic. This post will also be my last on the topic.

I note that you completely disregarded my advice. And have yet to admit it was rude (or apologize) for calling people irresponsible and for calling people here names, which is far more rude than disagreeing or putting in your "expert advice" in a belittling tone.

Sigh.. Since credentials and "experiance"(sic) seems to matter so much to you:

I work as a senior design engineer for a major firm that creates custom embedded systems, some of which actually use ARM chipsets (mainly v5). I hold dual bachelor degrees in CS and CE from RIT, and have worked in the embedded systems field for over 15 years, including contract and full time positions at HP, Nortel, and Xerox. I work on micro systems like the N900 daily, both in hardware and software, and know a thing or two about "ohms law" and how theory often differs from reality. Is that enough to pass your test?

I have, in fact, disassembled my N900 to do preventative maintenance on it for the micro-USB slot. Motherboard discoloration around warm surface mount components is normal for devices like this. It's quite common with such system, especially in sealed compact devices. I see it daily in my line of work and can tell you it's just not that significant.

None of this was the point of my post. So now that maybe you're paying attention, I'll state it again a bit more plainly:

What people are reacting to here is you behaving rudely. It may even be unintentional on your part, in fact I strongly think you're being rude without realizing it. The intent of my posting was to get you to re-examine what you have said with a different perspective. I asked you do to that mental exercise so you could better understand why so many people here are upset with you, and maybe even get you to think a bit more before you post in the future.

Most people here have no qualms with the idea of creating a better way to heat sink the CPU, or ideas on how to improve overclocking. Most have no issues with someone presenting new ideas, or discussing how to help prevent potential damage to the system from heat. They do take exception to be being called names and/or discredited for making improvements that most of the active community here rejoices over, which is exactly what you have done.

Hopefully my being a little more blunt this time has gotten my point across. Maybe you'll stop trying to "argue" with me and instead think about whats been said here. Not one person here has disagreed with the core ideas you've mentioned (preventing CPU overheating). Most of the negativity against you has been a reflection or reaction of your perceived attacks, rudeness, and belittling of others. (And yes, saying things like "I'm an engineer, and know better than you" is also rude and belittling, regardless of how true it may or may not be.)

As a fellow engineer, I leave you with this logic:
Assume most others in the room agree that you're the one that's being rude. Is it logical to think that everyone has formed a conspiracy against you? Or is it more logical to consider that something you've done (intentional or not) may have in fact been rude, or triggered that perception? Logic dictates (to me anyway) that reflection is the proper path here, thus I advise it to you.