More laptop buying questions - processor and hard drive

If I plug in a powered hub, it works fine on the same machine.

i have the same problem with my dmx controller (lights)....not sure why....but i just use a hub....why fight it

Because I'm a stubborn old engineer who has an inner need to know why something doesn't work as advertised?

1. I would like to be able to work without mains power - this is after all the supposed advantage of a laptop.

2. I like as few cables, connectors and plugboards as possible.

3. Because, having seen that 2 of us have the same effect (I won't say problem), others will know to watch out for it?

I still have the inner need to know whether it's a particular chipset, a feature of usb2 (why does the hub say it is usb2 when powered, usb 1.1 when not?), quirky design of the laptop or faulty on-the-edge design of the usb device.

But I'm just a grumpy old man today after being caught out by yet another unexpected (and unrelated to this) 'feature' of XPPro last night.

Because I'm a stubborn old engineer who has an inner need to know why something doesn't work as advertised?

Okay... quite reasonable as far as I'm concerned - but I'm going to throw some fuel on the fire, and sit back and bask in the warm glow...

3. Because, having seen that 2 of us have the same effect (I won't say problem), others will know to watch out for it?

I still have the inner need to know whether it's a particular chipset, a feature of usb2 (why does the hub say it is usb2 when powered, usb 1.1 when not?), quirky design of the laptop or faulty on-the-edge design of the usb device.

It's more than 2 people - we've had reports of this before. And however you want to describe it, the word 'problem' seems to fit quite well. The power situation with USB is as follows:

The supply voltage is nominally 5 volts. Any device requiring 100mA or less can be powered either by the host or the hub, as long as the total power availability hasn't been exhausted by other devices on the bus. A single device can draw up to 500mA from the bus - if the current is available. It's also quite well known that not all hosts can actually supply the 500mA necessary - especially laptops. It is not at all unusual for a laptop to limit the supply current down any one USB port to 100mA anyway. So, any device requiring more than 100mA to run should be externally powered or self-powered.

But that's not the end of it. Devices actually tell the host how much current they require - and this can vary according to whether they are switched on in self-power mode or not. This is how some devices can switch between power modes when they are unplugged from the bus, or the host shuts down. There is also an absolute limit for current drawn by a device in suspend mode, and that's 500uA. It is allowed to draw this from the bus if it qualifies as bus-powered. And how long before a device can go into this mode, you ask? Well, it needs to do it if it's seen no bus activity for a paltry 3mS - this is a part of the basic spec. Either a host or the device itself can initiate wake-up - and it's the need for all these additional signals that make USB1.1's paltry 12Mbits/s seem even slower.

Anyway, the bottom line is that USB bus power is an accident that's found somewhere to happen, as far as I'm concerned - it sucks. 500mA @ 5v is 2.5W, and that's not exactly a lot of power - the Traveler is reckoned to be quite a miserly device in power terms, and at its peak it requires somewhere between 9-12W. This falls within the FireWire bus powered spec, but there's not a laptop around which will even think about supplying this - that's why they all have 4-pin Firewire connectors, not 6-pin ones. FireWire, from this POV, is rather better thought out.

But as far as some USB devices failing from a bus power is concerned, all it takes is for the wrong power information to be relayed back down the bus, or for it to be misinterpreted for the whole thing to fail. Any hub that reports a mode change when not powered may not be telling a lie - it's quite possible for a hub also to run itself from the host, and in a lower speed mode, it will consume rather less power. But if it discovers that it can run from 100mA, and then proceeds to do so, despite the fact that the host is limited to supplying only this much, anything other than a self-powered low speed device isn't going to be capable of using it - even though it thinks that it should be able to. It's these ridiculous situations that are actually allowable in the spec that make a complete nonsense of the whole thing, and can easily make it fail to perform.

I'm back with thinking that the nearest system I can implement to SteveG's the better - given that his works.... and he knows of what he speaks.

My thinking tomorrow might be different!  (Er, no, not about Steve!)

The good bit about it is the Traveler - but I don't see why this wouldn't work with any FireWire-equipped laptop, quite frankly. I do understand the attraction of me saying that it definitely works with the HP zd7000, though - even though the fan noise is higher. But, as I said before, I regard that as an opportunity, not a problem...

My reason for wanting to reduce the fan system noise is simply that I don't like it!

Thanks to Steve for his extremely detailed answer about USB. The only trouble is that I still can't find any software that will tell me what is actually going on wrt power with a particular USB device.

I suppose I could make a sort-of USB 'breakout box' and set up some sort of measuring/recording system. And this is where I hit the feeling that "it's only a Soundblaster and not worth the effort".
Another reason for looking at the phenomenon in more detail is that I sometimes use an external HD, which comes with a split lead to take its power from 2 usb ports simultaneously. This works on some machines and not others, but doesn't correspond to the SB work/not work machines, and confirms Steve's statement that Firewire is better thought out than USB.

With regard to laptop noise, I do have some experience of using laptops in public events near audiences.

The P4 ones that I've used (Acer and Medion) were definitely too noisy to think of using in this situation. The fans ran almost continuously and were very noisy. The laptops also had bigger and heavier psu's.
The Centrino ones I've used (Acer and IBM) had fans that did not run constantly and were less noisy.  BUT the sudden appearance of a fan noise in a quiet portion of a play did attract the attention of nearby people. The level was actually fine if the fan didn't cut in and out - it's the change in background sound that makes it noticeable.
NB What follows is not entirely serious.... If I were doing this full time, I think I'd consider taking along some masking noise generator eg a small fan that ran constantly. A mixer with a fan in the psu would certainly mask the quieter laptops. Unfortunately my mixers are silent.

At present, I honestly don't know what's affecting the results at all. What I'll do at some stage is the same test sequence (it's not of my making) with this drive on a desktop machine, and see what the difference is.

I now have a slightly better idea of what makes a difference to the performance - and it's not so much the bus itself as the way the particular interface is integrated with the machine. Because the PCI bus on the desktop machine has a lot of stuff hanging off it, the performance with FireWire and USB2 isn't as good as it is on the laptop (the FireWire and USB2 ports on the desktop machine aren't integrated into the MOBO at all). On the laptop, there's a significant difference between the inbuilt FireWire port and the two ports on my Cardbus adaptor - there's a 30% hit on the sequential write rate, but not much difference in the read or random RW results, and it's the built-in port that goes faster (the previous results quoted were from the Cardbus port).

But it remains the case that the best results I can get (and this is consistent in all the tests) are from the USB2 ports on the laptop. How this actually works out with real world signals remains to be seen, though. However you look at it, you can stream everything that comes from the Traveler to an external HD without any trouble at all.

On the remote chance that anyone is even remotely interested, the laptop model that is heading to the top of my shortlist is the HP/Compaq nx8220 with Pentium M 770 processor (just over 2GHz).  A bit of research on various forums reveals some concerns over fan noise, though counter-arguments are made suggesting that this will be an issue for the faster processors of most brands and models.  In any case, since I am primarily going to use the laptop to make noise rather than record (and especially will not be recording where there will be an audience to disturb), I do not expect fan noise to be a big issue for me.

I was also looking at some Toshiba models, but there appears to be an opinion on these other laptop forums that Toshiba have been dropping the ball recently on quality.  But the nx8220 seems to have the feature mix I want compared to the Toshiba models available here.  Steve's comment about HP's ability to actually supply sensible installation CDs is also a quality-of-service factor that's important, and I don't know how Toshiba's rate on that front.

Great minds are thinking alike here....  though I haven't yet come across it with that 2gHz processor.

See http://www.xbitlabs.com/articles/mobile/display/sony-hp.html (you've doubtless come across the review at http://www.notebookreview.com/ which a search there will turn up).

Great minds are thinking alike here....  though I haven't yet come across it with that 2gHz processor.

I've spotted it at Harris Technology (though I gather a lot of people don't like them; certainly they are only offering it at RRP).

See http://www.xbitlabs.com/articles/mobile/display/sony-hp.html (you've doubtless come across the review at http://www.notebookreview.com/ which a search there will turn up).

Yes, notebookreview.com is one of the sites I did research.  I hadn't seen the xbitlabs.com review - thanks for that.  It appears to be for an older model in the nx82220 series; I certainly expect from the specs I've seen that the later nx8220 models can do significantly better than 1280x800.  If you follow all the threads here you may recall I'm a Dual Head junky (and therefore a screen resolution junky) and this is my first foray into laptopland.

http://www.cworld.com.au/ is looking good, at least for a slightly slower processor model.

I'm considering a Dell 600m laptop for possible broadcast engineering duties. I know Acer and IBM are realistically better choices as I have read here on these forums. The Dell however, has the legacy serial connector that I may need to interface with many broadcast related equipment.  What other factors should I look for in a laptop for this purpose? or would this laptop be adequate? Or should I just consider an older laptop rather than a new one. Any suggestions would be appriciated.

Ill throw in a couple of my personal experiences:

Stay away from any Compaq computer.  I have had horrible experiences with their desktop models and laptop models.

My experience with HP's is that they are only decent computers, work alright but aren't anything special.

Dell computers have worked much better for me then the two former.

I have worked with one IBM laptop, and it was indeed very nice.

Isn't it Dells that have big trouble with power-supply noise on audio?  They're OK on batteries but not on mains?  Generally I have a prejudice against them on the grounds that the Dell name comes up a lot in problem-related posts, though that could simply reflect the number they sell.

I wouldn't worry about getting a machine with a serial port - I think USB to serial converters are very inexpensive?

I've just ordered the HP8220 or whatever it was we were discussing above.  The resolution and viewing angle on the screen is amazing (apart from other plus factors).  I'll order the Traveler tomorrow.  There's a concert coming up in a cathedral on 7th July with string quartet, choir, solo singers, and later a different choir with the Salvation Army band which might make a suitable testbed.

Isn't it Dells that have big trouble with power-supply noise on audio?  They're OK on batteries but not on mains?  Generally I have a prejudice against them on the grounds that the Dell name comes up a lot in problem-related posts, though that could simply reflect the number they sell.

Wouldn't that depend on the rest of your gear?  I mean, an external firewire box would be different from a PCMCIA card, which would differ from the horrible internal audio.

I've only used my Dells for working on audio that already existed as data, so I can't help with recording particulars.  But I CAN say the built in audio is crappy (as usual on laptops).

I'm pretty sure that it was a Dell laptop -- don't know the specific model -- that had horrific "motorboating" noise when we plugged in a line-level signal from a mixer.  Connection was with a shielded, but unbalanced, audio cable that worked fine just a short time before with an IBM laptop.  Once the AC adaptor was unplugged, the audio quieted down to an acceptable level for the purpose of a live Webcast... which is to say, so-so quality.  I hope it was just a fluke, and not representative of their quality.

Personally, I'm very happy with my Dell desktop unit, but with an m-Audio card installed.  A lot of co-workers are being issued Dell Latitude laptops, and I don't think there are an unusual number of problems... but they're for typical business use, not audio.

A FireWire or USB interface effectively isolates the audio converter from the laptop power supply - but using a direct input to the machine doesn't, of course. I wouldn't expect to have any particular noise problems with a Dell laptop when using an external converter, but with a direct connection to almost any laptop, under certain circumstances, all bets are off.