USB Power hack with DC/DC Converter - am I crazy?

[chconnor]

Hi! So check out this crazy idea:

I have a firewire device (an audio IO box) that I want to power from my laptop, but since PC laptops don't provide power over firewire, I wondered if I could power it via my USB ports.

I am very much an electronics beginner, barely familiar with Ohm's law, etc, but I'm a technical dude, so i get arrogant and think i can figure this stuff out. Don't let the following jargon fool you.

My idea:

Take four USB cables, plugged into the four USB ports on my laptop. On the other end of each of them, cut the head off, clip the data lines off (unused), take the power lines and wire them in series. 5V at .5mA max per connection (my understanding of max USB power) = 5V at 2mA, 10W max output.

This goes into a DC/DC converter to take it to 12 or 15VDC.

The box, according to tech support, comes in at "12V, 1.0 amp & 8W", which doesn't exactly line up unless 1A is just the max draw, but given other things I've read I think it's basically an 8W device, can take 12-24 VDC, and briefly draws extra current on startup resulting in like 10 or 12W.

Here's the DC/DC converter I was looking at:

https://www.mouser.com/search/Refin...CC10-0512-SF-E&N=1323038&Ntk=Mouser_Wildcards

Data sheet at:
https://www.tdk.co.jp/tefe02/ea335_cc_e.pdf

(the CC10-0512-SF-E is the particular one i'm looking at, either 12 or 15V selectable)

It's 84% efficient, so from 10W that's 8.4W, so that may well be enough to power the device... ???

Among 1000 things i don't know about what i'm getting into are the following concerns:

- if the 4 USB ports on my laptop are truly separate "hubs" and therefore each able to supply that much power in tandem
- even if they technically are, if they will actually supply that much power, given the power regulating algorithms that some laptops enforce (and since USB devices are supposed to "ask" for that much current)
- if the DC/DC converter will cause problems in the laptop's power system (either by drawing so much or having weird back-interference...) or otherwise hurt the computer
- if the DC/DC converter will cause problems in the audio device for similar reasons, especially since DC voltage conversion (i've read) can cause "noisy" power supplies...

I have a Dell Inspiron 6000.

Any thoughts, corrections, grave warnings, mocking responses, welcome.

Especially welcome are posts of the "here's a simpler, more elegant, more effective solution" nature.

Thanks for your time!
-c

 

[Nigel Goodwin]

This sounds fraught with danger, and you might well kill your laptop - and you'll certainly drastically shorten it's battery life.

Presumably the interface comes complete with a mains PSU?, are you wanting to use it in the field?.

BTW, what interface do you have, I'm currently looking for one for my daughter, although her laptop doesn't have Firewire.

 

[yngndrw]

I agree with Nigel, if you havn't got a socket to plug it into, your battery will be dead quite quickly.

Nigel: I have an M-Audio FireWire Solo, which is a really good recording interface. M-Audio do also make interfaces which run off USB:
https://www.m-audio.com/index.php?do=products.list&ID=mobileinterfaces

 

[Nigel Goodwin]

Nigel: I have an M-Audio FireWire Solo, which is a really good recording interface. M-Audio do also make interfaces which run off USB:
https://www.m-audio.com/index.php?do=products.list&ID=mobileinterfaces

That's one of the makes I've been looking at, also Tascam and Edirol.

As always, cost is a factor, and both Audio and MIDI would be nice!

 

[chconnor]

Thanks...

Thanks everyone for your consideration...

Yes, it has a mains PSU, and yes, this was for the occasional use in the field for remote recording, or use while on/in forms of transportation that don't have power outlets (I travel a lot). I have a couple laptop batteries, so I was hoping altogether they could go me some distance. Even if I could get an hour or so of life out of it it would be well worth the range of recording/working location options.

Of course the much more practical option is one of many rechargeable external battery packs that exist which could power the unit (although I still worry about "noisy" power...) This half-baked plan is also about the dorky interest of making it work from the laptop.

Other jerry-rigging options I'd be into considering are tapping into my extra laptop battery (Li-ion, Dell, 11.1V)... basically rigging my own external battery pack using that as the source... Not sure how/if drawing current from that battery in a non-sanctioned way would potentially affect it...

But I'm not quite ready to let go of my dream. Out of curiosity, what's the "might kill my laptop" risk? Is that from potentially drawing too much current from it? Or too much all at once? I was thinking of ways to stage up the draw on the USB ports rather than all of it kicking in at once... I was emboldened by various hacks like this: https://youtube.com/watch?v=L79RjOY-zMY
...which of course are different situations, but still. If it's just plain dumb to do, I can accept that , but I am curious as to why...

The unit is a Presonus Firebox. I did a heck of a lot of research on the options at that price area, and I have to say the Firebox came out on top. (There is certainly a less-expensive tier, many which are USB powered). I've been very pleased with it. There are a couple caveats: Most newer Macbooks can't actually power it via firewire due to a change in their powering scheme, and there is a problem with the firmware that means that 96Khz recording isn't "true" 96 until you upgrade the firmware (which is easy). 2 preamps (w/phantom), 4 analog in, 2 digital in, 6 analog out (really 8, if you utilize the headphone jack), 2 digital out (so altogether 6 in 10 out, if you count everything). MIDI in/out. Very flexible routing, zero-latency monitoring, 24bit/96Khz max (on all ins/outs). Well-constructed (though the firewire jacks are fragile-seeming). $300. On my 2GHz laptop I can get like 1.5ms reported latency, though I usually work at like 5-10. For probably a year there has been a rumor, backed up by tech support but still fuzzy, that there will "soon" be firmware to allow them to daisy chain with each other as well. If you don't go with some other USB solution, then sounds like you daughter needs a firewire card: For the firebox (and most other things, I suppose) make sure it has a TI chipset (they also recommend VIA chipsets, but others recommend against it). I successfully use the SIIG NN-PCM312 (TI chips).

Thanks!
-c

 

[Super_voip]

Usb

wiring the 4 USB ports in series would be a very big mistake, all the power for the ports comes from a single source, so wiring in series would simply short out the USB supply. No real damage should result as the USB ports monitor for just such problems and will shut down. Maybe parallel was idea, just not expressed correctly. If there is an option to power over firewire, why not simple make a power injection box, I have similar for USB.

 

[mneary]

Although drawing the maximum from four USB ports in parallel might seem valid, as a practical matter the computer probably is not designed for it.

I agree that an external battery dedicated to the peripheral is a better idea.

 

[chconnor]

Oops, yes, i meant "parallel". Sorry for that.

Ok, I'll let it go, grudgingly.

Just out of curiosity, if the powering from the laptop USB ports wasn't an issue, did I get the rest of the plan right? Just in terms of my math and the selection of the right kind of DC/DC converter, etc?

Thanks for your sage advice,
-c

 

[Nigel Goodwin]

Just out of curiosity, if the powering from the laptop USB ports wasn't an issue, did I get the rest of the plan right? Just in terms of my math and the selection of the right kind of DC/DC converter, etc?

Looks fine - I would suggest using your spare laptop battery to power it, or a small SLA battery, or even a few NiCd's/NiMh's.

 

[skipbogard]

You're not crazy, but you need to do some DERATING

You are on the right track...but missing the 'real-life' engineering details (e.g. realities of life) when it comes to supplier parts.

It is an engineering best practice in Power Design to "DERATE" components. You're familiar with the concept I bet...you just didn't know there was a name for it. Most of the good suggestions above hint that this is what needs to be done.

An example of derating you're probably already familiar with....

If you wanted to buy an ATX Power Supply and you added up all your PC components' power draw...would you try to buy a supply with exactly that power rating? Meaning, if you added up all your components & got 550 watts of power that all the devices would use (if all used at one time)...would you just buy a 550 watt power supply? Probably not. I hope not! Most basic "How to Build Your Own PC" books would tell you to buy a supply that is about 20% larger or more. The reason they might give is "Growth" (adding future disks or PCI cards, fans, etc.) But, there is a more important reason...

In effect, what they are really telling you to do is DERATE a supply by about 20%. If you were to Google "DERATE" you'd find out that using a rule of thumb like 20% (or even 40%) sort of makes up for: (1) reduced efficiency due to temperature extremes (2) reduced power due to aged components (3) fudge factor for marketing "best case" claims of power supply rating...blah blah. Hence, "future growth" is not nearly as important when buying a Power Supply as is "not working so close to the limits" of the power supply's claimed performance metrics.

If you don't want to age your laptop quickly, I would derate the current that all power systems under consideration can supply:

(1) instead of say 500 mA use 400 mA per USB port
(2) Also, derate the DC-DC converter you plan to buy
(3) As far as effect on battery drain...derate the battery too (the rated capacity is in mAh).

The derating concept sums up many of the concerns that people had above about you working so close to the limits (comments like "you're laptop xxxx probably was not designed for yyyy "). They were sort of telling you not to work so close to the edge!

- Skip Bogard

FYI...if anyone cares...the reason I'm here at this site looking at this thread......(and I expound here because it is a real life example related to the thread issues discussed above...)

I'm dealing right now with an ACER EZDock 4 Laptop docking station that has 6 failed USB ports as well as a failed PCMCIA port. Many people on the internet (including me) have complained that all Acer EZDock 4 ( EZ4 ) expansion units fail at about 15 months of use. Acer doesn't seem to have a fix for this either...and for many people Acer has refused to replace the units.

My current theory of why these USB ports fail (and the PCMCIA slot) is that one root cause of early component failure is Acer specs ALL of its laptop external power supplies to be 19 volts. I think they use this high input voltage to make their LCD screens brighter than competitive screens in stores (the 19 volts goes into an Inverter Board that then powers the LCD at a very high voltage....the higher the voltage the brighter the LCD screen...so given COTS components (COTS=Common, Off-the-Shelf parts...like LCD screens) ...an ACER 19 volt powered laptop will look better in a store than say a Compaq powered with a 18 volt AC/DC adapter)

I think that this is a marketing ploy (to use 19 volts in all Acer laptops) because if you do a simple experiment..... take the laptop off the AC-DC converter and let it run on the 12.8 volt (11.1 volt nominal) battery power...the screen gets considerably dimmer. It's a fact of life that the ACER machine I own cannot get it's LCD to be as bright on DC battery power as it can on AC power...no matter how much I fiddle with the Power-Saving settings to be minimal (little savings).

So what? What's the big deal about using 19 volts when plugged into the AC-DC converter? Who wouldn't want a brighter display?

Well, I think that the DC-DC converter boards they use in their laptops probably can survive 19 volts due to a speced max input voltage of say 18 volts (a common spec for mid-price DC-DC converters) . But the docking station station DC-DC converter I think can probably only survice about 13.8 volts (this is more typical than the 18 volt DC-DC converter max input) for el cheapo converter parts. So...the fundamental design problem is contention between:

(1) the marketing people who want to sell a bright display vs.

(2) the laptop people who want to reduce the cost of the DC-DC converter board in the laptop and so used an 18 volt max input spec vs.

(3) the summer coop student at Acer who designed the *@*@&#& friggin' EZDock 4 and used a cheap Chinese DC-DC converter board without any overvoltage protection and which worked just fine on their bench with a laptop battery fluctuating between 12.8 down to 8 volts (then the laptop shuts down below 8 volts)...and in final testing worked fine with a 120 VAC - 19 volt DC AC/DC supply.....

..but not for long!

This story should demonstrate to you the conflicted choices that an engineer has to make when marketing is beating the design & manufacturing groups up to cut costs.

It's worth noting also that Acer laptops tend to burn through INVERTER BOARDS and LCD displays! I wonder why.....maybe it's to make people selling replacement Acer parts on Ebay rich....

 

[Hero999]

Hi and welcome to the forum,

Great post, just a bit too late.

De-rating is good but often allowing for the maximum possible load being connected at 100% duty is impractical and results in overkill.

For example a domestic ring main might have 10 double 13A sockets around the house, so you might be tempted to use wire and a circuit breaker, capable of handling 260A as this is the maximum possible load. However, this would require very thick cable and a huge circuit breaker which would be prohibit idly expensive. In practise, we consider what appliances will be connected to the sockets and rate our components accordingly. For a start, half the sockets probably won't be used and the ones that are will be used for small appliances like TVs, radios and computers that draw less than 1A so we can happily use a 32A breaker and thinner cable.

Duty cycle is another factor that needs to be taken into account. At work we have transformers that convert 230V to 110V to supply sockets around the worshop for powering portable tools. They have a typical rating of 3kVA continious or 4kVA peak. This is becase people might use them to power large loads like heat guns which aren't used for long periods of time.

I would apply the same logic when building a PC. Most of the time the CPU isn't being streached and other power hogs like the DVD writer aren't being used so there's no point in buying a power supply that can give 500W continiously when one rated to 300W continious and 500W peak will do.

 

[Super_voip]

An interest observation, I have just completed a power use survey at work for PCs as the idea was to have the PCs shutdown automatically at night and use a Wake on LAN signal to bring them back up in the morning. The survey was to get figures on just how much power could be saved. The average ON (24/7) power consumption was 40Watts and it was calculated that if the project was to proceed and 1/2 the power could be "saved" the yearly cost reduction would be $1200, given the the monthly power bill is $10,000 the project has been shelved. The power figure is lower in Quad cores than in single or dual cores, the older the PC, the higher the power consumption.

 

[Hero999]

That's not true surely?

My old 386 didn't even have a heat sink.

Modern CPUs have a huge heatsink.

 

[Nigel Goodwin]

That's not true surely?

I would have said not - look at the PSU specs for a start, modern PC's are 300-400W, some older PC's were only 60W PSU's.

My old 386 didn't even have a heat sink.

Modern CPUs have a huge heatsink.

Many 486's didn't have heatsinks either.

 

[chconnor]

Thanks...

thanks much for the added info... -c