Author Topic: EEVBlog #285 - Isolated USB Lab Supply  (Read 62280 times)

0 Members and 1 Guest are viewing this topic.

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13695
  • Country: gb
    • Mike's Electric Stuff
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #100 on: May 31, 2012, 12:14:52 am »
Here's someone doing a project along these lines:
http://www.kickstarter.com/projects/210251816/usb-based-variable-power-supply-for-small-projects-0?ref=live

Well I'll be damned.
Testing leds with no current limit??
And what's the deal with D2 on the output after the feedback sensing divider?
No isolation, and less than 5V limit (which will drop significantly with the cable) doesn't make it very capable IMO.

Dave.

Total fail - 5v max is not very useful, no current set or measure, no output on/off switch, knob sticks out too far, and how the hell do they need $14k to get this piece of crap made..?

But hey, it's Open Source Hardware TM so it must be good ;)
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline caroper

  • Regular Contributor
  • *
  • Posts: 193
  • Country: za
    • Take your PIC
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #101 on: May 31, 2012, 12:17:29 am »
That looks like they have more marketing than engineering talent.

Offline Psi

  • Super Contributor
  • ***
  • Posts: 9889
  • Country: nz
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #102 on: May 31, 2012, 12:19:01 am »
All this usb 100mA/500mA talk is pointless.
Add a dip switch to limit the output to 100mA, those who want to follow the spec word for word and limit themselves to 100mA can do so.

Actually, would be nice if there was a dipswitch for USB3 mode, which can supply 1A.
Or usb3 charging sockets which can do 1.5A.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline adrianblack

  • Contributor
  • Posts: 29
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #103 on: May 31, 2012, 12:19:45 am »
A curious situation happens with my iPhone. The proper iPhone power cube is marked 5 V 1 A and doubtless the iPhone takes this much current from it when charging.

If I plug the iPhone into a regular USB port on a computer it gets recognized as a USB device and also charges, but it doesn't shut down the USB port. So I presume it limits itself to 500 mA in this case.

If I plug the iPhone into a USB charging port on a particular Duracell battery charger, it overloads the port and shuts it down. At this point I do not know if the phone is trying to pull 1 A and causing an overload, or if it is trying to pull 500 mA and the USB port can't supply it.

I presume if I want to find out what is going on I need a USB breakout cable and to insert an ammeter in the circuit. Maybe I'll get round to doing that at some time.

Phones and devices that use more than 500mA look for particular resistance values on the data lines so they know to go to high-power mode. There is no standard to this but most Android phones will accept the resistance offered by Apple chargers and go into full charge mode. If they don't recognize the value, they just drop down to a slower 500mA charge.

 

Offline hlavac

  • Frequent Contributor
  • **
  • Posts: 536
  • Country: cz
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #104 on: May 31, 2012, 12:20:27 am »
Anyway, tell us more about the current prototype!
I know people think LED displays are a waste of limited power, but somehow I feel a power supply needs LED displays ;) LCD does not feel right for some reason...

What can be done about limiting the losses on the LED displays? Low voltage rail amounting to LED+switch drop and a miniscule resistor?
Or some funny stuff with inductors and PWM power? I have seen some pretty small SMD 10uH inductors with 1Ohm resistance in 1206 size, but it would take something like 5MHz PWM frequency to use them as a lossless LED current limiter... and the EMI would be a mess I guess.
Good enough is the enemy of the best.
 

Offline Psi

  • Super Contributor
  • ***
  • Posts: 9889
  • Country: nz
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #105 on: May 31, 2012, 12:24:20 am »
There is no standard to this but most Android phones will accept the resistance offered by Apple chargers and go into full charge mode. If they don't recognize the value, they just drop down to a slower 500mA charge.

There is an agreed standard http://www.usb.org/developers/devclass_docs

But i dunno if anyone is actually using it.
It's more about charger supported usb sockets i think.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline adrianblack

  • Contributor
  • Posts: 29
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #106 on: May 31, 2012, 12:24:47 am »
Here's someone doing a project along these lines:
http://www.kickstarter.com/projects/210251816/usb-based-variable-power-supply-for-small-projects-0?ref=live

Well I'll be damned.
Testing leds with no current limit??
And what's the deal with D2 on the output after the feedback sensing divider?
No isolation, and less than 5V limit (which will drop significantly with the cable) doesn't make it very capable IMO.

Dave.

Yuk -- your very early prototype looks 100x better than that thing. Surprising it has gotten as much funding as it already has.
 

Offline Psi

  • Super Contributor
  • ***
  • Posts: 9889
  • Country: nz
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #107 on: May 31, 2012, 12:26:26 am »
What can be done about limiting the losses on the LED displays? Low voltage rail amounting to LED+switch drop and a miniscule resistor?

Isn't that much of a problem. Can dim the leds with data from an LDR/photodiode.
Could also switch off or dim the display after x seconds of no control adjustment.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #108 on: May 31, 2012, 12:27:50 am »
I know people think LED displays are a waste of limited power, but somehow I feel a power supply needs LED displays ;) LCD does not feel right for some reason...

That's what I felt too, hence the LED displays.

Quote
What can be done about limiting the losses on the LED displays? Low voltage rail amounting to LED+switch drop and a miniscule resistor?
Or some funny stuff with inductors and PWM power? I have seen some pretty small SMD 10uH inductors with 1Ohm resistance in 1206 size, but it would take something like 5MHz PWM frequency to use them as a lossless LED current limiter... and the EMI would be a mess I guess.

I've got a novel solution that saves power and cost, that I'm sure not everyone will agree with. Many will positively hate the idea :P (at least they'll forget about the USB negotiation  ;D)

Dave.
 

Offline adrianblack

  • Contributor
  • Posts: 29
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #109 on: May 31, 2012, 12:31:02 am »
I've got a novel solution that saves power and cost, that I'm sure not everyone will agree with. Many will positively hate the idea :P (at least they'll forget about the USB negotiation  ;D)

LOL, think of all the billions of stupid USB "accessories" like laptop fans, coffee mug warmers, Christmas lights, etc. that people use every day... 500mA with no negotiation and they work on pretty much every USB port on the planet.
 

Offline adrianblack

  • Contributor
  • Posts: 29
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #110 on: May 31, 2012, 12:34:58 am »
There is an agreed standard http://www.usb.org/developers/devclass_docs

But i dunno if anyone is actually using it.
It's more about charger supported usb sockets i think.

Seems to mostly focus on also switching between USB OTG and normal mode ... It also doesn't allow more than 500mA for portable devices anyway.

I think an extension of the USB 3.0 spec does allow for ports putting out 900ma. It'll be great -- we'll be able to have coffee mug warmers than may actually work!  :D

 

Offline rolycat

  • Super Contributor
  • ***
  • Posts: 1101
  • Country: gb
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #111 on: May 31, 2012, 01:23:59 am »
Absolutely - must have 4mm sockets and binding posts - a possible compromise if space is short is 4mm sockets plus some sort of spring terminal - like the sort used for speakers on some amps.

I have considered the dual 4mm socket and spring terminal idea, as it does save protrusion space. But ultimately 4mm binding posts are just too irresistible.

Dave.

You can get some which are much lower profile than normal, only around 10-12mm projection, such as these: http://www.digikey.com/product-detail/en/111-2223-001/J587-ND/241133

The downside is the lack of colour coding, but if you can ignore a big plus and minus sign on the device, you can ignore the colour of the jack, too..

Here are some inexpensive low-profile binding posts with colour coding:
http://cpc.farnell.com/pro-signal/psg02051/4mm-binding-post-gold-red/dp/AV15599


 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #112 on: May 31, 2012, 01:28:43 am »
Please think about mechanical stability too. Micro usb is flimsy. And my experience with surface mounted usb connectors is that they are easily peeled of the board unless there is an external mechanical block like a case..... If you get the micro , find one that has through hole chassis pins. The usb pins can be smd , but the metal sleeve should have pins.


For cost saving on the led displays take a look at liteon multiplexed display with 4 digits. Easier to rout since the muxing is already internal and cheaper. Can often be found for around 1$ ..
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #113 on: May 31, 2012, 01:32:19 am »
Absolutely - must have 4mm sockets and binding posts - a possible compromise if space is short is 4mm sockets plus some sort of spring terminal - like the sort used for speakers on some amps.

I have considered the dual 4mm socket and spring terminal idea, as it does save protrusion space. But ultimately 4mm binding posts are just too irresistible.

Dave.

You can get some which are much lower profile than normal, only around 10-12mm projection, such as these: http://www.digikey.com/product-detail/en/111-2223-001/J587-ND/241133

The downside is the lack of colour coding, but if you can ignore a big plus and minus sign on the device, you can ignore the colour of the jack, too..

Here are some inexpensive low-profile binding posts with colour coding:
http://cpc.farnell.com/pro-signal/psg02051/4mm-binding-post-gold-red/dp/AV15599


Ah, nice.. Normally I don't bother looking at CPC (Or Farnell, for that matter).
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #114 on: May 31, 2012, 02:22:27 am »
As for the isolation on usb. If you use the ftdi232 you can have that sit onthe vbus. You only need to put two optos to cross into the isolated section where the cpu is. Cheap optos will do. You just run at 9600 baud. Cy127 for example. In smd.
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline codeboy2k

  • Super Contributor
  • ***
  • Posts: 1836
  • Country: ca
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #115 on: May 31, 2012, 05:16:34 am »
Well I'll be damned.
Testing leds with no current limit??
And what's the deal with D2 on the output after the feedback sensing divider?
No isolation, and less than 5V limit (which will drop significantly with the cable) doesn't make it very capable IMO.

Dave.
Yeah that's what I saw too.  No current set/limit/readback, no isolation, and only 5V out, max... if that.
I've seen USB ports actually put out 4.8V, so they might have a 0-4.5V variable supply with a readout.  Whoopee!

The writeup talks about charging batteries, so I assume D2 is to protect it in case someone leaves a fully charged
battery connected and pulls the USB power connector.  They can't put the voltage sense after the diode because then it would feed back
into the (unpowered) regulator. But I don't think it would harm it. With the diode there as they've done it, they would have
to account for the (unknown) diode voltage drop in their micro.  But now the accuracy has all gone to shit.

Nice to see them using a Renesas R8C/2K. I like any design bold enough to break away from the AVR's, and PIC's.
However, for all that they are doing with the micro (just an ADC to provide a voltmeter), a PIC might have been cheaper.

I don't see the value of that supply.

« Last Edit: May 31, 2012, 05:32:31 am by codeboy2k »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #116 on: May 31, 2012, 05:49:54 am »
Yeah that's what I saw too.  No current set/limit/readback, no isolation, and only 5V out, max... if that.
I've seen USB ports actually put out 4.8V, so they might have a 0-4.5V variable supply with a readout.  Whoopee!

I doubt a single unit will ever be able to give out 5V and any significant current.

Quote
The writeup talks about charging batteries, so I assume D2 is to protect it in case someone leaves a fully charged
battery connected and pulls the USB power connector.  They can't put the voltage sense after the diode because then it would feed back
into the (unpowered) regulator. But I don't think it would harm it. With the diode there as they've done it, they would have
to account for the (unknown) diode voltage drop in their micro.  But now the accuracy has all gone to shit.

Nice to see them using a Renesas R8C/2K. I like any design bold enough to break away from the AVR's, and PIC's.
However, for all that they are doing with the micro (just an ADC to provide a voltmeter), a PIC might have been cheaper.

In the comments they mention maybe being forced into using an NEC processor to add data logging or something.
And they talk abut another design that goes to 20V?

Dave.
 

Offline madworm

  • Frequent Contributor
  • **
  • Posts: 371
  • Country: de
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #117 on: May 31, 2012, 05:52:08 am »
I like the idea of such a thing. If it is reasonably priced, small enough and in deed comes as a diy kit, then I'm in.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #118 on: May 31, 2012, 06:08:27 am »
I like the idea of such a thing. If it is reasonably priced, small enough and in deed comes as a diy kit, then I'm in.

Well, it's almost all SMD, so a kit can't really be done. But I could let you solder the binding posts?

Dave.
 

Offline peter.mitchell

  • Super Contributor
  • ***
  • Posts: 1567
  • Country: au
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #119 on: May 31, 2012, 06:14:26 am »
I thought I should really mention this:

Voltage drop across a standard usb mini/micro cable of 1m length is horrendous!

I did a bit of design on making external batteries for consumer electronics equipment that charge from USB ports(generally iPad/tablet devices) and this was my biggest issue. My device could pump out 5v@2.5A no problems for charging an iPad but on the other end of the Apple USB cable it had dropped to a measly 3.7v! My battery pack also had similar issues when charging itself; it would charge very slowly so I intentionally used a more "rare" usb socket, the USB B socket and included a more bulky cable.

Even at 500ma this can raise issues!

Also, about the whether it should be 100ma vs 500ma vs ???ma, spec for it to only draw 500ma at most but leave a footprint beside the resistors to put in a small trimpot, that way you stay inside the rules, but people who want to push it can, and can use it with devices like the iPad charger to their full advantage.

How about backlit LCD instead of LED? Means you can use it in bright sunlight easily AND in the dark, also saves power.
 

Online tom66Topic starter

  • Super Contributor
  • ***
  • Posts: 6680
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #120 on: May 31, 2012, 06:28:30 am »
Dave, how isolated will the power supply be?

Safe to connect output side to mains powered device?

Or just safe to tie two in series?

If it's the latter you probably wouldn't need to worry too much about USB isolation as the output could be floating and measured using a diff amp... you need a diff amp for current already anyway so it's not too much additional cost.

Just a thought.
 

Offline madworm

  • Frequent Contributor
  • **
  • Posts: 371
  • Country: de
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #121 on: May 31, 2012, 06:41:54 am »
I like the idea of such a thing. If it is reasonably priced, small enough and in deed comes as a diy kit, then I'm in.

Well, it's almost all SMD, so a kit can't really be done. But I could let you solder the binding posts?

Dave.

Duh... that is trivial stuff ;-)
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #122 on: May 31, 2012, 06:50:08 am »
Also, about the whether it should be 100ma vs 500ma vs ???ma, spec for it to only draw 500ma at most but leave a footprint beside the resistors to put in a small trimpot, that way you stay inside the rules, but people who want to push it can, and can use it with devices like the iPad charger to their full advantage.

People forgot get that isolated DC DC converter cost so many $/watts, and it's not often linear. To design the thing worst case around some of the figures being thrown around, it could double the BOM cost or more.

Dave.
 

Offline G7PSK

  • Super Contributor
  • ***
  • Posts: 3859
  • Country: gb
  • It is hot until proved not.
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #123 on: May 31, 2012, 07:16:40 am »
USB.3. will provide 900ma. and since 2007 most usb are at the battery charging spec. so will provide 500ma without any jiggery pokery. So why not make the unit to USB3 standard also if you want to run from battery there is a plethora of battry usb phone and mp3 ipod etc power booster/chargers on the market which will enable mobile use without a computer for the power supply and totally gets around usb power limits. 
« Last Edit: May 31, 2012, 07:26:10 am by G7PSK »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVBlog #285 - Isolated USB Lab Supply
« Reply #124 on: May 31, 2012, 07:20:00 am »
Dave, how isolated will the power supply be?

Standard 1KV

Dave.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf