Live on Kickstarter: Tiny USB 2.5W V/I controllable power supply

[brrup]

Hello,


Update 05/2021: Project Relaunch Now

The first campaign in 2019 failed, so more than two years later after finishing the project in my spare time, there is a relaunch now!

https://www.kickstarter.com/projects/507351807/sirubox-the-v-i-controllable-power-supply-for-your-pocket

---

here is a link to my first Kickstarter project I started some days ago, maybe it will be interesting for some of you:

I developed a tiny USB 2.0 variable power supply with controllable voltage and current:

• Voltage from 0 to 15V with current up to 600mA (2.5W)
• Controllable via onboard web server (no driver installation for many OS)
• Remote control is simple using the Rest API

Bruno

[Markino]

Wow!!! Great job man!!!
I’m really interested to buy one, can you explain the difference from the early bird and the special offer?

[brrup]

Hi, Thank you!
Cool if you like this tool. There is no difference, the Early Bird is just limited in quantity but reduced in price.

[PlainName]

That's a jolly clever idea and implementation.

Just a couple of questions...

Is there any security on the web server? What's to stop someone else whacking up the volts to blow my project up?

Can the web server be modified to allow setting of volts/current without actually changing the output until so 'apply' button is pressed? I am thinking that clicking spinners on edit boxes could lead to tears given a mis-click or wheel scroll, so being able to hold back changes until they are known to be correct could be useful.

[brrup]

Thank you.
Generally, the web server is running on the device itself, your computer will not share it to the outside would. The siru.box adds as a separate virtual ethernet device when its is connected via USB using the RNDIS protocol. (It is the same concept mobile internet sticks are using).
There is a nice RNDIS example project for STM32 discovery board, if you are interested: https://github.com/fetisov/lrndis .

But if there is someone sitting in front of your computer, or hacked into your computer, there is no additional security on the web access planned. This would also complicate the simple Rest API access for scripts.

To prevent form accidentally changes, there are different ideas in my mind. There will be a setup area to adapt the handling to what you want.
Possible options might be:
[  ] Show confirmation dialog on changes
[  ] Show additional apply button

If there are better ideas, I will try to do my best to implement them.

Another option here will be to setup a startup voltage/current. (On default, the output will be disabled.)
With this startup setting, you can power the device by a micro USB battery charger without being connected to the PC.

[PlainName]

Ah, excellent. I got the wrong end of the stick and assumed it used wifi or something to be like any other IoT kit and available to any passing browser (my phone, for instance). I see now that it isn't like that

[Terry01]

Very cool indeed! 

[Zucca]

Nice!

All the EE should have one of this in the Laptop bag!
In the next version please consider to crank up the current to 2A, almost all the USB ports today can deliver 2A.

Don't know what to do, either buy it now and hack it. Or wait for the next >600mA version...

[grizewald]

Excellent idea and great implementation.

Backed. 

[MadTux]

Looks kinda nice, especially for quick hack jobs where you don't want to go to your lab.
Price is good too.

But it looks like common GND with your USB, which would be a big NO for me.
It's not safe for your computer and not really usable for general purpose work.
Go with full galvanic isolation (by optocoupler or DC free protocol over inductor), even if it costs more, it makes the difference between a toy and a lab instument, IMO.

[tsman]

Any chance of a USB CDC serial port interface? Could be an easier to use alternative to the RESTful interface. Making a composite USB device to handle the RNDIS network interface and the CDC serial interface might be a bit too much hassle though. In that case, maybe a basic telnet interface instead?

My other minor request would be some way of changing the background colour on the web interface. Bright yellow needs

What zucca said about increased current capability.

[retrolefty]

I too think it's a great idea. But like already posted, if the device does not isolated ground from the PC to the DUT then I wouldn't want to use it.

[0xdeadbeef]

The siru.box adds as a separate virtual ethernet device when its is connected via USB using the RNDIS protocol. (It is the same concept mobile internet sticks are using).

I wonder if this could be an issue in a company environment. Usually admins in a company aren't happy it you try to add ethernet devices to your PC. Most certainly using mobile internet sticks is forbidden in most companies as this would circumvent the firewall. So chances are that installing virtual ethernet devices is forbidden or even impossible in many companies.

[pylo]

Galvanic isolation is the best ofc, but has significant disadvantages (cost, physical size, limited current capability). You can get away instead with proper protections, meaning output over-voltage, over-current, and reverse-current protections. Considering the lack of isolation otherwise, I'd say these are mandatory or else many users will ruin their computer's ports.

[PlainName]

What's the issue with DUT ground? I realise that if the DUT ground is not PSU negative then there could be issues, but you wouldn't connect PSU positive to some random track and expect it all to survive, right? It's not like a BNC socket is going harbour 240VAC on it without you knowing.

[mtdoc]

Very clever. Good idea. But overpriced IMO.

[Markino]

I have couple of questions before goes on with my order:
- I was supposed that the max current of a generic USB laptop port is 500mA... am I worng? How can you reach 600mA 15V? Sorry I'm a beginner
- I'd like to use your power supply connecting it to my MacBook 12" through an usb-c/usb adapter or with an usb-c hub. Will it works fine?
- Is there any rick for my MacBook? May I damage it in case of reverse polarity conenction, over voltage or.....

[tsman]

- I was supposed that the max current of a generic USB laptop port is 500mA... am I worng? How can you reach 600mA 15V?

It doesn't. Look at the graph marked "Current output limitation at 2.5W" and you'll see the maximum current drops as the output voltage rises. 2.5W / 15V = ~166mA if this PSU was 100% efficient which it can't be.

I'd like to use your power supply connecting it to my MacBook 12" through an usb-c/usb adapter or with an usb-c hub.

USB-C is backwards compatible if you have a suitable adapter. If other USB-A devices that want 5V at 500mA work okay with your adapter or hub then so should this PSU.

Is there any rick for my MacBook? May I damage it in case of reverse polarity conenction, over voltage or.....

brrup will have to answer this one and state what protection circuitry is part of the design. This is what most of the other replies are concerned about.

[Markino]

Thanks for response!
Yes, as first I'd like to understood if is there a good protection/de-coupling to prevent damages on my laptop

[OwO]

Yeah the price is a little on the high side, but otherwise it's a nice idea; be on the lookout for chinese clones popping up on aliexpress before your campaign even finishes!

[Markino]

Anyway... can't wait! I've just ordered one Early Bird so I'm a backer of your project man! 

[brrup]

You all are right, GND is on USB GND. Having it isolated would be cool, but it also would increase the costs, and maybe the size. I decided to start with the simplest variant, my budget and time is very limited.

Cool, thank you Markino! 

The answer was already given, with USB 2.0 we are limited to 500mA @ 5V = 2.5W.
There is a graph showing possible output current down on the Kickstarter page.

I really like the idea having a 15W version based on Type-C, however its the same argument, for me this can only be the second step, and it will again come with additional costs in components.

There are several mechanisms to prevent the USB Host from damages:
On software level, I got good results with an additional interrupt handler on output shortcut situations, that guarantees that input current always fits the USB 2.0 specification.
Additional, when USB input voltage is falling down a specified mark (e.g. 4.75V) the output current is reduced automatically. (A message will be shown that there is a problem with your input). When falling down again, the output will be disabled.
For the situation things go really wrong, I added a PPTC resettable fuse to protect computer from consequences.

Maye I can add a video, showing the noise on USB line with the oscilloscope while doing shortcuts on the output. Will do this next week.
 
As RNDIS is a Microsoft concept, you have to install the HoRNDIS driver on Apple products: https://joshuawise.com/horndis
I never tested your usb-c/usb adapter combination, but as siru.box acts the same way all other USB devices do, I can't find a reason why it should not work.

[Markino]

Thank you so much Brrup and keep up the great job!

[timgiles]

Just backed - great project

[agehall]

One can never have too many power supplies! I just ordered one! Great work!

[PlainName]

with USB 2.0 we are limited to 500mA @ 5V = 2.5W

Clearly not applicable to V1, but couldn't you use one of the off-the-shelf fast-charge chips to get 1A or more at 5V?

[OwO]

Also in practice most PC USB ports are fine delivering 1-2A, and since you have the current auto-backoff feature, maybe allow drawing up to 2A by default

[Kasper]

Cool idea. Nice job. You mentioned injection molding. Did you look at urethane cast, vacuum forming or off the shelf enclosures?

[electrodacus]

Just made a pledge on the naked version
Take a look at my current Kickstarter https://www.kickstarter.com/projects/electrodacus/solar-bms-and-digital-mppt-for-pv-electricty-and-h  and especially look at how I done the enclosure for my devices as that may be a good solution for your's also.
Your devices looks to be very small so the PCB case should work and cost about as much as the plastic molded one but look better (just my opinion about the look).
Hope your project will be successful but is not looking that great at this point in time.

[Kilrah]

Also in practice most PC USB ports are fine delivering 1-2A, and since you have the current auto-backoff feature, maybe allow drawing up to 2A by default

+1 on that, it's mandatory to default to 500mA max, but there should be "expert" software switches to allow for 1A/2A limits since most ports since years support fast charging.

[brrup]

Many thanks to all my new supporters!

I was thinking a lot about how to increase output current.

In my opinion, USB-BC (USB Battery Charging 1.5A) and USB-PD (up to 100 W) are not that common that it would be okay to increase costs/size for people having it, especially in a first version.

My focus was on USB 3.0 that would allow 900mA but only after the device requests for it. Hope I not overlooked it, but I was not able to find a valid answer how to request 900mA when having a USB 2.0 device, all ends in conclusion that I must have native USB 3.0 support on device then.
Some of the PCs I tested here did active limiting to 500mA when using the USB 3.0 port. It depends on the Host chip, some are more tolerant. It seems to be no problem to shortly get more (I think because mobile USB hard drives need enough initial current to run).

The answer will be USB Type-C, here the concept is nice, they separated USB transmission protocol and power setup mechanism: Up to 15W, there is no need of using USB PD specification, the output current is configured by a simple passive resistor setup for both, the cable and the device. This allows using USB 2.0 together with 15W. When needing more than the 15W again, the USB PD protocol must be implemented.

Yes, I would really like to add this hidden switch to enable more power for people that know what they do, but the final hardware design does not allow this.
I calculated/simulated/tested all for 500mA now, increasing to 900mA or more would be a significant difference and requiring other components.

So sorry, you have to wait for the next version that hopefully will come one day.

Hope your project will be successful but is not looking that great at this point in time.

Your are right, it is not looking that great for the moment. There are too many people on world, that seems not knowing about my project.
Yes, your PCB case style is really cool, never seen this before.
Good luck to you.

Did you look at urethane cast, vacuum forming or off the shelf enclosures?

The truth is, I did not. There are online supplier that are focused on injection molding for smaller prototype quantity. They use aluminum molds that are much better to process automatically. When calculating the total costs for 500 samples, it ends in something between $2 and $5 for small cases. This is still a lot for something that will cost some little cents for larger amounts, but it was okay for me.

[Kasper]

Did you look at urethane cast, vacuum forming or off the shelf enclosures?

The truth is, I did not. There are online supplier that are focused on injection molding for smaller prototype quantity. They use aluminum molds that are much better to process automatically. When calculating the total costs for 500 samples, it ends in something between $2 and $5 for small cases. This is still a lot for something that will cost some little cents for larger amounts, but it was okay for me.

That cost is not too bad then. Based on your funding goal, I figured it was higher.

Finding a small batch method could still help to reduce your goal on the next try. Going the other direction, I was recently told steel molds are better than aluminum if you plan to tune the design because steel molds are easier to weld and cut.  That advice was from my most expensive quoter though so...

Ideally your mech design can be built with any method but at the least it is nice to have 1 small batch method and 1 large batch method.

3DHubs has a lot of info on different materials and processes and will auto quote simple builds which I found really helpful when I started learning about enclosures.

[electrodacus]

Hope your project will be successful but is not looking that great at this point in time.

Your are right, it is not looking that great for the moment. There are too many people on world, that seems not knowing about my project.
Yes, your PCB case style is really cool, never seen this before.
Good luck to you.

Did you look at urethane cast, vacuum forming or off the shelf enclosures?

The truth is, I did not. There are online supplier that are focused on injection molding for smaller prototype quantity. They use aluminum molds that are much better to process automatically. When calculating the total costs for 500 samples, it ends in something between $2 and $5 for small cases. This is still a lot for something that will cost some little cents for larger amounts, but it was okay for me.

For how small your device is and 250 to 500pcs the PCB case may look better (I'm biased about that ) and cost about the same likely around 2 to $3.

[Kilrah]

To me PCB casing looks like an "amateur"/"can't afford more" solution. If he can get an injection moulded plastic case for about same price it's a no-brainer.

Too bad about the higher current capability

In my opinion, USB-BC (USB Battery Charging 1.5A) and USB-PD (up to 100 W) are not that common

All my computers in the past years have suppported USB-BC. It's rarely mentioned but they do. Would imagine the vast majority out there does.

[electrodacus]

To me PCB casing looks like an "amateur"/"can't afford more" solution. If he can get an injection moulded plastic case for about same price it's a no-brainer.

Have you seen my devices made out of PCB ?
I think they look better than what I can do with injection molding and since they are FR4 they are also fire proof. It allows for capacitive touch buttons also at no additional cost.

[brrup]

On of the big players for small batch series injection molding is https://www.protolabs.com/ I think. There you can upload your 3D model, and the next day you get an online offering about the costs.
The big drawback is, if you later want to order e.g. 15.000 parts, the price will be still at > $2 per part.
As I understand aright, the aluminum mold must be replaced by a new one after 15k to 30k samples, but manufacturing is > 50% faster (=lower price). Steel is much hardener, you can produce 100k or more with the same mold.

[grizewald]

It's a shame that this didn't reach the funding goals, but to be honest, I didn't expect it to as the goal was unrealistic.

@brrup: Do you have any plans to still make these, despite the Kickstarter not being funded?

[ANTALIFE]

OP, you should fund this via Crowd Supply. I'm certain you will get more traction there

[brrup]

Now, after more then 2 years, it was possible to finish the project in my spare time!
I have 300 units prefabricated ready for shipping.

The last days, I relaunched the Kickstarter campaign with a funding goal much lower than the first time - so I hope this time it will work!

Here is the new campaign link:

https://www.kickstarter.com/projects/507351807/sirubox-the-v-i-controllable-power-supply-for-your-pocket

[PlainName]

Mine just arrived - thanks!

But... it doesn't work. When I plug it into my PC (W7) I get a popup about the driver being installed, and Windows goes off to Windows Update to find it, but then it exits as nothing applicable being found. Of course, with no driver there is no network and no web page.

Have I missed something, like some specific driver I should manually install? Couldn't find any instructions anywhere, and the website is 'coming soon'.

[PlainName]

Fixed. RNDIS needs to be installed manually, and rather than paraphrase the process here, this is where I found excellent instructions:

https://developer.toradex.com/knowledge-base/how-to-install-microsoft-rndis-driver-for-windows-7

Gosh, it's a long time since I last saw those!

[PlainName]

And having played with it some, it's an excellent piece of kit. Brilliant for shoving in the laptop bag.

A suggestion, though: when the range is being used, the graphs would be better covering only that range rather than the possible fill scale. For instance, when simulating battery power, the range might be set to 2-3V but the graph scale remains 0-15V. It would be much better to show just 2-3V. (Having said that, it's still possible for the voltage to drop below the minimum if current limiting takes place, so perhaps an even better solution would be to scale to the visible min/max values, plus a little for padding.)