50-400V Isolated DC-DC converter for $7

[methods]

Oddly - if you try and find a low power (say 5W) DC-DC converter for over 100V you will be disappointed to find that not only are they not commercially available - but the chips are not even out there to roll your own.

Here is the irony... Every single switcher on earth is a DC to DC right? If you buy 10 and dissect them guess what you will find? You will find 10 DC-DC converters with a full bridge rectifier on the front end + maybe a choke and a fuse.



So drum roll please................... Each and every one of us already has what we need in our house right now

I happen to be in the market for a 5W DC-DC that can take an input range of about 70V to 150V.
I need 12V as my output - but you can just as easily find 5V, 18V, 24V, etc.

I down selected to one particular model - here are some highlights

* 50V to 400V input (I quit testing at 320V DC)
* 12V out @ 500mA
* DC Isolated
* $7 on Ebay shipped from the US
* 400V input caps
* Integrated fuse
* Output filtering
* Input noise filtering
* Low profile (aka flat transformer)
* High quality IR switching fet
* Good isolated PCB design
* 800mA @ 10V loaded hard
* 12V +/- 0.3V @ <600mA
* >5W continuous output

For the prototype work I need to do over the next few months it makes a hell of a lot more sense to just buy one of these off the shelf for $7 and use it than it does to spend countless hours trying to design a new buck circuit from scratch (that will doubtlessly cost MORE than $7). Maybe you have the time and expertise.... but me? pffftttt.... I am a good firmware engineer and a shit analog engineer.

If you need to go to lower voltage? 10V - 80V DC-DC converters are as common as herpes on Ebay. You just have to know how to search.

Anyhow - I am burning this one in. I potted it completely to get worst case heat. On my final unit I will probably let the electrolytics poke up out of the potting. I will probably snip that pesky LED too! (what a waste of energy...)

-methods

 

[Jeremy Harris]

These look similar to the $7 ones from Ebay that I posted about yesterday. I'm using them to drive three cheap Turnigy balance chargers for a low power, plug and play, charging solution, but suggested that they'd probably be pretty good as DC DC converters for higher voltage systems.

Jeremy

 

[yopappamon]

Do you find the capacity is significantly reduced when operating at the lower end of the input range? That's been my experience.

 

[grindz145]

Right on Methods.

I've seen a few people do this now, and it makes a ton of sense. Using economy of scale to our advantage.

 

[methods]

These look similar to the $7 ones from Ebay that I posted about yesterday. I'm using them to drive three cheap Turnigy balance chargers for a low power, plug and play, charging solution, but suggested that they'd probably be pretty good as DC DC converters for higher voltage systems.

Jeremy

Yea... This is my first technical post in a very long time and I have not had time to read the forum in nearly a year. I was worried that by now this may be old news... which is good news

-methods

 

[methods]

Do you find the capacity is significantly reduced when operating at the lower end of the input range? That's been my experience.

Yes - but on a case by case basis. This particular unit was still able to put out 11.55V @ 460mA
With higher current switchers it seems to be worse - like Laptop bricks

-methods

 

[amberwolf]

On yours, does it make any difference to power capabilities to bypass the input rectifiers?

FWIW, it's not old news, but rather very helpful. The more technical you can get with it here, the better.

Others *have* used celphone chargers, laptop adapters, etc., to power stuff off their packs, just typically not at as high a voltage as your tests show. I use a laptop adapter in series with my pack to help my CFL headlight and taillight brightness (and make them start faster):
http://endless-sphere.com/forums/viewtopic.php?f=2&t=14193&start=0

 

[methods]

On yours, does it make any difference to power capabilities to bypass the input rectifiers?

My original plan was to bypass the input diodes but at the last minute I decided that the reliability of a polarity independent input outweighed the savings.

If I decide to run this supply further than the first run of prototypes then I will probably measure the gains from eliminating little things like the 2 diodes drops, the output LED, etc. For a quick napkin calculation:

Input draws around 75mA with a 6W load on the output and 100V on the input
Assuming the two diodes drop 700mV each that means we are burning 1.4V * 75mA = ~100mW
meh - (shrug) I would rather have bi-polar input than save 100mW

Efficiency is roughly 80% (napkin number)

To put this into perspective... The power source is a 25S 80Ah pack ~= 7.4kwh and my average load is about 1W so I am good for about a year. Oh yea - and the primary load is a 1,200A Kelly controller... so... that puts things into perspective.

(I draw a small fraction of a Watt when the vehicle is OFF)

-methods

 

[Doctorbass]

It's great to see there is still some demand :lol: :lol

I mentionned that fact in some thread about the same thing that 2 years ago. ( i'll try to link both if i can find it on the thousands of threads :?


On my side it was for higher power.. like 5A at 12V...

I talked about the fact that many laptop and old LCD 15" monitor switcher can starte at voltage as low as 20Vdc at the 120V ac input.

I also tested that some will start but will be able to hold their output voltage anc power at a bit higher voltage than their start thershold.

I recommand the use of the 12V to 19Vdc at 30-90W output switcher.. they generally work for 35Vdc and higher input.

Also all 5Vdc lithoium stuff charger work really well ( generally those who are specified for 100-240V)


http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=7887&hilit=laptop+power+supply

http://www.endless-sphere.com/forum...6499&p=97909&hilit=laptop+power+supply#p97909

Doc