/diy/ - Do It Yourself

[Leonar No.677650]

I want to make my own "Poors man UPS"
I need this on a WiFi Access point

So far, I figure it out how to connect the Access Point, the Super Capacitor and the Wall wart. But I think that if I leave it as "it is" (without any resistance).

Could this setup charge the super capacitor? Do I need to add some resistors somewhere?

I want it to be as simple as possible

No Diodes, regulators, transistors or whatsoever. I just want it to keep it simple.

And yes... I know... it won't be efficient, but it will still be a poor's man UPS.

Think of it as a Joule thief... it is far from being efficient, but we still use that circuit like EVERYWHERE.

Thanks for reading.

[Anonymous No.677653]

>>677650
alright friend

A "joule thief" is a retarded self resonating dc-dc converter

Adding a super cap in parallel with your voltage source does absolutely nothing for efficiency.

The fact that you want us to think of the two as the same is worrying as it proves your incompetence

the cap shown has a maximum voltage of 2.7v and would most certainly be damaged at 5v

You would have to calculate the estimated current draw from an empty capacitor and choose a current source/ dumb resistor that fits the spare power your wallwart has to offer after supplying the access point


Really the only thing you need to know is OHMS LAW and how to apply it in your invention.


Im not gonna do it for you, its much more satisfying doing it yourself as its your project.

[Leonar No.677654]

>>677653
I want to make my own "Poors man UPS"

Maybe I should write it in CAPS

POORS MAN UPS

>Adding a super cap in parallel with your voltage source does absolutely nothing for efficiency.

>efficiency

>the cap shown has a maximum voltage of 2.7v

Of course I will be using either 2 caps or a caps that could handle 6volts at least

>Really the only thing you need to know is OHMS LAW and how to apply it in your invention.

Thats all I wanted to know

now I know I must only add a resistor that could handle the given wattage and Volts * Current Usage.

>proves your incompetence

chill up buddy

here, have some random map so you can calm down

=)

[Anonymous No.677662]

>>677654
I didnt mean to come off as rude/offensive


I quickly drew this schematic and it would probably work assuming the access point would work on 5v - the schottky voltage drop

AND there is no diode to prevent backpowering the wallwart which could potentially be a problem

[Leonar No.677665]

>>677662
Hey man, don't worry =)

I have been googling diagrams on how to achieve this and yes, I have seen people using many other components, not just resistors.

I just wanted to know if it was possible to achive something as close as possible, using only resitors. (Its my personal challenge)

Kinda like making your own FM Mic transmitter using only 1 transistor.

[Anonymous No.677671]

Even with super capacitor, you probably only have 2-3 seconds backup power.

Just take a look of how much current your WIFI Access Point will consume.

[Leonar No.677676]

>>677671
Well, my wall wart gives me 5 Volts @ 1 Amp

I have tried connecting 5 Volts @ 667 mAmps, and it does not work.

So My guess is that the AP consumes something in between 667mAmps to 1 Amp. So, my worst case scenario is that it uses 1 Amp.

2 or 3 seconds wont be enought for my purposes =) But if that is what it is, then I can't do anything about it.

https://www.youtube.com/watch?v=EoWMF3VkI6U

[Anonymous No.677680]

>>677650
>3000 farad capacitor

Here's what's going to happen if you do this:

1. That supercap will appear as a dead short to your power source until it charges, during which time it's effective resistance will slowly increase. Whether your power supply survives this is anyone's guess.

2. Your supercap is rated at 2.7V. Your supply is probably at least 5V, if not 9V or 12V unregulated. Supercaps really get upset when your over-volt them, you'd have to put several in series.

3. Assuming you survive all the above, it'll take HOURS for the thing to charge up, during which time no wifi.

4. Do the math on the total charge vs. current draw. Operating time will probably be measured in minutes, if that.

5. Expensive!

If you really want to do this, you need a charging circuit that regulates the current to the supercap bank so the wifi can operate while it's charging, and a changeover circuit that cuts it in (quickly!) when AC power fails.

Really you can do what you want to do (fashion a DC backup supply) using a NiMH or even SLA battery, charging circuit, and changeover circuit, and potentially have HOURS of backup time for your wifi AP, and save $$$ in the process.

[Leonar No.677683]

>>677680
I connected a large capacitor into my 5 Voltage regulator ( I was messing around with a random AVR Circuit in my protoboard).

My circuit was doing nothing but blink a LED.

I don't renember the capacitance that I was using, but I am pretty sure it was DEAD before I power up my circuit.

When I powered it up, it instantly started to BLINK.

But when I disconnected it, it would keep blinking for more than 2 minutes.

That's my inspiration on looking for something as simple as a paralell capacitor in order to power an access point on power offs.

I understand EVERYTHING >>677680, that you are saying, but I will give it a try anyways. I am 99% sure that you are right, but I want to see it on real life for myself =)

If for some 1% reasons it works, I will post the results here. I'll try to be sure to empty the capacitor before doing this.

[Leonar No.677685]

>>677683
I'll try doing that
(assume I am using 2 caps in series)

[Leonar No.677686]

>>677685
R1 must match the amperage that my Wifi access points needs (which is 1 amp)

so, R1 will be 5 Ohms

R1 CAN'T be higher than 5 Ohms, or it won't deliver 1 Amp

[Leonar No.677689]

>>677686
Now I can see the problem....

Either I find how to power my access point using 500mAmps or find a 2 Amps Wall Wart.

Otherwise it won't power up until the capacitor is charged.

Let's assume that my Access point works with 500mAmps.

Would this circuit >>677685 be enough to power up my Access Point and charge my Capacitor?

Lets assume that I am using 2 capacitors (5.4 volts) and R1 = 5 Ohms.

[Leonar No.677694]

>>677683
You where right

I was wrong

That 3kF capacitor looks like a short circuit

It will take LONG time to charge

[Anonymous No.677711]

>>677694
There's another problem you'll have that just occurred to me: If your power supply is 5V *regulated*, then when your bank of supercaps drops *below* about 4.75V, your AP will shut down. You need more than 2 supercaps in series, and charged from a higher voltage supply, and run through a regulator (preferably a switching regulator for best efficiency and therefore run-time) otherwise your supercaps will get only slightly discharged before your AP shuts down.

[Anonymous No.677713]

>>677694
>>677711
Note that I'm not saying 'hurr durr this is dumb don't do it', I'm saying 'it's possible but there are design hurdles you need to overcome to make it work properly and preferably optimally', and they're not earth-shakingly bad problems, they just take some thought.

[Leonar No.677714]

>>677711
Then I guess I will also need

*grabs sunglasses*

a Joule Thief

*Puts on sunglasses*
*Random crows yells YEAAAHHH!!!!!!!*

[Leonar No.677715]

>>677714
I will just drop the whole idea

I tried to make it work using only resistances... and it won't be useful.

It will take HOURS before the AP can even power up.

If we "assume" that the capacitor is alredy charged up, it would work on the first shut down. But if the shutdown lasts more than 1 hour, it will take 20 hours for the Access Point to go back online.

so no

[Anonymous No.677716]

>>677694
>>677711
>>677713
..and I'm being Posty McPostsAlot here..

I'm thinking what you'd be better off doing is designing a complete replacement for the wall-wart your AP is running off of, centered around using 3 or 4 of those supercaps in series. You'll need about a 2-amp capacity DC supply, current regulation for the supercap bank charging, and a buck-boost switching regulator for the 5V 1A output. A simple Schottky diode between the supercaps and the switching regulator should be fine for automatic switchover between AC power and backup power. Look to Linear Tech for nice, easy-to-use switching regulator ICs. A simple current-limiting resistor should be sufficient for current-regulating the charging of the supercaps, since they're rather well-behaved about being charged, and a simple linear supply with a voltage regulator for the AC-to-DC input.

[Leonar No.677719]

>>677716
I wanted to use Ultracapacitors because there is no "danger discharge zone" on them. Unlike rechargeable batteries, you can discharge them till 0.00000 volts and charge them again and NOTHING will happen. And the charge/discharge cycles are way more than batteries charge/discharge cycles.

I was thinking on using this as a long time solution for my devices, since I won't be able to physicalle be "there" in order to change the batteries.

[Anonymous No.677720]

>>677650

>"Poors man UPS"
>super capacitor

ha ha

what

[Leonar No.677721]

>>677720
lol
in tijuana you can get ultra capacitors for cheap

I live in San Diego btw

[Anonymous No.677725]

>>677721

The pictured caps hold 1458j each. A little over half a Wh. A crap-tier, low-capacity AAA battery holds more energy. A decent one holds double that.

A small UPS battery (usually a ~5Ah 12VSLA) holds 60Wh, or 216,000j (or 148 2.7V 400F capacitors).

This is why there is no commercially-available UPS that uses supercaps. It's not practical unless you can get them essentially for free.

[Leonar No.677797]

>>677725
The only "advantage" I see with using caps, is that you don't have to replace batteries.

Think on a scenario that you may need to install an Access Point on a rural area, so you have NO ONE to contact when you need to replace batteries.

That's the only advantage I see so far while using Supercaps.

Expensive $$$$ and
Low capacity are the 2 main disadvantages that I could live with... as soon as I don't have to replace anything, I will be ok with that.

Think on it as a Install and Forget it.

Zero maintenance or whatsoever

[Anonymous No.677798]

>>677797
In such a case, just use a real UPS. It'll switch off before it fucks the battery.

[Leonar No.677799]

>>677798
a real UPS with Supercapacitors?

been there done that with normal UPS

they last 2 years TOPS

I need it like, forever!

I don't want to return to the old installation place.

[Leonar No.677800]

>>677798
further reading:

http://serverfault.com/questions/29556/what-is-the-lifespan-of-an-typical-ups-battery

on this particular installation area, light goes off once per week, and it lasts like 30 minutes.

The maximum time we have had a black out, was 1 hours tops.

[Anonymous No.677818]

>>677800
So spec the UPS such that a typical outage uses <30% of the available capacity.

Supercaps aren't magic; they wear out too. Yes, they behave better with deep cycling, but they still wear out.

[Leonar No.677826]

>>677818
Supercapacitors can last 500,000 to 1,000,000 charge/discharge cycles

That's 9589.04 years in my case

Maybe I am missing something on that math... How else can they wear out?

As far as I know, supercapacitors don't have an expiration date. They just wear out if they are not correctly managed or if we keep using them.

What could possibly go wrong?

[Anonymous No.678725]

Perhaps I'm missing something, but wouldn't it makes more sense to use an AGM car battery than a supercapacitor?

>Battery tender with greater output than your application needs
>Large 12v battery because cheap, plentiful, high capacity
>step down voltage converter, maybe regulator too

Under that small of a load, in the comfort of a controlled indoor environment a quality car battery could last a decade.

[Anonymous No.678737]

>>678725
This is what I personally use now.
Well, not that aggressive toward reliability.
Power 'blinks' here too often.
Annoying to have to wait for gateway to restart and reconnect.
I have a gateway that runs off 12vdc
I have a garden-tractor battery on a charger/maintainer.
I have the power cable for the gateway connected to the battery through a bridge rectifier.
(didn't need but two diodes for the voltage drop but already had 25A bridge)
During last ice storm power was out for three days but I still had internet access from battery power.
I recharged my laptop from an auto adapter for my car.

[TOR USER No.678744]

>>677650

OK, all the diagrams here are getting ridiculous.
Your best bet is to dump the capacitors and get a rechargeable battery of some sort and placing a diode between the battery and the wall wort so that when the power goes out, the battery isn't trying to power the wall wort, wasting battery and potentially damaging it. If it is a 5v 1amp access point (that seems about right) capacitor of any size is not the way to go. Battery is the way to go. Get a large 6 volt battery for like a lawn mower or a scooter or something. Additionally you can find old ups throw away when companies move out and then its just a matter of buying batteries. The batteries are the cheapest part of the ups and no one ever replaces them. I have had at-least 4 of these I purchased for less than 20 bucks.

[Anonymous No.679379]

>>678725
This. Although I'd use deep cycle instead of car battery or even li-pol with solar cell. Lead acid batteries hates being discharged, if you let them discharged for a while (few days/weeks) they are loosing capacity incredibly fast.

I have older sealed lead acid from dead UPS and I'm currently building power regulator for my 9V router. Basically it is just cheap step down regulator from ebay, digital volt-meter and several connectors in a small box. Volt-meter so I can use it as a low power variable power supply.