[Alexander] sent in his instructable on upgrading the inefficient linear regulator in a DX3 RC radio transmitter to double the battery life. I’m actually more interested in the switching regulator (high res image) than the radio – this could be really useful. $15 to replace a $1 component hurts a bit – but it has the caps and coil needed to support the switching regulator.
I am a CEng student, I know what is inside 7805 and I know enough to make another. Switching regulators could be power efficient but linear regulators does not burn all the power which is more than you need as the producer told in website. It simply puts Vin to a bottleneck so it drops down, there is realy simple circuit to do this. If I am wrong, please only professionals corrects me.
So the battery life sould not be increased that much, not even close to it.
No, really. Linear regulators are horribly inefficent when the voltages get up in the 2x and 3x range.
Switching regulators (what the DimensionEngineering part is) are so vastly superior–It’s amazing they aren’t in more electronics. Oh yea, that’s right–the price.
I wrote the article, and yes–I can attest that the battery life is at least double. All with a switch of one part.
Cost of the radio: 260
Cost of the regulator: 16 (SH included)
So worth it. Think about this in your robots. You can ’stack’ the regulators to get more watts out of them. It’s all on the DimensionEngineering website. http://www.dimensionengineering.com/DE-SW033.htm
Posted at 7:33 pm on Apr 5th, 2007 by Alexander
@erik:
It has a little heatsink on it, and if you are really maxing it out you may want to put a little fan by it.
Posted at 8:26 pm on Apr 5th, 2007 by Alexander
I can confirm that linear regulators are generally less efficient, as they’ll drop whatever voltage they need to regulate and release it as heat. the 7805 has a dropout voltage of 2V–that means if you want 7V, the maximum efficiency you’ll get is about 70%. In the real world, it’s a whole lot less than that since the voltage drop is generally higher with more current, and that you’ll want to work with a higher input voltage than the minimum as leeway, especially when you’ve got a plug-in power supply which might change voltage with load. Most switching supplies are usually about 80%+ efficiency, with some going into the 95% range. They (well, at least the boost regulators) also have the advantage of having a higher output voltage than their input voltage.
Posted at 9:08 pm on Apr 5th, 2007 by viktor
There are numerous sources for switching regulator replacements for 3 terminal linear regulator. The TI PT5100 series are available from DigiKey for
Posted at 9:21 pm on Apr 5th, 2007 by Paul Bilke
Yeah it sucks how so many battery operated devices skimp on the psu..
A canon I have that uses AAs dies from low voltage, yet the batteries still have a lot of energy left… if they could somehow throw a boost-buck in there it would last a good bit longer and drain the batteries to nothing.
Posted at 10:57 pm on Apr 5th, 2007 by japroach
The manufacturer may have used a linear regulator to reduce EMI, kind of important with a transmitter. Hack probably voids FCC compliance.
Posted at 1:35 am on Apr 6th, 2007 by rob
The vast majority of battery operated devices either do use some form of switching power supply or run directly from the battery (the best way). Perhaps this radio (I didn’t spend much time at the website, it’s horribly slow and bloated) uses the direct battery voltage to power the output transmitte and just used a low current regulated voltage to power some small, sensitive digital circuit…
If the entire device really is powered from the linear regulator then it may be because it requires large ammounts of current sporadically. Batteries (well, most types) are excelent at providing large ammounts of current and linear regs are very capable too, even if they are inefficient. That linear regulator probably beats that switching regulator in areas of max curent handeling, speed in responce to transient current surges (which I’d expect to be common in a transmitting radio), and possibly quiescent (ie. no load) current.
These days it’s very common to see small switching regluators in cheap battery powered devices that cost an order of magnitude less than this radio. I don’t think it’s wise to jump the the conclusion that a linear regulator was chosen on the basis of cost alone, particularly considering the retail price of the radio.
Posted at 3:54 am on Apr 6th, 2007 by crimes
that’s an excellent mod
ti has a chip that’s 93% efficient switching regulator with 3 amps capacity, and is available in 3.3, 5, and 12 volts, as well as 1.5 I believe.
http://focus.ti.com/docs/prod/folders/print/pt6304.html
samples are free, and the chips themselves are cheap, and available on digikey
I’ve used a number of them in parallel to create a car pc power supply, that was over 250 watts capable and small in dimension.
Posted at 4:57 am on Apr 6th, 2007 by Alex
That is true, switched mode regulators make power lines full of spikes, because of their nature. Switching on and off transistor puts a lot of noise into enviroment. That’s probabaly why linear regulator was used in the device.
Posted at 5:11 am on Apr 6th, 2007 by Caladan
>Fucking STOP with the instructables! I can read Instructables already, it’s just plain lazy and unoriginal to link to them everyday.
You might read it everyday, but some of us don’t as there’s a lot on there we can’t be bothered with. It’s kind of handy to have the signal filtered from the noise by having relevant entries posted here.
Now how easily can this be applied to laptops or do they already, given the overall price, have decent regulators in them?
Laptops already have similar circuits. I don’t think a laptop manufacturer would mind putting in a part that costs more and would increase their battery life even 25%. If you saw two comparable laptops, and one boasted an extra hour of battery life for $15 more, which would you buy? If battery life is critical, then they probably have already addressed the issue in some way. The DX3 has a longer battery life than other transmitters already, and if it goes dead, it’s time to go back inside anyway.
Posted at 8:50 am on Apr 6th, 2007 by Mike
There a cheaper source of switched mode PSUs to be found in practically any car charger for cell phones or whatever. These things usually come free with a new phone, or can be had for next to nothing on Ebay.
Almost every single one I’ve looked inside uses a MC34063 1.5A switchmode chip soldered to a tiny PCB with all the coils and caps, etc. To change the output voltage, you just need to swap out the feedback resistor for the appropriate value.
As a bonus, they also have the current limiting function and can withstand anything up to 40V at the inputs.
If the original part did not have a heatsink, the switcher will not need one. The switching regulator will require less cooling. @#2 (ToreUyar), a linear regulator wastes power, and that is manifested in the heat it produces. Current flows constantly from the source in a linear regulator, providing a constant load. In a switching regulator, current flows occasionally from the source, so the amount of time that it is actually wasting power is quite low, and the amount of power that it wastes when it does is lower than the linear regulator, thus the power savings. ToreUyar, it’s time to put down the beer and pick up the books.
Posted at 9:53 am on Apr 6th, 2007 by pocketbrain
Good call on the PT5100 series. Although for this app, the footprint looks a bit to large at over an inch square. TI Datasheet: http://www.ti.com/lit/gpn/pt5103
Posted at 11:34 am on Apr 6th, 2007 by thetable
Is it me, or do i have to copy the ID, go to instructables and type it in their search to get to this page?
I’ve emailed instructables that their policy SUCKS MAJORLY! and i should be able to directly link to a project!
Posted at 11:57 am on Apr 6th, 2007 by xFred
xFred, something is flaky with instructables at the moment with the “?ALLSTEPS” page. Try http://www.instructables.com/id/EOXE7JRF03GBB2D/
Bryan, http://focus.ti.com/docs/prod/folders/print/pt5101.html has a link for getting samples for the pt5101 5v jobber. Getting samples from TI has gotten a little trickier of late thanks to some people taking advantage of the situation. Don’t request too many samples for fear of getting black listed, and IIRC, they wont’t send them to @gmail, @yahoo, @hotmail, or other generic email accounts.
Posted at 12:45 pm on Apr 6th, 2007 by thetable
Heh it was a nice surprise to see our product here.
I would like to respond to a few people’s comments:
“Switching regulators could be power efficient but linear regulators does not burn all the power which is more than you need as the producer told in website. So the battery life sould not be increased that much, not even close to it.”
This is mostly incorrect. The efficiency of a linear regulator is a function of the input voltage minus the output voltage. In this scenario, we have a 3.3V output voltage and the input voltage is above 10V. So you are looking at around 33% efficiency, which is abysmal. A switching regulator is almost mandatory here. Now, if the regulator was fed from a lower input voltage, like 4.5V, then the linear regulator would have been a good, cost effective choice. However from 8x AA…
Linear regulators are great for certain things and will always be a staple part of anyone’s electronics kit. But for scenarios where you have a high input voltage and you are drawing more than 100mA, a switching regulator is worth strong consideration.
TI’s switching regulators are nice, but they require you to add your own external capacitors which is kind of a hidden cost.
“9. The manufacturer may have used a linear regulator to reduce EMI, kind of important with a transmitter. Hack probably voids FCC compliance.”
This would be often be the case with many switching reg designs. However with the DE-SW0XX series, the EMI it puts out is almost unmeasurably low. This is because the core of it was designed to be placed in R/C environments which are insanely sensitive to RF noise. Also, whether the RF module on the spektrum receives 3.3V from a linear reg or a switching reg doesn’t really make a difference, it is still 3.3V so there is no change in radiated power so there is no reason for it to change its FCC compliance status. Can check it out on a spectrum analyzer if you like.
“If the entire device really is powered from the linear regulator then it may be because it requires large ammounts of current sporadically.”
Not really. I’ve measured it on the DX6 (different radio, similar technology). There is a constant stream of data being transmitted, and the separate RF module that the regulator powers draws a pretty constant 250mA or so. I would guess that the choice of a linear regulator was a function of ease of manufacturing, time to market, cost of components, cost of engineering a switching regulator etc.
“Now how easily can this be applied to laptops or do they already, given the overall price, have decent regulators in them?”
Laptops already have very good, efficient voltage regulation solutions in there already. I wouldn’t recommend even trying to improve them.
“It has a little heatsink on it, and if you are really maxing it out you may want to put a little fan by it.”
The DE-SW0XX series is designed to continuously supply a full amp and in this application it is supplying about 1/4 of that. You could probably rip the heatsink off the regulator entirely and it would still be fine. However what you have suggested is a good precaution for any engineer to take. Always keep an eye on temperature in enclosed environments, and add airflow if necessary!
Posted at 1:15 pm on Apr 6th, 2007 by Hisham from DE
#8: I don’t read instructables, so I rather like that he links here.
#13: poorly designed switchers do spew EMI throughout the power system, but linear regulators are actually legislated illegal in the EU since they’re so inefficient. Most switchers these days are very “quiet” and in fact significantly help power factor due to their use of a small input capacitor and their ability to draw current over more of the input wave (talking AC->DC here, but DC->DC is still DC->AC->DC :-)) Running at a higher frequency means smaller magnetics as well, and easier filtering.
Posted at 3:05 pm on Apr 6th, 2007 by akmixdown
The recompower version is smaller than the TI and cheaper but only 0.5a
Posted at 4:06 pm on Apr 6th, 2007 by paul bilke
@Bryan
Actually, there are other boost-converter (or buck-converter) circuits out there that are cheaper than this one regulator. Ladyada has a kit (mintyboost) that you can buy that fits into a altoids tin.
http://www.ladyada.net/make/mintyboost/
It’s essencially what you are talking about, but just in a easy-to-purchase kit form.
Also, I didin’t link the TI part. That was someone else. :P
Posted at 4:24 pm on Apr 6th, 2007 by Alexander
I just want to say thanks to hisham for the through response
Posted at 10:46 pm on Apr 6th, 2007 by Wolf
The problem with using custom switching circuits is that you’ll need a whole lot more design to get it right. There are probably a thousand more things to consider when designing a switching supply, especially EMI/RFI and noise at the output. Both of these problems are often not correctable by using a better switcher IC or even by changing the circuit–often times PCB layout becomes the most important thing to consider. In a case like this, although a switching supply might suit the scenario, it’s not entirely necessary, it’s too complex if you’re designing it, and time-to-market / cost, as hisham said, becomes a consideration with the dimension engineering module.
and laptops do already have switching supplies. in fact, switching supplies are probably everywhere where they’re necessary (don’t underestimate engineers =D). MP3 players, the print server i took apart, laptops, computer PSUs, etc–they’re everywhere (except for hi-fi audio, since inefficiency makes things sound better =D). We just don’t see switchers much in the hobbyist world simply because they’re too complex and often not worth the gain in efficiency.
Posted at 11:23 pm on Apr 6th, 2007 by viktor
So would it be possible to use a TI PT5103 switching regulator 3.3V/1A, instead of the Dimension Engineering regulator? I will be using it in a Spektrum DX6 which has more room in the transmitter as it is for R/C Aircraft.
Posted at 7:54 pm on Apr 8th, 2007 by Deranged
I build combat robots here in Australia and I currently use a Spektrum DX6 transmitter and receiver. I recently bought the DE-SW033 regulator for my Spektrum and I believe it is well worth it’s price. At a combat event, our robots have to be ready ever 30mins for the next round of fights, I have two Electrifly Triton2 chargers shoving 7amps back into my NiCD’s are each fight, and I then usually have to top up the Spektrum every second fight as well… With the DE-SW033 I only have to charge my Spektrum up once per day now at a full sized event.
I also bought a DE-SW050 to use as a Battery Eliminator Circuit (BEC) for powering my BR6000 receiver and Victor 883 speedcontrollers (Yet to test this out :P) and I also bought the variable DE-SW0XX model for using on prototypes I build. Nothing worst then trying to build a quick proto and you only have a bunch of 7805 regulators and no 7812’s left!
The DE-SW033 was a much cheaper solution then going and buying a 2500mAh NiMH pack for my spektrum too, so $15 well spent in my books!
Posted at 10:56 pm on Apr 9th, 2007 by Knightrous
Hack a Day serves up fresh hacks each day, every day from around the web and a special How-To hack each week.
Powered by WordPress
Dude, this would be awesome as part of a USB power kit. I mean, the complexity of the task was once a bit daunting to get a switch-mode regulator done at home, but this just fixes our woes in one package.
Posted at 5:51 pm on Apr 5th, 2007 by Alan Rager