flyinrog

My Feedback: (1)

I have been looking for something for 1/2a..since I have a few leckys going anyway I figure maybe use a lipo through a volt reg. and fly all day on a 11.1 volt lipo through one that goes down to 5 volts...and since all of you gentlemen in the 1/2a forum are the most knowledgable men on the planet..I thought hey ask these guys first....not a capacitor, but a regulator...thanks, oh links are good too...Rog

ProBroJoe

Well... Castle Creations makes a nice BEC which would do just what you're looking for.

http://www.castlecreations.com/products/cc_bec.html

Or... you could just buy an el-cheapo brushed speed controller and use the BEC circuit there...

http://www.bphobbies.com/view.asp?id...8&pid=NESKY159

There are physically smaller ESCs, but I like that this one claims a 1.5A BEC circuit - should power just about any 1/2A model with overhead to spare.

No reason why an LM7805 voltage regulatot wouldn't work, especially w/11.1 V in, but using a BEC intended to do just want you want takes the guesswork out of buying and rigging a 5V regulator and associated wires/connectors. Plug and play my friend, plug and play!

Hope this helps,

-Joe

Mr. Mugen

I use CC Bec's for reg's in my big planes. They are easy to use and work great. They weigh very little which would be good for 1/2a. Price is not all that bad either.

Kevin

http://www3.towerhobbies.com/cgi-bin...&P=SM&I=LXSWL3

Check them out.

Andrew

Ah so, lekky boy, you wanta regulator?

There are a couple of factors to be considered. What voltage differential is acceptable, Vin-Vout, and how much current will the regulator need to carry. There are a host of low drop out (LDO) regualtors that will work with either 3S (11.1v) or 2S (7.4v) lipos, but finding one that will handle enough current to run with one or two stalled servos is a bit more difficult. Lastly, what type of mount package do you want to use?

I tend to use the TO-220 package instead of the TO-263. The TO-263 is a surface mount device and I have trouble when soldering small parts. The TO-220 is easier for me to handle and I can cut the tab and legs off after making up what I want.

If weight is a real consideration, then you would probably want to use a 2S instead of the 3S. Either way, pack capacity is the key since the lipos tend to hold their voltage, then drop off pretty quickly. Most of the ESCs have a low voltage cutoff level - the regulators will just continue to dropoff until something stops working, so you would need to monitor your voltage level at the pack rather than the RX.

Although the regulators are available in both adjustable and fixed output voltage, there is no need to use an adjustable Vreg it just adds two additional resistors and most Vregs can be had in a fixed 5.0v output.

So, what's my pick? The [link=http://www.national.com/ds/LM/LM1084.pdf]LM1084-5.0[/link] is a fixed 5.0v regulator rated at 5A - more than most other LDO regs. Add a couple of 10 uf tantalum caps, connectors for your lipo and receiver, heatshrink to cover it and you're ready to go. Unless you're going to use digital servos or have 8 servos on your plane, 5A should carry your current need (actually it should be much less, but in this situation, there is no weight penalty for capacity).

The 1084 will hold 5v with a differential of about 1.25v. You might squeeze 8 amps out of it, but as the diffential gets smaller, the current rating also falls off.

[link=http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=LM1084IT-5.0-ND]Digikey[/link] has the LM1084 in stock, 5.0v fixed in a TO-220 package. The downside is that it's DIY. For me, I usually have most of the parts in a junk box somewhere and can usually cobble up something at about 10% of market price.

DeviousDave

Be careful with this approach, ESC's (especially cheap ones) can't handle large numbers of servos. The current draw ends up being too much-ChiCom servos in particular are a problem as they use hot motors in their servos to get the transition speeds up. It's the startup current the motors pull that is hard on the BEC.

Three servos is asking as much as you can from a typical small ESC.

MJD

My Feedback: (1)

I have a handful of MC7805's for exactly this sort of thing. I couldn't find any caps even close so I made a hybrid - half schematic and half photo. How informative.. .

But anyways flyingrog you can see how big one of these little regulators is, i.e. not too big, although a couple of small capacitors need to be added as mentioned. I should patch one together and weight it, to see how much benefit there really is for small models. The parts required would likely total about the cost of a 6" servo extension that you could cut in half to make the cable for one of these. 2s Lipo is fine with these as they need about 2V over output voltage minimum at the input.

In the end flyingrog, the question is do you win out over a small NiMH pack for example, i.e. capacity, weight, complexity, etc.

These same regulators, with the addition of little more in the way of ancillary components, can be turned into wonderful little current regulators for overnight charging of tiny NiCD and NiMH packs. I have a 20mA regulator for example, that plugs in series with a Futaba wall charger and charges my micro/mini packs at a constant 20mA, i.e. ~C/15 for the 300mAh and ~C/11 for the 220mAh.


MJD

combatpigg

My Feedback: (3)

Now don't these gizmos need to be bolted to some metal to achieve their rated capacity?

MJD

My Feedback: (1)

Yup that's right. But we only need a fraction of their rated capacity, and the loads are sporadic. A 1/2A model likely goes through maybe 200-250mAh per hour or so (i.e. average load 200-250mA), so the regulator is required to deliver about 1/4 of its capacity. I think the 7805 for example is likely fine as is with the external caps and connectors (Andrew whatchoo think?). Besides, a bit of heat dissipation could be added for a gram or so I am sure if needed. A 1" x 1" coupon of thin aluminum to which this is attached would provide a reasonable amount of heat dissipation for light loads.

MJD

Andrew

CP -

Sometimes, but the causes of heating can be surprising (well, at least to me). Generally, I think of heat generation being associated with current. Voltage regulators tend to heat when the voltage differential is large - a voltage drop from 20v to 5v has to be dissipated somehow and that's heat. For our use, dropping a 3s or 2s lipo to 5v may cause it to get bearly warm. For a 2s lipo, the drop is so small that a LDO reg might be a better choice. I put together another circuit to drive a foam cutter that uses a power transistor. This will drive a 36" bow - this particular MOSFET tends to heat as the switching frequency goes up, not so much with the current it carries.

You're correct, but for the loads imposed by most 1/2A applications, we can probably get by without it. Many of these Vregs will run at junction temperatures of 100C or higher.

BMatthews

For working those 3 pin jobs from a 2 cell 7.4 Lipo I wouldn't worry about any heatsink other than perhaps placing the back tab out in the open but still inside the model. But for a 3S pack at 11.1 I'd want to put the back tab out in the airflow or add a heatsink if it's inside the model. Buty either way do NOT bury them in a wad of foam rubber.

The issue is that these units contain a thermal shutdown circuit so if they overheat they shut down until they cool back down to the temp they are happy at. I don't need to fill in the rest do I?

Back in the day I ran a speed control that was passing 25'ish amps through 3 MOSFETS. All I did was put the MOSFET tabs through a hole into the outdoors. The passing wind kept them wonderfully cool and I never had an issue over about 7 years of flying and probably 80 to 100 flights before one day I did a gentle landing but with the prop down a bit and the motor's gear housing plastic snapped into 5 pieces.

Sticking the tab or exposing the whole backside of the heatsink to the passing airflow would ensure you don't need a heatsink and that it won't go into thermal shutdown.

And CP, I don't want to see you bringing up the fact that I mentioned the regulator's backside.......

Andrew

MJD -

I doubt if it would need a sink - the voltage differential is small and current demands are relatively low and sporadic. The 78xx series will hold a peak current of around 2.4 A. If we were running more than 3 servos, I would probably be inclined to look for a slightly higher current rating, just to manage a stalled servo more than anything. For 2s lipos, I think one of the LDO Vregs might be a better choice.

BTW, I cobbled up a constant current charger like you - primarily to form charge NiMH packs. None of my chargers would drop low enough for forming my 350 and 400 mAh packs.

flyinrog

My Feedback: (1)

PBJ thats the first one I looked at too, but thought there were more out there, I like Berg stuff and CC stuff so I will probably go that way...thanks....Rog

planebuilder66

My Feedback: (8)

Well, I have good luck with my triton JR, it goes as low as .1 amp, but its interesting to see this application, I just might try this and a 3 cell lipo around 800-1000 capacity. Maybe in my herr little extra setup, or the HOB bonanza, well see.

MJD

My Feedback: (1)

Same here, 100mA is as low as mine go. Charging 220 and 300 MiMH packs at normal rates when new is pretty hard on them, at least I would expect a far shorter service life. I made a few of them so I could swap if need be, rather than fast charge. Considering how little charge my Blink used up in setting up the controls, bench racing, then two test flights with full tank, I am not worried about capacity with the 300 for sure, and I have yet to fly the 220's. The extra weight of the 300 helped balance the extra servo and throttled engine weight. It's still <8 ounces though.

MJD

MJD

My Feedback: (1)

I admit I chainsawed the balsa piece here, or at least it looks like it. Actually it is very punky wood and not worth a hoot.

Anyhow.. I think this is how I would mount one of these. Low profile, and with boundary layer airflow there would be some cooling benefit over the more passive environment internally. If I liked them and wanted to make a few, it would likely be easiest to premake these on a standardized 1/16" ply plate.

Oh boy, another project!.. []

MJD

BMatthews

Andrew and MJD, post up a schematic of what your charging circuit is. I've done constant current chargers for some pretty small packs and had no trouble getting 78xx regulators down to 20 ma. I've also got a low power 78L05 running about 5 ma into a resistor to act as a precision voltage drop for a battery cycler circuit I built up years ago.

MJD

My Feedback: (1)

Sure, will do.

http://www.datasheetcatalog.org/data...r/DS009063.PDF page 9, top left

http://www.datasheetcatalog.com/data...8/MC7805.shtml page 22, top left

I tried copying them into a document but lost too much resolution to read the small print.

The main difference between the two circuits is that the fixed voltage regulators use the output voltage in the calculation, whereas adjustable regulators attempt to maintain 1.2V across R1 and the resistance values are based on that. Sometimes the quiescent current, Iq - the current that flows between ADJ and Vout no matter what - is factored into the equation. In that case of low current regulators, as in 20-30mA, you can compensate for Iq by subtracting 3mA (.003A) from the desired output current and selecting the R1 value from that. With larger currents it is typically ignored, since the difference with or without is often less than the variation in output current that results from normal tolerances in resistor value.

So, for the fixed voltage regulator MC7805 (5V) which I use as an example since I have a bunch, and for a desired output current of 30mA (.03A), R1 is determined by:

Iout = 5V/R1 +Iq* *optional, if you care about 3mA
.03A = 5V/R1 + .003A
.027A = 5V/R1
R1 = 5V/.027A
= 185 ohms

For adjustable regulators, simply use 1.25V in the same equation.

The resistor must be of sufficient power capacity to handle the current:

Power in watts = I^2R = current x current x resistance = .030A x .030A x 185ohms = 0.17W

So a 1/4 watt resistor would do the job, though I would typically choose a bit higher capacity for a benchtop device (rather than say for an onboard voltage regulator where weight is a concern).

Does this help?

MJD

Andrew

The charger I use came from [link=http://www.uoguelph.ca/~antoon/]Tony van Roon's[/link] site. It actually uses a MOSFET with a precision pot rather than a Vreg. Mine doesn't have a builtin ammeter - I just set the rate with my DMM in circuit, then pull it out and charge directly to the batteries.

MJD

My Feedback: (1)

Exactly how I've done it, so far. I have a nice old analog ammeter from my dad's collection, with a full scale of 200mA (and it is plenty accurate enough for this job), and this is eventually going into a 5 or 6 channel current source charger I'm casually cobbling together. It will have a selector switch to monitor whichever output I need to adjust. I may also incorporate a timer in each channel, but not sure. We'll see as it progresses and how much ambition I have ([sm=lol.gif] - my wife laughing at that comment).

MJD

MJD

My Feedback: (1)

BTW Andrew - please forgive my overly fun youth ravaged memory, but was it you that sent me that Doculam last year?

MJD

Andrew

Yup.

flyinrog

My Feedback: (1)

I bought a Swoop lecky from Hobby People, I dont think they have them anymore but it is a stryker type or space scooter plane a delta rear motor....well it is a brushed motor but a single cell 1300 lipo runs motor and 2 servos..well ol Rog ran it into the ground after a few flights,,,in pieces it has a rx and tx and can run off either that lipo or one of my regular nicads .....so this system is slated to go into my blink the lipo runs through the regulator then the rx, the nicad runs through the rx,, not both but either or....Rog

MJD

My Feedback: (1)

Whan you say "goes through" do you mean that is the order the components collide in a crash?

flyinrog

My Feedback: (1)

I said "runs through" but yes I guess it did[&:] but with the front mounted bee in it, it should go motor first then servos last