DIY BEC

[Mark Millward]

Has anyone any experience of making a bec  or voltage regulator with an lm317 5v 1.5a or338 5a chip.?. Are there any interference issues on 35 meg ?

All advice appreciated

  Mark

[Mark Millward]

Oops should have said 317 or 338 voltage regulator not chip

 Mark

[Brian Parker]

These are linear voltage regulators and, as such, if suitably installed, will not cause any problems BUT being linear they dissipate power through heat and are therefore inefficient when compared to switching regulators. They also require an input voltage of 2.5volts above the regulated output voltage.
 They do not have many devotees on this forum but I use 5volt fixed regulators on none powered slope soarers if using Lipos.

[Bert]

Mark

 

You need a low dropout regulator. I use two  LM2940CT in parallel on 2S Lipo for 2A at 5V without any problems at all.

 

Bert

Edited By Bert on 28/11/2009 12:44:11

[Mark Millward]

Thanks Bert   Will this prevent the 2s lipo from going below 3V ?

  Mark

[Former Member]

[This posting has been removed]

[Peter Beeney]

       Mark,   I wondered if I might add a few comments to the thread. Judging from your post, it would seem you might be a bit unfamiliar with voltage regs. The two example you have quoted are variable voltage regulators, I believe. Is there any reason you particularly wish to use these?                                                                                                       
    When lipos first appeared on the scene, I used a 6V, 1A fixed voltage regulator to drive a standard SPAD model, with 5 servos. Using a 2 series pack. With some very exhaustive ground and air testing this proved to be well within any overload limits. I was unable to get anywhere near the temperature required to shut it down, that’s around 150 degrees C. Indeed it was hardly getting warm. I used two different makes of packs, from two different suppliers, both batteries were 9% down on capacity from day one, and then one pack suddenly and completely failed overnight. So I went off that idea a bit at the time. Lipos are much more reliable now, I guess!                                    
    I installed mine in the switch harness, and as Eric says, this needs to go on the receiver side of the switch. When I tinker about with these types of chip, (all these components are colloquially known as chips), I solder short tails to the legs of the components. The legs are relatively quite fragile, sometimes just bending them about a little bit can break them off very easily; usually very close to the body; and then it can be the very devil to get a connection. Another way would be to bend the legs at right angles just once, fit the legs through the holes in a small piece of Veroboard, (or strip-board) and then solder it on. Then you can fix the various wires at will. 
    I would consider that a 1A regulator, when it’s hanging out in the open, in the wiring harness, will power the radio in a normal size model without flinching. A larger model, with digital servos, or a helicopter perhaps, may well require more power.                                  
    As Brian says, there won’t be any interaction with 35 MHz, or any thing else, come to that. Also there is no point in up-rating to 3 or more cells for more capacity. This will only add to the overheating problem too. Some arithmetic will show this; as will the physical evidence, too.  And you also need to keep an eye on the pack, to prevent it going below 6V. That’s 3V/cell! I fitted a flashing led to my model, to warn me if I’d left it switched on. It actually worked in anger, once, when I spotted it ticking away in the back of the car, just as it was getting dusk.                                                                                
    Electrically, these are quite tough. You are unlikely to zap it if you get a misconnection. Once in operation it’s very difficult indeed to break it. It simply shuts itself down for a short while. This saves the regulator but can result in the model trashing itself! As as been related in various other threads!
 
    If you require any other info please come back.            PB
 

Edited By Peter Beeney on 29/11/2009 21:29:17

[Bert]

If you don't use a low-dropout reg on 2S lipo you will not have a problem with the lipos going below 3V - because the reg will dropout long before you get to 6V and the plane will have crashed by then!

 

Seriously - the reg won't stop you over-discharging, but you can check the voltage every few flights, it's what I do.

 

Bert

[Tim Mackey]

Although not in the true spirit of the thread - these days one can buy purpose made regulators "ready to go" for little more than the price of the components.

Now I too like to fiddle about with circuits and components, and have done so for 40 years, but the price of stuff has fallen so dramatically, that I am sure many people these days opt to buy ready made on most things. Besides, IMO, constructing home made gizmos is not as much fun when most of the stuff is based around boring little ICs..... bring back the OC71 PNP germanium says me, or even a nice warm glowing Mullard

I fear the price and availability issue is the main reason we don't see much "E" in the mag these days too.

[Tim Mackey]

Posted by Bert on 30/11/2009 07:00:52:

If you don't use a low-dropout reg on 2S lipo you will not have a problem with the lipos going below 3V - because the reg will dropout long before you get to 6V and the plane will have crashed by then!

 

Seriously - the reg won't stop you over-discharging, but you can check the voltage every few flights, it's what I do.

 

Bert

 

Or add in a simple voltage monitor / alarm circuit which will sound when the supply battery reaches a preset level. I use such devices on my transmitters, as the inbuilt alarms are both too quiet to be heard in a howling wind on the slope, and also they activate too late to be of use - as I use Lipo in the Tx.

I find this simple "dead bug style" microscream very useful.

[Brian Parker]

 

To add to Eric’s ‘increasing the voltage with a resistor in the common line’. High current voltage regulation can also be achieved using the 78 fixed regulators with the addition of a power transistor (base to IC input).
 Also note that the Manufacturers stated drop-out voltages are for maximum rated output. If not driven hard, drop-out voltages are lower.

 Care is needed if mounting a heatsink when using a resistor in the common line, as the heatsink will not be at ground potential.

Yes Timbo, agreed, Oh for a return to the warm glow of the pentode, the tetrode and the triode 

 The ‘E’ is just about alive on the forum even if dead in the Magazine.

[Mark Millward]

Thanks very much all of you . My own electronics was based arround industrial digital and  TTL but that was years ago.  I think I will have a go at one or two of the suggested circuits . Has anyone tried a few diodes as a simple solution, a mix of germanium and silicone comes to mind

   Mark

[Former Member]

[This posting has been removed]

[Tim Mackey]

I had thought of doing this with servos - the Spekky rx units are good up to 9V or so, but 7V servos are stupidly expensive. I am sure there are reasons why it is not a good idea, and doubtless some of the electronic gurus will be along to tell us .

[Former Member]

[This posting has been removed]

[Mark Millward]

Phil    Could you post a little drawing of that circuit ?

  Mark

[Former Member]

[This posting has been removed]

[Mark Millward]

  Burst out laughing at that one  Phil. The 3055 and zenner. To think I have qualifications in electronics but cant remember a thing.

  Regards     Mark

[Former Member]

[This posting has been removed]

[Brian Parker]

 

I have used series silicon diodes to reduce voltages in the past (with no problems) but proper voltage regulation is preferred.
Phil, shan’t stand on your toes regarding 3055 circuit but here is a circuit for Timbo’s vintage OC71. (A voltage regulator that will also provide constant current (a few mA)).

 

 Referring back to the initial post, and in the interests of putting back the ‘E’, here is a basic switched regulator circuit using the 317 regulator in conjunction with a power transistor.

 

Some experimenting may be needed for the exact number windings around the ferrite ring (don’t cross over the windings), but it will oscillate at over 1K at light load and is good for 5A+ (use a 2Wminimum 10ohm resistor on the base of T1).
The base capacitor in the diagram actually reads 50microfarad (apologies for poor handwriting).

[Brian Parker]

Also, in the spirit of not spending more than necessary, a ‘gash’ length of household mains solid core cable can be used for the coil. Use the centre uninsulated earth wire, it is about 1.3mm dia (17SWG). Make sure the windings are not touching then give a coat of varnish as an insulator.
(Keep the red and black wires for making up batteries from single cells).

[Former Member]

[This posting has been removed]

[Peter Beeney]

    Just to recap on my previous post, and to get some more facts and figures on the 2S Lipo and the voltage regulator, I thought I might run a little test procedure past myself, similar to what I’d done the first time around. This post does contain lots of boring facts, figures and rubbish.  
    This is a 1A, 6V reg, I’m not sure of the drop-out voltage, let's say it might be 1.2 - 1.5V. I’m not too worried about that though, it’s the battery top end that we might be more concerned with.
    I used a constant 1A discharge rate, trying to ensure that the regulator was working considerably harder than it would be under flying conditions. I assumed a standard battery working range from 8.4V down to 6V. The regulator was out in the open, with no heat sink. At 8.4V, discharged for half an hour at one amp, it’s hot to the touch, but not too hot to to hold. The output is a constant 6V. Or to be pedantic, it’s just under 6V; but that’s exactly the same as the open circuit voltage anyway. Dropping down to an input of 6.8V, the output is down to 5V. At input 6.4V, output 4.8V.  Right down at 6.0V, output 4.4V. Below this, the output voltage dropped away rapidly, but the 1A current output was maintained constantly. I ran it for about half an hour at all of these settings. At no time did it appear to be be overloaded or overheating. When these things get up to around the 150 degree C shutdown level, they are finger-blistering hot!! One of the tests I did originally, was to connect the wing up and then constantly stir both sticks for about 5 minutes, again something you couldn’t do in the air. It didn’t even get warm! However, I’m not really keen on this sort of caper, I tend to think it starts to wear things out!
    The drop-out voltage is the minimum voltage between the input and output at which the stated voltage and current can be maintained. I personally consider this is never really going to be a big issue, with me anyway, because I’m extremely unlikely to be taking the receiver battery volts down to anywhere near flat.
    Low drop-out 6V, higher current output regs might be a bit thin on the ground, so if we would really like a higher output type, one answer might be a low drop-out 5V version. And then bump it up to around 6V with the help of a couple of small silicon rectifier diodes. Then you could take the 2S battery down to 6.6V and still have a 6V supply. The KA278R05C is a 5V, 2A, low drop-out regulator, the drop-out voltage is 0.5V at full rated current, less at flying currents. Of course, there is absolutely no reason why it can’t be used at 5V anyway; it’s just that I try and raise the rx supply to 6V whenever possible.

    With the greatest respect, I feel that I might need to approach some of the comments with a degree of caution. But I will make an observation about Phil’s suggestion about lowering the battery voltage with the use of diodes. I’ve just checked 5 1N4501 general purpose silicon 3A, 100V diodes at 500mA and the average forward volt drop is 0.8035V.  Lets call it 0.8V. Using this as a de facto standard, then 3 in series would result in a total drop of 2.4V. Connect this to a fully charged battery, 8.4V, and the rx voltage is 6V. Just right. However, when the battery is down at 6V the rx is at 3.6V. I wouldn’t want to go there! Admittedly, the battery is flat at this point anyway. But if you never wanted to get below, say, 4.6V at the rx you would not have to let the battery volts get below 7V. Using just one diode would narrow this ‘step-down’ voltage a bit, thus our 8.4V starting point would be 7.6V at the rx. This may be acceptable. But at the lower end, the 6V battery would equate to 5.2V. Much safer!
    I’d guess that 4 in series might just be a little bit too ambitious!
    And can I add please, none of this is at all intended as a criticism, merely as an observation. You pays yer money, and you takes yer choice!  

    Hope some this is of interest, at least.                     PB

[Tim Mackey]

Yes Peter very interesting mate. One observation I would make. I use Lipo in all my stuff, slopers, IC, and so on, including my DX7 transmitter. However I dont use the "normal" 3V per cell minimum level for the radio batteries. The 3V ( or even a tad lower TBH ) is fine on normally high discharge level currents, but remember that for instance, a 2cell Rx battery at even 7.4V has only around 20% remaining, and at 7.3V   < 10%. Rapid voltage drop off , or "dumping" occurs from these levels, and thats something I dont want happening on my radio gear!  I normally work on the level of approx 3.7V per cell as the lowest I let 'em get, The DX7 inbuilt battery alarm on my 3 x cell set activates at approx 10V, although the manual states 9V - way too low for my Lipos, so my microscream alarm which I fitted is set to sound at 11.1 V in my  3 cell tx pack. The following is extracted from a  earlier posting years back.

 

In testing, the LiPo pack took 36 minutes to drop from 10.6V down to 10V, with the first 1/10th of a volt lost in 12 mins, and the last tenth lost in only 3 minutes. This is to be expected, as we know pretty well any battery will tail off rapidly towards the end of its charge state. Although it would be technically safe to fly for around half an hour then from the meter showing 10.6V it would be wiser to stop quite a bit sooner at say 11V or so at the latest. The LiPo is pretty well empty at 11V anyway. The tx continued to transmit perfectly well with the alarm sounding right down to a meter reading of 9.7V ( this only took a further 5.4 minutes incidentally ) when I chickened out and shut off the tr.

Reaching this lower limit of 11V still however takes around  7 hours continuous useage from a full charge of the 2500m/a pack.