IR remote circuit running out of steam?!

[scag]

Hello
I have made the IR circuit found at this website.
http://www.lirc.org/improved_transmitter.html
But the electrical company that i buy my components from (Maplin) said they do not sell a 4000uF capacitor.
I had from a previous attempt to build this device a 4700uF capacitor and all works except; the signal became weaker and weaker every time a signal was sent (the brightness of the IR LED became less) untill i waited a very short time and then its back to full strength. (for around 2-3 button presses and then it starts to lose strength).<p>Do you think its due to the 4700uF amount? Do you think i can get away with less than 4000uF? Do you or anyone else know of the code for this component or know other UK based companies that supply it? <p>Hope this makes sense
Many thanks
Mike

[sofaspud]

4000µF is not a standard value for capacitors. The circuit description allows 20% tolerance for components so 4700µF is within that range. But it sounds like your circuit might be operating as designed, because the capacitor will discharge rather quickly when powering the IR LEDs.
If you don't need the visual indicator LED, I think it could be removed to conserve its current. (I'm a little confused when you say, "the brightness of the IR LED became less" unless you have very good eyesight.) Or perhaps the cap you have is leaky. You could try replacing it with another that is rated for low leakage, either 3900µF or 4700µF.

[jollyrgr]

A couple comments. You can "see" the brightness of IR LEDs by looking at them with a solid state video camera. Most home consumer video cameras from digital "still" cameras with video view finders, X10 wireless cameras, web cameras, etc. pick up the signal from a IR remote and LEDs and can be seen with the associated "monitor" be it the LCD viewfinder, a TV screen, or a computer display.<p>But asside from that I can see an issue with this circuit. You are expecting to get power for it from a serial port signal. This is a "questionable" design as you cannot depend on one serial port to be like another. In other words this circuit may work extremely well on one computer but not another.<p>I will make a couple suggestions. Replace the 4700uF capacitor with a larger one or even parallel a larger capacitor with it. (This suggestion is given in the NOTES section when adding more diodes.) Increase the value of the 22 Ohm resistor (on the emitter of the PNP transistor) to something that gives a balance between useful range and run time. This resistor is what is "limiting" the current bursts through the LEDs. Monitor the voltage on the 4700uF cap. See how long it takes to charge fully (well, reach steady state, as a capacitor never really fully charges). Send some commands and see how the voltage changes and how long it takes to drop VS how long it takes to charge. Roughly the ratio of charge time constant to discharge time constant is 100 to 1. In other words it takes almost 100 times longer to charge the capacitor as it does to discharge the capacitor. If you lower the value of the 2.2K resistor you can lower the charge time. Try something like 1.5K would be as low as I would go. A capacitor is said to be charged after five time constants. Unless my calculations are off the time constant of the charge circuit is 10.3 seconds so it will take about 52 seconds to fully charge. Neglecting the duty cycle effects on current draw this means the capacitor will be discharged in half a second. You get more "on" time because of the pulsed duty cycle.

[jwax]

How's does it work at all?
"GND" on the schematic should be +V, and GND should be at the emitter of the NPN.
A smaller cap (2200 MFD?) would settle the delay issue.

[philba]

the reason it works is because DTR goes negative. This is powered off the rs232 connection and the 4K uF acts like a battery. The orientation of the electrolytic threw me at first. The + going to ground is correct as that is the most postive rail.<p>Increasing the resistor on the PNP transistor is the right idea as to much current will deplete the cap. Or your could remove the .5K+LED leg. <p>Frankly, the group of LEDs is a bit odd - I'd have gone for a single high powered one. <p>Another approach would be to use a different RS232 lines for power (CD?) and avoid the whole DTR power thing but he just wants to tweak what he has.

[jwax]

Now I see! Thanks philba! You got that scag?
I agree with jolly- iffy design.

[scag]

Wow, I am very pleased with the response. I hope i can to all your level of competency.
Firstly this is my first ever attempt at electronics, and this device is to extend my PC as a Personal Video Player, as i need the PC to change the channel on my Freeview box via IR.<p>Ok so let me try and respond to what everyone has said sofar.
Generaly i think this design is 'iffy'! sorry its the only one that i have come across that works (mostly).<p>sofaspud you said 'visual indicator LED'
Yes this part did not work so have not included it, maybe its the size of the LED, dont know. Its not important.<p>Honest when a IR signal is sent the x2 IR LED do light up and i can see the light with the naked eye. Is that wrong?<p>Sofaspud, next time im back at Maplins i will ask about a low leakage cap.<p>Jolly Roger you say the design is questionable due to differing serial ports. When you say it may work extremely well on one computer but not another, what is it that you refer to not working? Is it the strength of the signal resulting in the distance away from the TV for example, or just flatly not work at all??
increasing the capacitor requires increasing the diodes which to the un initiated seems crazy, which i think philba agrees by saying 'the group of LEDs is a bit odd'.
Increasing the 22 Ohm resistor sounds easy but not only do i not know how to test for a balance between useful range and run time. I dont know what/how to tell if any of my left over resistors is greater than 22 Ohm as they are all in K's measurements. On a previous schematic that i tried building www.lirc.org/transmitters.html the resistor 5-100 i bought a 47 Ohm resistor (based on someones advice). Would this be OK or is it too great?
Sorry i am unable to monitor the voltage of the capacitor as dont have the equipment or experience all i can try is physicaly swap one component for another. I dont have any spare resistors between 2.2K and 1.5K, there either 100K, 0.5K, 15K or this 47 Ohm resistor. If you think this is worth a try again i will purchase next time im at Maplin.<p>Jwax
A smaller cap (2200 MFD)? On the Maplin website the capacitors http://www.maplin.co.uk/Module.aspx?Mod ... =&doy=29m4
They give values in pF (Pica Fared??) Would you be possible to tell me which capacitor in their list corresponds to what you advise??<p>Philba
yes the other approach of getting the power from somewhere else. I need to make this as simple as possible as some friends want to use this device aswell. Also the software 'WinLirc' is made for the serial device. I personaly would like to plug into the USB, but i thought try and get this to work first. Not going to run before i can walk!!<p>Jwax, Have i got that?? MMMmmm I think i have.<p>Well I hope i have given your input a proper response and have not just messed up my chances of making this device. Let me know.<p>Many thanks to all
Mike <p>[ April 29, 2005: Message edited by: scag ]</p>

[Edd]

Greetings Mike how’s everything in Wales?
With your comment on the IR emitter diodes , seems like more clarifying info for us is in order.
Like the color of the secondary visible emission that you are seeing, as well as the brand and part number of the IR units used.
In all of my uses of different units, and at pretty decent drive levels, I have not seen any that were visibly emitting. That was even the purpose of the shunted visible Led display shown in the unit.<p>73's de Edd
[email protected] ..….....(Interstellar~~~~Warp~~~Speed)
[email protected]….......(Firewalled-Spam*Cookies*Crumbs)

[philba]

Yeah, ditto that on the visible LED - I've never seen an IR LED that was visible to the naked (or clothed, for that matter) eye... I bet you got a regular LED slipped in your order.<p>The problem with different computers (laptops, especially) is that some of them don't output "true" rs232 voltage levels. Most of the time it works but when you are counting on getting something on the order of -12V (my guess), then it can be a problem if all that is output is 3V.<p>[ April 29, 2005: Message edited by: philba ]</p>

[sofaspud]

A couple of quick additional comments:
You could try putting 2 capacitors in parallel to increase the amount of stored energy. Yes, they will take longer to charge, but it sounds like you want to be able to switch through several channels so the discharge time may be the more important factor.
About the 2 IR LEDs... you might experiment with using just one IR LED, as philba mentioned. That should provide a little more operating time from the capacitor. I'll also join in saying I've never seen an IR LED emit visible light.

[rshayes]

When the LEDs are on, they are being driven with about 30 milliamps. This current will be discharging the capacitor. The charge is replaced through the 1K and 2.2K resistors in series with D2. The circuit will need about 4.4 volts on the capacitor to work (.7 + .7 + 1.5 + 1.5).<p>The RS-232 output itself should go negative somewhere between 6 to 12 volts. I forget the exact levels, but they are fairly high. There is some current limiting in the RS-232 line drivers, so this voltage may drop when loaded. Overall, the DTR line may wind up at about 8 volts negative. If the capacitor voltage swings between 5 and 8 volts, the average capacitor voltage will be about 6.5 volts. This leaves about 2 volts average across a total of 3.2K. The charging current, when DTR is negative, is about 600 microamps. This means that the LED can only be on about 2 percent of the time.<p>Reducing the size of the capacitor doesn't change things much. It recovers more quickly, but it also discharges more quickly. The allowable duty cycle won't change.<p>The RS-232 interface has several other outputs as well as DTR. You could get more current by adding 3.3K resistors from these outputs to the negative side of the capacitor and programming them to be always negative.

[scag]

Hi Edd
Yes Wales is wet today which is a shame as its a 3 day weekend. Still should be able to get out though. I gather you have been over to our fine land? Did you enjoy it?<p>You ask for 'the color of the secondary visible emission', its red.
The brand and part number of the IR units used is
http://www.maplin.co.uk/Mproduct.asp?oc=CY88V. though mine are pink color.
I hope this is enough<p>Hi philba
Strange about seeing the IR LED light up, i cant imagine that regular LED's have been slipped in as the device does work??!!??<p>Output voltage levels? I always thought this was 5V but everyone can tell im only going on what other people say.<p>Ohh i should say that the DTR pin is not used on the Serial port but the TxD (pin 3) as i was told would get better results.<p>Hi sofaspud
Yes to be able to quickly send more than one IR code is important, as not only to change the volume quickly is something we all do but also some channels are 3 digits long. (3 IR codes need to be sent as if the 3 buttons were pressed on the remote). I will try doubling the capacitors, but i am aware its the trade off between time charging and the amount of energy to send IR codes. Though if the initial voltage problem beeing given via the Serial port is changable on different PC's then surly this will further affect how long the capacitors (x2) will take to charge.<p>Yes i will also try with only 1 IR LED.<p>Hi stephen
Firstly thanks for the figures, due to having little electronic knowledge this is a bit out of my reach to know how to remedy it. I can try what you said in your last paragraph.<p>The RS-232 other outputs, i am now using the TxD, Pin 3.<p>Adding 3.3K resistors from these outputs to the negative side of the capacitor
What outputs? do you just mean the negative end of the capacitor? How do you 'programe them to be always negative' ??<p>Thank you ALL i hope to get this little nifty device, that i am calling 'TrX', roaring away very soon.
Many thanks
Mike <p>[ April 29, 2005: Message edited by: scag ]</p>

[scag]

Oh No!<p>I have tried my device with only one of the IR LED that did not work, no IR signal sent or the IR LED did not light up!!!
I then put both IR LED back together as before and now nothing. No signal sent and neither IR LED light up.<p>Interestingly i put together the LED and resistor that bridges the x2 IR LED and when a IR signal is sent the LED does light up. Meaning that power is getting to the IR LED.<p>Do you think i have burnt out the IR LED's?? Is there any way to check?? <p>Another interesting thing is this LED and resistor ive mounted them on a board with wires on each end so i can touch a possitive and negative part of the schematic and the LED lights up. I thought the capacitor discharges itself. if i place the two wires on this test LED (thingy) at each of the possitive and negative wires comming from the capacitor it stays lit??? It stays lit even when i unplug the from the serial port???<p>Oh blast im lost!! Any advice??
Mike

[rshayes]

I assume that you are using a 9 pin serial port on a PC. This has a subset of the RS-232 signals.<p>Three of these are output signals from the computer to the peripheral device.
Pin 3- TXD (Transmit Data)
Pin 4- DTR (Data Terminal Ready)
Pin 7- RTS (Request To Send)<p>TXD is the data stream. Data is sent a byte at a time with start and stop bits added.<p>DTR is negative when idle, and is driven positive to indicate that the computer is ready to send data.<p>RTS is negative when idle, and is driven positive to indicate that the computer wishes to send data.<p>The last two pins are usually set or reset in a register that is loaded by the computer. How these bits are controlled differs depending on the operating system. Programs running under DOS, such as Quick Basic programs, have fairly direct control of these pins. Other operating systems get more complicated.<p>If you are lucky, the software that you are useing may leave these pins is a negative state. Possibly you can check this with a voltmeter. If they are, they can be used as power sources in the same manner as the TXD pin is used. In effect, you are paralleling three power sources.<p>If you are not sure that these pins will stay negative, a diode can be used on each pin to prevent a positive level from discharging the capacitor. A 3.3K series resistor would duplicate the 2.2k and 1K resistors on the TXD connection.<p>This should allow the LED to be illuminated about 6 percent of the time rather than 2 percent of the time. It will also shorten the recovery time by a factor of 3.

[scag]

Hi stephen<p>Yes i am using the 9 pin serial port on my PC.
Yep i know each pin, but i did not know exactly what each pin did, thank you.
Do you think i was right now using TXD instead of DTR as the original schematic showed??
The software that i have been using is WinLirc
http://winlirc.sourceforge.net/<p>I gather what you are saying is that WinLirc leaves the pins in a negative state as it all was working up untill i tried using x1 IR LED.<p>What diode should i use on each pin to prevent a positive level discharging the capacitor, one like D1 and D2 on the schematic?
The 3.3K series resistor, are you saying to actually do away with the 1K and 2.2K resistors before and after the D2? If so where should the 3.3K go before or after the D2?
I bet i have not understood what your saying, sorry!<p>At the end of the day i dont need the LED, trying to sort this out within the schematic is it just causing more problems??
Many thanks
Mike