Built this little amp using an LM4871. I added a power indicator light (at the top of the schematic) and it has given me all kinds of grief . First, my dumb a$$ wired it in series with the 5V power and got a horrendous buzz on the output . Figgered that out and rewired it to as the schematic shows. That little bugger now drowns out any signal input with a nasty 1.97mHz wave. Cut the lead to and and it works fine. Can a guy use a blocking cap to filter that out? Hope everyone has a happy and safe 4th!

What are the specs for LA1? Voltage & current.

Wiring it as you have, the circuit sees 5V minus the LED voltage. It the LED is white, that leaves you less than 2V to run the rest of the circuit. U1 may be oscillating?

Rich

Here is the LM4871 spec pdf: http://www.ti.com/lit/gpn/LM4871. The LED is a 1.8V yellow I believe. I am a little confused as to how that LED takes away voltage to the IC?

You need decoupling caps of 0.1uf and/or .01uf on that power rail somewhere.

If the supply voltage is 5 V DC and the LED drops 1.8V, there is only 3.2 V left to run the IC.

Why not just connect the LED from 5V+ to ground using about 100 ohm resistor to limit the current?

Rich
PS: Yes de-coupling caps would be good.

I must be missing something, wouldn't be the first time . 5V comes in, one branch goes to the 220 ohm dropping resistor then the LED. Another branch goes to pin 7 of the IC. I am lost on how that LED steals 2 volts from the IC?

If you posted the original circuit, that's how it drops the voltage.

Yes, I originally put the LED in series with the 5V. I figgered out my error and it is wired like the schematic I posted, now. Still get the 2 mHz signal coming out the amp. Removed the LED, amp works like a champ. Don't need the LED, try trying to learn why it still screwed up my circuit and how to remedy it.

First add a 0.1 uF supply bypass capacitor, preferably ceramic, to U1. Make the connections as short and direct as possible. Then, are all the other components around U1 connected with short, direct leads? If not, make it so. Is it laid out so that no output comes near any input, either for one stage or for the whole device? If not, rearrange it. Is whatever you are using for a load match what U1 wants to see? If not, fix it.

Please show the rest of the circuit. Where does the +5 volt arrow go.

LEDs are very complex devices and come in many forms. I have had similar problems with interference when using cheap LEDs. The bulbs made for direct replacement for 6.3 ac bayonet type fixtures do not produce radio interference as far as I have experienced. White LEDs can be the most complex devices. The white light can be produced using a number of ways, including the blending of red, yellow, and blue. I would suggest you try using a red LED instead of white and see if that produces radio waves.

Norm

Individual LEDs do not oscillate or produce radio waves. They are just diodes with a higher Vf. Not rectifiers, as the reverse voltage is very low, so don't use them on AC.

LED replacements for AC incandescents can have various circuit elements built-in, like a rectifier or bridge. LEDs designed for use on 120 VAC have a high frequency driver circuit which can cause all sorts of RFI.

I believe your oscillation is occurring in the amplifier circuit, either from low Vcc or lack of de-coupling and bypassing.

Rich

I know of two ways to produce white light from an LED. Most common is to use a blue/UV LED and coat it with a phosphor. The UV excites the phosphor, just like in a fluorescent tube, emitting a white light with a good bit of blue color, depending on the phosphor. Unlike a fluorescent, there is no ionized gas in an LED to cause the noise typical of fluorescents (and mercury vapor tubes).

The other way is to combine 3 LED chips(red, green, blue) in a single package and balance the RGB colors to give white. That's the hard way to do it.

https://www.ledsmagazine.com/smart-ligh ... s-magazine

Rich

PS: there may be a third way to get white light. Drive a huge current through a blue LED and as the internal wire bonds vaporize, you may get nearly white light for few milliseconds

Winner, winner, chicken dinner! The LED I used was from a box of parts I picked up a while back. Oddly enough, they were labeled "free" in the ziplock baggie. Free musta meant cheap crap led, who knew that bugger could cause such a mess... Which leads to a question, how does a guy know when buying LED if its a "quality" LED???

https://en.m.wikipedia.org/wiki/Light-emitting_diode

The Wikipedia discussion implies that LED technology has been changing rapidly and there is no way to know if a cheap LED is just old technology or the result of inexpensive new improved production methods. I wonder if you put a RF choke (1mH) between the device and the rest of the circuit if that would help. There may be some switching action in the device that produces the RF.

Norm

Whatever floats your boat!

Rich

I once built a 2 stage transformer coupled AF amp. Whatever I played through it came out on the TV screen as interference bars (this was in the analog days around 1967).
As best I could figure, one of the junctions was oscillating in the VHF range.
Even some of the Edison incandescent bulbs could oscillate!

Bob

The transistors you are using, both separate and in the IC, are capable of operation in the hundreds of megahertz. Therefore, the circuit must be wired as if it were for 200 MHz even though the desired signal is audio. Proper bypassing, short leads, inputs separated from outputs, etc. are a must. Otherwise, it is likely to produce its own signal at whatever frequency things happen to resonate at.

In other words snivets.

Try a cap across the LED and see what that does. If it is making noise, that will bypass it to ground. Something like .01-.1uF.