Boosting the gain of this amp Moderators: adafruit_support_bill, adafruit

[shobley]

Hi,

I recently built this amp circuit to boost the antenna signal for a crystal radio demo. It works, but I was wondering if I were to add another transistor stage could I boost the gain of the amp further?

http://i770.photobucket.com/albums/xx347/wcirco/f5fa74a5.jpg

...and what would be the implications of doing that to the signal/noise?

As always - any ideas greatly appreciated.

Steve

[zener]

I am wondering if a different transistor might be better for a first try at improving it. Something spec'd more for that application.

[shobley]

Can you suggest a good alternative - the 3904 was the best one I had at home.

[zener]

well, this isn't my area of expertise (if I have any...) Maybe something like this:

http://semicon.sanyo.com/en/ds_e/ENN733D.pdf

DigiKey has it for .39

It is tiny so get a couple if you are going to try it.

Anyway, to answer your original question, there is a fast way to find out!

[stinkbutt]

If you want more gain, you have three options.

The first is using a transistor with a higher gain as Zener suggested. h(FE) for the 3904 is probably about 100, whilst the 2839 looks to go up as high as 160+. That's certainly one way to take care of it.

The second is to use a Darlington pair. It's basically using two transistors, where one just feeds into the base of the second one, so the second amplifies the output of the second. It's cheap (since you probably have a second 3904 lying around) and it's powerful: You can go from a gain of 100 to a gain of 10,000 (100 * 100 - yeah, that's the math.) That's it's problem, though: It's powerful, but perhaps too powerful. If you need a gain of, say, 250 instead of 10,000, then you're overshooting. By a lot. Using a darlington is a bit, ah... ham-handed.

The final option is to use an op-amp. An op-amp will basically allow you to construct a circuit with any gain you see fit. For example using this circuit:



...will give you a gain of Rf/Rin. So just set Rf at 10K and Rin at 10 Ohms, and you've got your 1,000 to 1 gain. Or whatever you feel like setting the gain at. Or you could use a POT, putting the middle pin at the junction between Rin, Rf, and the "-" terminal and you've got an adjustable gain amplifier.

Seriously, op-amps are awesome, and if you want to do any real audio work you'll need to use them. However, there are drawbacks. First, you need to go out and buy it. They're not expensive, but you may need to wait a while. Hopefully your Radio Shack has one in that "cabinet-of-random-crap" that they keep in the back. Usually they do. Secondly, and more importantly, op-amps require two voltage supplies in addition to ground. V+ and V-. V- is the NEGATIVE power supply. So you need to be able to generate a negative voltage reference, equal in magnitude to your V+. Now, mostly people don't really generate negative voltage. Mostly they just cheat, by setting a virtual ground with a voltage divider or using the dropouts from diodes. I imagine Adafruit cheated in their wave shield, actually, because they use an op amp in that circuit and I can find no negative power rails. In fact, you can see for yourself they just use the regular ground for the negative supply. (It's that pair of op-amps in the upper right of the image.)

[zener]

Are we amplifying RF or audio? I was thinking RF so I don't know if an opamp can handle that. But maybe. Maybe you need a high bandwidth one.

[stinkbutt]

Are we amplifying RF or audio? I was thinking RF so I don't know if an opamp can handle that. But maybe. Maybe you need a high bandwidth one.

I assumed he was using it to amplify the audio. Still the TS922 can operate at 4 MHz, which should be good for halfway up through the AM broadcast band. For higher frequencies I found about a dozen op-amps on mouser that were all < $5 and had bandwidth > 200 MHz, which should be good for the entire FM broadcast band as well. That pretty much covers every use of radio that this guy's likely to be using this reciever for, unless he's trying to literally listen to his microwave oven.

[shobley]

Thanks for the replies.

I'm probably going to go the darlington pair route, I did put together a simple amp using a TL082 (4Mhz bandwidth) and that seemed to work ok, but I needed to split the power to get it to work.

[stinkbutt]

Thanks for the replies.

I'm probably going to go the darlington pair route, I did put together a simple amp using a TL082 (4Mhz bandwidth) and that seemed to work ok, but I needed to split the power to get it to work.

If you find the gain is too high and you're clipping your signal you can probably just use a garden-variety voltage divider to reduce the output voltage of the first transistor, like this:

darlington.jpg (6.25 KiB) Viewed 2002 times

I've never seen this done, but I can think of no reason why it wouldn't work. It may not matter one way or the other, however. It would surprise me if a radio antenna signal somehow generated more than -2.8V of signal, (which is where you start clipping in that circuit) even after amplification.