I made a am transmitter. The signal is very weak because of the low voltage so I need an op amp to amplify it. I am not sure which op amp has the speed for my 1 MHz transmitter. Which op amp would be fast enough to handle 1 MHz and amplify the signal up to 12 volts.

 

Non, you need to use a powerful transistor output to boost the signal, can you post your circuit ?

 

My schematic is the same except that the input to the oscillator is 5v.

 

Can you give us more details about the oscillator please. A part number or a datasheet would be good.

 

The part number is 27861. I bought it from jameco.

 

The part number is 27861. I bought it from jameco.

You have already got a thread going about this which was answered by Ron H, the maximum supply is 5Volts !!
thread

 

...you need to use a powerful transistor output to boost the signal...

+1
That oscillator is meant to run TTL (small) loads. I played around a bit with one to get the most signal out of it but as you may know, 20 feet or so is about all you'll get with a reasonable length of antenna (less than 10').

It produces a square wave, which is a mix of many frequencies. I didn't see much spread to other frequencies when I built this but amplifying a square wave will likely put you in violation of local laws regarding radio broadcasting. It's likely to interfere with your neighbors AM radio reception.

 

It produces a square wave, which is a mix of many frequencies. I didn't see much spread to other frequencies when I built this but amplifying a square wave will likely put you in violation of local laws regarding radio broadcasting. It's likely to interfere with your neighbors AM radio reception.

Well, if you actually look at one on a scope it's usually a clipped sine. You do get high frequency content from the clipping but it's not as bad as a square wave. They are usually 1-transistor oscillators diode-clamped to the supply rails and work over a reasonable range of supply voltage. With a suitable band-pass filter on the output I would happily modulate one in the way suggested, but not push it to 9V.

I would be inclined to reduce the supply to 5V (although I suppose if it's survived this long at 9V it's probably fine), clamp the 8-ohm side of the transformer with antiparallel diodes (maybe LEDs, could then modulate the supply in the range 3-7V which seems sensible). Then feed a 1 MHz LC filter and a 1-transistor amplifier with a tuned collector load running off 12V or so.

Edit: Had a play in LTSpice, see attached.
​

I changed the audio transformer to 1:1 (1000:8 ohms is about 11:1 in voltage) and dropped the oscillator supply to get more modulation with a sensible audio source. 40 mW output with second harmonic 35 dB down.
​

 

... it's usually a clipped sine

I don't think it's that bad. I mean the rise/fall time is less than 2% of the pulse width - less than 10ns out a 0.5µs pulse at 1 MHz - so it ought to look awfully like a square wave.

You can measure the actual current draw from the device you have, and then you can get much more amplitude modulation of its power supply voltage by careful choice of the transformer in series with the power.

In the project I linked earlier, I used capacitive coupling instead of the transformer and had good results in getting as much signal as possible.

 

I don't think it's that bad. I mean the rise/fall time is less than 2% of the pulse width - less than 10ns out a 0.5µs pulse at 1 MHz - so it ought to look awfully like a square wave.

I'll confess I didn't really look at the specifications of that oscillator, it seems a lot closer to square than the ones I've used in the past. On those the datasheet gave a fairly wide window and actually measuring them they were about the middle 50% of a sine - I'll see if I can dig out a part number.

 

It may depend heavily on the loading? But I would think the users of these things would demand a pretty square clock signal from them.

 

It may. I have to say I was very suprised when I discovered how far from square the ones I had were. On the other hand so long as it's monotonic and consistant between cycles it probably doesn't matter.

 

Non, you need to use a powerful transistor output to boost the signal, can you post your circuit ?

I have no knowledge on this area. I was wondering whether something like this can be used? It is quite expensive though.

http://www.linear.com/product/LT1210

 

I have had sucess misusing CATV amplifier modules similar to this one: www.nxp.com/documents/data_sheet/BGY887B.pdf‎

None of them are specified down to a megahertz but they'll probably work fine. You can get about 15 V pk-pk from them before they start to distort too much. They're about £30 each from Digikey.

 

We're two pages in and you still haven't said what you really want to do. What sort of range do you want? Are you aware that you will likely be breaking local laws if you continue?

A single transistor RF amplifier is all you need to take you into a range that is likely not legal. As is, the crystal with a long antenna has a range of maybe 20 feet or so and is probably fine. Of course range depends a little on the receiver as well.