Hi all,
Just read all the recent posts- very intresting. Have been on other things including domestic duties. Also someone had this light controller and I spent a long time just trying to figure out the circuit diagram!
Few days ago I had tested your version of full wave rectifier radio on my very sensitive 32 ohms earpiece and I heard almost same level of audio from half rectifier and from full rectifier.
I added a tuning circuit for AM (LC) same as crystal radio and took its output to your both circuits. Full wave rectifier played same level of audio, I don't know why. And I didn't get audio from transistor amplifier. Maybe I made mistake.
Not too strange about the full wave detector sounding no louder than the the half wave, now that I have thought about it: like the eye, the ear has a logarithmic response, so doubling the power is just perceptable (3dB). To sound twice as loud you need 10x the power. I think all that is right. If your Xtal earpiece* is truely voltage driven, as opposed to power driven, the audio power would only double so you wouldn't hardly notice that. On the other hand, if it is power driven like a loudspeaker, the volume would go up 4x. You would notice that. If you use the 32 Ohm earphone you will swamp all effects out with the low impedance. Could I suggest staying with the xtal phone for the time being. We can do a hifi version, once the front end detector is sorted. (* appears to contradict physics but the low coupling efficiency, about 3%, is the cause)
If I remember correctly, the full wave detector should have less noise though, for a given signal, that is because the wanted signal is coherent and the noise is just er... noise. I think the noise improvement is around √2. No, doubt the experts will will be able to advise on this if I have got it wrong.
To develop the circuit you really net a meter that can handle up to about 500Hz or better still an oscilloscope.
Hi again,
How about a synchronous receiver?
Great idea, were you thinking about useing RF FETS?
I started looking at a switching demodulator using a compartator which is another approach, but not sure how noisy. If it works it should have a threshhold of about 1mv, compared to the about 150 mv for Ge and Schottkey diodes and probably about 100mv for the biased transistor detector. Now that we are allowed to use a power line for the xtal radio, all sorts of approaches are possible.
A front end amp is probably the best way forward. I think the frequecies for AM go up to 30Mhz.
Not to derail this thread, Willen Raj K, but just abrief comment:
Many years ago I also "discovered"
that a 1N4148 could be used as a temperature sensor. Tested a big bunch of them between 0º and 100ºC. Still have the lab tubes in my drawers.
Having heard about their linear response and how many mV they will change per ºC I found they showed a really non-linear response.
For any practical purpose I think is better you test them to see how much they can go apart from a straight response. IIRC (done it more than 15 years ago) the highest separation from a straight line occurred somewhere around 55º or 60ºC.
Those notes and my circuits from that time are lost. That moved me to never take again anything out of my home.
Additional and curious note: I was truly amazed to see that the temperature of the water (furiously boiling) in the kettle, varied some degrees depending where the sensor was located. Hot experience that some of my fingers will never forget.
You make me smile- I have done similar experiments way back. I was testing a circuit built with some expensive components (new fangled transistors) and read that the gain went up with temperature so I heated it all up with my mom's hair dryer. The gain sure went up, but so did the transistors. That cost me about 3 months pocket money.
The voltage depends on the semiconuctor bandgap which should decrease fairly linearly with temperature. Have a look at this:
**broken link removed**
What you have probably seen is the nonlinear effects from the bulk resistance ( I think that is the term) , which cloud the issue, more so as the current through the diode is increased. The 1n4148 is a fast diode with low leakage but I think it has a relatively high resistance. The 1N4001 has a low resistance but a high leakage which would also cloud the issue. Try putting less current through the diode: 10uA would be good if you have a high impedence digital volt meter. The emitter base junction of a small signal transistor should be quite good for temperatue measuerements. Try both the base/emmiter junction and then the base/collector junction.
Water can never get hotter than 100 degrees C, it's a pysical impossibility, unless it is pressurised that is (no doubt you know that). What you were measuring, I suspect, was small bubbles of steam.
Nice reading your posts and talking to you all, but need a 48hr day