Lou

Your second schematic doesn't look very healthy, I'd stick with the first. The 1K resistor has to be inserted between the output of the opamp and the gate of the mosfet. cheers

Hi If you want a clean supply, it is probably a good idea to avoid switching regulators, despite the amount of heath the circuit will generate. Capacitors are OK for improving the noise performance, but chokes risk destabilising the circuit; I'd avoid them. An improvement to your circuit would be the addition of a ~1K resistor in series with the the gate of the mosfet; although not strictly necessary, it will help in reducing potential parasitic oscillations. Lou

Hi I suppose this modulator is intended to be used with a TV. In this case, all you have to do is to add the sound subcarrier signal (or more exactly, the sound intercarrier signal) to the video signal at the input of your video modulator. The sound subcarrier depends on your country: in Europe mainland it is 5.5MHz, in the US, it is 4.5MHz, etc. You can generate this signal using a small VCO modulated with the audio signal. Premphasis is needed and also depends on the TV standard. Cheers, Lou

Hi Here is a small primer on the subject to begin with: http://www.web-ee.com/primers/files/an1016.rev0.pdf

You could probably modify the design to extend its range: by adding a precaler or a supplementary counter stage. Lou

These would typically be plastic types: mylar (polyester), polycarbonates. Lou

You can use the following circuit with minor modifications (capacitors slightly larger and one or two more turns to the coil). http://users.cableaz.com/~cappels/dproj/FMVFO/fmvfo.htm In order to servo its frequency, you have to follow it with a frequency divider that brings the frequency to some convenient value: 100KHz fe. You feed this frequency and your reference derived from a crystal into a phase comparator and you use the output to control the VCO. Alternatively you could use a synthetised FM tuner: the local oscillator normally runs 10.7MHz above the received frequency which means that when the top frequency of 108MHz is displayed the oscillator runs at 119.5MHz. If you divide by two this frequency, you get pretty close to 60MHz. You can reach the exact frequency by replacing the crystal of the synthetiser IC with a slightly higher one. It needs not be exact because you can always correct the value by choosing another receiving frequency than 108MHz. Alternatively you could run the local oscillator directly on 60MHz, but this implies some changes to the oscillator itself in addition to the synthetiser crystal. Lou

Hi Dody What you intend to do is extremely complex and difficult! First, in order to detect simultaneously several frequencies in real time, you have to use some form of convolution process: a Fourier transform, f.e.. But , that's not enough, because each single vehicle will produce multiple frequencies, mainly harmonics of the motor's rpm. Therefore, in the mass of frequencies that are collected and identified, you have to find groups of harmonically related frequencies (that is frequencies whose respective phases remain stable in time). Once you've sorted all the frequencies present in groups, you're able to determine the number of vehicles by counting the number of groups. It's a computation intensive application for a DSP circuit... Good luck anyway!

Hi I think you'll need three detectors: one barrier type, detecting any ball including the black one and two reflective type. One will have a red LED as a source and will react to white and red balls, the other will have a green LED as a source and will only react to the white ball. By combining the informations from the three sensors, the program will be able to determine the color of the ball. Lou

Hi Here is something that's probably closer to what you're looking for: http://microsys6.engr.utk.edu/ece/bouldin_courses/652/tutor6.pdf Lou

Hi Robert I think it is perfectly possible to reduce the ouput voltage of these supplies to 12V, but you won't succeed by overloading them with a bleeding resistor. What you have to do is to tamper with the voltage feedback servo loop. I don't know specifically these supplies, but you should find the feedback circuit on the secondary side, probably feeding an optocoupler linking the secondary to the primary. The regulating circuit takes a sample of the output voltage with a resistive divider; try to locate it and reduce the value of the resistor going to the supply output (not the one to the ground). This should reduce the voltage. Make a first try by lowering it of 20%; this won't destroy anything and will let you know if you have found the right spot. If it's OK find the value that gives you 12V. If there's no change, try somewhere else. Maybe there is already a trimmer on this spot. All this will work for 12V but probably not for much lower voltages; you'll be too far off the design values. Lou

Hi You can try connecting one side of your NE-2 to ground via a 100pF capacitor and use the other end to probe various points of the transmitter (maybe the antenna) until you find the right spot. Lou

Hi A frequently used solution for this type of problem is a fast, free-running counter (either in hard or soft form) that is read when you need your random number. Although it isn't a true random number generator, it is generally sufficient because the statistical repartition is OK and you have no way of knowing or influencing what the next number will be. Lou

Hi Have a look at this page : they have simpler, cheaper circuits and the transformers are fully specified. http://www.zetex.com/3.0/3-8-1a.asp Lou

Thanks Oli and MP, but what I want is precisely to avoid a separate log converter; I know that the DVM chip is able to implicitly perform the conversion, but I don't remember the trick. Lou

No, it's digital: the ADC displays dB's instead of mV, and I did obviously search the web as well as specific resources, but without results.

Hi Here are some suggestions: - Use the video modulator IC of a VCR recording circuit. You may have to adapt slightly the component values because in a VCR, luma and chroma are recorded separately, but it should be OK. In addition you can also use its playback counterpart for demodulation. - Use a VCO IC of the 74S124 family. - Use the VCO section of a high frequency PLL such as the NE564. If you use the same circuit as a demodulator, you will compensate most of the systematic non-linearites. Good luck

Hi You could use a rotating transformer salvaged from the head assembly of an old VCR. You could feed the primary with a 250Kc power square-wave and rectify the secondary output with Schottky diodes. I guess you'd be able to transfer a fair amount of watts in this way.

Hi I'm afraid a P-channel JFet is essential in this circuit: a JFet is a depletion device, conducting in the absence of any control voltage and I see no other device capable of that (except a P channel depletion MOS, which is even harder to find). Using another method would require an in-depth redesign of the circuit. A 2N3820 might be suitable despite its higher resistance, and is probably easier to find than the original types. Maybe you could try... Good luck

Hello Many years ago I've seen an application of a converter IC (ICL7106 style) displaying a dB output without any explicit log converter. The circuit was based on a special connection of the input, reference and integrating pins, but I do'nt quite remember how it worked or where it was published. Does anybody have details about such a circuit? Thanks