The AM Forum

Not sure if the subject is the right term for what I'm trying to do but here's what I want to do.

I'm trying to take a small DC voltage derived from an IF amp and bump up the rectified Vdc by 10x then through a voltage divider to drive a small volt meter to be an S-meter for the Retro75.

I want to amplify or convert a 0 to 0.5v voltage to 0 to 5.0v voltage (very low current) using a single power supply in this case the 12v used to power the Retro75.  It doesn't matter if it's inverting on non-inverting amp because I can simply change the polarity of the input voltage by flipping the rectifier diode at the Retro's IF amp.  I'm thinking an op-amp won't work or at least the configs I've tried, because the output or offset will be essentially 1/2 the rail supply voltage when the input is 0v when using a single power supply.  Any ideas or thoughts on how I can do this with a minimal amount of parts? I've tried a couple of op-amp configs. but no go.  Need some inpoot.

Why not a simple transistor(s) voltage amplifier?

Sounds like you want to peak charge a cap to drive the meter. Add a series diode and a cap after the op amp. The meter coil will discharge the cap. Using just an op amp the meter average voltage lands around zero. A peak charge rectifier will block the negative swing of the op amp. The cap value will depent on the meter resistance and sensitivity.

This is like trying to gin up an S meter for a rush box. IF you just rectify the IF, or simply take the DC term from the detector (top of C15), you will be quite disappointed in the performance of the meter.  Most S meters take off a voltage from AGC controlled stage(s) in the radio. It might be the AGC voltage itself, or screen voltage of an IF stage. The point is that it is in a control loop and the voltage will not be linear since the AGC function is not. Without too much detail, the AGC voltage is controlling the gain of multiple stages and the effect is that the higher the level, the less change in AGC voltage to maintain about the same output. this make the meter scale expanded at the low end and compressed at the top and thus easy to calibrate in pseudo dB.

If you want to do what you are contemplating, besides amplifying the voltage, you will need a log function. how about a log amp to start with. look at some samples in national semi's application notes.

Does the Retro75 have an AGC loop?  Sample that.  You'll get a more meaningful reading, and that's how most S meters work anyway  8)

I haven't seen a schematic - is one available?

It may be possible to get rid of the apparent 1/2 B+ offset by returning the meter to the wiper of a pot connected from + to GND.
Adjust the pot to zero the meter. The pot will be set, then at the same voltage as the op-amp output, about 1/2 rail if that was what is wanted. The meter will have to then go to full scale at half the + voltage, but that should be relatively easy to do by adjusting the series multiplier resistance on the meter to make it as sensitive as necessary.

A shows that. It is a bridge.

B may be better in practice and the zero balance pot is before the op-amp. In practice, the 1K balance pot would have a resistor above and below it to restrict its range to the general area (about +/-0.25V of the nominal value) needed because it will be very touchy if given to the full 0 to B+ range. Use a rail to rail input and output op amp for best result. The 100K and 10K resistors are like any op amp circuit to set the gain in this case to 10. The 20K resistor makes the 100uA meter have about a 2V full scale response. Any values can be changed to suit the operating voltage (5V in this case), my only intent was to show how to deal with the DC offset that a single supply has. Omitted are any capacitors and the like to modify the response time and bypass unwanted higher frequency signals. I think I drew it correctly but it is late, corrections welcome.

Hi Pat,
I was able to get something working and lo and behold we essentially came up with similar circuits. I'm going to try yours this weekend. Thanks.

Steve, no agc loop. 
Bob