Maker Pro
Maker Pro

10MHz peak measurement

I have the simple problem of trying build a circuit that turns a
~8-15Mhz, ~100mV-2V sine wave into a DC level (for read out by an ADC).
I've tried just rectifying and putting an RC to hold the DC level, but
I don't seem to get a flat response over the specified frequency
range. The sine wave will change amplitude over seconds, but no faster
than that. Any ideas, or am I just not good enough at making
rectifiers with caps (which is very possible).

Jesse
 
M

Mark Jones

Jan 1, 1970
0
I have the simple problem of trying build a circuit that turns a
~8-15Mhz, ~100mV-2V sine wave into a DC level (for read out by an ADC).
I've tried just rectifying and putting an RC to hold the DC level, but
I don't seem to get a flat response over the specified frequency
range. The sine wave will change amplitude over seconds, but no faster
than that. Any ideas, or am I just not good enough at making
rectifiers with caps (which is very possible).

Jesse


Hi Jesse. What rectifier are you using? At these frequencies, the
diode switching speed may be an issue. And what value and type of cap?
Some caps are good for high frequencies and some aren't. Are you using
a resistor in series to the cap? If not, the instantaneous current is
probably limiting the effective "smoothing." It sounds like an RC
filter might be the key. Or even better an op-amp precision rectifier,
do a search for "TI Op Amps for Everybody" and you should find an
excellent TI website and resource. This would probably work better but
be more complex.

Perhaps you could try a rectifier and 4th-order RC filter network,
like this:

AC in D1 100k 100k 100k 100k DC out
<-------|>|---\/\/--+--\/\/--+--\/\/--+--\/\/--+----->
FAST |+ |+ |+ |+ (buffer with a
--- --- --- --- unity-gain op amp)
--- --- --- ---
| 10nF | 10nF | 10nF | 10nF
| | | |
----- ----- ----- -----
--- --- --- ---
- - - -

You could feed this a 1kHz 50% duty cycle squarewave signal and get
a nice, smooth DC voltage out. Exact values and methodology left to
experimentation.

-M
 
M

Mark Jones

Jan 1, 1970
0
Mark said:
Hi Jesse. What rectifier are you using? At these frequencies, the diode
switching speed may be an issue. And what value and type of cap? Some
caps are good for high frequencies and some aren't. Are you using a
resistor in series to the cap? If not, the instantaneous current is
probably limiting the effective "smoothing." It sounds like an RC filter
might be the key. Or even better an op-amp precision rectifier, do a
search for "TI Op Amps for Everybody" and you should find an excellent
TI website and resource. This would probably work better but be more
complex.

Perhaps you could try a rectifier and 4th-order RC filter network, like
this:

AC in D1 100k 100k 100k 100k DC out
<-------|>|---\/\/--+--\/\/--+--\/\/--+--\/\/--+----->
FAST |+ |+ |+ |+ (buffer with a
--- --- --- --- unity-gain op amp)
--- --- --- ---
| 10nF | 10nF | 10nF | 10nF
| | | |
----- ----- ----- -----
--- --- --- ---
- - - -

You could feed this a 1kHz 50% duty cycle squarewave signal and get a
nice, smooth DC voltage out. Exact values and methodology left to
experimentation.

-M


Oh and I forgot to mention, that the voltage of an AC waveform will
change as it is rectified into DC. For instance, if you rectify a
12vAC signal into DC, it would be pulsating DC with a peak of about 17
volts. AC is termed as the RMS (root-mean-square) voltage measurement
of a sinusoidal waveform. 12vAC is 17vPP (peak to peak.)

See http://www.tpub.com/content/neets/14179/css/14179_175.htm for
more info.

That said, it is possible to perform mathematical functions by using
only op amps...

-M
 
Y

Yukio

Jan 1, 1970
0
You are describing the function of a AM RF detector operating ~ 10 MHZ , a
little pickier than a garden variety 455 KC AM Detector . Your requirements
for a Wide-Bandwidth is contrary to several generations of Ultra narrow band
designs, so you must back-engineer any modern circuit to increase the
bandwidth. Things like TRF amplifiers come to mind as opposed to HiQ IF
amplifiers. Just build an AM detector that will demodulate the local AM
station and go from there. Most OP Amps I've seen quit around 100KC so you
would have to find one that would work to around 20 MHZ . Diode Detectors
can be found that will work all the way into the GigaHertz range, the output
of which should in the ~ subsonic range ie 10's of seconds per Cycle, duck
soup for standard garden variety OP amps. Would a Voltage controlled
Amplifier help ? ie take the error Voltage from a PLL circuit monitoring
the Incoming Signal Frequency to adjust the gain of the Output amplifier to
track the frequency shift and flatten the output response

Yukio VE5YS
 
D

David L. Jones

Jan 1, 1970
0
I have the simple problem of trying build a circuit that turns a
~8-15Mhz, ~100mV-2V sine wave into a DC level (for read out by an ADC).
I've tried just rectifying and putting an RC to hold the DC level, but
I don't seem to get a flat response over the specified frequency
range. The sine wave will change amplitude over seconds, but no faster
than that. Any ideas, or am I just not good enough at making
rectifiers with caps (which is very possible).

Jesse

See this thread of mine from a few years back:
http://www.google.com.au/groups?hl=...lm=d9sC5IC25PE8Ew5m%40jmwa.demon.co.uk&rnum=2
(hope the link works)

I used an AD8037 high speed clamping amplifier in a novel full-wave
precision rectifier configuration from an AD app note. This gives a
precision peak detector which converts the sine wave into a DC value
equal to the peak of the waveform. It actually worked to more than
20MHz over a voltage range similar to yours if memory serves me
correctly.

Dave :)
 
Top