I need a basic wideband linear amp for my lab function generator. Hoping for an output of up to 24Vpp at 50mA. Can anyone point me to a suitable circuit?
Thank you,
Mark Ferris
DC-20MHz amp ckt.
Dec 28, 2007
14 Replies
12V and 50mA implies a load resistance of no lower than 240 ohms. Lab generators commonly have a 50-ohm series output resistor, to accurately drive 50-ohm coax cables, etc. So, if the generator is loaded with 50 ohms, the output amplifier sees a 100-ohm load, or 120mA for 12V. But if the generator output is shorted, as it often is, even if for a short while, the amplifier will see a 240mA load. So I'm not sure your 50mA spec is sufficient.
No.
Since 24VPP is ~ 8.5VRMS, then if the amplifier is a true voltage
source:
E 8.5VRMS
R = --- = --------- = 170 ohms
I 0.05A
What has RMS got to do with anything? The OP specified the performance in terms of peak-peak voltage. So it is the the peak voltage that matters.
[...]
John Devereux
A THS3001, running from +-15 volt supplies, could do that. Maybe you could talk TI out of an eval board.
John
Many thanks for your suggestion. Looks like a versatile chip. According to the data sheet, the closed-loop output impedance varies with frequency from 0.1 to 20 Ohms. One of my applications necessitates a 1K Ohm impedance match averaged across the 100Hz to
1MHz range. In terms of design prinicples, what would be the simplest way of achieving this?Mark Ferris
Mark Ferris schrieb:
Maybe a xDSL driver? TI makes horrible fast devices.
- Henry
www.ehydra.dyndns.info
It's less than 0.1 ohms up to your 1MHz! It's still only 1 ohm at
20MHz.
A 1k resistor fed from the chip output? But are you sure you want this?
John Devereux
Everything.
I think it is reasonable to assume the 50mA is the current at the peak. If it's not, then "all bets are off" anyway - it's explicitly a
*function* generator, so you definitely can't assume a sine wave. It seems perverse to take the 50mA as being rms, when the OP is clearly thinking in terms of peak output.Perhaps Mark will confirm.
John Devereux
OK, here's a slightly different tack.
With all the DC-20MHz function generators in use, why hasn't someone designed a general purpose buffer/waveform amp to suit? Something basic on a PCB, +/- 12V supply, usable output current and impedance, etc. IOW something other than the lab grade, multi-thousand dollar things. I know why they are expensive, but not everyone needs it for home bench use.
What other choices do we have re: a design tradeoff between practicality (cost effectiveness) and features. A schematic, a kit ... anything.
Mark Ferris
Discrete design isn't very hard. 2N3904's and 06's wull suffice, plus maybe some power transistors if you'd like the output class A.
Tim
-- Deep Fryer: A very philosophical monk. Website @
formatting link
A cheap one? Because there\'s no money in it.
I agree.
50mA is inadequate for a function generator anyway, whatever the definition. It may serve fine for an output impedance of 1k, with no load, but sooner or later, sooner I think, it won't serve well at all.
Geez, if 50mA is inadequate for a function generator, just think how
much more inadequate it would be for an amplifier strapped to its
output.
Perhaps what the poor, confused, OP is really looking for is an
attenuator ?
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required