[off topic] opamp question

[McBain]

sorry for this off topic question, but i dunno where to ask (lemme know if u know better forums)

i'm making a funtion generator (which will be used also to test audio equipment, so mebbe still on topic . It is based on the MAX038 IC. The design is almost ready, but i have some problems with the output stage.


a proposal of the output stage to accompany this text is at http://mcbain.studentenweb.org/schema.jpg

The MAX delivers 0...2Vpp (sine, square triangle). This is the in input.
I want 2 outputs
1) a high voltage ac with dc offset
2) a low voltage "pure" ac output

for (1) i feed te innput signal thru a pot (to change amplitide ) in an opamp summer. The other source for the summer is a dc voltage. Dc is not amplified, ac is 12 times amplified. So i should arrive at a voltage of 0 ... 24 Vpp adjustable by the amplitude pot added with an independant adjustable (?) dc offset of -15 ... +15. This signal is fed thrue a unity gain buffer with large output current (to be able to drive 50 ohm bnc cables)

for (2) i just need the input signal, feeded thru an unity gain buffer. I guess 2Vpp is enough for audio testing (or isn't it ? )

i have problems combining the 2 circuits. The DC offset cant arrive at the MAX output pin (i thing the IC wont like that) and the dc offset cant arrive either at the the low ac output, so i added an extra 2 buffers.

i really dont know if this will work or if there are other methods (i dont want to use a switch, and a C to block DC wont pass the 1 Hz ... 20 Mhz range). I'd like your opinion. And please dont noobize me here, i just wanna learn something.

another last thing, is 2Vpp enoug for audio testin ? otherwise i would add an extra opamp to amplify the signal (to example 4 V pp). If i do so, can i replace the opamp buffer with a non inverting amplifyer or do i put the extra amplifier behind the buffer ?

best regards,
McBain

 

[jan.didden]

waveform generator

McBain (Ed?),

Don't use a summer. Feed the AC to the neg input, DC to pos input, scale/gain appropriately etc. Should do the trick.

Jan Didden


PS "Lemme, dunno, mebbe"? What kind of dictionary do you have in Belgium?

 

[tiroth]

Neat IC. Square wave isn't very..mmm...square...at 20MHz. Looks like a nice easy function generator, but with minimum 0.75% distortion will have limitations for audio testing.

 

[jan.didden]

waveform gen

Ed,

Just reread your post, I asume you want the two outputs separately available, isn't it?

Jan Didden

 

[McBain]

neat ic

neat ic indead

and no, at 20 Mhz the square isnt very square anymore

but take a look at the commercial function generators, the cost a couple of $100's and only get to a couple of Mhz's (seems the MAX038 does very good up to 10 Mhz and above)

if you guys are interrested, i'll set up a webpage about it, with pcb's, when the project is finished

the gen will feature
* waveform: tri, square, sine
* freq 1 Hz ... 20 Mhz
- range (3 decades for the moment)
- course
- fine
* duty cycle adjustment (with this it is possible to get a more pure sine waveform) with defeat switch
* output 1: 0...24vpp adjustable+ -12...12 adjustable dc offset
* output 2: 0...2 vpp adjustable only ac
* ttl output

this on a board of about 10 * 8 cm (4 * 3.5 inch)

cost: $20 for the max, a quad opamp IC, a couple of high speed opamps, 2 unity gain buffers, some pots, switches and R's and C's

so nowhere near the cost of the commercial (mostly less good) function generators

best regards,
Ares Lagae

 

[McBain]

janneman, indead
but wont the inverting amp instead of the summer solve the dc offset problem on the max output pin and the other (low ac) output ?

 

[McBain]

will this solve the problem:

http://mcbain.studentenweb.org/schema2.jpg

or do i need to add an extre buffer between the connection to the buffer for the low ac output and before the differential opamp ?

 

[tiroth]

Your post caught my eye b/c I just recently "discovered" the '038 and was considering whipping up a generator. One thing that might be considered is driving a comparator for the square wave output. Done properly, you might get near 40MHz of clean square wave output.

 

[jan.didden]

waveform gen

will this solve the problem:

http://mcbain.studentenweb.org/schema2.jpg

or do i need to add an extre buffer between the connection to the buffer for the low ac output and before the differential opamp ?

Well, that depends on the value of the amplitude pot wrt the input resistor of the diffamp. Because of the input resistance, the ampl pot may not be perfectly linear. On the other hand, if you plan to use output level metering that is of no importance. On the gripping hand, if you have an opamp buffer to spare put one after the ampl pot.

My Heathkit waveform generator has a switch on the ampl pot with which you can defeat the offsett (just groung the DC input to the diff amp). I found that very usefull. There is only one output so you don't have to switch cables all the time, you can leave the DC pot as it is and quickly test with or without offset. It also makes the circuit a bit more simple. I strongly recommend you do something similar.


Jan Didden

"Things should be made as simple as possible, but not simpler"
ALbert Einstein

 

[Havoc]

Easiest thing to square up the output is by putting a FF in series, but this will halve the output frequency. This if it is important for you....

 

[joeblutz]

Mr. McBain,

Don't worry about the DC offset voltage appearing at the input of your circuit. As long as the OPA30 summer does not saturate it will keep its inverting input at essentially zero volts regardless of the offset voltage you select. Remember the inverter operates by maintaining a virtual ground at its inverting input.

For the same reason, no DC will appear at your second "pure" output. the OPA30 buffers aren't needed.

LOL

Joe