[Garbz]

The opamp will amplifiy differences between it's inputs, not in relation to a ground. The problem is the output voltage would have no reference and thus DC on it. Both of the rails should have a resistor to ground on the input that give them each a voltage drop related to ground.

 

[SnoopyRocks]

Quote:

Originally Posted by t52 here's my approch to i/v conversion: some explanation: the i/v conversion resistors r1 and r2 and c1 form a first order lowpass tuned ad 80khz (hope the values are correct, did some quick calc only), followed by another two lowpasses (r4/c3 nd r3/c2). then i use two opamps as buffers, followed by a summing opamp. i think this config could be quite promising as we have a second order lowpass BEFORE all opams, thus removing a lot of the high freq spectrum and lowering the opamp's slew rate requirements... would like to hear some comments, regardles of ez wants to use this design (it's his dac after all ) of course the filters could need some tuning for pulse response by simulating this circuit - i don't have a sim software though. can anybody recommend some free software?

Your circuit won't work as you desire. First of all, I/V is entirely accomplished by R1 and R2. Since they are 25 Ohms, the DAC must output 40mA peak to get 1V peak output swing. Can it really supply this much? I would be surprised if it could. The output DC voltage is set by the DC output current of the DAC into R1 and R2, as well as OpAmp offset voltages. More importantly, the summing amplifier will not work because of the extra feedback resistor (R7) to the positive input, which causes instability due to positive feedback. Even with this removed, the summing function won't be symmetric since the gain from the positive side is 1+R8/R6, while the inverting input will be -R8/R6. (i.e. gains of +7.67 and -6.67) An extra resistor, sized appropriately, from the positve input to gound is needed to fix this. IMO, your circuit needs some work.

Virtual ground confusion:
What is meant by calling an OpAmp input a virtual ground is that the amplifier forces it to a DC voltage (analog AC ground), while giving the node a low impedance. It does not mean that it is absolute ground.

 

[t52]

Quote:

Originally Posted by SnoopyRocks Your circuit won't work as you desire. First of all, I/V is entirely accomplished by R1 and R2. Since they are 25 Ohms, the DAC must output 40mA peak to get 1V peak output swing. Can it really supply this much? I would be surprised if it could. The output DC voltage is set by the DC output current of the DAC into R1 and R2, as well as OpAmp offset voltages. More importantly, the summing amplifier will not work because of the extra feedback resistor (R7) to the positive input, which causes instability due to positive feedback.

you're right, this was a drawing mistake as i was in a hurry. right side of r8 has to be connected to ground. i also increased the resistor values of the subtracting amplifier to reduce the load on the buffers before them.
here's the new schematic:

regarding output swing: the pcm1794 puts out +/-7.8mA peak, that would make +/-195mV peak, summed 390mVpp or 138mV rms, the following subtraction amp has a gain of 6.7, so the output swing would be 0.92v rms.

Quote:

Originally Posted by SnoopyRocks Even with this removed, the summing function won't be symmetric since the gain from the positive side is 1+R8/R6, while the inverting input will be -R8/R6. (i.e. gains of +7.67 and -6.67) An extra resistor, sized appropriately, from the positve input to gound is needed to fix this. IMO, your circuit needs some work.

hm, take this picture as a reference:

i connected the inverting input of the subtractor to ground to assume it's input is 0v all the time, now assume +1v on the positive input of the subtractor... the noninverting opamp input sees +0.5v divided by resistors r2 and r4.
to shift the inverting input to (almost) same potential the opamp has to output +1v - the inverting input now sees +0.5v also, so gain on both sides is 1, only differing in polarity... so i don't think your formula is for the gain is correct, they are correct only in single inverting/noninverting amps where the "other" input i s always referenced to ground.

 

[t52]

hey, i got news... just got an email from a friend who is a ee student and they have EWB available at the campus, so he could simulate ez's subtraction amp for me - he used 24 ohms resistors, think they were standard values and two current sources with +/-5ma, have a look at the results for yourself...

 

[ezkcdude]

I guess I failed to mention that I had modeled the circuit quite a while back using TINA-TI (which you can download freely from the TI website). That's how I chose the various values. For a peak-to-peak current of 7.8 mA with a 1 kHz sin wave, the RMS output is 1.58 V.

 

[t52]

could please anybody that's more expreienced in ee comment on the simulation image? i'm constantly wondering what is actually flawed - the implemetation of the i/v and differential summing amp or my friend's simulation results?!?

 

[ezkcdude]

Quote:

Originally Posted by t52 could please anybody that's more expreienced in ee comment on the simulation image? i'm constantly wondering what is actually flawed - the implemetation of the i/v and differential summing amp or my friend's simulation results?!?

You may be right. I just did a sim. If you change the I/V resistor on the inverting side to 35R (R24 or R30), then the voltages are within 1mV or so. 35R should be a common enough value, so maybe that's the way to go (but still 25R for the other side). I'm still of the opinion that it doesn't matter. If I'm right, though, it won't matter either way. If you're right, then we err on the side of caution. Still, I too would like to hear opinions from the peanut gallery.

 

[ezkcdude]

I've got almost all the parts to start building, and I should get my final order by Friday or Saturday. If I do, I will start Saturday, and I have three options for the op amps: AD8033, AD8065, and OPA134. What would you choose for this design? I'm leaning toward 8065, unless someone can sway my vote.

Whatever the outcome, I'll post some pics when I finish the build.

 

[ezkcdude]

I need some help! I finished the board today, and checked the regs first to see if the voltages were correct. It was looking good, as the positive rail was all o.k. 3.3, 5, and 12 were all o.k. and not overheating or anything. Sadly, the -12 reg is at 0V. It's not due to the power supply either, because that is reading +/- 15V. The only thing I can figure is that I burned the LM317 while I was soldering it to the board. Maybe I should have used a paper clip! So, any advice? How do you desolder a LM317? Short of doing that, I could use the negative rail from another +/-12V psu, but then I will have two different grounds going into the board. I should also say, I know it's not the design that is flawed, because one of the testers definitely had the negative rail working. Well, I'll say one thing for the board. It's pretty, and not too difficult to build, at least, compared to what I thought.

 

[Rescue Toaster]

Err, you used a LM337 for the -12v and a LM317 for the +5v right?

Because in the schematic on the page in your sig, they're wrong. If you put a 317 into U11's spot it'll die horribly.

 

[ezkcdude]

Quote:

Originally Posted by Rescue Toaster Err, you used a LM337 for the -12v and a LM317 for the +5v right? Because in the schematic on the page in your sig, they're wrong. If you put a 317 into U11's spot it'll die horribly.

Sorry, I meant LM337 for the -12V rail, not LM317. The schematic is wrong.