1st time I need exact V gain... Check my calculations?

ExpAnonColin · December 30, 2003, 06:35:48 PM

I have a square wave at .2v, and I want it to be around 10. So, I take an op amp, and make the gain 50 (.2*50=10), so I have an initial resistor of 1k and a resistor in the feedback loop at 47k (9.4v would be OK too). Then, for the inverting input I have -15v, the Vs 15v, and the Ground 0v. Will this work? It's the first time I need an exact amount of voltage gain.

-Colin

Joep · December 30, 2003, 07:16:37 PM

Yes, this will get you there, I think you are using an inverting amplifier right? Then the gain is -R2/R1 (where R2=47k and R1=1K)

Looks like this:
http://www.ecircuitcenter.com/Circuits/opinv/opinv.htm

In a non-inverting the gain is 1+(R2/R1), like this:
http://www.ecircuitcenter.com/Circuits/opnon/opnon.htm

Hope this help.

Joep

PS to get it more precise use two 100k parallel for R2.

ExpAnonColin · December 30, 2003, 07:27:33 PM

Right, that's what I was using-but I was planning on using the noninverting input (I don't want a gain of -50v

)

-Colin

Jason Stout · December 30, 2003, 07:32:08 PM

A gain of -50 just means that the amplitude of the signal is 50X greater and the phase of the signal is 180 degrees out. Inverted signal, gain of 50.

ExpAnonColin · December 30, 2003, 08:35:10 PM

Quote from: Jason StoutA gain of -50 just means that the amplitude of the signal is 50X greater and the phase of the signal is 180 degrees out. Inverted signal, gain of 50.

Oh. I didn't realize that the - meant inverted in terms of gain.

-Clin

Joep · December 30, 2003, 08:43:55 PM

Yeah, that's a bit confusing, but a - in gain means 180deg out of phase. Absolutely the signal will be 50 times greater. You can use both if phase doesn't matter.

Did you noticed the difference in gain for inverting and non-inverting. You have to add 1 to the gain with a non-inverting circuit. This does only matter if the ratio between R2 and R1 is little.

Good luck!

Joep

R.G. · December 30, 2003, 09:28:55 PM

Any time you need exactly ...something... you have to start thinking about tolerances. Lets play with some numbers.

First, this 9.4V to 10V you want out - is that peak, or peak-to-peak?

And that 0.2V in - how close to 0.20000V is it, exactly?

If you determined that you needed 47K and 1K for the gain, you get a gain of 1+(47/1) = 48, and then a 0.2V input gives you 9.6V. Fine so far.

But the 47K isn't perfect. For ordinary 5% resistors it can be +/- 5% in value, or from 44.65K to 49.35K. Likewise, the "1K" can be from 950 to 1050 ohms. So the gain can be ( and will be, for some resistors) from 1+(44.65/1050) = 43.5 to
1+(49.35K/950) = 52.95. There is actually a tolerance on that "1" in those equations, but it has to do with the gain of the opamp and it's truly insignificant at low frequencies.

So you get 8.7V to 10.38V if your 0.2V is perfect, and has no tolerances itself.

So if your "10V" level matters, you really need to use some carefully chosen 1% resistors, or if your 0.2V input varies and you really, really need 10V, you probably have to resort to a trimmer pot.

The study of tolerance changes in circuits is almost as dismal as economics. True worst-case design of circuits is part of what is responsible for the huge costs of military electronics. It's really expensive to make sure that for any possible combination of parts, the outputs will all be in the specified range.

ExpAnonColin · December 30, 2003, 11:40:39 PM

RG-it doesn't need to be perfect, it needs to be withing 1v. It's about .2v, I'm looking at it with an oscilloscope. I guess I shouldn't have put "exact" in the title, I should have put this is the first time I have to worry about the amount of gain I'm doing because I want it to be within 1 v. Thanks for the calculations, and I'll definitely take these things into consideration if it doesn't work out.

Could someone verify with me that I should be putting gnd into the inverting, -15v into the -Vs, and 15v into the V?

-Colin

ExpAnonColin · December 31, 2003, 12:57:54 AM

Wait, got it working

Turns out I needed only a gain of 2.2, I was reading my oscope wrong. Silly me.

-Colin

Brian Marshall · December 31, 2003, 02:52:28 AM

I love opamps, i always use them even when i probably shouldnt. I have yet to design anything original with fets or bipolars, although i did build a fuzz face with some extra knobs a long time ago, and tweeked it a bit.

Anyways.... If you want a 10 volts, i assume you mean RMS voltage right... or even peak to peak doesnt really matter if it is a square wave. you will need a 20 volt power supply at least... probably more like 25 volts. maybe you just mean 10volts from peak to peak...

are you making a practice amp or something?

ExpAnonColin · December 31, 2003, 03:27:07 AM

Brian-I'm boosting a 8038's square wave to the triangle's frequency in an EHX frequency analyzer.

-Colin

Brian Marshall · December 31, 2003, 08:05:51 PM

hmmm so you are going to get yours done before mine... i dont know the frequency analyzer at all.... what does it do with a triangle wave... does it have an LFO? Run off of AC or some higher voltage?

I know that it's a ring mod that's about it. Just seems wierd to me that there is that much voltage going on in that pedal.

Brian

ExpAnonColin · December 31, 2003, 08:08:49 PM

Quote from: Brian Marshallhmmm so you are going to get yours done before mine... i dont know the frequency analyzer at all.... what does it do with a triangle wave... does it have an LFO? Run off of AC or some higher voltage?

I know that it's a ring mod that's about it. Just seems wierd to me that there is that much voltage going on in that pedal.

Brian

Power supply is 40v, and I made some mistakes in my original calculations

Check out http://experimentalistsanonymous.com/mods , I'm all done.

-Colin

Brian Marshall · December 31, 2003, 09:12:59 PM

bastard!!!!

oh well... i'm still working on my kitchen... i cant f'ing wait until this is over, and i can get back to work on some stuff.

I have mine on paper, and it is opamp based, but i'm thinking about tabling it for now. Of all my ideas this one was the most complex, and I'm all about instant gratification right now.

Brian Marshall · December 31, 2003, 09:14:37 PM

oh wait now i get it.... the triangle wave is the modulation frequency right... so you moded it to be square. right?

ExpAnonColin · December 31, 2003, 09:14:59 PM

Quote from: Brian Marshallbastard!!!!

oh well... i'm still working on my kitchen... i cant f'ing wait until this is over, and i can get back to work on some stuff.

I have mine on paper, and it is opamp based, but i'm thinking about tabling it for now. Of all my ideas this one was the most complex, and I'm all about instant gratification right now.

Hey, Brian, I'd like to talk to you via email about an idea of mine. Send it over to colin@experimentalistsanonymous.com

-Colin

Brian Marshall · January 01, 2004, 03:10:45 AM

check your email...

I finally got some space in my garage, and hacked together a single stage jfet overdrive. Not bad... first time i have ever used a jfet as a distorter.

used a 2n5457 that i didnt even know i had. wooohoo!!!

ExpAnonColin · January 01, 2004, 03:49:15 AM

Quote from: Brian Marshallcheck your email...

I finally got some space in my garage, and hacked together a single stage jfet overdrive. Not bad... first time i have ever used a jfet as a distorter.

used a 2n5457 that i didnt even know i had. wooohoo!!!

I emailed you back!

-Colin

News:

1st time I need exact V gain... Check my calculations?