Help: Bias Problem of Simple Dual Opamp Boost

[thehallofshields]

So I'm building a Beavis Board type of thing, and I wanted to save myself the Breadboard space of input and output buffers, so I drafted up a boost circuit using a TL072 that looks like this:



But once I built it, I couldn't get the damn thing to Bias correctly. Redid my joints, tried 3 TL072's, and I've still got way high voltage at each output.



Pin 1 - 8.89
Pin 2 - 1.32
Pin 3 - 3.87
Pin 4 - 0.00
Pin 5 - 3.87
Pin 6 - 3.87
Pin 7 - 8.89
Pin 8 - 9.50

Maybe someone experienced could take a gander at what might be happening.

[aron]

Are you sure you have the feedback loops going to ground? I have no idea why pin 1 and pin 7 are so high.

[thehallofshields]

Yes. I'm reading the correct resistances from the Switch Commons to the Ground Rail.

I think my electrolytics are all oriented correctly.

I think I should be getting a little less Voltage Drop across the 470k resistors. My Voltage Divider is about 4.8 so they are dropping about 1V.

[MaxPower]

Something looks strange to me.

http://www.electronics-tutorials.ws/opamp/opamp_3.html

Note how Rf and R2 are wired? Compare it to how R7 and R8 vs Var1 and Var2 are wired in your schematic. Maybe that's the problem?

[aron]

What I usually do is pull the IC out of the socket and then measure the socket pins and see if you are getting the correct voltages at the right pins. See if there's a short or something.

[Seljer]

In theory, practically no current should flow through the 470k resistors, it can't flow through through the 1uF caps and the input impedance of the TL072 is in the range of gigaohms to teraohms.

In practice, 470k is a large enough resistor that if you have a cheaper multimeter, the meter's own internal impedance (if its only a couple of megaohms)  pulls it down enough to throw off the reading by a full volt or so. So you have to keep that in mind when measuring a live circuit with high value resistances.


edit: also, I'd check the feedback resistors/pots, if the output voltage is high and the negative input is low that means theres a DC current flowing through there somehow, and again, it can't flow through the capacitor and the input impedance of the opamp is really high so it can't be flowing into there, either meaning the feedpack path is broken or something is shorted and letting DC pass to somewhere.

I concur on aron's suggestion of measuring the voltage without the IC in the socket if possible, and perhaps also measuring the resistances directly from the socket pins (including resistance to ground from the negative input/output and from the output to the negative input)

[R.G.]

What is the purpose of R7 and R8?

[duck_arse]

^ what he said, only later.

[GibsonGM]

OT, but what program did you use to draft the schematic?   I've seen "that look" around for years, and kinda like it but never pursued it.  Sorry for OT question; LT Spice is awesome, but doesn't lend itself to a 'drafted' look as well....

[thehallofshields]

What is the purpose of R7 and R8?

I was trying to get a Gain Range as close to 1-10 & 1-20 as possible. The idea being that the input gain is very critical in distortion circuits and I want to be able to fine tune the gain to get the most useful sweep out of a conventional input volume pot.

[thehallofshields]

OT, but what program did you use to draft the schematic?   I've seen "that look" around for years, and kinda like it but never pursued it.  Sorry for OT question; LT Spice is awesome, but doesn't lend itself to a 'drafted' look as well....

This is a free program called Circuitmaker. I don't think that it can run the complex simulations like Spice can, but its easy to use and comes with Potentiometers and stuff that LTspice needs imported.

[thehallofshields]

Something looks strange to me.

http://www.electronics-tutorials.ws/opamp/opamp_3.html

Note how Rf and R2 are wired? Compare it to how R7 and R8 vs Var1 and Var2 are wired in your schematic. Maybe that's the problem?

Well, I don't think the 3rd Resistor to add series resistance should be a problem. Its just like (251/13.5) + 1 and (251/28.5) +1. It was just numbers play to get close to 20 and 10. I'll jump it with alligator clips and see if anything changes.

[thehallofshields]

What I usually do is pull the IC out of the socket and then measure the socket pins and see if you are getting the correct voltages at the right pins. See if there's a short or something.

I don't know why i didn't do that. I'll try this out tomorrow.

[slacker]

The 3rd resistors are fine except that they're in the wrong place. Look where Rin is connected to the opamp in the link Max Power gave. The equivalent of Rin in your schematic is the C5/R9/R11 combo.

[PRR]

A simulator may be handy to check that gain IS 9.82.... but that's not the problem.

Any busy circuit builder "should" (IMHO) be able to finger-check that drawing and verify what it should be doing. For DC, both amps are unity-gain, biased toward half of 9V, and output pins should sit at 4.5V.

A glance at the voltages suggests U1's C5 is acting like a short. (Would help to know what Var1's value was at the moment.) U2's problem is unclear. The fact they both have the "same"(?) problem suggests a common error.

The + In pin will always read low due to meter loading. Check that it isn't dead-zero or something else totally bogus, but if Out were sitting at the right place then +In is at the right place except when you meter-poke it.

Also wonder what the 1.5K do. They are fine if you need a jumper, but are a distraction when wondering what is wrong.

Since it is wrong, and still in prototype, simplify. Yank C5 so you KNOW(?) that leg to ground is open (unity gain). Jump Var1 (pots are not 100% reliable) with something small, say 5K. Examine the pointless 1.5K, maybe take the NFB directly to the +In pin just to see if there are bad contacts.

> Redid my joints

As a general thing, I don't like this "fix". If you made a bad joint already, you can make a bad joint again. I think you should LOOK at the joints. Remove blobby solder. See that the solder "wets" both lead and pad/lug, and is smooth between.

[GibsonGM]

OT, but what program did you use to draft the schematic?   I've seen "that look" around for years, and kinda like it but never pursued it.  Sorry for OT question; LT Spice is awesome, but doesn't lend itself to a 'drafted' look as well....

This is a free program called Circuitmaker. I don't think that it can run the complex simulations like Spice can, but its easy to use and comes with Potentiometers and stuff that LTspice needs imported.

Thanks...just looks 'groovier' than LT Spice, ha ha.  I'll check it out.

[thehallofshields]

Okay. Got around to pulling out the IC.

Socket 1 - 0.00
Socket 2 - 0.02
Socket 3 - 3.90
Socket 4 - 0.00
Socket 5 - 3.90
Socket 6 - 3.90
Socket 7 - 0.00
Socket 8 - 9.50

I'll have to inspect whats going on between 5 and 6.

[thehallofshields]

All right. Voltages are still way off, I can't seem to figure it out.

Does anyone have any last minute suggestions before I trash this board and start over?

[antonis]

In case of use a Veroboard check that the cut between pin 3 & 6 is clear enough...

(not by just getting a glimpse of the cut..)

P.S.
I have no intention to belittle you but sometimes we tend to assert the same result despite the number of checks..

[thehallofshields]

Just a general question: With Dual Opamps does the Bias of one side of the Opamp affect the other? Is DC offset from Pin1-3 affect Pin5-7?