OP275 op-amp low end issue

[jojokeo]

I'd like to know what's up with OP275 op-amp chips compared to others??? I've not experienced what appears to be an overly high & detrimental signal sensitivity issue in using them before. (please read below)

I have a fairly simple dual opamp build going that was originally loosely based on a fairly simple & typical overdrive schem using the non-inverting inputs and a OP275 op-amp. The first stage uses a couple 1n4002 diodes 100pF cap w/ 10k & 470k pot (typical gain loop) with a 2k2 & .1uF cap and 6k8 w/ 1uF for bass response. This goes into a second stage similarly but very low gain using 4k7 w/ single LED & 1nF cap in gain loop and a 2k2 w/ 22uF off inverting stage for bass there but low gain 2nd stage.

Things were going well in tweaking over last weekend and I thought I was finished. I gave things a day for my ears & myself to refresh. Then the next day I plugged my single coil strat in and things were confirmed very nice how I wanted.

Then I plugged in a dual humbucker strat - and WTF? Picking the low E firmly on the initial attack I get a distorted (the bad kind) noise that lasts less than a second until the signal voltage settles just a bit from that firm pluck. It also does it slightly less on the A string also! I can adjust the Gain, Vol, and Presence (feedback loop) and it doesn't really affect this from happening.

I've made a few various adjustments to no avail. Then I put things back how I had it and then used a TL082, TL072, JRC4558, OPA2134, etc. and NONE of these have this low end anomaly like the OP275 does?! I've tried 3 or 4 other OP275s and they all exhibit the same low end distorted anomaly and poor behavior. All the other chips sound great and work "normally" - as you would expect.

Now I can use one of the others but I've spent days tailoring the circuit around the OP275 since it has nice gain & break-up characteristics along with the high end qualities I like for this circuit. Any ideas what this is, why it's happening, and what I can do to solve it?

Thanks for taking the time to read all of this,
Keo

[Transmogrifox]

I think you need to post a schematic of your circuit before any of us can make meaningful comments. 

A hint from the datasheet:

Output Voltage Phase Reversal
Since the OP275's input stage combines bipolar transistors for
low noise and p-channel JFETs for high speed performance, the
output voltage of the OP275 may exhibit phase reversal if either
of its inputs exceeds its negative common-mode input voltage.

With 9V PS, the internal 7.5V zeners won't keep you out of trouble here, so it's one possibility if one of the inputs is being driven toward ground.  Datasheet didn't state common mode input range that I could see, which is the missing link.  Probably about 1 or 2 diode drops.  Would guess if within 1V from ground it could be a problem.

Overload or overdrive recovery time of an operational amplifier
is the time required for the output voltage to recover to a rated
output voltage from a saturated condition. This recovery time
is important in applications where the amplifier must recover
quickly after a large abnormal transient event. The circuit shown
in Figure 10 was used to evaluate the OP275's overload recovery
time. The OP275 takes approximately 1.2 ms to recover to V
OUT
=
+10 V and approximately 1.5 μs to recover to V
OUT
= –10 V.

Another possibility.  If either op amp is being driven to the rail then it might be hanging on one polarity for up to 1ms causing some nasty noise.

An oscilloscope would help a lot in diagnosing this problem.

[jojokeo]

I think you need to post a schematic of your circuit before any of us can make meaningful comments. 

How do I upload on this site?. The things that you mention is ONLY for the OP275? Every other dual op-amp chip I've tried works fine in this circuit - albeit some sound and perform better than others.

[aron]

The read me before you post tells you a few ways you can upload a picture.

[R.G.]

The things that you mention is ONLY or the OP275? Every other dual op-amp chip works fine in this circuit.

The things that TMF mentions apply to every opamp. What he says, subtly, is "read the datasheet for your opamps".

I'm very, very certain that every other dual opamp chips does not work fine in this circuit, only the ones you've tried. Every opamp is different, and has quirks of its own. Phase reversal when the opamp inputs are stressed in some way is very common. What changes is how susceptible the chip is and to what condition.

There's a common acronym for the proper response here. It's "RTFM", standing for Read The Fibbling Manual. Or datasheet in this case.

Running into the need to read the datasheets is the first step in getting out of beginner electronics.

[jojokeo]

The things that TMF mentions apply to every opamp. What he says, subtly, is "read the datasheet for your opamps".

I'm very, very certain that every other dual opamp chips does not work fine in this circuit, only the ones you've tried. Every opamp is different, and has quirks of its own. Phase reversal when the opamp inputs are stressed in some way is very common. What changes is how susceptible the chip is and to what condition.

I am trying to better understand & learn this info from the datasheets but am challenged and why I'm here posting a question with my low end issue. I appreciate any & all help for you all. I wasn't sure if my issue with the OP275 was slew rate related or something else as I've not had something like this occur in the past.

*update - the OP275 doesn't work correctly after numerous biasing changes, bass response changes, signal input limiting to a degree (which is definitely not wanted), etc... all the other chips (and more) I've mentioned and tried all work and sound as they should. IF someone has anything relevant to say about this specific op-amp then great but apparently there's not much experience or knowledge of the nuances of this one?

[jojokeo]

Was hoping some may have similar experience and/or input on this particular op-amp?

[PRR]

> the OP275 doesn't work correctly after numerous.....

I think you need to post a schematic of your circuit before any of us can make meaningful comments.

[jojokeo]

I did PRR and got nothing. It's a moot point now since I had to abandon that POS chip. It appears nobody can answer basic info either? At this point the schematic makes no difference if you read my prior update. I know and understand basic op-amp circuitry amost as well as tube amp design. BTW, nice to see you here also. 

[PRR]

I can't think of any "good reason" for the troubles you describe. The OP275 has a fine reputation for fine audio. Opamps in general have potential bass out the wazzoo, controlled by the NFB you apply.

I was hoping that a glimpse of the schematic you are working from might trigger some thought why a OP275 would work any different than your other opamps. I am curious. But if it is not to be, then oh well.

[blackieNYC]

Do you think if you would have attenuated the input level with a voltage divider, you would have seen the distortion disappear, which would just mean you've hit something too hard? And despite the fact that this is a $acred Cow HiFi Butler-something Amplifier (I'm surprised at your findings too),TMF thinks the data sheet is saying the OP275 could be a suspect, in how it compares to other chips, for this particular circuit and symptom you are observing.  But it's not working for you so by all means put it aside. I'm sure your patience was maxed out. You seem to be an experienced tech. Hopefully you can try it again sometime, in a similar circuit gain-wise.  Maybe in a circuit that works well with another op amp.  Then you'll have a concrete example of what these experts are saying.  And your question certainly has drawn the finest, right up to the point where I chose to come in. And we non-experts would love to have someone write a post that says "headzup - the op275 doesn't work in the U1b stage of the Ganektikazoink. I think I know why."
You know PRR from another forum? This is like running into one of his old high school buddies. Tell us some embarrassing anecdote.

[jojokeo]

I'll post it again, but again this is what I had that was causing the issues. I changed biasing and Vref a number of various ways & values along with the bass response removing R18 & C10, etc. Rolling back the guitar vol almost halfway didn't really help nor did changing the three controls through their rotations - all with humbuckers. I could get the similar slightly reduced response with single coils in the same manner. I switched out a number of input caps C1 but there's a point where it makes no sense to continue going smaller and losing bass response further. In the gain loop I tried LEDs, mosfets, silicon diodes changing the Vf and in the 2nd op-amp also, along with a few gain value changes there as well. Even the hard clipping diodes post op-amp A, C13, C6, etc...so yes, I got "maxed out" you could say needed to move on. I like a challenge but I'm not a glutten for punishment!

The reason I like challenges like this is because 1) I'm stubborn, 2) don't give up easy, 3) anal as can be, but most importantly 4) this how I (we) all learn - and I mean really learn. Books and equations only teach a person so much. And this is where I have my anecdote for PRR - he always has a way of making even the most highly technical thing understandable and rather simple by finding a usually funny abstract example to get his point across. His way of thinking is burned into the back of my mind and on many occasions a golden quip will surface such as "quit over thinking the damn thing" (but said in the way only he does best). With all the time spent humming and hawing a person could've already built & tested it or changed something out about 10 times already. But it usually involves something like a car, barn, house wiring, furnace, or any number of other references he'll use in similar situations...I respect the man immensely and see him on many a forum unselfishly giving his time helping others.

Currently I went back to much of this schematic with a few tweaks here and there. Bias wise I'm using a 470K to amp B and 270K to amp A. I switched out the Vref resistors from 47k's to 10k's and as said earlier removed R18/5k6 & C10/1uF (too much bass response along with too much increased distortion). I was able to remove the stuff at the end - R14, C3, & C16. I voiced things mainly using the OPA2134 now but it also sounds really good with these: TL082, JRC4558D, RC4558P, NE5532AP - yet all in their own way as they should. I'm usually not a fan of the TL072 and this proved itself again as a dud (fizzy that cuts out abruptly/noticeably as the note sustains). But that damn OP275 just pukes up it's guts with this low end artifact crap and is like nails on a chalkboard, it's not even close in this circuit. I'm pretty sure I got them from one of two guys on ebay I've had good experiences with named Tayda & Polida and supplied most of the others if I'm not mistaken which all work as they should. Many builders buy from these guys I'm pretty sure?

[PRR]

Now I see the schematic.

I'm mystified.

But you *should* put a 0.1uFd cap directly _AT_ the chip's power pins. The TL072 is fairly tolerant of power pins wandering uncontrolled. A 100u cap within a few inches is ample. The OP275 is not that much faster, but *may* be more susceptible to wobbly support. The datasheet does clearly show 0.1uFd caps on some of the suggested circuits.

I was just looking at a house. The cellar stairs were frighteningly "loose". Aside from a busted step, I finally realized the stringer was just undersized for its length, and had been supported mid-span by a hanger, which had been cut (some strange stuff happened in this house). Oh, and there was dog chew-marks on the railing. For TL072/OP275 thinking: the springy stringer probably didn't bother the dog, but unsettled me.

[jojokeo]

Now I see the schematic.

I'm mystified.

But you *should* put a 0.1uFd cap directly _AT_ the chip's power pins. The TL072 is fairly tolerant of power pins wandering uncontrolled. A 100u cap within a few inches is ample. The OP275 is not that much faster, but *may* be more susceptible to wobbly support. The datasheet does clearly show 0.1uFd caps on some of the suggested circuits.

I have the 100uF mounted 2 holes (using stripboard) away from pin #8. Interestingly, I redrew the current schematic changes made recently and noticed the missing 0.1uF and added it. In recreating the current layout these will be both within one and two pins away from pin #8 however the ground will rely on a jumper to get down to pin #4. This should suffice? I will give things a go later this evening and report back. Thanks as always (and for the visual too) 

[samhay]

Two thoughts, neither of which are much better than hand waving.
The OP275 can draw quite a lot of current. How much does the supply voltage sag through R2?
Some op-amps get a bit funny about driving caps to ground and you have a capacitive load on the output of the second op-amp. It's a tiny load, but what happens if you lift the 1n or 4n7 cap?

[amptramp]

I would bet if you tried the LF356 op amp, you would get the same results as with the OP275.  This is one of a few op amps that has the nasty effect that when you exceed the common-mode input limits, the output can reverse, causing all kinds of audio distortion.

The common-mode input range operation from ±15 VDC supplies is ±11 volts for an OP275.  This means you are out of the usable common mode range if a signal goes within 4 volts of the rail.  You are running on a single supply of 9 VDC.  So your usable common-mode input has to be 4 volts above ground and 4 volts below the positive rail or between 4 and 5 volts DC.  You have no trouble with a low level of input since you are biased at 4.5 VDC but once you add a humbucker with high output, you exceed the input signal range on the input peaks.  The second stage is no better off - you have diode clippers to ground that clip at a 1.3 volt level meaning signal excursions can range from 3.2 to 5.8 volts at the non-inverting input to the second stage.  Both stages appear to have a problem.

Oh, I just noticed - you have a series protection diode meaning your supply is actually 8.3 volts.  I am surprised you are getting this to work at all because your common mode range just went from one volt to 0.3 volts.

The TL071 series has the same worst-case values of ±11 but typically is -12 to +15 when running from ±15 VDC supplies.  At 9 VDC, this translates into an input range of 3 to 9 volts DC.  Biased at 4.5 volts, you would have to get 1.5 volt peaks on the negative side of zero to upset a TL071 and I doubt any pickup, single coil or humbucker, would be capable of overloading it.

You solution may be simple - swap out the 9 VDC supply for something 12 volts or more.  You can get 15 VDC, 16 VDC, 18 VDC and 24 VDC supplies.  Make sure you have enough margin with the capacitor voltage ratings (I use 60% derating for this type of thing) and the circuit will behave the way you want.

[jojokeo]

The OP275 can draw quite a lot of current. How much does the supply voltage sag through R2?
Some op-amps get a bit funny about driving caps to ground and you have a capacitive load on the output of the second op-amp. It's a tiny load, but what happens if you lift the 1n or 4n7 cap?

Both very interesting observations. When R2 was 100r I had measured 8.8v & didn't check for sag but - I've since changed this to 10 ohms. Because the op275 had a lot of high end I added that last cap to shave some off using a decade box to find a value I liked. I hadn't done anything different overnight when I thought the circuit was a done deal but then the low end thing (and frustrations) began...

[jojokeo]

Oh, I just noticed - you have a series protection diode meaning your supply is actually 8.3 volts.  I am surprised you are getting this to work at all because your common mode range just went from one volt to 0.3 volts.


You solution may be simple - swap out the 9 VDC supply for something 12 volts or more.  You can get 15 VDC, 16 VDC, 18 VDC and 24 VDC supplies.  Make sure you have enough margin with the capacitor voltage ratings (I use 60% derating for this type of thing) and the circuit will behave the way you want.

I will be removing the 5817 series diode & simply using a std 4002 for protection in the standard manner to negate voltage drop. And along with changing R2 to 10r my supply will be higher and less chance of current limitations as the prior post mentioned. Along with adding the 100nF to power pins this all should ar least help?

Thank you for all the detailed info this is much appreciated. I hope to first eliminate what it is I'm experiencing and then if sucessful I will attempt to repeat it. I want to find out exactly what is causing it if possible?

On a side note when the circuit was working initially it had a low end slight modulation (best way to desribe it?) that was only most perceptible playing A5 & B5 power chords. It would last a second or two at most then settle out. It would've likely maybe not have been noticed by the non-discerning musician? But it botherd me enough to feel the need to address it. I initially used LEDs in the gain loop and thought it was the signal clipping point, then removed the extra large bass cap tied to the gain loop...anyway , I just thought I'd mention that characteristic because I think this all may be related & relevant?

[PRR]

> when you exceed the common-mode input limits, the output can reverse

Did you peek at the specsheet? That's cheating!!

[jojokeo]

> when you exceed the common-mode input limits, the output can reverse

Did you peek at the specsheet? That's cheating!!

I saw and read this but must admit to having difficulty understanding everything exactly but I think I get the gist of it? However, there is a 6k8 series resistor on the non-inverting input already. Will this not suffice anyway or does it need to be as close to that suggested value as possible and simply using a 3.9k standard value? I can't imagine that this is that touchy?