Can I feed TL074 or TL072 with ±18V supply?

[Agung Kurniawan]

Hi everyone
as I say, can I?
Im going to build guitar amp using some TL074, TL072 and J201.

I feed the J201 stage with +18V, and because there's no lot of space I want the power supply section just have 2 regulator on it. so then the power supply voltage I have is ±18V.
will it be problem to feed TL074/072 with ±18V?

[Ice-9]

A 30 second search on Google for the datasheet shows the max voltage supply is +/-18v so yes you can.

Datasheets are the first thing you should look up when trying to determine what you can/can't.  

[Agung Kurniawan]

A 30 second search on Google for the datasheet shows the max voltage supply is +/-18v so yes you can.

Datasheets are the first thing you should look up when trying to determine what you can/can't.  

I have look the datasheet, I just hesitant cause I hit the maximum value. Anyway thanks

[induction]

The +/-18V max value quoted assumes a bipolar power supply, so the maximum voltage difference between the rails is actually 36V. You should be just fine at 18V.

[Ice-9]

I have look the datasheet, I just hesitant cause I hit the maximum value. Anyway thanks

Ok, I see, so it would be benficial to make sure your PSU has a good regulated +/-18v output so as not to go above this. If you are worried it would also be easy to add a couple of regulators to bring it down to +/-15v

The +/-18V max value quoted assumes a bipolar power supply, so the maximum voltage difference between the rails is actually 36V. You should be just fine at 18V.

The OP actually said +/- 18v in his post not 18v. 

[R.G.]

What I9 said.

As a bit more education in datasheet-reading: the phrase "Absolute Maximums" means just that. The manufacturer of the device is telling you "This device will work OK up to this line in the sand. If you cross the line  by one tiny bit, you have invalidated our warranty, and we will not say anything positive or negative about how it will behave. It might work. It might work for a little while, then die. It might convert its packaging into vapor and expel bit and pieces with great speed. It may even act like it's working, but fail erratically, or in subtle ways. So if you value your jobs (you engineers, you) you will not ever let the items listed below exceed the absolute maximum, even for an instant. Bad Things may happen".

On a chip with a power supply voltage specified at +/- 18V, it is bad practice in commercial design to use it with +/-18V supplies because there is a tolerance on even nominally "18V" regulators. They may wobble over 18.0000000V a little at times. A cautious engineer would not use them over, say, +/-17V. A bold engineer would use them at 18V, knowing that there will be some field failures happen. A wild and crazy engineer would use them at 19V or 20V because the engineer had tested several in the lab and that batch worked OK at 20, so they all should do that. Right?... er, right??

If you have +/-18V and it's for your use, and you don't mind them maybe failing at some random time in the future and having to fix the unit or scrap it, fine, use the +/-18V supplies. If you like reliability and predictability, put in +/-15V regulators at about $0.25 each.

[amptramp]

I had a similar problem a while back with a device that was rated at 14 VDC maximum powered by a 15 VDC supply.  I had two options: add a regulator to bring it down to 12 volts or add a green LED in series with the supply.  I used the LED and it dropped 2.05 volts, enough to stay within the required absolute maximum.  The decoupling capacitor was on the IC side of the LED but voltage came up slowly enough to not put a large charging current through the LED.

I would suggest the OP do the same.  LED's are cheap.

[Mark Hammer]

I concur with brother Keen.  The question that begs asking, though, is what purpose running them at +/-18V serves.  There ARE legitimate purposes, but not everything requires max operating voltage.  A great many commercial audio products do just fine and receive rave reviews with a simple +/-15V.

So, would the extra few volts make a real difference in performance, or is it simply a matter of needing more than +/-15V for something else, and +/-18V is the most you can use to power the op-amps AND something else, using a single power supply?

[PRR]

> is it simply a matter of needing more than +/-15V for something else

What he said: "I feed the J201 stage with +18V".

JFETs are hard to manage at 9V, easier at 18V, so I get this. He needs one 18V supply.

The opamp side "could" be re-rigged (Vref, caps) to run on just the one 18V. Or he could rig +18V/-12V which is only 30V total; the opamps won't mind. Or he could add dropping components (diodes, resistors) in the opamp supply feeds; I have done that.

Or he could ask the JFET if it minds running on 15V or 16V. Likely it can, with small reduction in maximum output (the opamp can boost that back).

I've run opamps on UN-regulated +/-17V for years. They did not fail. Your results may be different.

I used to run power chips 20% overvoltage. They did fail about every decade.

There is a 1980s studio console which contracted for "selected" BiFet chips to run on +/-20V. They were fine when new. Anecdotal reports 30 years down the road are that they seem to show more dead chips than comparable gear worked inside commodity rating, and the general consensus is to back-down the rails to +/-17V to prolong the original chips and allow commodity replacements. The 2dB reduction in max output is not worth blow-up trouble while recording music live.

[Mark Hammer]

Well-conceived, well-stated.

So, would it be problematic to whip up a +/-18V supply and throw a 1N400x diode or two between the power rails and the supply pins of the op-amps to drop the supply down a smidgen for safety, while still being able to provide 18V for the J201s?

[amptramp]

Well-conceived, well-stated.

So, would it be problematic to whip up a +/-18V supply and throw a 1N400x diode or two between the power rails and the supply pins of the op-amps to drop the supply down a smidgen for safety, while still being able to provide 18V for the J201s?

Or as I mentioned, an LED on at least one power rail.  I have used a red LED to get a stable 1.88 volt drop and a green LED to get 2.06 volts.  But I also have another piece pf equipment that used ±12 volt rails to drive high-frequency transistors rated at 12 volts and three of the transistors were dead.  I backed the supply down to ±11 volts.  So far, so good.

[antonis]

Diodes & Leds are good (zeners are better in some cases) but a humble resistor may be the best solution..

For a typical working current of about 1.5 mA for each one amp, a 180R - 220R for quad and double value for dual amp shouldn't be fine..??

[bool]

A working solution for TL07x would be to add inline standard silicon 1N400x diode + 100R/1W resistor per each rail. This also implies that a capacitor (usually a 22uF per opamp) should be wired from each rail to Gnd. (dual opamp: 47uF; quad: 100uF).

Zeners would be better; but more expensive and also hotter.

[Agung Kurniawan]

why I want to feed the Opamp on -/+18V:
1. There's not enough space for the regulator and I want the pcb to be small ass I can

2. this amp that I going to made have distortion channel with crunch mode. this channel has the sound character almost the same like tube distortion. I take Mesa Boogie MkIIC+ for the sample sound and comparison. especialy John Petrucci guitar tone. this channel using Jfet, Opamp and some zener diode. and they sound the best using +18V. the end of the opamp stage on this mode use bipolar power supply. and then back to first point when I have no space for the regulator

any sugestion for high-voltage low noise opamp? I need both dual and quad package

[Mark Hammer]

How much worse will it sound at 17 or 17.5V?

[Phoenix]

why I want to feed the Opamp on -/+18V:
1. There's not enough space for the regulator and I want the pcb to be small ass I can

2. this amp that I going to made have distortion channel with crunch mode. this channel has the sound character almost the same like tube distortion. I take Mesa Boogie MkIIC+ for the sample sound and comparison. especialy John Petrucci guitar tone. this channel using Jfet, Opamp and some zener diode. and they sound the best using +18V. the end of the opamp stage on this mode use bipolar power supply. and then back to first point when I have no space for the regulator

any sugestion for high-voltage low noise opamp? I need both dual and quad package

Can you show us a schematic of what you're doing? I can understand your wanting to keep +18V for the Jfet part of the preamp, but it's highly improbable that you'll actually NEED +/-18V for the op amp section, especially if there are any clipping diodes. You just simply don't need that kind of headroom unless you're running straight into a ~100W solid state power amp or driving a tube power amp.
You can probably get by with +18/-15V or +18/-12V as has been suggested already.

[Agung Kurniawan]

why I want to feed the Opamp on -/+18V:
1. There's not enough space for the regulator and I want the pcb to be small ass I can

2. this amp that I going to made have distortion channel with crunch mode. this channel has the sound character almost the same like tube distortion. I take Mesa Boogie MkIIC+ for the sample sound and comparison. especialy John Petrucci guitar tone. this channel using Jfet, Opamp and some zener diode. and they sound the best using +18V. the end of the opamp stage on this mode use bipolar power supply. and then back to first point when I have no space for the regulator

any sugestion for high-voltage low noise opamp? I need both dual and quad package

Can you show us a schematic of what you're doing? I can understand your wanting to keep +18V for the Jfet part of the preamp, but it's highly improbable that you'll actually NEED +/-18V for the op amp section, especially if there are any clipping diodes. You just simply don't need that kind of headroom unless you're running straight into a ~100W solid state power amp or driving a tube power amp.
You can probably get by with +18/-15V or +18/-12V as has been suggested already.

I haven't make the schematic yet, but it look like amptweaker tight metal with switchable input stage from opamp and Jfet and I have subtitute some component.

and you're right. I got big headroom that give me nice punchy mid frequency. and this one gonna straight in to 100Watt SS power amp.




is that the diode you all talking about guys?
and is there any specific affect if I use bipolar but different voltage as sugested?

[anotherjim]

That's exactly as done in many solid-state amplifiers. A pair of resistor + zener droppers to make either +12/-12 or +15/-15 for the op-amp based preamp & tone controls from the higher voltages the output power stage uses.

[ElectricDruid]

If you want to keep the +/-18V rails, you should find an op-amp that can cope with a bit more voltage.

TI have a page of high power/high voltage op-amps:

http://www.ti.com/lsds/ti/amplifiers/op-amps/power-op-amps-overview.page

Maybe you could find something there?

Edit: Analog Devices OP42 is another option - +/-20V max supply.

The trouble with this is you're into specialised chips rather than general purpose stuff. They're much harder to find and more expensive too.

Tom

[poiureza]

Hello,
I know this is an old thread but I cannot seem to find a more recent one.

My TL072 datasheet from July 2017 states :
Absolute Maximum Ratings
VCC+ VCC-     MIN -0.3V MAX +36V

To me that's a very confusing info.
One could read it as : "absolute lowest voltage has to be above -0.3V" which basically says "no bipolar PS, put VCC- @0V and VCC+@36V to get the most out of it"
Or it could read "Minimum VCC+ = -0.3V and maximum VCC- = +36V" which lakes little sense to me

I know neither is correct as I know it to work perfectly with a bipolar -9V/+9V supply.

I'm just curious on how to read this in the correct way ...