I personally always push the supply near the limit of the 5532/34 op-amp which are maximum rating to +/- 44V to around +/- 42V which gives me a higher headroom of 3V peak compared to the standard +/-36V seen in so many design. I used this technique in my precision preamp of Douglas Self and everywhere possible when I use 5532/34 op-amp in audio field.🙂
First of all it's +_22v, second you have to be aware of the output load including the feedback network that can fry your op-amp if it's not properly adjusted.I preffer for op-amps to be buffered when driving a hard load at max supply...
+/-18V is fine with these, anything more starts to enter iffy territory, its unlikely you really need the extra headroom, and in fact it can fry equipment down-stream that's using a lower voltage (the standard opamp 100 ohm output resistors aren't going to protect very well if you try to push 20V into a +/-15V circuit for instance). Running at +/-18 means the output will be about +/-16V at worst. 18V is a standard one for regulators and DC-DC converters too.
The slightly reduced operating temperature will be less stressful for the devices long-term too.
The slightly reduced operating temperature will be less stressful for the devices long-term too.
So what would circuit do you prefer to get into hard clipping the first ones at the input or the last ones? I personally use the highest rails op-amp at the input and the lowers higher power rails at the output after attenuation and diode input protection .Scratchy potentiometers, dodgy rca sockets can get the first circuits into saturation or the higher power ones ...If the first ones are delivering the full burst signal into some protection diodes for the next stage , nothing should be at risk, i think...plus i don't use just 100ohms resistors...1...10k is fine most of the time, even in inverted topology with bipolar inputs.I need those resistors anyway to limit the bandwidth with simple first order RC filters that i am using in any single audio ciurcuit conservatively.
Last edited:
Supply limit take 2
Thing not said is that all supply pins are fed thru 10 ohms limiting resistors with 3 bypass capacitors to first increase PSRR, SNR and limit current consumption. Aside this the load involve hooked to OP-AMP are of high impedance (>2K) at output. The input so far is commonly drove by small signal appliances like radio, auxilliary, TV, etc. CD is usually limited by resistor ratio to reduce dynamic.
Thing not said is that all supply pins are fed thru 10 ohms limiting resistors with 3 bypass capacitors to first increase PSRR, SNR and limit current consumption. Aside this the load involve hooked to OP-AMP are of high impedance (>2K) at output. The input so far is commonly drove by small signal appliances like radio, auxilliary, TV, etc. CD is usually limited by resistor ratio to reduce dynamic.
Can you not see the error in your figures...
+/-44v is 88 volts in total. +/-36 volts is 72 volts in total. All these are way in excess of the devices maximum ratings.
Your 10 ohm resistors may provide a small benefit with regard to filtering but the current consumption of a particular chip is pretty much constant with respect to the supply voltage.
It's important you understand these factors rather than carry on misunderstanding them 🙂
+/-44v is 88 volts in total. +/-36 volts is 72 volts in total. All these are way in excess of the devices maximum ratings.
Your 10 ohm resistors may provide a small benefit with regard to filtering but the current consumption of a particular chip is pretty much constant with respect to the supply voltage.
It's important you understand these factors rather than carry on misunderstanding them 🙂
Supply limit take 3
Sorry for misinterpreting supply voltages...! When I say +/- 44V it mean total excursion from + to -. It's like saying +22V and -22V reference to 0V, not 88V. The same for +/- 36V, again total excursion from rail-to-rail of 36V, not 72V, meaning +18V and -18V reference to 0V.
Sorry for misinterpreting supply voltages...! When I say +/- 44V it mean total excursion from + to -. It's like saying +22V and -22V reference to 0V, not 88V. The same for +/- 36V, again total excursion from rail-to-rail of 36V, not 72V, meaning +18V and -18V reference to 0V.
The devil is always in the detail 🙂
If you run the opamp on -/+22 volt rails (so 44 volts in total) then the output can only swing (excluding losses) -/+22 volts if measured from the zero volt line.
It can't be -/+44v because the output can not be in two places at one time. It can not swing 44 volts positive or 44 volts negative.
If you measured from the negative rail then the output can swing to positive 44 volts only (using the negative rail as reference point).
If you measure from the positive rail then the output can swing to 44 volts negative only. (using the positive rail as reference point).
If you run it on a single rail (so a single +44 volt supply) then you can achieve -/+22 volts by AC coupling the output and with the opamp biased to the midpoint of plus 22 volts.
They are all subtleties but they are all so important 🙂
If you run the opamp on -/+22 volt rails (so 44 volts in total) then the output can only swing (excluding losses) -/+22 volts if measured from the zero volt line.
It can't be -/+44v because the output can not be in two places at one time. It can not swing 44 volts positive or 44 volts negative.
If you measured from the negative rail then the output can swing to positive 44 volts only (using the negative rail as reference point).
If you measure from the positive rail then the output can swing to 44 volts negative only. (using the positive rail as reference point).
If you run it on a single rail (so a single +44 volt supply) then you can achieve -/+22 volts by AC coupling the output and with the opamp biased to the midpoint of plus 22 volts.
They are all subtleties but they are all so important 🙂
Read...
I wrote nothing about what the op-amp would drive but about the supply of it. If you read my last lines, my supply voltages are referenced to 0V. I rarely work with single supply meaning having to create an artificial 0V that's about Vmax divide by 2.
I wrote nothing about what the op-amp would drive but about the supply of it. If you read my last lines, my supply voltages are referenced to 0V. I rarely work with single supply meaning having to create an artificial 0V that's about Vmax divide by 2.
Read...
I wrote nothing about what the op-amp would drive but about the supply of it. If you read my last lines, my supply voltages are referenced to 0V. I rarely work with single supply meaning having to create an artificial 0V that's about Vmax divide by 2.
I wrote nothing about what the op-amp would drive but about the supply of it. If you read my last lines, my supply voltages are referenced to 0V. I rarely work with single supply meaning having to create an artificial 0V that's about Vmax divide by 2.
I have read, and I'm not trying to argumentative 🙂 so this is my last word on it.
So if you mean total excursion from one point to another then that is a single quantity of 44 volts. It is not -/+ which really does imply a total of 88v.
So the possible excursion at the opamp output is plus 22 volts or minus 22 volts or -/+22v not -/+44v.
So if you mean total excursion from one point to another then that is a single quantity of 44 volts. It is not -/+ which really does imply a total of 88v.
So the possible excursion at the opamp output is plus 22 volts or minus 22 volts or -/+22v not -/+44v.
Yes...
You got it... actually my precision preamp and my electronic crossover are playing while writing.... these 2 equipment uses +21 volts on positive supply pins thru 10 ohms resistors and -21V on negative pins supply thru 10 ohms resistor, on each op-amp. So I can say that I'm running at 1 V on each polarity near the maximum allowed. So all audio PCB I built and builted have that possibility of extending supply. If for a reason of other op-amp I can't entend, I just place a jumper instead of zener diode on ref pin of regulator. But I have so many parts of 5532/34......
You got it... actually my precision preamp and my electronic crossover are playing while writing.... these 2 equipment uses +21 volts on positive supply pins thru 10 ohms resistors and -21V on negative pins supply thru 10 ohms resistor, on each op-amp. So I can say that I'm running at 1 V on each polarity near the maximum allowed. So all audio PCB I built and builted have that possibility of extending supply. If for a reason of other op-amp I can't entend, I just place a jumper instead of zener diode on ref pin of regulator. But I have so many parts of 5532/34......
Last edited:
Most of the op amps circuits I have seen are run "at or under" recommended voltage levels. There was only one manufactured device that ran them at the edge. The Ashley stereo line mixers (and likely their other mixers) are running TL072 op amps right at the maximum point of +/-18V. The TL072 are used in the gain and output stages (2043 used for input stage). The Ashley mixers have a bloom type of sonic signature that is similar to some tube preamps. I don't know if this is a result of running the op amps right at the edge or something else in the circuit. Has anyone else heard this sonic signature when running op amps right at the maximum?
Supply near limit
Well I never heard of that amplifier but looks like it sound amazing.... the reasons I see to push supply limit are more on better PSSR, better headroom because not every amplifier have same gains. Like me I use the Load Invariant designed by Douglas Self. It has moderate gain and why not try different? So far it gives satisfaction.
Well I never heard of that amplifier but looks like it sound amazing.... the reasons I see to push supply limit are more on better PSSR, better headroom because not every amplifier have same gains. Like me I use the Load Invariant designed by Douglas Self. It has moderate gain and why not try different? So far it gives satisfaction.
For plain amplification headroom shouldn't be limited by the supply, that implies too much gain, and gain before attenuation (a recipe for headroom issues)
For something more complex like filter/EQ sections then supply headroom may be an issue, and by a large factor, so perhaps the correct approach is to properly analyze headroom and scale signals accordingly, rather than run all the chips hotter than is necessary.
For something more complex like filter/EQ sections then supply headroom may be an issue, and by a large factor, so perhaps the correct approach is to properly analyze headroom and scale signals accordingly, rather than run all the chips hotter than is necessary.
Supply near limit
That's interesting but this preamp has also tone control that can clip easily signal, a nice way to limit clip would be to compress signal but distortion will increase.
On next week, I will run bench test with different supply ( suggested and near limit), without and driving loads and I'll measure average temperature over time (<=4 hours). That's a good point and I have on hand heatsink made specially for OP-AMP that I did use on similar project at work. Because some OP-AMP will naturally draw more current and will heat like many parts of Analog Device and at suggested supply. The temperature sensor will be infrared type of good quality.
That's interesting but this preamp has also tone control that can clip easily signal, a nice way to limit clip would be to compress signal but distortion will increase.
On next week, I will run bench test with different supply ( suggested and near limit), without and driving loads and I'll measure average temperature over time (<=4 hours). That's a good point and I have on hand heatsink made specially for OP-AMP that I did use on similar project at work. Because some OP-AMP will naturally draw more current and will heat like many parts of Analog Device and at suggested supply. The temperature sensor will be infrared type of good quality.