Transformer VA/current for 6xNE5532?

[s_c_leach]

Hello All

I am building a simple HP/LP filter that uses 6xNE5532 and using PSU with LM317/LM337 to give -/+12v. Initially I had spec'ed a transformer that is 25VA but would prefer to use 7VA giving 233ma per rails (15v) due to space restrictions.

I've checked the data sheet for these op-amps and they use 16mA, but I am unsure how much current these needed with a source? do the opamps use require more current when driving a load? if so how much?

I will be using the opamps between source and 2xamplifiers at line level.

Also the HP/LP filter is going to be used so that I can break-in speaker drivers so doesn't have to be to hi-fi - so doesn't have to have much headroom on the psu.

Thanks for any help in advance

Stephen

[codegen]

From datasheet, they use typically 6ma, and 16ma maximum, they dont require more than that. 7VA will do fine for 6 NE5532 behind LM317/337 i think.

[s_c_leach]

Brilliant, thanks for the help

Stephen

[janneman]

Originally Posted by s_c_leach

Hello All

I am building a simple HP/LP filter that uses 6xNE5532 and using PSU with LM317/LM337 to give -/+12v. Initially I had spec'ed a transformer that is 25VA but would prefer to use 7VA giving 233ma per rails (15v) due to space restrictions.

I've checked the data sheet for these op-amps and they use 16mA, but I am unsure how much current these needed with a source? do the opamps use require more current when driving a load? if so how much?

I will be using the opamps between source and 2xamplifiers at line level.

Also the HP/LP filter is going to be used so that I can break-in speaker drivers so doesn't have to be to hi-fi - so doesn't have to have much headroom on the psu.

Thanks for any help in advance

Stephen

I think you should also look at the max output level in the design load. For instance, 10VRMS into 600 ohms (admittedly extreme) requires an additional current from the supply to the opamp standing current.

jan didden

[s_c_leach]

Hey Jan,
The output will be going into a standard amplifier, so the input impedance of the amplifier would be the design load? if we assume that 2v peak to peak and 1k input impedance then that's 2.82 mA.
The circuit i'm using is from Rod Elliot 12dB / Octave Linkwitz Riley Crossover.
Or is the output impedance the 100R on the output of the circuit? I don't really understand what this resistor is used for.
It's been a while since i've done this sort of thing.

Regards

Stephen

[AndrewT]

The quiescent current quoted in the specification is for Zero output current from the opamp.
Janneman confirms that.
Any extra current fed to the load must come from one or other supply rail.
Short lived transient currents can be fed from smoothing capacitance and on pin decoupling.
Output current from the opamp is current to charge and discharge capacitive loading, current to feed into the feedback network, current to feed a resistive load, current to feed an RF attenuator.

The sum of these can easily exceed 10times the the next stage resistive input current.

Each half of a 5532 could draw ~8mA of quiescent current. 1mA to the load and allow 9mA more to feed the capacitances and feedback loop. Total per 5532 is ~36mA per dual chip.

Now to your 7VA 15+15Vac transformer.
The AC output current into a resistive load is <=233mAac.
After feeding a capacitor input filter the DC output current is ~<117mAdc.
Since your opamps will never be drawing 36mA continuously from both rails concurrently you do not need to be concerned with the heating effect of drawing a continuous 117mAdc from both rails of the transformer. Go ahead with the 7VA.

Warning !!!!!!

230:7VA, EI, 15+15Vac transformer may have a regulation approaching 30%.
the maximum voltage from the transformer could be
254/230 * 15 * 1.3 = 21.5Vac
after rectification and smoothing the maximum DC voltage on the smoothing capacitors could be ~+30,-30Vdc. You dare not use 25V capacitors. Use 35V or 40V or 50V capacitors.

Check the power dissipation of the regulator chips with 29Vdc at input and 12Vdc at output and passing a reasonable average current. You may need a heatsink for each regulator. At a average and continuous 50mA current, each regulator chip must dissipate ~ 850mW.

[AndrewT]

Originally Posted by s_c_leach

if we assume that 2v peak to peak and 1k input impedance then that's 2.82 mA.

most CD players will give ~3Vpk into a 10k resistive load. i.e. 3mApk. The EQ circuit may have to pass that current through to the power amp.

[s_c_leach]

Thanks Andrew for your answer. Looks like I need to go back to to do a bit of study on Power supply theory :-S. I have some spare 63v caps lying around and was planning to use them anyway so no problems there.

So with this design it would be better to use a 12-0-12 transformer, giving me
min 12*1.41=16.9Vdcv
max 254/230*12*1.3 = 17.2Vac * 1.41 = 24.5Vdc
reducing max dissipation to 625mW

One other question I have
for 9-0-9*1.41 would give 12.7Vdc, but from the datasheet I need approx 1.5volts
across the regulator for it to work, if we count for some regulation at 50mA
Or is this to close for comfort?

Thanks for you help
Stephen

[AndrewT]

12Vac is normally selected for regulated 12Vdc.

The minimum voltage is 216/230 * 12 = 11.27Vac.
After rectification and subtracting a 0.7V diode drop you get a minimum DC voltage of 15.2Vdc with ripple superimposed. In this case you need to estimate the bottom of the ripple voltage.
try guessing at 1Vpp of ripple.
You have a worst case lowest voltage being fed to the regulator of ~14.2V
That is just enough for the regulator not to drop out when passing 100mApk to 200mApk output current.

15Vac simply gives more head room above the drop out region in return for extra reg dissipation.

Because you average currents are more likely to be of the order of 50mA the worst case lowest voltage is very unlikely to ever happen, But the design process allows for worst case highest voltage and worst case lowest voltage.

You can insert reasonable operating voltages to find what the regulator may have to cope with in normal operating conditions.
I would do that to check total heat being released inside the case to ensure all the various heatsinks are going to cool adequately for all "normal operating conditions.

BTW,
you can measure absolute maximum voltage from your 15Vac transformer. You don't need to guess.

[s_c_leach]

Perfect, thanks for you reply, not a problem to go with the 15-0-15 transformer so I will go with that one as it takes into account all situations and I will ventilate the case well.

Thanks for you assistance

Regards

Stephen