Adding 2 extra ne5532 IC's to circuit drops VCC dramatically

[j_flanders]

I wanted to add a spring reverb to my Vox Pathfinder 10, a small SS amp.

I used the schematics from Rod Elliott (ESP):
https://sound-au.com/articles/reverb.htm (Figure 5a for drive and Figure 9 for recovery.)

It all works and the reverb is decent but adding these two dual ne5532 opamps changes the power supply to all opamps.
Where all opamps in the preamp previously got +13,6V / -13,6V they now all sit at +3,9V / -3,9V.
The preamp now has less headroom, the distortion characteristics have changed etc.

Since both positive and negative supply voltage have changed by the same amount it seems unlikely I made a mistake somewhere and that the reduction is purely because of the current these two extra circuits draw?

So what are my options to prevent this from happening and get everything back around +13V/-13V?
Mess with R27 ,R28?, voltage regulator?, chargepump?

Here's a schematic of what I did:
https://i.postimg.cc/NLh4QDWF/pf10-reverb-problem.jpg



Here are my veroboard circuits that I created:
https://i.postimg.cc/XNZ67vR6/Screen-Shot007.jpg


I'm using an 8AB3C1B tank (8 Ohm in, 2k Ohm out)

[antonis]

Make R27/R28 220 - 270 R and C26/C27 150μF..

Original circuit (without extra 5532) draws about 8.5mA [(20-13)/820] so additional about 10mA current draw (5532 two op-amps) result into 3.9V..

P.S.
Of course, you can power 5532 just before R27/R28 if you're sure for +/- 20V not exceed +/- 22V..

[j_flanders]

Awesome! Thanks. Gonna try it later on today. I'll report back.

[j_flanders]

P.S.
Of course, you can power 5532 just before R27/R28 if you're sure for +/- 20V not exceed +/- 22V..

Right, that's even more convenient. I'll measure that voltage first. I have a couple of NE5532's to spare as well.

I was also thinking about changing the NE5532 in the recovery circuit to an TL072 because the spring tank's output coil (like a guitar pickup) is directly connected to the input of the opamp there and the low input impedance of the NE5532 might load it down.

I previously tried increasing the 2k2+2.2nF (similar to guitar tone pot) and the 220k (similar to a guitar volume pot) but I couldn't hear much, if any, of a difference. But if the NE5532 has an input impedance of 100k that would impose a rather low maximum there.

But I see the TL072 has a maximum supply voltage of +18V/-18V which would rule out either the proposed solution of tapping directly into the 20V supply or the opamp.
I see there's an TL072H version with +21V/-21V maxima but that's rather close that 20V.

[antonis]

You can place a resistor in series with Vcc & Vee, taking into account each op-amp typical supply current (e.g. 1,5mA for TL07X series) for voltage set at desired (safe) level..

P.S.
There are various reports for powering 5532 at higher than +/- 17V is not recommended..!!!

[j_flanders]

I took the 'gamble' and went for the easiest solution: tapping into the 20 volts supply. It measured 20,1 volts.
So far it works very well. The amp sounds back like it did before.

Should I be surprised that I hear little or no difference as far as reverb goes? Vcc and Vee now measure around +/-19,5 volts where it was around 3,9volts when I tapped from r27,r28. Or is the current still the same and the only difference that it does not 'load down' the power supply as much?

There are various reports for powering 5532 at higher than +/- 17V is not recommended..!!!

I googled really hard before I tried the 20volts on the NE5532's but couldn't find any reports about that.

The ic's are in sockets, so they're easy to swap out if you can suggest a dual opamp that likes +/-20volts, has higher input impedance than the NE5532 and can drive the reverb input coil equally well or better.

[MikeA]

There are various reports for powering 5532 at higher than +/- 17V is not recommended..!!!

I googled really hard before I tried the 20volts on the NE5532's but couldn't find any reports about that.

There are a few things still beyond Google's reach...  from Douglas Self's Small Signal Audio Design (fair use.)
   

[j_flanders]

Thanks Mike. So, they're getting hot for only a little extra headroom. Good to know.
I'll measure the temperature and see how hot they get.

[antonis]

The ic's are in sockets, so they're easy to swap out if you can suggest a dual opamp that likes +/-20volts, has higher input impedance than the NE5532 and can drive the reverb input coil equally well or better.

You don't need higher input impedance 'cause for U1A is dominated by R1 (10k) and U1B (the right one) can easily manage DC offset in case of  R2 made up to 470k..

P.S.
Maximum supply voltage should be considered safe only when output is far from maximun rated current..

[j_flanders]

I'll measure the temperature and see how hot they get.

In my guitar room where it's currently 17°C the 4558's measure around 22° C and the 5532 around 36°C. (amp chassis open and the amp was on for an hour or so.)

You don't need higher input impedance 'cause for U1A is dominated by R1 (10k) and U1B (the right one) can easily manage DC offset in case of  R2 made up to 470k..

U1A takes the output of U2A of the Vox preamp so high input impedance isn't critical there.

Unfortunately I don't think I understand your other two remarks.
Input to U1B is the tank's output coil.
Maybe I got this wrong but I thought that the NE5532's input impedance was around 100K, so using higher value resistors for R1 and R2 makes little difference because they're paralleled by the 100k of the opamp input.

What would the input impedance be if I'd use 1M for both R1 and R2 for example? 500k//100k?

P.S.
Maximum supply voltage should be considered safe only when output is far from maximun rated current..

I don't know what that means in this particular case/circuit.

Anyway, thanks a lot for your help so far. I'm glad it works now, and without setting the house on fire.

[antonis]

Unfortunately I don't think I understand your other two remarks.
Input to U1B is the tank's output coil.
Maybe I got this wrong but I thought that the NE5532's input impedance was around 100K, so using higher value resistors for R1 and R2 makes little difference because they're paralleled by the 100k of the opamp input.

That's a very common misunderstanding..

100k input impedance is considered the differential input resistance and not the resistance from either input pin to a supply rail...
It barely affects the circuit 'cause the inputs are generally at the same voltage (thanks to negative feedback..)
When you wrap them with feedback, op-amp acts to keep the difference between (+) & (-) as small as it can. This, in turn, raises the effective impedance.
From the datasheet:
"All characteristics are measured under open-loop conditions, with zero common-mode input voltage, unless otherwise specified."
100k (or so) impedance represents the effective resistance between the positive and negative inputs (+ & -), for the amplifier in open-loop conditions.

P.S.
By raising R2 value up to 470k you create about 74mV, additional to 20mV across 100k, bias offset (200nA X 370k) but it shouldn't be a problem 'cause stage DC gain is unity (thanks to C2) and output is DC isolated (thanks to C3)..
(94mV lower headroom isn't an issue, is it..??)

[j_flanders]

That's a very common misunderstanding..

I was reading this topic yesterday which is where I got that misunderstanding:
https://www.diystompboxes.com/smfforum/index.php?topic=114212.0
Someone asked how to calculate the input impedance or load the guitar sees at the input of an effect.
This is what R.G. replied in 2016:

The 1M to ground pulldown would make the input impedance 1M for audio if it was the only thing there.

[...]bias resistor of 510K to Vref and the input of your opamp. [...]

The opamp itself has a "high" input impedance at its non-inverting input.
JFET and MOSFET opamps have impedances so high that you may as well ignore them as an open circuit.
The very popular NE5532 has an impedance there of about 100K. 

If the opamp is a FET version, the load can be simplified to just the bias resistor.

There's some back and forth going on further on in the thread. You replied as well. In the end it wasn't clear to me who or what was right.

[antonis]

Just copied part of Paul's answer here: https://www.diystompboxes.com/smfforum/index.php?topic=105347.msg949082#msg949082

Yes, there is a number on the NE5532 data sheet showing input impedance as 30K worst-case. Not an issue. Under NFB this increases by the excess gain. NE5532 GBP is 10MHz, so it has gain of 1,000 at 10KHz, the top of the guitar band. Say you close-loop at gain of 3.5. You have 300:1 of excess gain. 30K*300= 9Megs intrinsic input impedance.

[antonis]

Maybe a schematic should be more helpful.. 



Ri = differential input resistance (30k to 100k for NE5532)
A_OL= Open loop gain
β = Feedback fraction, R2/(R1+R2)

[anotherjim]

Does this explain why a 5532 or it's like don't substitute well in FX built with 100k or higher Vref divider resistors? When it powers up and before negative feedback can take effect (and that may be weak with a 1M gain pot), that 100k between inputs pulls Vref out and it can't stabilize.
Not an issue with bi-polar supplies of course.