Author Topic: Virtual ground push-pull  (Read 5880 times)

0 Members and 1 Guest are viewing this topic.

Offline oliver602Topic starter

  • Regular Contributor
  • *
  • Posts: 120
  • Country: gb
Virtual ground push-pull
« on: April 15, 2019, 01:16:58 am »
Hi,
Im trying to make a split supply for a diy function generator kit I got from banggood.com. it needs +12 -12v +5 and Gnd.

Im using a boost converter module to step up a battery to 24V and plan to make a virtual ground and the +5v using 2 opamps.

I've been simulating the circuit in EveryCircuit. I tried adding a push-pull setup on the virtual ground opamp to increase maximum current and now the simulation will not run.

Anybody like to point me in the right direction?

Any tips on buffering the output of the ground opamp in this circuit?

http://everycircuit.com/circuit/4693683140820992
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2763
  • Country: us
Re: Virtual ground push-pull
« Reply #1 on: April 15, 2019, 02:49:32 am »
Maybe something like this

   
« Last Edit: May 01, 2019, 07:33:03 pm by MarkF »
 

Offline oliver602Topic starter

  • Regular Contributor
  • *
  • Posts: 120
  • Country: gb
Re: Virtual ground push-pull
« Reply #2 on: April 15, 2019, 03:22:35 am »
Interesting! What does R4 on the opamp output do?
 

Offline MarkF

  • Super Contributor
  • ***
  • Posts: 2763
  • Country: us
Re: Virtual ground push-pull
« Reply #3 on: April 15, 2019, 04:15:31 am »
The op-amp in is providing the virtual ground. 
However, it's output current is not very large. 
The transistors are used to boost the op-amp output current. 
The resistor, R4, determines how much boost current is provided.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: Virtual ground push-pull
« Reply #4 on: April 15, 2019, 05:26:15 am »
R4 also helps with cross-over switching distortion when a load varies from source to sink and back. With the op-amp's output at say 1/2Vcc plus 0.2V neither transistor would be on.

You can get it in an IC TLE2426 The 'Rail Splitter' Precision Virtual Ground.


 

Offline oliver602Topic starter

  • Regular Contributor
  • *
  • Posts: 120
  • Country: gb
Re: Virtual ground push-pull
« Reply #5 on: April 28, 2019, 02:20:18 am »
I'm having trouble getting my circuit to work correctly. I have removed the push-pull transistors to help debug.

If I put a 1K resistor between +12V and VGnd the voltage across the resistor drops to about 3V... VGnd gets pulled up.

If I put the resistor between -12V and VGnd, it works as expected.

With no resistor the voltages are as expected and there is no voltage across R4.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Virtual ground push-pull
« Reply #6 on: April 28, 2019, 07:53:18 am »
Check Q2. Is it really a PNP transistor? Could you have used NPN by mistake? Try substituting it with a different one.
 

Offline soldar

  • Super Contributor
  • ***
  • Posts: 3595
  • Country: es
Re: Virtual ground push-pull
« Reply #7 on: April 28, 2019, 08:04:10 am »
If you are going to be converting I would just generate all voltages separately. It will be more efficient and makes more sense.

I am pretty sure you can get commercial PSUs which will output the voltages you need.

You can probably salvage one from some old electronics.
All my posts are made with 100% recycled electrons and bare traces of grey matter.
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #8 on: April 28, 2019, 09:57:43 am »
As I've been using opamp rail splitter too, so I've done a simulation.
Interestingly, with any opamp I get these <1mVpp oscillation bursts at the points where the load's current waveforms cross each other (it means the current I1=I2).
The oscillation starts just before the crossing happens.
Any idea where it comes from?
Does the opamp's cross-over distortion create that bursts?
« Last Edit: April 28, 2019, 10:18:57 am by imo »
Readers discretion is advised..
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Virtual ground push-pull
« Reply #9 on: April 28, 2019, 10:21:51 am »
Try adding a 1nF capacitor between U1's output and negative input. This should tame the oscillations. Then tweak the capacitor value for best performance.
« Last Edit: April 28, 2019, 10:23:24 am by Kalvin »
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #10 on: April 28, 2019, 10:27:25 am »
Nope..
Readers discretion is advised..
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Virtual ground push-pull
« Reply #11 on: April 28, 2019, 10:29:21 am »
The oscillation isn't much of a surprise, as the load is capacitive. Increasing the ESR of the capacitors may help, at the expense of poorer transient response.

You could also try connecting and inductor with a resistor in parallel with it between the capacitors and feedback node. The circuit, consisting of the inductor and output capacitors, will have to be over or critically damped, otherwise it will ring and oscillate.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Virtual ground push-pull
« Reply #12 on: April 28, 2019, 10:33:35 am »
Try tweking the 68 ohm resistor at the output a bit higher, like 100 - 150 ohms. Does it make things worse or better?
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #13 on: April 28, 2019, 10:52:10 am »
@Zero999: the both caps had/have 20mOhm ESR (in all sims above and here).

« Last Edit: April 28, 2019, 11:08:28 am by imo »
Readers discretion is advised..
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #14 on: April 28, 2019, 10:55:55 am »
@Kalvin: 50ohm-1k step 50ohm, the bigger the worse.
Readers discretion is advised..
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #15 on: April 28, 2019, 11:07:20 am »
1ohm ESR with 2x47u caps.
Readers discretion is advised..
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #16 on: April 28, 2019, 11:37:55 am »
Rail Splitter v2
Vgnd = 12V, 5.7uVpp ripple

PS: with the currents and voltages as in this simulation I would use beefier transistors..
« Last Edit: April 28, 2019, 12:04:08 pm by imo »
Readers discretion is advised..
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #17 on: April 28, 2019, 12:38:17 pm »
And finally the Rail Splitter v3.
Does not work with all opamps (ie with LM358)!

V3A is with emitter protection resistors. The higher the emitter resistors the larger the Vg "ripple".
« Last Edit: April 28, 2019, 12:55:56 pm by imo »
Readers discretion is advised..
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Virtual ground push-pull
« Reply #18 on: April 28, 2019, 01:04:40 pm »
Just use one of the electrolytic cap models on the output, no need for separate resistors then.  Stuff often oscillates if you use a perfect capacitor model in spice.
 

Offline soldar

  • Super Contributor
  • ***
  • Posts: 3595
  • Country: es
Re: Virtual ground push-pull
« Reply #19 on: April 28, 2019, 01:22:50 pm »
As a curiosity I will mention that you can use a complementary pair audio amplifier as a rail splitter / virtual ground.

Any of these will do it
https://www.eevblog.com/forum/projects/classic-40w-audio-amp-with-discrete-components/
All my posts are made with 100% recycled electrons and bare traces of grey matter.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Virtual ground push-pull
« Reply #20 on: April 28, 2019, 01:27:09 pm »
Just use one of the electrolytic cap models on the output, no need for separate resistors then.  Stuff often oscillates if you use a perfect capacitor model in spice.

It can oscillate in real life too.
@Zero999: the both caps had/have 20mOhm ESR (in all sims above and here).
All of those are under damped, so should oscillate.

Use a squarewave to test whether it really will oscillate.

The inductor makes the load look like a 1R8 resistor at high frequencies, where it's going to oscillate. R3 can't be removed or made too high, otherwise it will resonate with L1, C1 and C2. Increasing C1 and C2 will also cause oscillation. The inductor could be omitted, but the DC current would drop across R3, which might not be a problem, at low currents.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Virtual ground push-pull
« Reply #21 on: April 28, 2019, 02:40:29 pm »
Here is my shot at the circuit. The C3 and C4 are used to tweaking the step response and stability. The transistors' quiescent current is well below 1uA. The circuit is stable at least with C1 and C2 ESR in range 0 ... 50 milliohms.
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Virtual ground push-pull
« Reply #22 on: April 28, 2019, 03:20:31 pm »
Pulsing loads and RS v3A.
4.7u 47u 470u caps 50mOhm esr.
« Last Edit: April 28, 2019, 03:32:23 pm by imo »
Readers discretion is advised..
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Virtual ground push-pull
« Reply #23 on: April 29, 2019, 07:08:43 am »
Just use one of the electrolytic cap models on the output, no need for separate resistors then.  Stuff often oscillates if you use a perfect capacitor model in spice.

It can oscillate in real life too.

Quite, though my point was that zero ESR/ESL capacitors don't exist in reality.  Simulations will never be close to reality if you don't use models of real components.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Virtual ground push-pull
« Reply #24 on: April 29, 2019, 09:24:45 am »
It's easy to select capacitor models which fairly closely match reality.

The good thing here is that simulations will tend to use lower ESR capacitors, than what's available, therefore model the worst case.

It appears the class-B output stage was what was causing the instability, as the op-amp had a step change in its output, when changing from sourcing to sinking and back again. The above circuits are more class AB, which is more stable, but it's probably still prudent to test with different capacitor and ESR values, at different currents.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf