[rds]

Any idea why we don't see (buffered) precision shunt regulators being used to create virtual grounds?

 

[Avro_Arrow]

You mean like the TL431?
Maybe we are too used to using TLE2426 or
just a pair of resistors....
 
 
Quote:

Any idea why we don't see (buffered) precision shunt regulators being used to create virtual grounds?

 

[cobaltmute]

It would seem that the advantage of the TLE2426 or the resistor divider is that they track the rail as it droops on battery.  If you have stable rails on line power, it would seem fine to use a shunt.  But then again, if you are on line power, I don't see why you just don't use a dual rail power supply.

 

[rds]

 
[size=medium]

Quote:

It would seem that the advantage of the TLE2426 or the resistor divider is that they track the rail as it droops on battery.  If you have stable rails on line power, it would seem fine to use a shunt.  But then again, if you are on line power, I don't see why you just don't use a dual rail power supply.

The advantage of single supply besides battery usage is simplicity.  Many of the most well known DIY amps use single supply from the wall (ie the MMM and the PPA).  The problem with using a voltage reference in those designs is that people want to be able to adjust their power supply voltage, which would mean they would have to do some calculation for their voltage reference.
 
An advantage to regulating a battery would be to keep the voltage steady.  An unregulated battery will droop with varying current demands.  By regulating you lose some efficiency in exchange.  
It seems that normally people use large rail capacitance to try to overcome the droop.  Regulating can also have the advantage of requiring much lower rail capacitance (small size in portable devices).
​

[/size]




Quote:

You mean like the TL431?
Maybe we are too used to using TLE2426 or
just a pair of resistors....
 
 

 

Actually I was thinking of the TLV431 (its low voltage counterpart), as it is one I'm familiar with.
 

 

[cobaltmute]

The problem with regulating a battery supply is drop-out.  How much headroom do you need for the regulator versus how much output swing are you going loose?
 
The virtual ground issue is one that I keep struggling with.  I've got a bunch of Cisco PoE adapters:  48V - 300ma.  Sounds like a good supply to make a simple discrete amp, but how to do the virutal ground....

 

[nikongod]

With a battery powered portable I would be very cautions about anything that makes the amp less efficient. Unless you can prove that its an improvement (class-A biasing opamps for example) its probably just wasteful.
 
For stationary stuff, regulators are nice. I guess the question becomes why not do away with the battery and regulate an inefficient, but VERY well thought out wall power supply? When you look at it from a truly maximum effort standpoint battery power supplies are simply cheaper than solving the problems of building a real power supply. 
 
 
 
TLV431 is not rated for particularly high maximum voltage (7V absolute max, 6V recommended). It wont work in most headphone amps.
 
I think as mentioned people have historically used the TLE2426 or resistor divider because it splits the rails symmetrically every time. I happen to think people place far too much emphasis on maximum voltage swing, but symmetrical rails gets you the most swing out of most push-pull SS circuits.

 

[beerguy0]

Quote:

With a battery powered portable I would be very cautions about anything that makes the amp less efficient. Unless you can prove that its an improvement (class-A biasing opamps for example) its probably just wasteful.
 
For stationary stuff, regulators are nice. I guess the question becomes why not do away with the battery and regulate an inefficient, but VERY well thought out wall power supply? When you look at it from a truly maximum effort standpoint battery power supplies are simply cheaper than solving the problems of building a real power supply. 
 
 
 
TLV431 is not rated for particularly high maximum voltage (7V absolute max, 6V recommended). It wont work in most headphone amps.
 
I think as mentioned people have historically used the TLE2426 or resistor divider because it splits the rails symmetrically every time. I happen to think people place far too much emphasis on maximum voltage swing, but symmetrical rails gets you the most swing out of most push-pull SS circuits.

If you are adept at working with SMD, there are numerous miniature LDO regulators that can deliver 150-200mA, (easily enough to run a couple of opamps) with dropout voltages of 200-300mV, and can handle up to 16V in. This definitely opens up new possibilities for portable designs.
 

 

[rds]

Quote:

TLV431 is not rated for particularly high maximum voltage (7V absolute max, 6V recommended). It wont work in most headphone amps.

The TLV431 is the low voltage version of the TL431, which can output between 2.5 and 36 volts.  I mentioned the TLV431 because I was using it in a low voltage application where I needed a 1.65 reference voltage (for a 3.3V supply).