Why negative LDOs are more expensive than positive LDOs?

[zapta]

I am looking for a fixed -2.5v LDO in small quantities. On Digikey, the least expensive is TPS72325DBVR at $2.99 while for +2.5V the least expensive is MIC5504-2.5YM5-TR at only $016. That's more than x10 more expensive. Why?

[wraper]

Because demand and production volume is not nearly as high.

[DavidAlfa]

Negative voltages are pretty rare in everyday electronics.
Easily, sales of positive Vregs are 50x the negative ones.

[TimFox]

A related question:  before retirement, I needed a reference for a negative voltage (from -12 V power), otherwise similar to an ADR02-type "series" three-terminal positive reference regulator.
(Voltage accuracy not critical, but good temperature co-efficient.)
At that time, I could find absolutely no negative-output devices of that type, and had to use a two-terminal (shunt reference) that could be used with either polarity.
(No, the circuit could not be changed to use a positive reference, unless I wanted to add a precision op amp and two really-good resistors to invert the voltage.)
A quick check today at Mouser showed no negative three-terminal references.

[PCB.Wiz]

I am looking for a fixed -2.5v LDO in small quantities. On Digikey, the least expensive is TPS72325DBVR at $2.99 while for +2.5V the least expensive is MIC5504-2.5YM5-TR at only $016. That's more than x10 more expensive. Why?

You chose a combination of negative and -2.5V, so have hit the 'rare' adder twice.
lcsc shows many negative regulators, but the cheapest are 79xx series, fixed voltage, and Torex 3.3V parts do show at better prices.
 

[wilhe_jo]

Depending on your design, just take a +5v and a +2.5v reg.

Your GND is -2.5v, the +2.5v rail gets 0V and the 5V reg gives you +2.5v

3.3V  5V and 12V give you several very cheap options.
An adjustable Lm317 is very cheap as well...

Another way is to introduce some "virtual GND" with some opamp.

So no need for negative in a lot of cases...hence the negative once are expensive.

73

[SiliconWizard]

It's all relative.

[langwadt]

Depending on your design, just take a +5v and a +2.5v reg.

Your GND is -2.5v, the +2.5v rail gets 0V and the 5V reg gives you +2.5v

except now your "0V" can't sink any more current that the load on "-2.5V"

[zapta]

I think that with positive +2.5 and +5V LDOs, the common rail will not be able to sink from the +2.5 rail, but something like this end of life IC would work TLE2426.

My circuit needs to be floating from the +5V supply and have +/-2.5V rails so I am using this D050505T-1WR4 for +/-5V and then +/-2.5v regulators for the regulated rails.

[langwadt]

how much current? a TL431 maybe?

[zapta]

about 10-20ma. Yes the TL431 may work and is less expensive.

I already placed the order at DigiKey. ;-)

[Konkedout]

I think you have problems if your +2V5 reg output (which is serving as ground) needs to handle ground return current from the +5 regulator.  But I do like LM317s for what they are good at.....

[Siwastaja]

Usually the linear regulator is only small part of the total cost of having a negative supply. Where is it coming from? A center-tapped transformer, for example, tends to be more expensive, too.

If lowest cost is important, you would usually try your best to modify the topology so that negative supply is not needed (e.g. by biasing all signals to halfway between GND and Vcc), even if that comes with some performance tradeoffs. When budget is not limited, then throwing a dual supply at all analog problems makes design easier.