Hi all-

One of the engineers at my work is suggesting changing from a LDO to a series pass transistor for voltage regulation as a cost reduction. What are some pitfalls to doing this other than the obvious ones such as less accurate, no short circuit protection, no thermal protection, etc.

Thanks

 

Hi all-

One of the engineers at my work is suggesting changing from a LDO to a series pass transistor for voltage regulation as a cost reduction. What are some pitfalls to doing this other than the obvious ones such as less accurate, no short circuit protection, no thermal protection, etc.

Thanks

Your linear regulator *CONTAINS* a series pass transistor. NPN emitter follower in a standard positive regulator, LDO uses a PNP with its collector feeding the load - this configuration eliminates the Vbe overhead.

 

Hi Ian,

Thanks for your reply. Yes, I understand that the LDO contains a series pass transistor. I guess my question is really why not always use only a transistor which is cheaper than a packaged LDO.

Thanks

 

Hi Ian,

Thanks for your reply. Yes, I understand that the LDO contains a series pass transistor. I guess my question is really why not always use only a transistor which is cheaper than a packaged LDO.

Thanks

Its not cheaper after you add all the other components. Less solder joints = less man hours and less connections to fail.

The 3-terminal variety all include current limiting and thermal shutdown.

 

It takes added circuitry to control the series pass transistor.
How can the sum of those additional parts, plus assembly cost, be cheaper than using an LDO?

I think that "engineer" doesn't know what he's talking about.
He sounds more like a bean-counting manager than an engineer.

 

It takes added circuitry to control the series pass transistor.
How can the sum of those additional parts, plus assembly cost, be cheaper than using an LDO?

I think that "engineer" doesn't know what he's talking about.
He sounds more like a bean-counting manager than an engineer.

There can be good reasons for using discretes - but I'd have to think about it...............

 

It takes added circuitry to control the series pass transistor.
How can the sum of those additional parts, plus assembly cost, be cheaper than using an LDO?

Actually, the NPN transistor and zener diode are cheaper than the LDO (by pennies, of course). To your point, assembly may add some cost, but they only focus on BOM cost.

 

.... the NPN transistor and zener diode are cheaper than the LDO (by pennies, of course).

It also requires at least one resistor.
And that circuit will have much poorer regulation and voltage output accuracy.
But if pennies are important and the poorer performance of the zener-transistor circuit is tolerable, and there's no possibility of accidentally shorting its output, then you may have no choice but to bow to the bean counters.

 

It also requires at least one resistor.
And that circuit will have much poorer regulation and voltage output accuracy.
But if pennies are important and the poorer performance of the zener-transistor circuit is tolerable, and there's no possibility of accidentally shorting its output, then you may have no choice but to bow to the bean counters.

Every once in a while you find a chip that's cheaper than the minimalist discrete solution.

Its more luck than anything - but there's a few out there.