I would select a linear -- non-switcher -- PSU which can
supply more than enough current , add more electrolytic
and ceramic caps with short leads  

AND FINALLY:



 

----------------------------
Please read FAQ240-1032: How to ask a meaningful question
My WEB: <http://geocities.com/nbucska/>

"An amplifier" is a bit vague. The PSU recommendations for a 1GHz-5GHz amplifier are quite different to those for a DC-1GHz amplifier.

Hi,

Well, I apologize for the lack of information. I´m testing a LNA up to 1.5GHz.

Thanks for your answers.

Should I assume from "1.5GHz" you mean DC to 1.5GHz? If so you really shouldn't power it directly from a bench power supply UNLESS it is a manufactured part with an inbuilt feedthru and regulator. In this case the feedthru/regulator will do all the work for you and the power supply requirements are not especially critical. If this is a "homemade" amplifier from MMICs then you should use your own linear regulator with short leads on a little pice of board right up close to the amplifier. At 1.5GHz you are susceptible to several mobile phone wavebands.

The fact that your existing PSU is too noisy suggest the PSRR of the amplifier is not very good anyway. Are you sure the noise is coming from the wires and not coming from the power supply as say spurious magnetic fields?

Make sure you put 10nF to 100nF across the feedthru into the amp package to minimise HF noise.

I have used an Agilent E3631A for this sort of amplication although the amplifiers in question didn't work below 1GHz. There is also a lower power triple output supply which I couldn't find on their current website.

Obvioulsy these are not "recommendations" because you still haven't given enough details and frankly other local situational requirements may be relevant (do you need the power supply for other applications as well, how much current does it need to supply, what power rail(s) does the amp have, and so forth).

Hi,

Thanks for your comments, I found them quite useful.
Well I´m testing LNA amplifiers, AM, FM and DAB bands. I test them inside a Farady chamber. The amplifiers are for mass production but are still in the development phase. They have a regulator, using a zener.

I have also found in the Agilent website that the E3631 PSU has very good noise parameters.

Could you please explain me what is a feedthru?

PSRR issues, well that´s a good point and I´ll check that too.

Thanks!

Are you sure your noise problem isn't your zener?  Zeners make GREAT broadband noise sources.

Peter

Leiser:
The FT is a tubular capacitor; the inner electrode has
one lead which is coaxial with both electrodes and
protrudes at both ends . The outer electrode
of the C is a solder plated ring  to  be soldered
around to the perimeter of the hole in the shield.

Its effect is equivalent to a perpendicular wire going
through the hole. On each side of the shield all around
the wire there are a bunch of capacitors connecting
the wire to the rim of the hole in the shield with
infinitesimally short leads. ( i.e. with L=0 )

----------------------------
Please read FAQ240-1032: How to ask a meaningful question
My WEB: <http://geocities.com/nbucska/>

Batteries are good and quiet if you really want to keep the power supply switching noise away.
If all else fails, try that.

kch