Commercial High Electron Mobility Transistors

I have to design a low noise amplifier in frequency range 10Khz to 10MHz. I am planning to use GaAs transistors for that. Any suggestions, which one to look on and where?

sure you want use a FET?
"Rumors" says that FET (also low noise HEMT) have higher 1/F noise than BJT. At so low freq., 1/F noise may be dominant.
Also, for the common source configuration, it's practically impossible to match the input gate. This because the gate is an "open circuit".

Have you taken in consideration the BJT ? (or HBT)

Thanks Sergio,
Another complication is that i have to cool this amplifier to cryogenic temperatures, thus i started with GaAs devices becaouse of Silcon freeze-out. Moreover, the commercial GaAs devices are typically HEMTs but for very high frequencies (>GHz). Can i still use these HEMTs for my freq. range? I have a high impedance source thus i was more inclined to use FETs. No never checked out the HBTs.

warning:
GaAs transistor have strong gain up to few GHz and you will have problems with stability. Be carefull.

Thanks Bouli,
This is exactly i am struggling with now STABILITY. I made a simple inverting amplifier circuit using 1 such GaAs HEMTs. Used all surface mount capacitors. And as i am trying to measure the DC charachterstics by simply connecting 2 power supplies at the gate and drain, the drain current rises normally with increasing drain voltage. But then after few volts it drops..and when i connected a spectrum analyzer at the drain, i see oscillations at 5 Ghz.
I have made this prototype in house by using a hand drill and was very happy to see my board (ps attachement) as it looked cleaner than my previous efforts.
So...hmm these devices oscillate even when trying to measure their dc characteristics.
Do you have any suggestions?

show me schema and bouard picture. then I wiil be able to advice you more.

I haven't been loged in. I havent seen that picture.
I'm not suprised taht this don't work.
So few advice:
about measuring DC characteristics:
You should have proper RF load and source impedance to have transistor stable. For beginnig try to use 50Ohm load fot RF (R in series with C, separate DC) and connect DC surce by RF choke.

About home made printed board. You can try method with IRON.
Do some nice PCB with some CAD tool, and print it as a mirror on nice colour newspaper for women. Paper should be sliipy, and don't worry that there are pictures in it.
after that clean your board using alcohol and abrasive paper. Then put printed paper face down on prepared board and iron it about 2 min. Iron should have about 200 C ged. Dot IRON to much. Is good to put board on IRON and push paper to board with flannel.
Then put board into water for few min. and take off the paper.
Then put the board into vapours of alcohol (not into alcohol) for 1 min.
Then put the board to acid.
Using this method you can get boards with 10mil lines.

Hi buoli,
I have attached the circuit schematic and the prototypes i built, one with BNC and the other one with SMA.
For the dc measurements i connect the power supplies VGG and VDD. I have bypassed both these power supplies well with surface mount capacitors. I vary VDD and measure the drain voltage with a multimeter to calculate the IDS. The drain current rises normally to some point and then drops suddenly. If probed by the spectrum analyser the peaks at 5 Ghz, 10 Ghz, 15 gHz are observed, when this current drops.
For prototype board i make islands for every net int he schematics. Please see the arrows to know which island corresponds to which net. Let me know if youhave any more questions for me to help.

Hallo
Some advice:
1. Decrease islands area. You propably have too big capacitance to the ground, and for 5GHz you have shorted circuit to the ground
2. Make your board as small as you can. (In RF area)
3. Do some drills to GND in source area.
3. Connet DC supply by SMD inductors or SMD resistances (more then 1k)
4. Try to change Rgate (1.5 Ohm)
5. You can try stabilize transistor by feedback.

Thanks,
I should have my inductors in by tomorrow. I did try to look into circuits for feedback to stabilize the FET, but didn't find any. Do you have any reference or suggestion, i can try.