LNA oscillating or bias circuit not starting up

Hi,

I am building a S-band to L-Band downconverter and my front end, mixer and LO is working great, however when I switch my IF amp (BFP740 amplifier design) on, the entire spectrum is full of spurs, MASSIVE ones. but if I touch the input or output port of the amplifier circuit, everything works as it is supposed to, I can take away the RF and re-apply it with no problem, but if I switch that amp of and on again the spurs come back. I have even made a little box around it for better grounding, no luck. I think the amplifier switches on and off the whole time untill you touch it.

Please help!

It oscillates or Power Supply is not well filtered.
Connect a 1K or more/less resistor at the output/input ( on the AC path).It will probably stop to oscillate.If it doesn't, check your filtering of power supply..

If you measure the difference in frequency between the spurious outputs, this is the frequency at which it is oscillating. This will give you a clue as where your problem is. If its very low frequency then its bias or power supply, if its RF (IF) then its a RF decoupling problem.
Frank

chuckey said:

If you measure the difference in frequency between the spurious outputs, this is the frequency at which it is oscillating. This will give you a clue as where your problem is. If its very low frequency then its bias or power supply, if its RF (IF) then its a RF decoupling problem.
Frank

Click to expand...

I have an ultra low noise LDO and filtering on the supply, decoupling is good, have a 100pF and 1nF everywhere necessary. The amplifier by itself works perfectly, its only when it gets the input signal from the LO that it does this

- - - Updated - - -

BigBoss said:

It oscillates or Power Supply is not well filtered.
Connect a 1K or more/less resistor at the output/input ( on the AC path).It will probably stop to oscillate.If it doesn't, check your filtering of power supply..

Click to expand...

1K resistor to ground?

Your IF amp is unstable, have you decoupled the Vcc and bias lines with at least a .1 MF +10 MF to stop LF feedback, is the amp DC coupled? Have you a proper IF filter preceeding it? and interstage screening? You still have not let on at what frequency its oscillating, depending on your IF frequency, a ferrite bead on any gate lead will help to stop VHF/ UHF gain and hence oscillations.
Frank

avatarET said:

I have an ultra low noise LDO and filtering on the supply, decoupling is good, have a 100pF and 1nF everywhere necessary. The amplifier by itself works perfectly, its only when it gets the input signal from the LO that it does this

- - - Updated - - -



1K resistor to ground?

Click to expand...

Yes, 1K ( or try the other values ) resistor to GND from Input or Output port.The position can vary so your should try some critical points on RF path.

chuckey said:

Your IF amp is unstable, have you decoupled the Vcc and bias lines with at least a .1 MF +10 MF to stop LF feedback, is the amp DC coupled? Have you a proper IF filter preceeding it? and interstage screening? You still have not let on at what frequency its oscillating, depending on your IF frequency, a ferrite bead on any gate lead will help to stop VHF/ UHF gain and hence oscillations.
Frank

Click to expand...

Hi Frank,

Yes I have a 100 pF, 100nF and 1 uF on the supply, I have a very sharp BPF before and after the amplifier. looks like it is oscillating at around 1.4GHz, 1.8-2GHz and around 4GHz

- - - Updated - - -

okay, I have the amplifier running off a separate power supply and when starting up it draws half the current that it is supposed to, after touching it it draws the correct amount.

maybe it is your grounding issue, can you post your PCB and SCH?

avatarET said:

Hi,

(BFP740 amplifier design) on, the entire spectrum is full of spurs,

Please help!

Click to expand...

It`s funny transistor. BJT SiGe with ft above 26GHz and Spar up to 20GHz and exist @ different packages and some of them are sens to a right VIA s and at right place. It is quite important. Otherwise you can get a lot of effects.