Amplifier oscillates at 100MHz

[anli]

I think, Jan Didden is the most close to the reason of oscillations: probaly half of all different scopes I have worked with oscillate with shorted probe and probe gnd when max (and near) Y scope sensitivity is selected. Such oscillation takes place wile measurements too.

I don't insist on I have worked with the best scopes, but it is a reality of my life 🙂.

 

[Duo]

Yes, I have observed that the oscillation went away when the amplifier power was disconnected. (I suppose there is still the possibility that it is picking something up.

 

[jleaman]

Maybe build up a small pcb board and see if the problem goes away..

 

[Duo]

I'm pretty sure the scope isn't the problem. It's a high end scope and seems to be immune to a lot of the problems that most scopes I've worked with have.

 

[jleaman]

I'm pretty sure the scope isn't the problem. It's a high end scope and seems to be immune to a lot of the problems that most scopes I've worked with have.

What kinda scope is it. ? what is the frequancy ? Could it be your probe's ?

 

[Duo]

The scope is a Tektronix 2247A with original probes that came with it.

I think I'll try it later with some of my other scopes; though I'm far more likely to believe it's the breadboard and not a scope problem.

 

[amplifierguru]

You could contact the 'chief' amplifier engineer at Madrigal - he had a similar problem when I landed the revelation on him that LAYOUT MATTERED, after I advised a reroute of a Vas output PCB track away from the input diff'l pair on his rehash (of someone else's competent design) project amp. Immediately the THD dropped 2 orders of magnitude - to which he replied "Does layout matter?" . I asked "Who did the PCB layout?". "The draftsman!" "What criteria?" "Fit!?" He then asked how to do it. I explained that you need to identify all the sensitive low level areas/tracks and all the high current and voltage radiating tracks/blocks and keep them apart. "Can you identify them for me?". "Sure", I replied. I marked out the various sections of his 12" sq board and he hacksawed it up, connecting sections with 3-4" jumpers, fired it up while he moved the pieces around in 3D space!

Then it went up in flames and the engineering manager came racing out with a fire extinguisher and nuked it!


Seriously! You had to be there!

The *******s picked my brains for 4 months then didn't pay me!

 

[Duo]

Yes, I do understand that layout is of utmost importance in amplifiers. As I said, my PC boards work very very well, never any stability issues.

I tried this amplifier on the breadboard again at home today (I did the other one at the highschool) and this time it works perfectly as expected.

The only thing I did differently this time was spread the various stages of the amplifier farther apart. I also used a start ground instead of the rail in the breadboard. I found that 68pF was enough for the stabilization capacitor and the amplifier works well.

I still haven't determined exactly what the deal was with the 100MHz oscillation with the one at the school, but I intend to figure out the problem before I re-do the layout anyway. It would be interesting to know.

 

[amplifierguru]

So your 100MHz is gone?

 

[Eva]

100Mhz is too high a frequency for a global feedback loop oscillation. It must be local oscillation across one or two transistors and breadboard capacitance is the most probable cause

As you have found, this is a layout issue

 

[Duo]

Yes, that is what I figured it was.