Groundplane questions

No, but it's usually reserved for the power rail. In your case, there is very little attached to the power rail and what is lends itself to the power bus you have. I'd make that as wide as possible, though. I quickly looked at the layout and think the bottom pour is pretty much tied together, or can easily be. One reason to tie it to ground is if you want to mount this directly to a metal enclosure. If course that can be done with pads and stand-offs, too.

If you *don't* ground the bottom, watch that TO220 (7812) package. The tab is ground, so the pad under it should be cut out. If that portion of the pour is ground, it's a (small) heat sink. Perhaps widen the power trace at the top into a pour and make the rest ground?

What SMT?

Incorrect acro usage.

He probably means surface mount transistor.

Use a bit of common sense.

Oh... wait..

At a quick glance you may have a problem with capacitance to ground if you ground plane the bottom near the input, after all the input cap is only 0.3pF so it won't take much extra capacitance to affect the circuit.

7812 mounts on top (red) and that will be GRD.

Why? Bigger would spread any heat. I expect to make bottom pour GRD also.

I drew a red line to where the SMT will mount on the bottomside. It mounts to the three little pads.

Thanks, Mike

Ok, I see what your saying, the original also has the groundplane near the .3pf cap. I could eliminate the groundplane on both sides, this would only add about

1/8" of distance, because the grounded box is the next layer. I'll look at it, if there is another place for the mounting screw, it might help in reducing capacitance. Thanks, Mikek

Call Agilent. SMT is all over their site. They use it in place of "surface mount transistor". MikeK

"amdx" wrote in news:abb22$4d3b10a4$18ec6dd7$ snipped-for-privacy@KNOLOGY.NET:

No, they use it as the acronym for "Surface Mount Technology". That includes not only transistors but also capacitors, diodes, microcircuits and resistors.

--
Bob Q.
PA is y I've altered my address.

Hence my claim of incorrect usage, as you are using the term as "Surface Mount Technology" to refer to any such type component that gets mounted in this way.

That was once also "SMD" or "Surface Mount Device". That one IS correct.

It (SMT) was never really 'correctly' used to mean "Surface Mount Transistor", regardless of whether it was or is used that way anywhere.

It is similar, but not quite to the "battery" and "cell" problem.

.

red

I don't know about antenna input circuits. But if you have high input impedance and want to keep the input capacitance low, you may not want the ground plane right near the input. Is the circuit going to live inside a metal box?

George H.

I don't know about antenna input circuits. But if you have high input impedance and want to keep the input capacitance low, you may not want the ground plane right near the input. Is the circuit going to live inside a metal box?

George H. Yes it is in a box, Take a look at the circuit,

The final design is at the bottom, the last two photos show the input circuit best. Mikek

nd.

),

Yeah, I looked at that quickly. Are you going to hang the input up in the air?

George H.

Adding an extra ground plane just increases the stray capacitances to ground. If the circuit can tolerate this, this is just fine.

The extra ground plane reduces the effects of other stray capacitances e.g. to the box, so the PCB works the same on the table as well as when installed in a metallic box. For EMC issues, the extra ground plane is also fine.

The stray capacitances through the PCB will also have temperature coefficients as well as the humidity may affect this capacitance and at high frequencies (UHF+) the dielectric loss issues may be an issue. For this reason, through board stray capacitances should be avoided close to free running LC oscillator tank circuits, since those variable capacitances will change the oscillator frequency.

In order to oscillate at 5 GHz, the transistor must have fT in the order of several GHz, do you really need such fast transistors in a circuit intended for 10 MHz ?

Microwave transistors tend to oscillate if there is too much stray inductance between the emitter and ground and for this reason, microwave transistors are often flush mounted with emitter leads immediately soldered to the ground plane at the transistor package. I could not find any microwave transistors in that layout.

Yeah, I looked at that quickly. Are you going to hang the input up in the air?

George H.

Yes, the input is designed as a 0.3pf capacitor, with air as the dielectric. Although I do wonder about the size of the plates. Seems he could have made smaller plates and moved them closer together. With smaller plates the surrounding materials would have less effect. Mikek

The author used a BFR92A, a 5 Ghz surface mount transistor. He causioned it may oscillate and gave an alternate slower transistor, I have the BFR92A. So I need to design to prevent any problem. "How" ? Do I move the upper and lower groundplanes away from the area around the smt.

The lead is rather long from the emitter, could that be a problem? It doesn't go directly to ground.

Ya, seems this is a problem. In this picture you will see a red line going to the position where the smt mounts.

I have three little pads for it to lay on. I'm probably looking for trouble where there is none, but I only want to do this once. :-) Thanks, Mikek

BTW, I downloaded PCBExpress, for this PCB, it is a breeze to use, the learning curve is quick. 15 years ago I used another program, much more difficult and buggy.