From: Mike Monett
X-Mailer: Mozilla 2.02 (Win16; I)
Subject: Re: Amplifying stage with *negative* Voltage gain??
References: <firstname.lastname@example.org> <3DF263CE.3868@Spam.Bots> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <3DF8BEAE.6391@Spam.Bots>
Date: Thu, 12 Dec 2002 11:17:56 -0800
NNTP-Posting-Date: Thu, 12 Dec 2002 11:17:57 EST
Organization: Bell Sympatico
Paul Burridge wrote:
> On Thu, 12 Dec 2002 08:51:58 -0800, Mike Monett opined
> >This was one of my original questions - what are you using for an
> >oscillator, and why do you need an emitter follower?
> I built the oscillator using a 40m Xtal from a published circuit. My
> recollection is that you mustn't use the output straight from the osc
> itself or risk stability problems. Hence the idea of using a
> e/follower buffer stage. I then intend to make up a couple of voltage
> amplifying stages followed by a low power PA.
Yes, this makes sense from your other posts.
> >What kind of oscillator are you using? Can you post the schematic or give
> >us a url?
> I'm working on it right now.
> >What's Stripboard? If it's what I think it is, you'd be much better off
> >using copperclad as a ground plane, especially at these frequencies.
> It's AKA "Veroboard" (tm) if that's more to your liking. I just didn't
> want to advertise. :-)
> So far as a ground plane is concerned, I'm surprised such construction
> would be necessary at low HF such as 40m. UHF and beyond, then for
> sure, yes
Here's an example
This construction is useful for simple logic circuits and very low-performance
analog, such as low-gain audio amplifiers.
The problems are completely different than in simple audio amplifiers, where the
impedances are fairly high to get the desired gain. This rolls off the high
frequency response. In your work, you need to run at much lower impedances (50 ohms,
say) which greatly increases the bandwidth.
Nowadays, the most common transistors have ft's in the 250 MHz to 1GHz range. While
you may intend your circuit to run at lower frequencies, the transistors may have
At these frequencies, the strips provide excellent inductors. They form tank
circuits for unwanted parasitic oscillations, and provide crosstalk from one section
of a circuit to another. There is no way to provide good bypassing, and no RF ground
reference anywhere. You cannot provide a good ground reference for the scope probe
ground, which partially explains the confusing readings you posted.
Even on copperclad, grounding is a local concept. At VHF and UHF, the impedance of a
solid copper ground plane can easily generate fairly high voltages at different
points on the ground plane, especially in oscillators and power amplifiers. These
can have high circulating currents and generate amazingly large voltage differences.
The idea is to keep the currents localized with good layout and bypassing to
minimize crosstalk. When you are using transistors capable of gain at VHF or UHF,
you need to use the proper techniques to get them to do what you want, not what they
> >I saw in other posts you are interested in working at higher frequencies,
> >such as 40MHz. If so, you need to work with a ground plane for your
> Hmmmm. I probably would, but don't believe stray
> inductance/capacitance is likely to be a problem at anything below
> 50Mhz, if that's what your concern is. Still, I've been away from the
> hobby for some 20 years so stand to be corrected!
Just as an example, a 400 MHz colpitts oscillator I am working on right now has a
10pf tank capacitor and approximately 6.5nH tank inductor. The inductor is 2 turns
of #20 enamelled wire wound on a 1/16" drill rod. The leads are perhaps 3/16" long.
I had to use a microscope while winding the inductor to ensure a tight fit around
the drill shank.
The inductor is very small, but that's all it takes at these frequencies!