Hello,

I'm getting some concepts confused.

I'm building a 2 stage BJT amplifier, common emitter and emitter follower.
The behaivour of the circuit is quite accurate, as I measured values very near the ones I calculated on paper. The problem is the cutting frequency of the AC coupling capacitor at the input of first stage.

I use a 50 ohm signal generator, with a 1Mohm BNC scope probe to inject the AC component, trough a 100 nF capacitor. The stage input impedance is around 500 ohms.

The cutting frequency should be around 3 kHz, but the -3dB point, compared to max value attained at high frequency, is reached at 1.5KHz. It seems that the signal generator is not showing 50 ohm, but around 400 ohm, which forms a voltage divider with the input impedance (~500ohm) of the circuit.

Why? I tried to source a simple RC with a higher 2,2kohm and same 100 nF capacitor, and it corroborates it. The source is aorund 400 ohm output impedance, according to the voltage divider it makes.

Back to the bjt amplifier, at 3kHz, the theorical cutting frequency, the AC voltage in the base is -3dB the one present in the signal generator output (at the tip of the probe, to be exact). But NOT -3dB of the max value at higher frequencies, because the voltage on the tip of the generator's probe raises at frequencies near the cutting frequency, obviosly, because the capacitor impedance raises. So, the cutting frequency is the point at which the input is -3dB from the current voltage on the generator, but NOT -3dB from the maximun voltage at high frequencies. How can this be handled?

Thank you

Attached Images

bjt.jpg (47.6 KB, 24 views)

 

If it helps, I calculate the input impedance of the amplifier to be 553 ohms and the -3db frequency to be 2.879khz. You may want to try a resistor in series with the hot lead of the generator around 500 ohms.

__________________
The great thing about electronics is unlimited ways to do the job. The only limit is one\'s imagination. I generally think my way is best.
Show me a different way. I have an open mind.

 

If you're using a standard scope attenuator probe to connect the 50-ohm generator output to the 500 ohm input of an amplifier, I wouldn't expect you get get hardly any signal into the amp at all. If the generator is feeding the probe 10v, you should only be getting 555µv into the amp input. A scope probe, even a 1X non-attenuating probe, should never be used to inject the signal from an generator into an amp. Even a 1X probe may have several hundred ohms resistance from probe tip to BNC center conductor.

Dean

__________________
Dean Huster, Electronics Curmudgeon
Contributing Editor emeritus, "Q & A", of the former "Poptronics" magazine (formerly "Popular Electronics" and "Electronics Now" magazines).

R.I.P.

 

yes... I'm using a 1X 1Mohm scope probe. It exhibits around 400 ohms, yes. So, this is normal?
I have no other BNC probe.

 

First and foremost a 50 Ohm output of a signal generator MUST have a 50 Ohm load connected in order to operate IAW specifications. Check your owners manual.

 

I have no manual. I bought it used.
When I got it, a year or so ago, I thought it was used, appart from RF circuits, to generate variable voltage signal, to source a transistor circuit, for example. I initially used it as a digital clock generator (it has a TTL output, but only for square waveforms). But it definitely generates good sine, saw and square waveforms when sourcing a high impedance load (above 1k ohm, for example).

It has a 600 ohm output too. But the lower the output impedance of a voltage source, the better, so I always use the 50 ohm, and TTL output for driving IC clocks.

If this is not the correct tool, what kind of equipment is needed to source analog circuits, like a multistage transistor amplifier or OAmp ?
Maybe I'm doing some strange things, but I'm self-taught. I'm finishing electronic engineering, but it is very theory oriented (close to physics) and the real world analog labs start next year, my last, I hope. So thank you so much for helping me out to reach the practical knowledge! :-)

 

Take a look at this link.
http://www1.electusdistribution.com....d/impmatch.pdf
I think the tool you are using should be fine, just need to understand the issues with matching.
Instead of a scope probe for injection, try a cable with appropriate sig gen connector on one end, and pig tails on the other side. This may require soldering to your board under test.

 

I know about impedance matching, L networks, transmission lines... but working in audio frequencies... I usually forget about microwave and reflexions, and try to drive an impedance with a lower output impedance driver, so voltage is mostly droped in the load.
What kind of probe should I use to avoid getting around 400 ohm instead of 50 ohm output from the generator? a 50 ohm coax cable with a BNC connector, and some kind of tip in the other end?

 

Just a piece of cable from the generator to your input - an obvious choice is 50 ohm coax as it will fit in a BNC plug properly.

A scope probe is for the input to a scope (and perhaps a frequency counter), not for anything else.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

If I remember correctly we connected the 50 ohm to the generator and the other end of the cable was not terminated because it usually went to a much higher input impedance connection.
So essentially no probe is needed.

 

Thank you! Now it is clear. I'll get a coax cable and BNC connector, and try again.
I wander why, though, the scope probe exibits ~400 ohms when used with the generator, and 1 MOhm when used in a scope

 

It doesn't, I can't say I've ever measured a scope probe (it's not something there would ever be any need to know), but it's the impedance of the SCOPE that's 1MOhm - nothing to do with the probe.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk