For whatever reason, this circuit simulates just fine but has no gain in the lab. Why is this? The transistor seems to test fine. I haven't swept if for an IV curve yet, though.
On your breadboard, what's the collector DC voltage?
Tweak R2 for around 2.5VDC.
R3 may need a different value too, it's basically in parallel with R2.
The transistor may be cut off.
R3 should be >> R2 so it doesn't load down the voltage divider.
This is why an emitter resistor is typically used, instead of a base resistor.
Tweak R2 for around 2.5VDC.
R3 may need a different value too, it's basically in parallel with R2.
The transistor may be cut off.
R3 should be >> R2 so it doesn't load down the voltage divider.
This is why an emitter resistor is typically used, instead of a base resistor.
Last edited:
It was Vcc (5vdc) which indicated to me that the transistor was not turning on. I was going to swap transistors but I don't have another 2n3904.
I also recall that I forgot the input cap, so the input was dc coupled to the signal generator....I don't have access to the board at the moment, so I can't check if that was the issue. Does that sound like it was the problem?
I also recall that I forgot the input cap, so the input was dc coupled to the signal generator....I don't have access to the board at the moment, so I can't check if that was the issue. Does that sound like it was the problem?
Yes, the Q is cut off. Hard to say if the generator was the problem.
The DC voltage at the node with R1, R2, R3 seems to be too low, due to R3 being too small compared to R2.
Usually we want a transistor circuit to be insensitive to the value of the base DC current, since beta varies.
Then the R1/R2 voltage divider should be low enough impedance so R3 and the transistor don't load it much.
But with an emitter resistor, the divider sets the DC voltage across the emitter resistor
(less the 0.6V junction drop and small R3 drop), and hence sets the emitter current Ie = Ve / Re.
The DC voltage at the node with R1, R2, R3 seems to be too low, due to R3 being too small compared to R2.
Usually we want a transistor circuit to be insensitive to the value of the base DC current, since beta varies.
Then the R1/R2 voltage divider should be low enough impedance so R3 and the transistor don't load it much.
But with an emitter resistor, the divider sets the DC voltage across the emitter resistor
(less the 0.6V junction drop and small R3 drop), and hence sets the emitter current Ie = Ve / Re.
As drawn the bias point is too sensitive to the HFE of the 2N3904. Include a resistor in the emitter.
A quick estimate - the base is fed from ~1V and 10k that's 0.4V/10k so 40uA base current. With a typical HFE of 200 that's 8mA collector current so 1.6V across the load resistor. But HFE could be as low as 70 giving ~0.6V across R4. For a robust, repeatable design you don't want the bias point to be so sensitive to HFE, hence emitter degeneration is a thing.
A quick estimate - the base is fed from ~1V and 10k that's 0.4V/10k so 40uA base current. With a typical HFE of 200 that's 8mA collector current so 1.6V across the load resistor. But HFE could be as low as 70 giving ~0.6V across R4. For a robust, repeatable design you don't want the bias point to be so sensitive to HFE, hence emitter degeneration is a thing.
Yeah the Hfe I was seeing in the LTspice model was about 300 if I am interpreting it correctly. I am just looking at Ib versus Ic and it's a factor of 300 for DC.
I appreciate the suggestions. It really helped adding the emitter resistor. I didn't want to before because I found it hard to maximize the voltage swing with it there -- I am aiming for 3V swing. I am sure there is some math I am missing on that side.
I appreciate the suggestions. It really helped adding the emitter resistor. I didn't want to before because I found it hard to maximize the voltage swing with it there -- I am aiming for 3V swing. I am sure there is some math I am missing on that side.
The single transistor amps I've seen that don't use an emitter R tend to get the base bias from the collector via a resistor. I've not looked into how sensitive that arrangement is to HFE. Its certainly fewer resistors.
3V swing certainly should be achievable when emitter degeneration's used so as to obtain a predictable and stable bias point.
3V swing certainly should be achievable when emitter degeneration's used so as to obtain a predictable and stable bias point.
The real problem is the 5V supply. Is that in stone?
More DC voltage would certainly help to design a better circuit.
How much voltage gain is needed?
Do you want 3V peak to peak, 3V peak, or 3V rms?
Only 3V peak to peak is even possible with a 5V supply.
More DC voltage would certainly help to design a better circuit.
How much voltage gain is needed?
Do you want 3V peak to peak, 3V peak, or 3V rms?
Only 3V peak to peak is even possible with a 5V supply.
Last edited:
Yes, married to the 5v supply. It's being powered by a raspberry pi.
3vpp which the circuit appears to be doing per the simulation above thanks to your suggestions. I've no idea how much distortion is really there, however. Voltage gain over all needs to go from 100mVpp to 3Vpp, so 30.
3vpp which the circuit appears to be doing per the simulation above thanks to your suggestions. I've no idea how much distortion is really there, however. Voltage gain over all needs to go from 100mVpp to 3Vpp, so 30.
How do you feel about op amps? That would be pretty easy to do with a rail to rail type.
And it would definitely have low distortion. There are many, the cheap TLV9052 would work, for example.
And it would definitely have low distortion. There are many, the cheap TLV9052 would work, for example.
Last edited:
Voltage gain is roughly R4/R3 so about 50 in the schematic of post #6. To hit 30, increase R3 to 8R2.
I've used op amps so much I've forgot how to properly bias a bjt and I somehow ended up in a situation where all we have in the labs are bjts. I might just bring some op amps from home but I didn't want to give up that easily. If going for 3vpp swings the collector current so much that the transistor goes outside of its linear region, then I suppose I don't have a choice ha.
Distortion from a single transistor classA amp is relatively innocuous, subjectively, being low order.
I would suggest the last example here, if you use a transistor.
https://www.electronics-tutorials.ws/amplifier/transistor-biasing.html
https://www.electronics-tutorials.ws/amplifier/transistor-biasing.html
This (original) circuit has no DC feedback so it will only work if the transistor gain is exactly the simulated value. A useful circuit works with a wide range of transistor gain. You can use the emitter resistor, or you can feed R1 from the collector, with R1: R1/R2= (5/2-0.65)/0.65, R1 = ~28K, except I(R1) should be at least Ic/20 so that the transistor Ib is a small part of I(R1) current, ie R1 ~= 3k7, R2 = 1k3. If you don't like a 1K input impedance, then a 250 Ohm output is too low for a one transistor amp. BJT gain and JFET Idss are unpredictable. You can fiddle and select with parts to make a real transistor work, or you can make a circuit where the actual value doesn't matter much. Only resistors and certain capacitors are made with precisions.
The classic beginner circuit has no R2, and R1 is connected across the transistor C-B, where R1 = R4*HFE. If the gain is high Vc is low, if it is low Vc is high but it never goes completely off or on.
The classic beginner circuit has no R2, and R1 is connected across the transistor C-B, where R1 = R4*HFE. If the gain is high Vc is low, if it is low Vc is high but it never goes completely off or on.
Refer to a good electronics text book that explains how to bias an NPN transistor in Common Emitter mode. There are programs/apps that will calculate R1 R2 R3 and R4 for a specified gain, that's assuming your required gain is a lot lower than the transistor current gain (hfe). The resistor R3 provides the negative feedback needed to stabilise the gain and achieve linearity. These resistors R1 R2 R3 and R4 set the transistor DC operating point within the transistor's specification curves. For some applications the DC operating point bias current can be set very small (microAmps) whereas for other applications the bias current is set much higher (milliAmps). The resistor R5 is so high that it need not be included. As a general comment your resistors R4 and R3 are low values for what is essentially no output load (R5 very high). As a starting point set R4 to say 2K7 and then calculate R1, R2 and R3.
One interesting link:- preamp The equations for calculating R1 R2 R3 and R4 are more detailed than this link explains. The value of hfe doesn't matter as long as its much higher than your required gain. You can think of hfe as something like the open loop gain of an OpAmp.
Its worth knowing about the three main ways of connecting a bipolar transistor i.e. Common Emitter, Emitter Follower and Common Base. The Common Base configuration is seldom used but is very interesting and I have seen it used in very clever ways.
One interesting link:- preamp The equations for calculating R1 R2 R3 and R4 are more detailed than this link explains. The value of hfe doesn't matter as long as its much higher than your required gain. You can think of hfe as something like the open loop gain of an OpAmp.
Its worth knowing about the three main ways of connecting a bipolar transistor i.e. Common Emitter, Emitter Follower and Common Base. The Common Base configuration is seldom used but is very interesting and I have seen it used in very clever ways.
Last edited:
@rectifryer , as already said by others, emitter resistor is a must in such circuit to keep stable DC bias. In case you need a full circuit gain, put an electrolytic capacitor in parallel with the emitter resistor. You will have full gain and stable bias at once.
@PMA if you put a high value capacitor in parallel with the emitter resistor, AC gain will be maximum allowed by hfe and hfe varies a heck of a lot with temperature, particular transistor and DC Collector current. For AC you are operating without feedback in open loop. The capacitor you refer to is usually fitted in multi-stage amps and for these negative feedback to set gain is provided from output to input.
The point being that for the simple R1, R2, Rl, Re "Voltage Divider Circuit Bias" good circuit values of these resistors will set a precise stable gain (typically small compared to hfe). As the input signal gets more positive it turns the transistor on harder and the voltage drop across Re increases thus giving negative feedback to stabilise the voltage gain.
The point being that for the simple R1, R2, Rl, Re "Voltage Divider Circuit Bias" good circuit values of these resistors will set a precise stable gain (typically small compared to hfe). As the input signal gets more positive it turns the transistor on harder and the voltage drop across Re increases thus giving negative feedback to stabilise the voltage gain.
Last edited:
I think the main problem is the nonexistent input coupling capacitor (post #3). The signal generator just shorts the voltage divider to ground. Is the voltage at the collector reasonable when you disconnect the generator?
By the way, if the circuit did work, it would distort a lot, except at very low signal levels.
By the way, if the circuit did work, it would distort a lot, except at very low signal levels.
Last edited: