0

I am working on a school project and I am struggling to design what is supposed to be an easy CC-amplifier. All other parts of the circuit are calculated and should work. The power amplifier is what's remaining.

I know the fact I should start with RE (R9 on schematic) and base it on the load line (estimated to 4 ohm). I am however unsure how to do it and I can not seem to find a good formula anywhere on the internet on to calculate the remaining resistors (R7 and R8).

Can anyone help?

Schematic of the circuit

JYelton
  • 34,119
  • 33
  • 145
  • 265
VinFurr
  • 1
  • 1
  • Have you used a circuit simulator before? R9 is not far off what is needed I reckon but, simulation would be really useful here. – Andy aka Nov 09 '23 at 14:32
  • yes, i do have access to multisim, but i would prefer to do calculations instead of only trying and failing. i believe it will give better results and it would look better on the report – VinFurr Nov 09 '23 at 14:39

2 Answers2

0

The value of R9 - in conjunction with the DC emitter current - should be chosen so that the DC voltage drop VE across R9 is app. 50% of the supply voltage.

Then - as a next step - the voltage divider R7-R8 must be designed so that it can provide a DC bias of app. (VE+1.3) volts (two base-emitter drops for the darlington). That`s all. (You can forget the small base current into the Darlington).

Don`t forget that the parallel combination R7||R8 loads the gain stage (T1) and reduces the gain of this stage. This should be taken into consideration for selecting the niveau of both resistors. But the quiescent DC current through R7-R8 should be at least 10 times the estimated base current into R7 (otherwise it must not be neglected).

LvW
  • 26,610
  • 3
  • 23
  • 56
  • but doesnt R7 and R8 also depend on R9 and the DC current going trough it? it did that for the CE amplifier – VinFurr Nov 09 '23 at 20:18
  • Yes - as I have mentioned: R7 and R8 have to provide a voltage which is app. 1.3 volts ABOVE the voltage VE across R9. – LvW Nov 10 '23 at 10:40
  • do you have any thoughts about high or low R9 value? what will be the benefits of high or low value? – VinFurr Nov 10 '23 at 16:08
  • The feedback effect depends not only on the R9 value but also on the emitter current - because it is the VOLTAGE across this resistor what matters. This voltage drop should be roughly 50 percent of the supply voltage – LvW Nov 10 '23 at 16:41
0

Here is a simple step-by-step process:

Step 1. A good starting point is to bias the voltage at Q3 emitter at half the supply voltage, which in this case will be 10V.
So Q3Ve=10V.

Step 2. Next, we select the bias current for Q3 emitter. Let's stick with your existing resistor where R9 = 20Ω. So that gives Q3 emitter bias current of: Q3Ie = 10V / 20Ω = 500mA.

This means that, with no signal, the power in Q3 and R9 will be 5W for each - I do hope you have a good heatsink for both of these components.. :-)

Step 3. Assuming Q3 Ve is 10V (half of Vcc), we then know that the bias voltage at Q2 base is:
Q2Vb = Q3Ve + 2 x Vbe. Assuming Vbe =0.7V, this gives:
Q2Vb = 10 + 2 x 0.7 = 10 + 1.4 = 11.4V

Step 4. Next, we select the bias current through R7 & R8. Assume that Q2 & Q3 both have DC current gain of 100, then the current gain of the Q2-Q3 Darlington is 10,000. We use this to calculate the base current into Q2: Q2Ib = Q3Ie / 10,000 = 500mA / 10,000 = 50uA.

Step 5. Next, we choose the current in R7 to be much larger than this base current, anywhere from 10 to 20 times this is usually good. If there are no other considerations to take into account, then I like to use a factor of 20 times, so that means the current in R7 will be: R7i = 50uA x 20 = 1mA.

However, in this case, R7||R8 will have a loading effect on the previous stage, the Common-Emitter voltage amplifier, which will reduce its gain. The higher this factor, then the lower the values for R7 & R8, which means the more the loading will be, and the lower the voltage gain from Q1. You can reduce the effect of this loading by using a "boot-strap", which I will discuss later. For now, let's assume that the factor of 20 is OK.

Step 6. Next, we calculate the current in R8, which will be 19 times rather than 20 times the base current of Q2 (since the current into the base of Q2 comes from R7 but does not flow into R8):
R8i = 50uA x 19 = 950uA.

Step 7. Now we know both (a) the voltage across each of these resistors, and (b) the currents through them, so we can now calculate resistance values for R7 & R8:-

R7 = (20V - 11.4V) / 1mA = 9.6V / 1mA = 9.6kΩ

R8 = 11.4V / 950uA = 12.0kΩ

Step 8. Now we need to select actual values that you can buy in a shop, so I would suggest using 10kΩ for R7, and 12kΩ for R8. I would not be too concerned at this stage about making R7 to be exactly 9.6kΩ using some combination of series and parallel resistors. You will find that some of the assumptions we have made may have some errors, for example, the gain of Q2 & Q3 may be different to 100, and their Vbe may not be 0.7V (I suspect Q2 may be somewhat less than this, say, about 0.65V, depending on which transistor you use).

Comment
Please check the values you have used for the capacitors and inductor that connect the speakers to the amplifier output. Something does not look quite right there. For example, I would have expected C6 (80uF) to be smaller than COUT (22uF)?

Also, I suspect you will need some capacitors connected from the collectors of Q2 & Q3 to ground. Inductor L1 will filter the current out of the power supply Vcc (which is good), but will cause a voltage "ripple" to the rest of the circuit. To reduce this ripple, place filter capacitors (eg: one or two big electrolytics) from Q3 collector to GND.

Bootstrap:
I mentioned earlier that R7 & R8 present a load on the output of Q1. The R7-R8 bias network will present a load to any signal present at Q1 collector. What happens is that R7-R8 will present as a resistor in parallel with the resistor in the collector of Q1 "RC". The value of this resistance is calculated as R7 in parallel with R8, which we shall call "Rbias". Using the values above, we get: Rbias = R7 || R8 = 10kΩ || 12kΩ = 5.45kΩ.

Putting this in parallel with resistor RC, we get:
RC' = Rc || Rbias = 840Ω || 5.45kΩ = 728Ω.

Notice that the effective collector resistor of Q1 has reduced from 840Ω to 728Ω (about 14%), which will reduce the voltage gain of the common-emitter amplifier by about the same %.

You can avoid this loss of voltage gain by applying a method called "bootstrapping". We can use the output of the amplifier (at Q3 emitter) to supply most of the current required to drive the voltage at the bias network located at Q2 base. We do this by adding two new components:

  1. A new capacitor (say, 10uF) connected from Q3 emitter to the join of R7 & R8.
  2. A new resistor (about 10k) inserted between the base of Q2 and the join of R7 & R8.

Note that the capacitor from Q1 must now connect to the base of Q2. (By the way, you have two capacitors marked "COUT" in the schematic.) The diagram below will give you an idea, this is for a single-transistor CC stage, in your case you have a Darlington so you just connect Cboot to the emitter of the second transistor.

enter image description here

Source of image: https://www.petervis.com/Radios/bootstrapping/transistor-bootstrapping.html

What this does is allow Q3 to supply most of the current required by R7||R8 to cause the voltage at Q2 base to "swing" with the signal, thus relieving Q1 from this burden. This means that the loading effect of R7||R8 is dramatically reduced, thus reducing the voltage gain loss from 14% to 1% or less.

This new capacitor & resistor, let's call them "Cboot" & "Rboot", "pulls up the voltage at Q2base by its own boot-straps", as the old saying goes, hence the name.

Links to some good information on the "boot-strap" technique:-
https://sound-au.com/articles/bootstrap.htm
Effect of bootstrapping in amplifier circuit

Fabio Barone
  • 2,101
  • 1
  • 17
  • wow, this looks really helpful, thank you! i will look into it

    by the way, would you think that this script would be good for the calculations? https://pastebin.com/uEWQerfM

    – VinFurr Nov 10 '23 at 15:36
  • the capacitor at 22uF is thought to filter out frequencies below 20Hz, tanken from this tutorial: https://blog.mbedded.ninja/electronics/circuit-design/bjt-common-emitter-amplifier/

    while the ones at the end are mant to be low pass and high pass filter for the speakers

    – VinFurr Nov 10 '23 at 15:45
  • re: pastebin link: I would not go the to trouble of writing some code for the calculation steps, some paper & a calculator is what I used, perhaps a simple spreadsheet (Excel, or Google Sheets) if necessary. Depends if you are planning to re-do this calculation many times, or just once to build this probject. – Fabio Barone Nov 10 '23 at 20:39
  • @VinFurr Hi again, re: capacitors. Yes, I understand what roles the capacitors play in your circuit. However, that link you provided only discusses capacitors associated with a single-stage CE BJT amplifier, it does not discuss how to couple a low-impedance load (such as a speaker) to the output of an amplifier. Although these two topics seem simuliar, they have very different requirements. Also, connecting capacitors and inductors together may have severe consequences, suggest you read up about *resonance, eg: https://en.wikipedia.org/wiki/LC_circuit – Fabio Barone Nov 10 '23 at 20:46
  • @VinFurr This post may also help you with your design: https://electronics.stackexchange.com/questions/688745/why-am-i-getting-heavy-output-distortion-from-a-bjt-amplifier-when-driving-an-8/ – Fabio Barone Nov 12 '23 at 21:58