Sorry, I intended to write 0.47uF, which would have a cutoff frequency below 10 Hz.
1uF is fine, of course.
Yes, I normally use electrolytics for values 0.47uF and above.
No cork sniffing here.
At most, I sniff soft drink crown caps and can tabs
Easy, Electronics 101 schoolbook example, but with more realistic values than those usually shown there.
1) load resistor R2: 4K7
Nice midrange value.
Not too high to unduly increase output impedance, not too low to increase consumption or lower *input* impedance
The design is a compromise if we use *just* 1 transistor, so don't expect miracles.
2) Gain = R2/R3 so for 10X, R3=470r
3) choose collector current so collector sits at around 1/2 +V
In this case:
I choose +V around 25V (within the OP design limits) so Vc around 12V .
Note: my calculations typically use "Engineering" precision, which says that it's useless (except in a few special situations) to work with more than 2 or 3 significant digits, when parts often have 5 or 10%tolerance, and transistors a 3:1 gain spread, so I usually round to nearest voltage and parts values to nearest "standard series" value.
Back to calculation:
4) We have 12V (as I said, I'm always rounding values here) across R2 so Ic=2.5mA
So Emitter sits at 2.5mA x 470 r above ground .
Ve=1.2V
5) we need to bias the base.
It will sit at (0.65V + Ve) above ground= 1.9V
6) the base will need a biasing current of:
2.5mA/Hfe
I choose a cheap general purpose widely available transistor which I personally buy in bags of 500 or cases of 1000 (or "machine gun strips" of 1000) : BC547C
I also stock its brother, BC557C and use them **everywhere** .
Don't suggest you stock that many, but any DIY should buy at least 20 of each (for peanuts) and have them available.
VERY useful.
I use BC547B or C .
The letter indicates Hfe , 250 or 500 respectively.
If I design considering Hfe around 300, both work .
Your KT603 has a very poor and inconsistent HFe of "10 to 80" , not useful here.
The preamp stage would have *very* low impedance, require larger input cap, etc.
You save nothing.
As I see them , they are only suitable to, say, drive a Led or something similar.
Back to biasing.
The base current will be 2.5mA/300 (Hfe) = 8uA
A standard textbook compromise is to choose current through the biasing divider to be 10X the actual needed base current, so as to minimize its influence.
Fine, it works, but it needlessly lowers input impedance.
School typically designs something that works in the Lab and simulates well.
Fine with me.
But I design commercially so stuff will always be used connected to "something else" , so I'd find raising input impedance a bit a real advantage.
So I choose biasing divider current "only" 4X base current.
Which is a safe bet because the transistor I chose may err having *higher* than normal Hfe (I chose B or C
)Since we will *still* have the actual base current (Ib) flowing through it, they will add, so we have 4 x Ib + Ib = 5 Ib = 5x8uA = 40uA.
So: R2=22V/40uA=550000 ohms.
So a 560K resistor would be fine, but since I know I rounded values a few times along the way, I choose the next lowest value (meaning I allow for a slightly higher current) resistor value.
Which would be 470K .
So R5=470K.
Yes, some Simulator package might give a value of, say, 528K or 493K or whatever, but let's go back to the real world.
7) R6 is what's needed to put around 1.9V at Q1's base, one of the original conditions.
Of course I choose a "standard" value close enough.
8) Input impedance:
You will find a "reflected impedance" (maybe there's a better name for it) of Hfe x R3 = 300 x 470r = 140K
Which is in parallel with R6, so the actual *stage* input impedance will be around 40K .
9) input capacitor: whatever's needed to reach "low enough" (your choice) considering the input impedance.
Being a MI amp designer and builder, I usually cut below, say, 60Hz in Guitar amps and below 30 Hz for bass amps, to avoid unnecessary "mud", but 10 Hz is reasonable in the Music Reproduction world.
Won't get into discussions as to what *really* means Hi Fi, be my guest to set your own preferences.
10) I got more detailed than usual, because I wanted to be very clear, and at the same time explain why "Real/Commercial" World design often takes so many compromises.
It's a matter of competitiveness and diminishing returns.
Getting a little extra performance spending 10X the $$$ is a Commercial suicide.
The enthusiast or DIY dos not have those constraints.
Good luck .
And .... buy a few "good" transistors and leave those KT603 for less demanding uses .
Good luck again.
There is an example by Linsley-Hood in the modular pre-amp design intended for use with his Simple Class A amplifier on Rod Elliots Sound Pages website. It is a lengthy article and the example is shown in figure 10 towards the end. It will work from a range of power supply voltage.
see
The Class-A Amplifier Site - Modular Pre-Amplifier Design
You could also look up Paul Kembles pages, A Paul Kemble web page - index to 'sound' webpages. there are a few examples there notably in the Cambridge Audio P series amplifier preamp stages. There is plenty of other interesting material on this and Rod Elliots pages.
Michael J
This preamp has 10X gain.
If your power amp has 1.7V input sensitivity, using this preamp ahead of it means that it cam be fully driven by 170,V audio.
Practical use?:
the typical generic MP3Player/Pendrive fed from 1 AA battery has an earphone output of about 200mV, so it can now drive your amp to full power.
If you have a weak signal such as this one, put a 25K or 50K Volume pot, preverably audio taper, between preamp and power amp.
If you have a strong output CD player or some other line powered signal source , with up to 2 V signal, you will need the same volume control at the preamp input, or it will be overdriven even if you "lower the volume".
C2 is there because you have 12/15V DC at the transistor collector and you do not want it at the output.
The input cap defines the low frequency cutoff.
.47uF give you around 10 Hz cutoff; .1uF will give around 50 Hz (which may be acceptable with cheap speakers) and so on.
that's a maximum of 180W into 8ohms speaker, or 360W into 4ohms speaker.
Adding a 10x (+20dB) preamp in front of that, giving a sensitivity of 170mVac for 180W into 8ohms is unusual.
Last edited:
Fahey
this one
i use a a source an cd-player and stereo amp msd A200s as loudspeakers regulair 2 way system one 89 db.
this one
i use a a source an cd-player and stereo amp msd A200s as loudspeakers regulair 2 way system one 89 db.
Last edited:
If you download the program this .asc file of your preamp (unzip the folder) will run straight off.
You can see the amp is inverting. The response extends to many Mhz. Bottom end is poor due to the 0.047uf cap.
I forgot to add the decoupling cap after the 1K resistor which will slightly alter the AC characteristics... but I'll leave that for you to add
If you get stuck then ask.
You can see the amp is inverting. The response extends to many Mhz. Bottom end is poor due to the 0.047uf cap.
I forgot to add the decoupling cap after the 1K resistor which will slightly alter the AC characteristics... but I'll leave that for you to add
If you get stuck then ask.
Attachments
Download and install Spice and download and unzip my file and save it somewhere such as documents.
LTSpice is a long learning curve but I would recommend 100% Bob Cordells book "Designing Audio Power Amplifiers" which has whole sections on using LTSpice for complete beginners. It all makes sense
To get you started...
Open (run) LTSpice and click "file" at the top left and "open" and browse for the .asc file you have saved. Click it and the file will open.
Right click the circuit and select "edit simulation CMD" Among the options are "DC op pnt" Select that and the DC conditions of the circuit can be probed. Select "AC Analysis" to plot the response of the circuit. You set the limits and scale. Transisent allows you to probe the circuit as if you were using a scope. You have to put a stop time in such as 10ms.
You can use different models (but thats a more advanced option) but for your circuit it won't make any difference in either simulation or reality what devices you use as long as they are generally suitable NPN small signal type.
The book I mentioned will take you from zero to more thn pretty competent in easy steps. And Bob has all the circuits on his website as downloadable files.
LTSpice is a long learning curve but I would recommend 100% Bob Cordells book "Designing Audio Power Amplifiers" which has whole sections on using LTSpice for complete beginners. It all makes sense
To get you started...
Open (run) LTSpice and click "file" at the top left and "open" and browse for the .asc file you have saved. Click it and the file will open.
Right click the circuit and select "edit simulation CMD" Among the options are "DC op pnt" Select that and the DC conditions of the circuit can be probed. Select "AC Analysis" to plot the response of the circuit. You set the limits and scale. Transisent allows you to probe the circuit as if you were using a scope. You have to put a stop time in such as 10ms.
You can use different models (but thats a more advanced option) but for your circuit it won't make any difference in either simulation or reality what devices you use as long as they are generally suitable NPN small signal type.
The book I mentioned will take you from zero to more thn pretty competent in easy steps. And Bob has all the circuits on his website as downloadable files.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.