Mini amp circuit

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:At the (-) end of C3. I'll have to take some screen shots to show you. It's really wierd.
The voltage at the (+) end of C3 should look identical to the output from your previous screen shot. The only difference should be that at the (+) end, the AC waveform is riding on a constant DC voltage, ideally 6V.
The (-) end of C3 is your audio output with the DC removed by the cap, C3.
-=VA7KOR=- My solar system includes Pluto.
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:I went down to the local "The Shack" and, although they had LM741's, they didn't have complementary pairs of either sets of TIP's. They had '31's and 42's, but NOT '32's or '41's!
Typical...
Get yourself a Jameco catalog. They have lots of semis and cheap! Catalogs are free if you request one from the web site. I got one because I find their on-line search to be awful. It's absolutely user-hostile.
-=VA7KOR=- My solar system includes Pluto.
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

Bob,

Your suggestions have yielded much more efficiency from the original design adaptation:
1st = ~580mW
2nd = ~640mW
3rd = ~ 1.02 W
Current (4th) = 1.42 W

The changes I ended up with are:
R7 (Q2 emitter resistor) = 10 ohms
Q2 base to ground resistor = 110k ohms
2nd 2.2k ohm resistor added in series with R6
22 uF electrolytic added to the junction of R6 and new resistor (above), positive to said
junction and negative to the line running from R4 and C3 +.

Q5 dissapates ~350 mW, Q6 ~405 mW. Total circuit current shows as 4.961 mA. RMS
AC voltage at C3 negative is 3.355 V which, when worked backwards gives just about
a 9.5 V total swing, so it is using a fairly large amount of the available supply.

I have the schematic drawn, and I have the junction voltages turned on so that when
I run the sim, they will show. I just have to clean it up to show everything clearly, it is
a bit jumbled right now. I'll see what kind of time I have this weekend.

CeaSaR
Hey, what do I know?
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:Bob,

Your suggestions have yielded much more efficiency from the original design adaptation:
1st = ~580mW
2nd = ~640mW
3rd = ~ 1.02 W
Current (4th) = 1.42 W

Q5 dissapates ~350 mW, Q6 ~405 mW. Total circuit current shows as 4.961 mA. RMS
ac voltage at C3 negative is 3.355 V which, when worked backwards gives just about
a 9.5 V total swing, so it is using a fairly large amount of the available supply.
Fantastic! That is about all the output voltage swing you can expect. There will always be a voltage drop across the output transistors, and some loss due to the lower voltage excursion limit of Q2's output.

I was wondering how much power you were hoping for, because you are past the power limit for passive ambient cooling of the small plastic output transistors. At a minimum you should use some heat sinking like clipping them to a couple of square inches of sheet aluminum, with heatsink compound or silicon grease between the flat side of the transistor cases and the sheet.

If it were for reliable operation or more than an experiment, I would change the output transistors to a higher power type with a metal case or metal tab.
I have the schematic drawn, and I have the junction voltages turned on so that when
I run the sim, they will show. I just have to clean it up to show everything clearly, it is
a bit jumbled right now. I'll see what kind of time I have this weekend.
Good.
So. do you want to build a bridged output version? That would just require another amp driven in reverse phase. Output would increase by 4 times, and neither amp would need an output cap (C3). In that case, you definitely need heat sunk output transistors of a higher power type.

[EDIT] This is cool. Amplifier design by long distance! :mrgreen:
-=VA7KOR=- My solar system includes Pluto.
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

I've taken the liberty of re-adapting this design to use the previously mentioned 32V ps I have
and added a pair of 2N3055's at the output in Darlington configuration with the 2N4401's. The
only change needed was to the Q2 base/ground transistor - 36.3k ohms, a 33k + 3.3K in series.
This setup will push 12 W through a 8 ohm speaker while keeping all transistors well below their
limits (Q1@73mW, Q2@44mW, Q3 & Q4@<0.5mW, Q5@29mW, Q6@42mW, and the 3055's@<2.5W).
If I put a 4 ohm speaker in there, they pull a little more (Q1@75mW, Q2@41mW, Q3 & Q4@<1mW,
Q5@90mW, Q6@143mW, and the 3055'[email protected]). A 2 ohm speaker load puts the 4401's back in peril
again, but the output pumps to almost 38 W! Needless to say, I wouldn't mind seeing this built at
the 4-8ohm version.

I have most of the stuff here, so I hope I can try to breadboard it this weekend, at least the low
power version. If all goes well, the high power version may get tested too.

CeaSaR

[EDIT] And international design! Most excellent!

If all goes well I have an old speaker cabinet that would be great for a Recycled Electronics project.
The cabinet is in really good shape, and the driver just needs the wire fixed at the cone. I'm going to
try to fix it. If not, there are tons of used speakers at the local Goodwill. That's where my 12 V ps
came from. So lots of this would be recycled. I think it would be cool to do. :cool:
Hey, what do I know?
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:I've taken the liberty of re-adapting this design to use the previously mentioned 32V ps I have
and added a pair of 2N3055's at the output in Darlington configuration with the 2N4401's.
You might try adding a one more 1N4001 bias diodes in the diode series to make up for the voltage drop of the added emitter-base junctions of the 2N3055's. If there is not enough bias diodes, you get crossover distortion; too many diodes and excess current shoots through the output transistors. Excess unrelenting current generates excess unrelenting heat buildup in the output Xistors, even with no audio input. With the added gain of the 3055s, massive current will shoot through if the bias voltage is just a little bit too high. This is the kind of circuit where you want to make sure that no excess flows.

Make sure you have resistors from Base to Emitter on the 3055s. They will help the 3055's turn off. 250 ohms or so should work. The high (+) drive already had two transistors in darlington configuration? Now it has three, and the transistor current gains are multiplicative.

I think we're finished now. You are successfully making changes on your own. :grin:
-=VA7KOR=- My solar system includes Pluto.
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

Multiple replies here...
Bob Scott wrote: 1.) That is about all the output voltage swing you can expect...
2.) I was wondering how much power you were hoping for...
3.) ...heat sinking...
4.) ...I would change the output transistors to a higher power type with a metal case...
5.) ...bridged output version...
------
6.) You might try adding a one more 1N4001 bias diodes in the diode series...kind of circuit where you want to make sure that no excess (current) flows.
7.) Make sure you have resistors from Base to Emitter on the 3055s. They will help the 3055's turn off. 250 ohms or so should work. The high (+) drive already had two transistors in darlington configuration? Now it has three, and the transistor current gains are multiplicative.
8.) I think we're finished...
1.) I figured that the max swing I could expect should be at least supply - (~2 x output junctions), so 12 - (~2 x (0.6 x 2)),
12 - (~2.4) or ~ 9.6 volts +-, between 9.5 to 10 V. Just my empirical guess because you have to take into account the
output junctions and some loss due to other parts (resistors, capacitors and speaker inductance - the ~2 x's).
2.) Once the formula for power was shown to me, I did a quick calc to show at most 1.6 W, but that would really be pushing it.
3.) Yeah, I saw the current running through them in the sims and knew that they would need to be cooled somehow. I have
some sheet aluminum I could use if I wanted. RS has the "grease".
4.) I thought that later maybe some 2N2222A's - the TO-18 cased one's would be better.
5.) That would be easier to do with OPAMPs for the inputs. A dual for mono, or a quad for stereo. I know the LM324 is a quad,
is the LM1458 a dual? Gotta look that up. I'd love to see your take on that.
6.) When I wrote my last post with stats, I tried doing that to see what would happen. While the output was slightly
increased, I didn't like the increase in current through the Q3 - Q8 stages. AND, I set the input to 0 V (no input) and saw that
the current running through those transistors was doing nothing but heating them up. Looks like 3 diodes work very well in this
design, keeping everything working efficiently and cool.
7.) When placed in the schematic, it looks like the resistors are helpful when the speaker impedance drops to 4 ohms or less.
The sine wave becomes slightly jagged without them. At 8 ohms, the wave looks just beautiful. As for the multiplication, I'm
being lazy right now and letting the software do the work. I am learning though. :mrgreen:
8.) Well, almost. You had mentioned a current mirror, and I have seen some amps that use what look like differential inputs.
From what I can see, a current mirror takes place of the diode string and that could give you the ability to adjust the
quiescent current. Am I correct? What advantage would a differential input give you? Stability?

From a post further above - "Jameco Catalog", I have one of those as well as MCM, All Electronics and another one I can't
find right now. I have been able to find most things in the on-line stores, but paper is so much more convenient to look
through. Would you believe that MCM actually has the best stock of Ge transistors? It does! And at a better price!

Thanks,
CeaSaR

PS, anyone else interested in this topic may jump right on in. I welcome the participation.
Hey, what do I know?
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:4.) I thought that later maybe some 2N2222A's - the TO-18 cased one's would be better.
Apparently the plastic packages they use for transistors conducts heat quite well. I know it is hard to believe. Power transistors always have metal tabs for heat dissipation. I believe what I see. The 2N4401 is higher power and current than the 2N2222. 2N2222 would be a downgrade, I think. You'd best compare the data sheets, especially the current, power, and thermal resistance specs.
5.) That would be easier to do with OPAMPs for the inputs. A dual for mono, or a quad for stereo. I know the LM324 is a quad, is the LM1458 a dual? Gotta look that up. I'd love to see your take on that.
I think that the LM1458 is a dual 741. I think that the 1458, 1558, 4558 are almost, if not, identical. 1458 and 1558 were seen in lots of Asian cassette decks circa 1980.

The LM324 is a separate animal. It can operate at much lower power supply voltages down to +/- 3V or ~6V total between + and - PS inputs. An upgrade for this IC would be the now also obsolete LMC660. It has rail-to-rail outputs. All of these are old technology. The newer chips are less noise and higher bandwidth.
6.) When I wrote my last post with stats, I tried doing that to see what would happen. The sine wave becomes slightly jagged without them. At 8 ohms, the wave looks just beautiful. AWhile the output was slightly increased, I didn't like the increase in current through the Q3 - Q8 stages. AND, I set the input to 0 V (no input) and saw that the current running through those transistors was doing nothing but heating them up. Looks like 3 diodes work very well in this
design, keeping everything working efficiently and cool.
7.) When placed in the schematic, it looks like the resistors are helpful when the speaker impedance drops to 4 ohms or less.
In consumer equipment one of the diodes is sometimes replaced with a thermistor mounted to the heat sink, so the bias current tracks voltage changes to compensate for heating of the output transistors. ie: Bias current stays constant with temperature changes. Also, a variable resistor is placed across one of the diodes or thermistor for manual fine adjustment of the bias current.

The base to emitter resistors are used to keep the output transistors "off" when they need to be so shoot-through will not occur. See here: Even factory darlingtons have these resistors. This one is ".12K", 120 ohms:

http://www.fairchildsemi.com/ds/TI/TIP120.pdf
-=VA7KOR=- My solar system includes Pluto.
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:8.) Well, almost. You had mentioned a current mirror, and I have seen some amps that use what look like differential inputs. From what I can see, a current mirror takes place of the diode string and that could give you the ability to adjust the
quiescent current. Am I correct?
The diode string would stay. Remember the resistor we split with the capacitor added to the top of C3? It increased the high drive capability of the output stage, because the 2.2K resistor would run out of current as the voltage across it dropped. The current mirror would have done the same thing, differently. Constant current sources do not run out of current as the voltage across them drops.
What advantage would a differential input give you? Stability?
[Note to self] ?? I must edit this.
From a post further above - "Jameco Catalog", I have one of those as well as MCM, All Electronics and another one I can't find right now. I have been able to find most things in the on-line stores, but paper is so much more convenient to look through. Would you believe that MCM actually has the best stock of Ge transistors? It does! And at a better price!
Really? GE? can you point me to the URL of the page that has GE transistors? I have not seen one of those since the 60s, when they were GE-2, GE-3, GE-8 unreliable germanium devices, but suitable for teenagers to play with. I remember my Dad paying $2.50 for a GE-2 in a TO-5 can. Today that 1963 $2.50 is worth more than 10X or 15X $2.50. Imagine paying $37.50 for a 2N2222! :grin:
-=VA7KOR=- My solar system includes Pluto.
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

Ge, not GE. :lol: Germanium... :mrgreen:

CeaSaR
Hey, what do I know?
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

I was looking at the 2N2222A and saw the specs (from Circuit Maker):
Si 500mW 40V 800mA 300MHz GenPurp pkg:TO-18 3,2,1
and the 2N4401:
Si 625mW 40V 600mA 250MHz GenPurp pkg:TO-92B 1,2,3

I looked at 800mA vs. 600mA and thought it was "tougher", but I see that the power rating is 500mW vs. 625mW.
2N4401 wins by 125mW. I also assumed (I know, never do THAT!) that the metal can would be better at heat
disapation. I guess I was fooled by that.

As far as the LM's go, I think I have 1 or 2 to play with, but would rather go with the TL0 series or better should I
wish to keep these projects.

The Constant Current version I saw was a 3 resistor string - fixed on each end and variable in the middle with the
wiper feeding the base and the collector/emitter connected to the beginning/end of the string. I will have to look
that up for you to see it. I believe that one also had the diff. input.

BTW, I found I have a 15 pack of, I think, MPSA42's. It's hard to read the black letters on the black plastic case.
Looked them up and they are lower power, but I see them in use around the web. Those aren't in power situations,
though.

$37.50 for a 2N2222???? AAGGHHH!

CeaSaR
Hey, what do I know?
User avatar
Bob Scott
Posts: 1192
Joined: Wed Nov 20, 2002 1:01 am
Location: Vancouver, BC
Contact:

Re: Mini amp circuit

Post by Bob Scott »

CeaSaR wrote:I looked at 800mA vs. 600mA and thought it was "tougher", but I see that the power rating is 500mW vs. 625mW. 2N4401 wins by 125mW. I also assumed (I know, never do THAT!) that the metal can would be better at heat disapation. I guess I was fooled by that.
Oh, I thought there was a larger difference. That was my assumption. Never mind. :???:
I have noticed discrepancies with some of Jameco's catalog specs, though, compared to the manufacturers data sheets. eg: 1N914/1N4148 current listed as only 10mA in one of their catalogs, and the transistor data varies.
-=VA7KOR=- My solar system includes Pluto.
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

I looked up the data sheets for the 2N2222A and the 2N4401. They both have the exact same PD at 625 mW,
but the 2N2222A has an IC of 1 A while the 2N4401's is only 600 mA. Maybe there are a bunch of "not quite right"
pieces of info out there. :/

CeaSaR
Hey, what do I know?
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

In the process of getting the schem's compiled for the web. Will be up later today or tomorrow.

CeaSaR
Hey, what do I know?
User avatar
CeaSaR
Posts: 1949
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Mini amp circuit

Post by CeaSaR »

Here is the final for the 1.4 Watt Low Power Audio Amplifier. First is the clean schematic, second is a
blowup showing voltages/power output, third is a couple of scope shots - after and before C3.

Latest schematic. Please refer to this one for future discussions.
Image

Bob,
Notice the first scope shot labeled "@ C3 OUTPUT". See how the top of the waveform is ~ 4.3 V and
the bottom is ~5.0 V. That's what I was referring to earlier.

As for the Medium power version, I am still playing with the B-E resistors you mentioned, but I haven't
been able to get it to work quite right. Maybe I'll post it later.

CeaSaR
Hey, what do I know?
Post Reply

Who is online

Users browsing this forum: Ahrefs [Bot] and 29 guests