Re- Transformer phasing

Since the primaries are on separate cores, (two transformers) the phasing is irrelevant except that if both transformers are the same phase (start of primary winding to "hot" for example), then the secondaries will be "in phase" for series work - i.e. the "0" winding will be the same secondary wire on both units.

If one transformer is hooked up primary out of phase to the other (start of primary winding of each transformer on different wires), then likewise the secondary windings will be out of phase with one another by approximately 180 degrees.

For practical purposes, hook both transformers up with no load on the secondaries, connect one transformer wire secondary wire to the other, check no load voltage.  If no load is in the vicinity of 36 volts or better, you are good.  If not, hook together one of the same leads of the connection with the OTHER lead of the second secondary, should now read as above. 

Because the units are on separate cores, this will not damage the transformers, if they are on the same core - potential heating of the primary winding can occur with primaries are hooked up out of phase. Also, because the units are on separate cores, the secondary out of phase condition may not be exactly 180 degrees, and thus you would have a small but measurable voltage when the secondaries are hooked up "out of phase"

To verify phasing on a multiple primary transformer, you need a variable voltage source.
leaving secondaries under "NO LOAD" conditions, connect one end of a primary to one end of the second primary.  Apply a SMALL voltage (around 10-20 volts) to the combination and note the no load AC voltage on a secondary. Then, disconnecting the primaries, hook the first "one end" to the other end of the second primary.  Repeat voltage application.
Odds are that with one configuration you will have a small but repeatable voltage on the secondary (where the two primaries are in phase) and in the other configuration you will have virtually 0 volts on the secondary, (where the two primaries are out of phase).

Finally, and of MOST importance:
BE EXTREMELY CAREFUL when dealing with mains voltages!!!  Electrical current moves almost at the speed of light and Mains voltages have sufficient voltage and current capability to be LETHAL!!

ALSO:  when dealing with transformer leads, use a method such as brass brads in a board that will allow you to solder the leads into place on your bench, keeping them from flying around to contact you, your equipment or each other!  (old transformer engineering trick!)


Hope this helps!

 :tu:

Looks pretty good to me!

By putting the two secondaries in series, you have effectively created a 36V center-tappped secondary.


There will probably be some very minor difference in the absolute voltage level of the + to the  - side, as no two transformers are exactly alike.  However, it should be relatively insignificant in your applications!  With 10k uF on each side, it ought to work nicely!

If you talking about the buffered output - that is strictly optional, and will have no effect on the amplifier performance.  It simply allows an output that is electronically isolated "buffered" from the rest of the amplifier.  you could put a short across the buffered output to ground, and the amplifier will still run.

If you are talking about the FET guitar input buffer, then it will simply affect the input impedance of the amplifier - if you have another preamp that is between the guitar and power amp stage, don't worry about it.   Without the guitar input FET circuit, the input impedance of the amplifier is the potentiometer, in this case 50 kiloohms.

Casiomax,

Almost any solid state circuit can benefit from a firmly regulated supply.  Caveat - be certain regulator can deliver at least twice the expected full steady state load, and have a minimum of 1000 uF post the regulator to insure that it absorbs the fast rising currents until the regulator can catch up.

Secondly, bear in mind the circuit has a voltage gain of 100 as it is straight up.

With a 12 Volt supply, the absolute maximum producable signal by the amplifier chips is 12 volts peak to peak before the waveform distorts due to supply rail limitations. (actually a hair less than 12 V, but we'll say we have ideal amplifiers for this example)

12/100= 0.120 volts p/p maximum input before distortion.

I have measured 0.750 volts p/p out of my electric guitar.

As you can see, doesn't take much to drive the output to distortion.  The FET input of the original design maintains the high input impedance that guitars like to see in an amplifier and the 10-50k volume pot then serves as an attenuator.

lacking the front end impedance match, here is a high impedance attenuator schematic:

Do you have the DC ground of the tonemender tied to the DC ground of the amplifier?

Your quite welcome!

Just a thought for you - have you tied the DC ground of your system to the "earth" ground plug on the mains?

Second notion - do you have your preamp exposed to the air and not running on the same  DC supply as the amplifer (thus common ground)?  This could be the source of hum as well .  Try tying all circuit power grounds together.

With most of my FET input devices, due to the high impedance present, if I don't have the DC ground tied to earth, I get some hum - with a lot of metal at DC ground and the DC ground not tied to Earth, I get a LOT of hum when running on the mains.  I do recommend bringing all the grounds to a single point (google "star grounding") then tying that point to the earth ground.

The advantage of single point grounding for hum reduction was illustrated to me when I had two of my amplifiers daisy chained together through the "buffered output" of the first.  I had one in one room, and the second about 30 feet away in another room.  Each was plugged into the wall, but the wall circuits are on different breakers. When I had the second one plugged in, it hummed something awful.  Fortunately, it was a battery backup unit and I unplugged it from the wall, running that unit on battery - eliminating the ground loop between the signal cable and the two "earth" ground points - hum disappeared.

hope this helps!

RE: Bridging TDA2003

Indeed you can, take a look at the datasheet for the chip, some versions of it on the net have the bridge circuit included.

Casiomax,

Glad to hear you had a successful build.  I designed the amplifer to be squeeky clean in its original inception, and haven't tried putting any pedals other than my chorus in front of it.  The chorus pedal does NOT have a "shorting when unplugged" input to it and if I turn the amplifier on with only the chorus pedal plugged in and on, I also get HUMMMMM.

If you look at the third paragraph of the first post, you will see that I recommended a "shorting when unplugged" input jack to the amplifier.  Most commercial amplifiers (the good ones, anyway,) have this type of arrangement on the input to keep hum down when the guitar is not plugged in.

Teemuk has it right on the money!  I'm certain that the hum you hear is the high impedance input picking up the powerline frequency.

The noise should go away when the guitar is plugged in.  It is for this reason the "old timers" in the music scene recommend plugging in the guitar BEFORE turning on the amplifier - stops hum and "connection noise".

Hope you enjoy it! :tu:

Bismark,

Looking at your supply design, I suggest you double check the polarity of the "negative side" power diode - from your description of what happened, I suspect you applied positive voltage to the negative side (by having the diode in backwards)- would cause both effects you mentioned - the extremely hot 7918 and an explosive situation with an electrolytic.  It also would explain the misbehaviour of the 3886 chip.

The power supply you have built, if the schematic is closely followed, is indeed a +/- 18V supply.  However, it has an absolute maximum output of 1 amp (the current limit of the 7818 and 7918 regulators.)  This is good for amplifiers up to about 5 watts tops.

The power supply utilizes two half wave rectifiers to develop the  positive and negative voltages.  To keep within the current rating of your transformer, I wouldn't draw more than 1/2 amp from each side.

(With a 24 volt, 1 amp transformer you have 24 VA out.  24 VA can be likened to an absolute maximum of 24 Watts DC -  at the diodes. With a +/- 18V out, it is 36 volts end to end.  36 V * 0.6666 amps = approx 24 Watts.  This does not include losses in the regulators themselves or the inherent losses in capability when converting VA to Watts DC.  Preceding information brought to you by Ohms Law: WATTS=VOLTSxAMPS) 

This supply is good for things like op-amps and low power amplifiers, but NOT for a higher power amplifier such as the 3886.  Additionally, I wouldn't use LESS than 4700 uF on each side of the supply right after the diode, and given your transformer, you need a primary filter capacitor voltage rating of at least 35 volts to handle the peak voltages coming off the diodes. (24VRMS is approximately 34 Volts peak).

I agree with those who have gone before that a +/- supply from a center tapped secondary is best - also much more efficient than a dual half wave rectifier setup - and that will improve ripple issues and reduce "hum" in the audio.

I hope this helps!

:tu:

Casiomax,
Yes, you certainly can use a different preamp with this unit.  I used the FET as an impedance matching device only, to give me a higher "front end" impedance going into the amplifier.
I use this ampifier with both electric and acoustic (piezo pickup) guitars, and the EXTEREMELY high input impedance is a MUST for the piezo pickup.
The FET preamp section has NO gain to it, typical of a source follower circuit.  In fact, I have measured the output voltage at the high end of the Volume Pot and the voltage is approximately 90% of the input voltage.
The buffered output is simply an additional feature that I use to daisy chain amplifiers together. (or to go to a "house" system).
Good luck and good building!
:tu:

 
Teemu,
Thanks for the comments on making the design more efficient, if I build another one, I'll incorporate you suggestions - anything that simplifies building without sacarificing quality is stupendous :tu:

Noobie,
Sorry, being right in the center of the U.S., I don't deal with the EU.  The version of LM386 I get from Mouser is made by New Japan Radio, if that helps any!

 
Teemu,
Agreed that this is a lot for a little power return, but this amplifier grew out of an experiment of "can I do this?" and is now used as a sound amplifier for a scanner in my bedroom- but has been used in travel (due to small size) for a guitar amplifier in a hotel room!

I shared the schematic not to advise people to use this, but as a general learning experience that it CAN be done, given the right load - and it worked well for the 32 ohm speaker I had kicking around in the junkbox!

 
Noobie,

I would look at a single chip per speaker of the TDA2003 chip.  It operates quite well on 9-12 volts, and with a 4 ohm speaker at 9 volts will deliver about 2.5 watts into a 4 ohm load.

It is a bit high on current for the typical 9 volt transistor battery, but with 6 AA to D cells you'd be all right.  You also could look into using a sealed lead acid (my favorite for portable amps).

Half the power output for 8 ohm speakers!

Do a google search on TDA2003 for the data sheet, you can also look at the B.E.M schematic - it runs on 12Volts into 4 ohm speakers,giving me about 7 watts per speaker, maximum.  It is a real forgiving chip to work with, and the base circuitry is fairly simple.

good luck! :tu:

Teemuk,

The only 'bigger' amp I have right now is the "Bug Eyed Monster" shown elsewhere in this forum.  I am working on a 30 watt amplifier, based on a toshiba chip I harvested out of an automotive radio unit.

I build each amplifier enclosure from solid lumber and coat with three coats of polyurethane.  Pine is usually what I use, but I have recently come into possession of a couple of cypress boards - the next amplifier enclosure!

Sounds to me like an impedance mismatch causing the amplitude issue.  I have attached an image of a simple preamp, one multipurpose FET and one general purpose NPN transistor.  Overall gain is approximately 5, and the bias is adjusted to have output centered around zero (so no lopsided clipping).  9 volt battery powered.

I hope this helps you!