I use a star topology, with the connecting point inside the chip. The ground of the pot and of the input were connected to pin 5 of the chip (one of its four ground points). The pot was directly soldered to the stripboard and for its connection to the input I used a short (~3cm) shielded cable.
My guitar has no tone, only two volumes (one for each pickup). With them the circuit works fine.

Yes, that fixed it. Well, sort of. The pot worked alright, but at high volumes it would draw 40-60 mA and pick up hum from the environment. It was soldered directly on the stripboard, and connected to the input jack with good quality cable. Plus it was noisy when adjusted (rotated). I removed the pot and the circuit worked perfectly.
The pot was plastic and bought from an electronics store. Should I go for an expensive, quality metal pot from an electronic guitars shop ? Or is it something else? Could it be its value (100kΩ)?

Thank you for your replies.

Enzo, using ST chips is not written on any stone    It's just that I am familiar with their chips, have used them in the past. Any recommendations are welcome. My issue is that I create all my circuits on stripboard, so my choices regarding packages are limited. I have searched for a 5W or similar chip, but the only solutions were DIPs (unable to be heatsinked properly, so I would have problem dissipating all the excess energy and would face thermal shutdowns) and the next step would be the TDA2003 (being extremely non-efficient at such low power levels).

J M Fahey, this is exactly what I did. The speaker's label did not mention Ohms so I measured it with a DC multimeter at 6 Ohms. I admit that I was not aware that it is not the proper method. All my other speakers, from 3 to 8 Ohm, when measured provide their stated values. Why not this one? Is it because of the increased inductance? (the other speakers were max. 3 inch). If so, wouldn't the resistance settle after the initial inductance effect ?

Hello all

I recently bought a Celestion Bulldog speaker, made in UK for Vox, 15W, 6 Ohm. I want to build an amp head just for this beauty.
I am unable to choose between TDA2030, 2040 and 2050, does anyone have any recommendations?
What transformer ratings should I go for? (THE AMP MUST NEVER REACH > 15W 6 Ohm)

Thank you in advance 

Hello all

I have built a little 1W amp using a TDA7267 (schematic: http://www.st.com/stonline/books/pdf/docs/1569.pdf). It worked like charm. I then decided to add a volume control, so instead of the input capacitor at pin 4 I connected a 100k log pot between pin 4 and ground. I then connected the input capacitor at the pot's input. The layout is like this:

INPUT -||-|-----
               |
               |---- PIN 4
               |
               |---- INPUT GND + PIN 5 (CIRCUIT GROUND)

(sorry for the terrible ASCII scematic, I have no EDA software on this computer)

The circuit did not work anymore. Instead of 10mA of quiescent current it drew before, it drew 60mA afterwards. It produced no sounds at low volume and a low sound, heavily distorted, at high volume. I removed the pot and reverted to the original design, It worked normal again.

I have build this circuit twice with different stripboard layouts, different caps. I am fairly positive that the issue is the log connectivity and not somewhere else (e.g. faulty caps).

Any thoughts?

I have read that a major difference between guitar amplifiers and hi-fi is the bandwidth. Since a guitar produces sounds from 80 to 5000 Hz higher frequencies are not only unnecessary but also unwanted, since they can introduce interference and oscillation.

In a TDA2003 circuit this is controlled by the Cx-Rx network. I have changed the original values from 39 to 47 nF and Ω, respectively - haven't noticed any difference. What would be a correct set of values?

Hello

When using a metal chassis the safety ground must connect to it. What about connection from the safety ground to the circuit?

What if there is no safety ground available? Just the (+) and (-) terminals from the rectifier of the transformer?

What if I use a plastic box, either with safety ground or just with the two terminals?

The last two questions are mostly because of another amp I am building. Low power one, inside a plastic box.

Thank you for your reply.

Without any complex math, is there a hard and fast rule to relate the "correct" value of an input pot to the input resistance of a chip? What about opamps with very big input resistance?

Is there a resource online describing it with complex math? :-)

Hello all

I am looking for a chip able to produce around 1-2 watt on a single channel with no distortion. Unfortunately all chips I have found so far have very high gain, and for this low power it means heavy distortion. I tried installing resistors in series with the input, but it is not the optimal way as I practically amplify input noise. To make matters worse I work on stripboard and this limits my choices - SIPs are obsolete, DIPs cannot be heatsinked, Pentawatts are one category higher in terms of power and require higher voltages.

I would appreciate any suggestions.

Just built a TDA2003 amp on stripboard, works like charm :-) I have read the sticky post suggesting grounding techniques at the Amplifier Discussion subsection of this forum, and I have placed the chip in the middle of the board, with power supply & speaker connections and caps and the zobel network on one side of the stripboard, and all other connections to the other. I cannot imagine a better way to do this without pcb :-S The chip's quiescent current at 12 Volts with 1.6Ω load (two 3.2Ω speakers in parallel) is 40mA, meaning I have no oscillation issues.

My problem is that I have this audible hum which is reduced only when I 1) touch the heatsink, or 2) touch the shield of the shielded input wire, or 3) connect the circuit to my guitar and touch the strings. With the guitar connected and none if its metal parts touched, the hum gets really annoying. If that helps you, the hum is exactly the one I get from my Marshall MS-2 battery-powered amp. To make matters worse, I have not installed an input pot yet, and with its 40db gain the TDA2003 is unforgiving :-)

My question is: What can I do to reduce the hum?

I have not installed it in a case yet, and supply it by my bench PSU. I was thinking of connecting the safety ground cable to the screw holding the chip to the heatsink, seems to me like a good spot. What do you think? In case of a simple 2-wire connection (no safety ground cable) what should I do?

P.s.: If it is of any relevance, I have changed the values of some parts: C4 (output coupling to load) from 1000 to 2200 μF, C2 (ripple rejection) from 470 to 2200 μF, Rx from 39 to 47 Ω, Cx from 39 to 47 nF.

Thanks :-)

I hope I write in the proper subforum.

Does anyone have a schematic of the Marshall MS-2? I have searched everywhere in the site, including the links to other sites, but found nothing.

Hello all, I am new in here :-)

I have a question regarding the Pentawatt class AB models from ST (TDA2003, TDA2006, TDA2030, TDA2040, TDA2050). I want to include a volume control in the input, but I have 3 questions:

1) What type of pot do I use? logarithmic or linear?
2) How do I calculate its value?
Where do I place the capacitor?

Thank you in advance.