help in building TDA7294

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
ok
i m changing hte board again, i have some rubycon 330uf at work , and regardng the 2uf across the + - rail i will have to find it from some where idont have it with me,

dont take me wrong but - if we use anything greater than 330uf on power rail amplifier board it will give not so great sound then how come the data sheet shows 1000uf caps for the rails , i m not so well versed wiht this side of details so m asking
 
Last edited:
For reference, here is the power supply that has been running my TDA7294 for years.

Not mentioned in the drawing is that those 100nF at the output of the power supply board are cheap lossy polyester dip caps (mylar bubble cap) that are internally an RC (automatic snubber) because of their high ESR.

However, the 100nF's on the amplifier board's power circuit are high efficiency ceramic or polypropylene (low ESR).

My apologies for the simplistic appearance, but I would have replaced it if something outperformed it. :D
For reference, the 8800uF per each rail figure is about bare minimum size to support bass, so you may prefer to expand it to 5x 2200uF caps (11,000uF) per rail and get a bit better bass. This power supply is similar to a considerable number of discrete amps in the solid state forum. The capacitors are tightly nested together for noise reduction purpose.
 

Attachments

  • Ladderback Schematic.jpg
    Ladderback Schematic.jpg
    67.2 KB · Views: 433
Last edited:
ok
i m changing hte board again, i have some rubycon 330uf at work , and regardng the 2uf across the + - rail i will have to find it from some where idont have it with me,

dont take me wrong but - if we use anything greater than 330uf on power rail amplifier board it will give not so great sound then how come the data sheet shows 1000uf caps for the rails , i m not so well versed wiht this side of details so m asking

Possibly, the 1000uF in the datasheet are signal grade caps. Possibly, the main customer of TDA7294 uses 330uF caps and so ST cannot release that information on the datasheet. Probably, its both reasons.

You can really easily test it for yourself. :)

P.S.
220uF//2200uF at the amplifier board isn't horrifying. However, my own just worked better (clearer) when all of the big caps were at the power supply board, not the amplifier board.
P.P.S.
The power circuit that I tried to describe is like the majority of discrete amplifiers in the solid state forum. Have a look. :)
 
Last edited:
hey this is drawingh has all the information one needs, and its not shabby at all so relax, i m going to do the same for my amp, i had initially kept 3 per rail 4700uf/50 volts,
but now i will drop it to 2200 and keep 5 per rail, got lots of samwah SD series 2200 /50 v at my work place, also will go to market to see if i get nicichon or some other at 105degs low esr,
thaks for this info
 
one more question smps use low esr caps right, i know a good smps manufacture near myplace and he uses all nichicon caps or elna caps rated at 105degrees will see the series he is using, and take them from him - good idea or bad one?

Low inductance at the power supply board. So, the big caps can be cheap. Its fine.

Low ESR at the amplifier board, which is a better quality cap.
You might want to explore some options for:
100uF//100uF per rail at the amplifier board
220uF per rail at the amplifier board
330uF per rail at the amplifier board
. . . and so you could collect the caps for this test from your friend at the smps manufacturing.
 
Last edited:
Um, connect ESP's handy headphone jack circuit and then try the options I've mentioned for power circuit. You can also loop it through RMAA software.

The frequency response of TDA7294 is Really easy to level, and you won't need a tone control. This is true of most of the ST amplifiers and also Natsemi's LM1875.

I have a small value cap as a load at the input, in order to block frequencies higher than the audio band. That improves efficiency a bit and reduces hum. Sometimes I'll put a 1m load at the rca jack too.

I have a 4.7nF hyper-efficient ceramic cap (radio grade disc) in parallel with my NFB cap in order to wipe out some inductance from that cap. Mine also has a 3.3R 1/2w carbon resistor as a ground lift for the NFB cap in order to make it a "less attractive" spot for the return line energy from the speaker.

For loud dynamics, the speaker is hooked directly to the amplifier board; but, for a bit milder mids, you can ground the speaker at the power supply board.
Compare:
For TDA7294 (which doesn't "shout"), I like. . . both speaker wires connected directly at the amplifier board.
For LM3886 (which really "shouts"), I like to ground the speaker at the power supply board.

Input selector. . . just like at Decibel Dungeon. Mine is a 4x4 rotary (four pole, four throw) so that there's no difference from a straight wire (and no ground loops).
You can also use a three pole (left, right, ground) four position switch (four sources) if you have stereo build (amp has only 1 transformer, meaning that it has only 1 ground).

For two sources, a simple DPDT switch can do it (left, right. . . with two sources), but only on a stereo amp, not a dual-mono amp.

Its a really trouble free amplifier--just exactly like a discrete amp for power circuit; however, its not as good an amplifier as a discrete amp so you'll need to make that up by doing a really good job with the power circuit, making a seemly selection of NFB cap, making a seemly selection of input filter cap, and using good solid core wire for all of the small signal (input) audio. Then of course, you have results far better than average (because of effort and attention to detail).
 
Last edited:
Hi Friends,
My Logitech z5500 bass speaker is not working.The bass amplifier is bridged with two TDA 7294 chips.All the satellite speakers are working fine but no sound from the bass speaker.My control pod also working well with the woofer volume is showing but no sound from woofer.I have tested the woofer speaker,it's working fine.
I think one of the two bridged TDA 7294 chip might gone but i'm unable to find out the fault.
Plz let me know the testing procedure for the bridged TDA 7294 bass amplifier.Your help is highly appreciated

Thanks,
Sunny
 
Member
Joined 2002
Paid Member
Quick question re the TDA7294: I see that it has separate pairs of power pins, one pair for the signal stage, the other for the D-MOS output devices. Can I use separate supplies to feed each of these sections, or will this cause oscillation/stability problems?

Triophile
I hope my response does not come too late for you.
Make sure that the signal stage PSU comes on first. Otherwise, you will turn the TDA into an explosive device :mad:(wear protective glasses).

This means that the combination of regulated psu for the signal stage and unregulated psu for the output section will kill the IC.
You have to delay the unregulated psu.
(Something that datasheet does not mention)

And make sure that the grounds of the two psus are connected together.
My suggestion: Better start off with a common psu for both sections

Best Regards
George
 
Last edited:
I didnt want to make new threah so i ask here: Whats the difference between mute and standby? will both shut down the amp or what? In the datasheet of tda7294 reads that there is an option to make one command to both but nothing about what they actually do? please somebody help:D
 
I didnt want to make new threah so i ask here: Whats the difference between mute and standby? will both shut down the amp or what? In the datasheet of tda7294 reads that there is an option to make one command to both but nothing about what they actually do? please somebody help:D

The chip is still in operation for mute. It just simply not amplifying the input signal. For standby, the chip is not operating. You can see Fig 16 when the chip is in standby, the operating is simply zero. When it is in mute, the operating current a bit higher but not output. To make it a better graph, the operating current level should follow signal amplitude when the chip is not mute. There is a sequence for power up, standby and mute. The RC network in Fig 17 does that. Fig 16 tells you what Fig 17 does.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.