Hi,
I have bought some PCBs from AliExpress designed for making amps with 2
X TDA7293 each DIY kit TDA7293 Amp Amplifier Bare PCB Board Double Sided Circuit Boards For DIY-in Integrated Circuits from Electronic Components & Supplies on Aliexpress.com | Alibaba Group
The PCBs have very good quality and they have also printed the components they must be used, but only with the names (such as R1, R2,...C1,C2 etc) without any info about values of the components.
I know that I can find in data sheet the right values of components, but before I start to check line by line the pcb I like to ask IF someone have all ready use this PCB and can give me this information.
Thanks in advance and excuse me for any potential spelling or grammar mistake but English is not my language
I have bought some PCBs from AliExpress designed for making amps with 2
X TDA7293 each DIY kit TDA7293 Amp Amplifier Bare PCB Board Double Sided Circuit Boards For DIY-in Integrated Circuits from Electronic Components & Supplies on Aliexpress.com | Alibaba Group
The PCBs have very good quality and they have also printed the components they must be used, but only with the names (such as R1, R2,...C1,C2 etc) without any info about values of the components.
I know that I can find in data sheet the right values of components, but before I start to check line by line the pcb I like to ask IF someone have all ready use this PCB and can give me this information.
Thanks in advance and excuse me for any potential spelling or grammar mistake but English is not my language
Daniel has kindly found you some references for choosing component values. My identical board is on its way but evidently I have no experience in populating it yet.
I would say, use the information given in the datasheet such that you adapt the values according to your needs.
- Use the 22uF bootstrap capacitor suggested, you have no basis for estimating a better value.
- Use the 4R7/47nF Zobel network suggested in Daniels first link, you have no basis for estimating a better value.
- Use 100nF polypropylene decoupling capacitors for the small capacitors.
- As electrolytic power decoupling capacitors, use the biggest you can fit in the board with the voltage rating you need (normally 50V).
- Use gain setting resistors such that you have a gain between 10 and 40, preferably 20-30 for good stability.
- Choose the signal capacitors such that your cut-off frequency is around 5Hz.
- For mute and standby inputs, use the values suggested in the datasheet. They should work properly.
Last edited:
Just a few notes:
Basically, you don't set the the gain to what you want--you have to set the gain to what the amplifier wants.
Symptoms:
Gain too high may give a pleasant tone but really boring low resolution. Because, it actually is bored.
Gain too low may give a more revealing presentation along with uselessly 'angry' tone and excess heat. Because, it actually is angry.
In-between those extremes is what you'd want to aim for. That takes me a few hours.
Think of it like winding the clock, in that both too loose and too tight, sounds like consequences.
To make that selection process go faster/easier, relocate the gain divider directly on the chip--the feedback resistor right across the chip and the feedback-shunt resistor from the in- to the feedback-shunt-cap by most direct path. On the mini boards, this fits easily.
The decorous location with long traces that the board provides, hinders stability slightly, requiring compensation of higher gain; but/instead, since it is easy to mount those two resistors closer to the chip, the quality boost is not laborious.
Mind that diode! If it is frail, broken, or flipped, the amp explodes. I'd be wanting a part slightly more stout than what comes with the kit.
Ideally 33u, but the datasheet went for more broadly available values.
That can work; however, ceramic is tamer and easier at that locale. Polypro//electro makes for cap selection hardship. That case, the polypro is so efficient that the 100n value is too big. On a very sketchy guess (even for me), I think that 33n polypro might work without causing so much ringing in the electrolytic caps. But, if you're going by the book use ceramic at that specific locale. So, maybe not overspend at that spot.
The bigger the cap, the more difficult it is to find one of relevant quality at small signal locale. So, I respectfully suggest to relocate 'the biggest cap you can fit' to the power supply board (not the amplifier board). I think that a very fortunate choice of low-esr 470u would be the biggest cap you'd like to have on the power rails on your amp board. And, maximum doesn't mean best. This is not a 'gainclone'--your chip is significantly different. View documentation for discrete amplifiers (range is 220u to 470u).
With this amp 20 to 42 is the operable range, precisely corresponding to minimum versus maximum power voltage. 20 if bare minimum voltage or 42+ if maximum voltage. If in-between, choose proportionately. That also means: If you want hi-fi then don't maximize voltage.
Basically, you don't set the the gain to what you want--you have to set the gain to what the amplifier wants.
Symptoms:
Gain too high may give a pleasant tone but really boring low resolution. Because, it actually is bored.
Gain too low may give a more revealing presentation along with uselessly 'angry' tone and excess heat. Because, it actually is angry.
In-between those extremes is what you'd want to aim for. That takes me a few hours.
Think of it like winding the clock, in that both too loose and too tight, sounds like consequences.
To make that selection process go faster/easier, relocate the gain divider directly on the chip--the feedback resistor right across the chip and the feedback-shunt resistor from the in- to the feedback-shunt-cap by most direct path. On the mini boards, this fits easily.
The decorous location with long traces that the board provides, hinders stability slightly, requiring compensation of higher gain; but/instead, since it is easy to mount those two resistors closer to the chip, the quality boost is not laborious.
I have absolutely no idea. The funny thing is that I've never tried that. Seemed like too much bother.
Mind that diode! If it is frail, broken, or flipped, the amp explodes. I'd be wanting a part slightly more stout than what comes with the kit.
Last edited:
An impressively elaborate reply. That will give most of us guidance in designing amplifiers with TDA7293.
Until now I had the impression you had some links to (and knowledge from) TI. This is ST and you have a very detailed knowledge as well.
Then a general question, is the TDA7293 as problematic as some forum members claim or is it a useful alternative to the LM3886?
Until now I had the impression you had some links to (and knowledge from) TI. This is ST and you have a very detailed knowledge as well.
Then a general question, is the TDA7293 as problematic as some forum members claim or is it a useful alternative to the LM3886?
Not interfering with Daniel´s excellent advice, 
, just adding some practical use comments: many problems come from datasheet calling it a "100V - 100W" amplifier which it is not and people trying to reach that.
Its case is exact same as LM3886 so no way it can reliably dissipate more heat than its cousin, thermal resistance to heat sink is the same.
IF you use it "as an LM3886" class device, in this case with no more than +/-35V rails and pulling no more than 70W RMS ... by the way as suggested in datasheet body ..... , then it´s quite reliable.
Speaking from a Guitar amp maker point of view, more concerned with reliability and abuse than extreme Audio performance.
, just adding some practical use comments: many problems come from datasheet calling it a "100V - 100W" amplifier which it is not and people trying to reach that.Its case is exact same as LM3886 so no way it can reliably dissipate more heat than its cousin, thermal resistance to heat sink is the same.
IF you use it "as an LM3886" class device, in this case with no more than +/-35V rails and pulling no more than 70W RMS ... by the way as suggested in datasheet body ..... , then it´s quite reliable.
Speaking from a Guitar amp maker point of view, more concerned with reliability and abuse than extreme Audio performance.
Since their ideal applications are different, I feel that neither is a useful alternative to the other.
The LM3886 is very useful for sharp and clear TV and Movie audio. Indeed, it 'does what it says on the tin.'
The TDA7293 is like a basic discrete Fet-output amplifier shrank to fit into a chip. You are in charge of durability and tonality, just like with a discrete amp.
TDA7293 durability:
Voltage insulation at the heatsink is required
It is best to heatsink before soldering the chip
If the v- is lost, so is the chip (No screw terminals for power!)
Set the gain high enough for placidity (stable at full blast, not just 1 watt)
Don't put big power caps on the amp board
Do the power circuit like a discrete amp
A stable amp decreases heatsink expenses, considerably
To answer the question if TDA7293 is as problematic as some forum members claim, that answer is yes; if one used the online advices for LM3886 with a TDA7293, the result is disastrous, because those two chips are very different and because TDA7293's documentation is less effective (junk schematics, good text).
However, you can use the documentation and applications for an equivalent basic fet output discrete amplifier with TDA7293, and that works very well.
I would also suggest no more than +/-33vdc rails, because at the lower voltage you don't have to run the gain up so high to stabilize it. Sure it can run at higher, but there's workarounds and labor that aren't needed at lower voltage. It would just be more straightforward at the lower voltage.
Power.
You could make a good start with a basic discrete amp power supply, involving 4 of 10,000uF caps (2 per rail) and a big metal square bridge rectifier (the ordinary sort).
For embellishment, you could replace the bridge rectifier product, with 8 of 8a 600v Fairchild Stealth diodes for two discrete bridge rectifiers, supporting a dual-secondaries transformer. That simplicity will outperform the majority of snake-oil projects, and with much less labor involved.
For safety, let's use modest drainer resistors and light up an LED per each rail at the power supply board, for some indication that there's a charge present. Since the amp probably will get some fine tuning, then we DO need to know when it isn't safe to touch. Don't assume that you'll be mindful enough to use the multimeter every time--use the LED's on the power board!
As for the transformer, I'd think you'd like a rather heavy 22+22vac transformer capable of 4 amperes or more. For embellishment, have more amperes, not more voltage.
You could make a good start with a basic discrete amp power supply, involving 4 of 10,000uF caps (2 per rail) and a big metal square bridge rectifier (the ordinary sort).
For embellishment, you could replace the bridge rectifier product, with 8 of 8a 600v Fairchild Stealth diodes for two discrete bridge rectifiers, supporting a dual-secondaries transformer. That simplicity will outperform the majority of snake-oil projects, and with much less labor involved.
For safety, let's use modest drainer resistors and light up an LED per each rail at the power supply board, for some indication that there's a charge present. Since the amp probably will get some fine tuning, then we DO need to know when it isn't safe to touch. Don't assume that you'll be mindful enough to use the multimeter every time--use the LED's on the power board!
As for the transformer, I'd think you'd like a rather heavy 22+22vac transformer capable of 4 amperes or more. For embellishment, have more amperes, not more voltage.
Be very careful altering the gain.
These chip amps have a minimum gain before they start oscillating.
I reduced one to a gain of 7 because I used it in conjunction with a valve with a gain of 8 and it just oscillated and got very hot. It wasn't until I read the datasheet again I noticed the min gain of about 22.
I couldn't change the valve gain much so put a 1000pf across chip amp + and - inputs and that calmed things down.
These chip amps have a minimum gain before they start oscillating.
I reduced one to a gain of 7 because I used it in conjunction with a valve with a gain of 8 and it just oscillated and got very hot. It wasn't until I read the datasheet again I noticed the min gain of about 22.
I couldn't change the valve gain much so put a 1000pf across chip amp + and - inputs and that calmed things down.
Hi nick
the current for the power supply is mentoined the rating of the transformer/or SMPS
e.g.200VAC transformer with 22V and deliver 5 ampere each wounding
AS-2222 - 200VA 22V Transformer - AnTek Products Corp
if you compare with an 100VA transformer..then you will se he is abel to deliver just 2,3 amps....
the current for the power supply is mentoined the rating of the transformer/or SMPS
e.g.200VAC transformer with 22V and deliver 5 ampere each wounding
AS-2222 - 200VA 22V Transformer - AnTek Products Corp
if you compare with an 100VA transformer..then you will se he is abel to deliver just 2,3 amps....
Last edited:
It would work either way. But, if you want more linear (including sturdy bass support), the higher amperage transformer helps (especially in getting the steady voltage needed for high quality audio).
This looks pretty good for a stereo amplifier: AS-3222 - 300VA 22V Transformer - AnTek Products Corp
For sound quality purposes, even the 400va version would be useful; however, it is very huge and that's unnecessary unless you plan to drive 4 ohm speakers.
Daniel, thank you very much for answering.
My plan is to use those cards to make a 5.1 system in experiment of mix triode tubes and those ICs in supertriode circuit, and I will use 7 ICs, 5 for the channels and a pair for the subwoofer.
My diy speakers are 8 ohms but the sw is 2x2 ohms so I count it as a 4 ohms channel (using 2 ICs bridged)
for this system 2 transformers 1 KVA each do you think its enough?
Sounds like minimized audio quality, minimized durability, quick and permanent hearing damage also resulting in quick and permanent reduction in the capacity to enjoy music. So, that's the risks. Please explain the need of it? What is the venue size? A venue much larger than a house might need that much audio; but, it would be unusual to use chip amplifiers for it.
Initially, this seems like a case of application mismatch.
And these, albeit enjoyable, are not good enough to bridge them.
Also, you might like to know that more quality can do a similar level of enjoyment at less damaging output levels. So, perhaps, re-aim.
Last edited:
Modulus-686 | Neurochrome :: Audio
There's the highly linear 360W chip amp.
Looks like it would take a 500va transformer.
The nested feedback amplifier saves a little money there, because you don't need to upsize the transformer current capacity for linearity purposes--the linearity of a nested feedback amplifier, the master feedback loop, is run from the small signal amplifier on stabilized power, unaffected by the efforts of the 6 high current chips pushing your woofer.
Anyway, Neurochrome Modulus-686: 240W(8Ω); 360W(4Ω) @ 0.00025% THD+N that shows what it takes to do a good job driving 4 ohm load, at high power, with chips. For the subwoofer task, we didn't actually need the super-low distortion figures; but, the linearity of it, is necessary for accuracy.
That accuracy avoids bass monotony, by simply preserving the uniqueness of audio tracks (it has clean bass, excellent specs, no sonic signature of its own).
There's the highly linear 360W chip amp.
Looks like it would take a 500va transformer.
The nested feedback amplifier saves a little money there, because you don't need to upsize the transformer current capacity for linearity purposes--the linearity of a nested feedback amplifier, the master feedback loop, is run from the small signal amplifier on stabilized power, unaffected by the efforts of the 6 high current chips pushing your woofer.
Anyway, Neurochrome Modulus-686: 240W(8Ω); 360W(4Ω) @ 0.00025% THD+N that shows what it takes to do a good job driving 4 ohm load, at high power, with chips. For the subwoofer task, we didn't actually need the super-low distortion figures; but, the linearity of it, is necessary for accuracy.
That accuracy avoids bass monotony, by simply preserving the uniqueness of audio tracks (it has clean bass, excellent specs, no sonic signature of its own).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.