Hi everyone.
I built a 8ch amplifier with TDA7293 boards from ebay. My idea is to drive a 3 way system with a digital crossover and in the future add a 2 way central speaker. The board from china is the one that integrates the rectifier and the capacitor, so it can be connected directly to the transfomer output.
I choose a 2x30V toroidal transformer since the TDA7293 is specified to work up to +/-50V.
There began the issues.
I tested the amplifiers directly on a normal full band speaker and it sounded noisy on lower frequencies.
Then I started checking what was going on with the oscilloscope.
With no load on the output and with a sine applied at the input, the amplifier works well and clips at 30Vrms. It shows some distorsion from 20Hz to 50Hz but this is not the main issue.
Testing it with a 4 ohm resistive load, the amplifier seems to go in protection mode when the output goes over 6Vrms with frequencies from 50Hz up to 20kHz. From 20Hz to 50Hz it seems to start oscillating either!
I tried a 8 ohm resistive load and things went better. The amplifier showed the same little distorsion on low frequencies (20-50Hz).
For further test I used a variac to lower the main supply to 20-0-20V and the amplifier went stable also with a 4 ohm load.
I tried to add more caps on the rails (it has 2200uF normally fitted) but no luck.
The schematic is the same as the application note from ST, apart the input cap that is 1uF.
My conclusion is that the TDA7293 cannot work with power supply more than +/-27V max +/-30V but many projects use it with higher voltages!
Is there anyone that can give me a suggestion on it?
The only idea is to use a 18-0-18V transformer... But that's not worth!
I will try to post some pictures later...
Thanks!
I built a 8ch amplifier with TDA7293 boards from ebay. My idea is to drive a 3 way system with a digital crossover and in the future add a 2 way central speaker. The board from china is the one that integrates the rectifier and the capacitor, so it can be connected directly to the transfomer output.
I choose a 2x30V toroidal transformer since the TDA7293 is specified to work up to +/-50V.
There began the issues.
I tested the amplifiers directly on a normal full band speaker and it sounded noisy on lower frequencies.
Then I started checking what was going on with the oscilloscope.
With no load on the output and with a sine applied at the input, the amplifier works well and clips at 30Vrms. It shows some distorsion from 20Hz to 50Hz but this is not the main issue.
Testing it with a 4 ohm resistive load, the amplifier seems to go in protection mode when the output goes over 6Vrms with frequencies from 50Hz up to 20kHz. From 20Hz to 50Hz it seems to start oscillating either!
I tried a 8 ohm resistive load and things went better. The amplifier showed the same little distorsion on low frequencies (20-50Hz).
For further test I used a variac to lower the main supply to 20-0-20V and the amplifier went stable also with a 4 ohm load.
I tried to add more caps on the rails (it has 2200uF normally fitted) but no luck.
The schematic is the same as the application note from ST, apart the input cap that is 1uF.
My conclusion is that the TDA7293 cannot work with power supply more than +/-27V max +/-30V but many projects use it with higher voltages!
Is there anyone that can give me a suggestion on it?
The only idea is to use a 18-0-18V transformer... But that's not worth!
I will try to post some pictures later...
Thanks!
There are two heatsinks, made of standard aluminium 250mm x 100mm x 40mm, each one with 4 boards. Since it will be used on a multiamplification system, I considered that only the two boards (one for each heatsink) connected to the woofers as high heat producer. There is also a forced cooling system with 60°C threshold.
By the way, the tests mentioned above are with only one board at once.
The issue is there with only one board connected or with all boards.
By the way, the tests mentioned above are with only one board at once.
The issue is there with only one board connected or with all boards.
The sine amplitude is reduced, but not switched off. I believe that is related to a current limitation. Also temperature is rising quickly. With a thermal gun I saw that the external case arrives at 95°C - 100°C. The isolation with the heatsink is guaranteed by a mica washer foil and silicone paste has been used.
Is every board behaving the same way individually? Can you provide a link to the boards you bought?
I have a board with those chips, it also had rectifier and smoothing capacitors on board, I didn't have a problem with overheating, but it was a bit noisy until I removed the rectifiers and capacitors and used an off board supply. I suggest you try that. As a quick test there is no need to remove the rectifier and capacitors, just connect your DC supply to the AC supply inputs
Last edited:
Ok, I will try it as soon as I hook another rectifier and capacitors.
The transformer is rated 300VA. I know that is not sufficient for the whole 8 channel to work, but the problem is that with only 1 channel (for which the trafo would be sufficient) the issue remains.
To start, 7293 is an analog chip, NOT digital. So the seller is either a moron or a bald faced liar.
I wondered how these hucksters could build these boards cheaper than the cost of the parts, so I bought a few. The answer - used parts, cutting corners, parts so cheap I wouldn't bother picking them up off the floor. I had to massage the boards and add parts to get them to even work, and I can design and build amplifiers from scratch.
It's a shame that these are marketed towards amateurs, who would have limited resources to figure out why they don't work as advertised.
I couldn't even salvage $5 worth of usable parts from the boards I bought. Waste of money and effort.
In contrast is our own member Tom, whose boards are pricey but reflect the actual cost of research and production. No free lunch.
We all bought stuff off ebay, but this is how we learn (not to buy again), if the amp is 5$ but an tda7293 by itself is more than that , then how do you think is going to sound? Parts are not good for sure, I would remove all the components from one boards and replace with new ones.
See what's happening.
See what's happening.
Exactly. It is all a learning experience.
To be fair, I use a lot of used parts. But I would never misrepresent what they are. I almost never sell stuff anyway.
I took some 7293 chips out of a dumpster boombox and built a very nice power amplifier with them. I used nice capacitors, metal film resistors, a used transformer and rectifier. The performance still surprises me. Less than $20 worth of new parts.
Much more satisfying than buying junk off of ebay.
That's what I had to do to get the fleabay board to work. The chips worked fine. Not worth the effort, at all.
Last edited:
Thank you all for the suggestion.
As I said, I tried to replace the TDA7293 with a supposed genuine one, since RS Components is IMHO a certified seller. Price is less than 4$ for one piece. If you go to the source (that PCB costs 0,1$, I wouldn't spend more for 1000 pieces...) the cost could be like that.
Philosophycal discussions apart, fact is that this is an incredibly simple board designed according the manufacturer specifications.
I embrace scottjoplin suggestion regarding the rectifier and I will proceed on the replacement. There are no other active components or semiconductors excluded the IC itself, the rectifier and a 1N4148 on the mute/off circuit.
I wrote here to understand if there was something I didn't considered.
On ebay there is plenty of junk and plenty of almost good things, I do know it.
These boards have been bought bearing in mind that the only useful thing I need was the PCB even if the TDA7293 was a fake one, so I proceeded like that.
As I said, I tried to replace the TDA7293 with a supposed genuine one, since RS Components is IMHO a certified seller. Price is less than 4$ for one piece. If you go to the source (that PCB costs 0,1$, I wouldn't spend more for 1000 pieces...) the cost could be like that.
Philosophycal discussions apart, fact is that this is an incredibly simple board designed according the manufacturer specifications.
I embrace scottjoplin suggestion regarding the rectifier and I will proceed on the replacement. There are no other active components or semiconductors excluded the IC itself, the rectifier and a 1N4148 on the mute/off circuit.
I wrote here to understand if there was something I didn't considered.
On ebay there is plenty of junk and plenty of almost good things, I do know it.
These boards have been bought bearing in mind that the only useful thing I need was the PCB even if the TDA7293 was a fake one, so I proceeded like that.
A used, but genuine, chip is almost certainly better than a fake. This assumes that it hasn't been damaged.
A used, genuine chip will show evidence of being washed. A fake will probably have at least one glaring defect. The most obvious defect is that the heat sink tab will not be flat.
I don't know how long these chips can last, but mine came out of a unit that was manufactured 18 years ago.
You can damage the chip with haphazard disassembly. Short all the leads together before removing or installing the chip, or before removing ANY component from the board. The shorting wire should be the first thing you install, and the last thing you remove. 24-28 gauge wire threaded through the pins should be fine.
Thanks for the tip, would explain one of my past experiences with a chip amp.
I actually enjoy swapping out the garbage caps and tuning up the sound on lo-buck stuff, is frustrating though when the boards have botched circuits however. I got one where the polarity was marked wrong for the incoming power. I guess that cost extra to do correctly!
I actually enjoy swapping out the garbage caps and tuning up the sound on lo-buck stuff, is frustrating though when the boards have botched circuits however. I got one where the polarity was marked wrong for the incoming power. I guess that cost extra to do correctly!
The TDA chips are particularly vulnerable to damage (even from cold solder joints) due to its feature of having the voltage supply of the output stage and the voltage supply of the driver stages separated. Power up one without the other and it's toast.
Used correctly, they've very robust. No thermal runaway is a nice feature.
Used correctly, they've very robust. No thermal runaway is a nice feature.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.