The TDA2822 is used as the µBITx final audio amplifier to drive headphones or speakers. The kit provides wiring instructions to wire to a panel mount “see through” stereo jack.
Several constructors have found that this little IC has gone up in smoke, much to their surprise and frustration!
Reasons for failure
It is not entirely clear why the TDA2822 is failing. In some cases it seems to be the result of inserting a plug into the stereo headphone socket. Inserting a mono headphone plug in the stereo jack could result in a short from the ring to sleeve. Even inserting a stereo plug could result in shorts.
However, there are also reports of the device going up in smoke spontaneously. One theory, from Jim Sheldon W0EB, is that this is because loud pops or extra loud signals cause the 470µF capacitor (C77) in the output circuit to draw high current during charge up, damaging the chip. However, it could equally as well be a run of bad chips, or the fact that the device is running near its voltage maximum (the original chip was rated for 15v maximum).
The first batch of µBITx shows an FCI PI1 TDA2822M chip:
These FCI branded devices have yet to show up with issues.
Raj VU2ZAP tested the current draw from the FCI chip with normal audio use and current was 60-70mA. A dead short with varied drive shows the current draw was between 300 and 800mA. At 800mA the chip got hot! The current did not go above 1A at any time.
Reference
The original ST parts (now obsolete) claimed an Absolute Max of 15v, and also gave that as the maximum operating voltage. This agrees with the specs for the NJ2073D and NTE7155 clones. So running it at 12v should be legal, though nowhere else in the ST datasheet is there mention of operation over 9 volts. A bit of a red flag. ST continues to build the SOIC8 variant, the TDA2822D.
Other manufacturers are making 8 pin DIP packages. These clones may or may not be marked. They are readily obtainable on eBay and AliExpress at very low cost (you can buy 10 for around US$1).
It is likely that cloned versions made in China have been used in some second batch µBITx products. These items may not be an exact copy of the original and may not be as robust. They may, in fact, be a low voltage version of the chip that is incorrectly labelled.
Several constructors have confirmed that their Batch 2 µBITx come with a WX branded TDA2822M as shown in the photo below.
Not all Batch 2 µBITx have this chip installed. It is likely that all current Batch 3 kits are affected. These WX chips seem to spontaneously combust at some point when used in a circuit with 12v DC applied, as illustrated in the photos below:
Testing by the GQRP club suggests that Chinese chips purchased online may not take more than about 6v DC (at around 45mA), and get very toasty at the original’s rated maximum voltage of 15v.
First reports of fried TDA2822’s seemed to be due to a shorted audio output, e.g. when plugging a mono plug into a stereo jack. In series with a proper 8 ohm load, the 470uF cap should be fine. With a short to ground, there will be a quick surge current of unspecified amps from the TDA2822 till the cap is charged. Though if that’s a failure mode, it is not the only one. More recent reports suggest that these chips are being fried spontaneously, suggesting an over-voltage issue.
Protecting your TDA2822 against short circuits
There are several possible approaches to short circuit protection for your TDA2822.
To reduce current in rush, the easiest approach would be to insert a 4 ohm 1/2 watt resistor in the output line to the speaker or headphone socket. While this reduces audio output slightly, it also protects the chip against short circuits, and slows the current inrush to the DC isolating 470µF capacitor in the output circuit.
Another approach would be to reduce the size of C77, the DC isolation capacitor – perhaps to 100µF or even 47µF. Circuits for the TDA2822 often use a 100µF capacitor. Experimentation may be required if the audio begins to sound a bit constrained.
One or other or both of these fixes is recommended fix for FCI chips, but these fixes won’t address the issues with a WX branded chip.
Protecting your TDA2822 against high voltage failure
The best option for all chip types would be to reduce the voltage feed into this chip to bewteen 5v and 9v. Since the audio stage is connected to +12v (rather than to the relay switched RX line) makes this a bit easier to achieve. A regulator or buck power supply is recommended for the feed to the TDA2822.
If you have a WX branded part in your µBITx, a mod to reduce the voltage to this chip is considered ESSENTIAL. The first step is to cut the short trace on the back of the board into the square pad of C76 (near X2).
Add an LM7805 regulator, pin 3 going to that square pad of C76, pin 1 going to
the feedthrough at the other end of the cut trace, and pin 2 going to ground. Maybe glue the LM7805 face down onto the back of the board, with the leads in the vicinity of C76.
Ideally, add a 0.33µF capacitor (and realistically any value of capacitor from 0.1 to 10µF) from pin 1 to ground.
This modification has now been shown to work, and details on how straightforward this is can be seen in Raj VU2ZAP’s experiment where he adds an 78L09 surface mount chip.
Replacing your TDA2822
If you need to replace the TDA2822, it should not be difficult to find a replacement IC. If you don’t mind waiting these can be procured very cheaply from the Far East. If you want it more quickly, then they can be order through a local supplier.
Make sure that you also purchase a socket (or use machine pin headers). This will make it easier to replace the chip in future if it blows again.
Removal of the IC is easily achieved by snipping the pins above the board, and then removing them one by one using a soldering iron, solder wick and needle nose pliers.
Replacing your TDA2822 with an LM386 Module
You could use an LM386 module as a replacement to the audio amplifier on the µBITx board. LM386 modules are readily available on the internet for well under a US$1. Buy several so you have spares in the junk box. Source the audio feed from the volume control output and wire up the stereo jack or speaker to the output of the module.
Replacing your TDA2822 with an LM386 using a DIL socket adapter
The TDA2822 and LM386 are not pin compatible, but it is possible to make an adapter to plug into the TDA2822 DIL socket (using two further DIL sockets).
The pinout map supplied byClaude HB9CGL is as follows:
LM386 pin – TDA2822 corresponding pin
2 – 4
3 – 7
4 – 4
5 – 1
6 – 2
Howard K4LXY shows his “adapter” using the LM386 to replace the TDA2822:
Claude has left pins 1 and 8 of the LM386 unattached, and in this configuration the LM386 datasheet states that it has a voltage gain of 20, or 20*log10(20) = 26dB. The LM386 on the Bitx40 has a 0.1uF cap between pins 1 and 8 to increase the gain to 46dB. The original TDA2822 of the uBitx has a gain of 40dB, so you may want to try adding a cap between LM386 pins 1 and 8 if you need more gain there. A larger value cap (up to 10uF) would improve low frequency response. Howard decided to include this capacitor.
Thank you so much for posting this. I assumed I did something wrong as I just finished my uBitx build, and my wx tda2822 chip started smoking as soon as I plugged in the speaker. I assumed I did something wrong and was afraid I burned up the whole thing. I googled and found this and now I’m relieved that I can potentially fix this. I’ll probably do the LM386 fix. That seems to be the easiest and more of a surefire way of solving the issue. Thanks so much!!!
Just as a note: Apparently the latest round of shipped uBITX kits actually use a socket and a third maker of the TDA2822 (the very latest uBITX with a socket for the TDA2822 are apparently using YouWang UTC2822M). At least if these end up having similar issues as the infamous WX chips it should be an easy swap! 😀
Thanks for the post. I quickly found why my ubitx was blowing fuses. Guess what… My ubitx has the YouWang UTC2822M version of the TDA2822. Mine blow simply turning the transceiver off and on during an attempted calibration setup. If I take the extra effort to reconfigure the system for the LM386 chip, is there evidence that it will be more robust.
Hello,
This is DS4HDW from South Korea.
I just received ubitx by DHL. I worried about this TDA2822 problem, too. However, there is not a TDA2822 chip on the main PCB that I received. Is it means that the problem is solved in this new version?
You will have the new v4 board. So you should not worry!
Mike ZL1AXG
Can a 78L09 in a to-92 package work? 100ma from this to-92 package enough?
I wonder if the NJM2073 is a suitable sub for the WX 2822? No other changes required??
It is fully compatible I belieeve
I need help to solve a malfunction.
I have a version 3 kit with the famous TDA 2822 in the WX version which for the moment does not produce smoke and the output is good.
The problem is that in SSB, even with strong signals, the audio is meowing and not understandable.
The same problem with 8/16 and 20 ohm speakers.
What can I do?
Thanks for the appreciated advice.
73 de IV3XHM, Toni
Antonio
I saw that you have asked list members to comment …. Mike ZL1AXG
Maybe the batch of TDA2822M’s used ware of the low voltage variant with 5.5V maximum supply voltage?
This variant exists and is sold by the G-QRP club for example.
It was originaly designed for battery powered receivers.