Ashhar Farhan VU2ESE, designer of the µBITx, has posted to the BITX20 IO Group to say that there have been changes made to the wire up diagram to fix the errors. This is good news as there will be fewer mistakes made by constructors meaning more happy customers!
While many will have moved on to one or other of the other firmware versions available for the µBITx, those wanting updated factory firmware will be pleased to know that a bug fix release will be coming soon. uBITx.net will let you know when this happens.
A further comment from Ashhar suggests that he may use the PTT line as one of the keying lines (for the straight key). This will mean a change to the wire up in line with the approach taken by the W0EB software team.
Sajid, VA3QY asked Ashhar Farhan where he could source a new and reliable TDA2822 audio chip.
Ashhar Farhan has replied, “I am working on an audio amp. I am testing it live. I always personally use circuits on the air before recommending them. I hope that by Tuesday I will have something you can use. About 100 out of 4000 boards have had this problem. and less than 20 of them have reported the blow out.”
Ashhar Farhan VU2ESE, designer of the µBITx tells us the sorry saga of the TDA2822:
“Here is the story. This is going to cause a lot of heartaches. I chose TDA2822 after listening to a lot of bad press about the LM386. Upon looking at its harmonic distortions et al, it was found to be a reasonably good device. In, it went.
“Now, unfortunately, the TDA2822 production has entirely stopped. None of the standard suppliers, including mouser, newark, element14 have no stocks left. We tried to find other source of remaining stock from our reliable supplier who has been supplying to us in the past as well.
“He turned up with a batch. We tried the ICs in burn tests for. Ten of them on ten boards at full volume for a whole day. In retrospect, we should have tested every individual IC. That batch had these WX ICs as well. About 100 to 150 of these must have shipped, there is no way to know. The way it works is like this : Not all the ICs turned out by a factory are good. So work very well, some not so well, some don’t work at all. These wafers make it to the silicon industry underground where they are cut out from the wafers and packaged and sold as low grade versions of the same parts. We got a a hundred or so of these lemons.
“We finally located a source of brand new, high quality TDA2822s that are currently being shipped. They cost almost five times as much as we were paying while this was in production. Well, such is life.
“In the meantime, we have to look for alternatives to the TDA2822. I am highly inclined towards making a discrete device amplifier. We will never run out of discrete devices, the audio would be so much better. Does the gang have any ideas?”
Initial feedback suggests that there will not be a consensus reached on a replacement. Some want to stick with the TDA2822 (assuming a reliable source of affordable devices can be found), some would prefer the ability to add in a module of the builder’s choice, and others would like to see the device replaced with discrete components. David N8DAH suggests the drop in replacement – NJM2073D-ND
Ashhar Farhan VU2ESE says:
“HF Signals is keeping the same price for the uBITx because they are able to manage within that price for now. if component prices go up, they might revise the prices.”
What excellent news for purchasers!
Jerry KE7ER says:
“Farhan recommends 100ma for the push-pull IRF510’s on the uBitx. More is better, you get into a more linear region of the FET. But more means more power getting sucked from the battery if portable, that power gets dissipated as heat in the FET’s. I’m going with 100ma.
“With a 500ma quiescent drain current when using a 24v supply, each FET is dissipating 24*0.5 = 12 Watts. That’s a lot of heat. Back off to 100ma and it is only 24*0.1 = 2.4 Watts. Which still seems plenty.
“You want to be very careful when adjusting IRF510 gate bias trim-pots, take it too far and the IRF510’s might be trying to dump 100 Watts or so.”
Note that a Mitsubishi D16HHF1 in push-pull requires a higher bias current setting (perhaps 250 mA or 300aA).