Double checking after damaging 2 audio boards

Status
Not open for further replies.

DaQue

Well-known member
I have managed to blow up 2, maybe 3 (one went missing but I think it had the same problem) specifically they no longer show up on the I2C bus.
To be clear the signals on both pins SCL and SDA look perfect on the scope still directly on the Teensy pins at the board board and on the audio board at the pin inputs. I
I am using a level shifter using BSS138 to isolate the 2.8 volt chip sharing the buss with the audio board. The 2.8v side had 470 pull ups because of very slow rise times. I may have found that issue and before the next board arrives anyone know if that would cause the codec to blow its I2C inputs.

I will try some larger pull up resistors on the 2.8v side maybe 4.7k and check the rise times I guess and just use the 2.2k already on the 3.3v side.

I I most likely goofed up and just plugged the teensy in backward or something that got them but $60 gone I thought I might be missing something.
 
Hi DaQue,
The damaged audio boards you mention are not the PJRC's Audio Adapter boards, are they?
If so, can you share info on your audio board/audio chip? Is it a homebrew audio board at which the Teensy [3 or 4?] can plugged in? It apparently runs an I2C bus at 2V8?

Paul
 
Hi DaQue,
The damaged audio boards you mention are not the PJRC's Audio Adapter boards, are they?
If so, can you share info on your audio board/audio chip? Is it a homebrew audio board at which the Teensy [3 or 4?] can plugged in? It apparently runs an I2C bus at 2V8?

Paul

Yes its a home made PCB and it has worked at times but before I can get everything working at once the I2C quits working even with the port scanner.
The chip is I am working with is an ISD5116 made by Winbond sometime near 2002. Very old analog solid state voice storage chip used mostly in pagers and answering machines. The spec on its digital lines are up to 0.3V above the 2.8v it runs on so I have a sot23-3 2.8v regulater and the isolation fets. The Teensy still talks to the ISD5116 but not the Audio board even when my adaptor board/5116 are removed.
 
PS: My PCB has jumper pads and could just be moved to the alternate i2c pins but I wasn't sure if 2 I2C ports at once would cause issues.
 
PS: My PCB has jumper pads and could just be moved to the alternate i2c pins but I wasn't sure if 2 I2C ports at once would cause issues.

If separate i2c ports are independently pinned and programmable in code they should not cause issues.

Paul made a demo breakout that has some half dozen i2c devices mounted with a set of switches to decide which of three busses they are wired to. And it also has Qwiic connectors for each of the 3 i2c busses to add more devices.
 
If separate i2c ports are independently pinned and programmable in code they should not cause issues.

Paul made a demo breakout that has some half dozen i2c devices mounted with a set of switches to decide which of three busses they are wired to. And it also has Qwiic connectors for each of the 3 i2c busses to add more devices.

I will took a look at 16 and 17 and they are free I might as well use them. Thanks!
 
Did a quick scan of the ISD5116 datasheet. Are you powering the chip with Vcc=2.8V? Is there a reason why you do not power it with 3V3?
View attachment 25829

Paul

Thanks for taking a quick look at it. That's above and beyond.

Yes the ISD5116 pretty well isolated on a board subassembly that plugs onto my board. It is behind a 2.8v regulator on that board in a section with a soldered metal shield. . The 2.8v of the board I plug into is not routed off on any the the 30 pins routing signals off .

I will figure it out. I am pretty sure I am going to use the 2nd I2C on the bottom at this point. This is a great project to do at work when things get slack this time of year. The boss didn't bat an eye when I asked for another audio board, they buy the parts and I just try to get it working.

Fun fact from that data sheet says the chip is at address 80 on the I2c bus but is actually at 40 when you scan it. As near is I can tell they used the r/w bit as part of the address in 2002, or at least differently than how Arduino handles addresses now.

IMG-3371b.jpg

IMG-3372b.jpg
 
Last edited:
Status
Not open for further replies.
Back
Top