the soldering on the top pic looks attrocious.
Give it a good scrub with a toothbrush and dish washing detergent.
Then look to see which soldered connections are still good and which are shorted and which components fall off.
The lower pic looks a lot better. Shiny joints that appear separate from their neighbours.
Give it a good scrub with a toothbrush and dish washing detergent.
Then look to see which soldered connections are still good and which are shorted and which components fall off.
The lower pic looks a lot better. Shiny joints that appear separate from their neighbours.
He's referring to the post just before the one with embedded pictures. The one you replied to telling me that all forum members can attach pictures. Regardless, I'm glad you think the board with the flowed solder looks good.
Sent from my SM-N900V using Tapatalk
I wasn't trying to confuse you. I meant his first attempt to post those two pictures (post 19) had an explanation of the soldering progress.
Last edited:
Are the PGA3211s running reliably for you? Another fellow and I have been experimenting with them for a while trying to figure out why some will lock up and/or burn up on volume changes. We seem to have cured it by changing the write speed of the code and moving the pull up resistors on the SPI lines right beside the PGA2311. I'm always interested in other's experience with these.
Last edited:
Internal pullups are around 22K in an Atmega328. One or two devices will work fine with I2C if they are near by. If you are using a bunch of devices it usually won't work reliably though. I haven't found a definite reason for the PGA2311 failures but we seem to have them cured with external pullups and altered write timing.
I am also using an Arduino, specifically an Uno. No pull-ups on SPI, never seen anyone do that before. Before getting to this stage I was using a single PGA on a poorly routed bread board and never had an issue once I worked out the correct SPI settings. Maybe I was lucky but none of the reference designs I followed included them either.
One caveat for future users of the PGA series volume controllers: The analog voltage rails must come up before the digital supply. This is not in the datasheet but I was having issues with bleed through of the 5v line to the analog rails... after some reading I found a post on TI's forum where someone on their end confirmed that analog supplies should lead. It might not be an issue with linear supplies but I'm using a little SMPS and the bleed through was screwing with it's ability to reach it's target voltage. On the bread board I added a series resistor and big cap to slow the rise of the digital 5 volts but the PGA's digital side draws so little current that I'm now powering them from an IO pin 500ms after the system powers up.
I can definitely hear an improvement running these in differentially rather than single ended. I'm very happy I went that route.
One caveat for future users of the PGA series volume controllers: The analog voltage rails must come up before the digital supply. This is not in the datasheet but I was having issues with bleed through of the 5v line to the analog rails... after some reading I found a post on TI's forum where someone on their end confirmed that analog supplies should lead. It might not be an issue with linear supplies but I'm using a little SMPS and the bleed through was screwing with it's ability to reach it's target voltage. On the bread board I added a series resistor and big cap to slow the rise of the digital 5 volts but the PGA's digital side draws so little current that I'm now powering them from an IO pin 500ms after the system powers up.
I can definitely hear an improvement running these in differentially rather than single ended. I'm very happy I went that route.
A little trick we learned from using Cirrus version is to put a 10uF cap on the digital power and feed it through a 10R resistor. This will delay the digital power up enough to be able to run all from a single supply if desired.
It's interesting some people are having better luck than others with the PGA231x than others. I'm starting to think the code changes we made fixed the issue. My first problem board was one of those EBay red board three input selectors. PGA2311 Stero Remote Volume remote Controller Preamp | eBay I passed it off as a fake IC but the real IC I replaced it with did the same. A few breadboard versions and a preamp I built also went up in smoke. I began working with Linuxworks who did the AMB LCDuino front end. He was experiencing the same issue and we worked together to get it running reliably.
If you don't need remote control, stick with a stepped attnuator or relay based attenuator.
I wouldn't waste my time or money on any PGA23xx design, nor do I understand why anyone else would.
Been there done that. Not only does the sound of those chips blow compared to a inexpensive stepped attenuator from China, then you have to deal with their tendency to "lock up" and shoot the volume to full blast for no apparent reason possible damaging your speakers! Isn't that oh so lovely??
For a remote controlled option, look for a MUSES 72320 design if you want a volume control chip that sounds literally "transparent" and doesn't have all the glitches/problems of the TI PGA23xx series.
Too bad TI doesn't discontinue these chips and replace them with something comparable to the NJR 72320!
I threw my PGA23xx volume control in the trash where it belongs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.