PGA2310 ground question
I'm building a 5.1 channel amp for my PC. The PCB contains five TDA2050s for the main channels, two LM3886s for a dual VC sub and three PGA2310 chips for volume control, as well as the control logic to generate SPI commands (from an normal up/down switch).
Connecting this lot together on a double sided PCB is a pain but I've decided on three power planes: digital ground, analogue power ground (for localised resovoir caps) and signal ground. All three planes will be connected with a star.
My question: should the analogue ground on the PGA2310 chips connect to the signal ground or power ground?
Attached is a work-in-progress pic of the Mark1 version. It worked by there was a noticible hum on some outputs. I traced the cause to the power electrolytics sharing a ground plane with the signals. I want to do it right this time!
The PGA2310 requires a digital ground which is the return for the +5V logic supply and this should be connected to your digital ground. Also, each one of the analog signals have a ground reference which I believe is labelled AGND. These should be the same potential as the signal references which are being amplified. It will probably be your signal ground.
I would also like to tell you about the PGA4311 which is a four channel version of the PGA2310, however, the analog section runs from +/- 5V and not +/-15V.
I can tell you from experience that these chips are rather nice. They work well as long as the control signal is good. It is not an SPI signal rather a data stream which is shifted in more like a serial to parallel shift register with a load. Once you have the signal timing correct, which isn't too difficult. They are rather sweet and can control volume over a large range or levels from -95.5dB to +31.5db(which usually is too loud any way). Good Luck and keep us informed of the progress.
The Analog and Digital grounds should be connected at a single
point, as stated in the Datasheet, p. 11.
I guess you have the DS, otherwise it's available at:
I've built a preamp myself based on the PGA2310, but I'm sorry to say that I wasn't impressed with the sound. I used ADG406 muxes + OPA627 (Gain -1) + PGA2310 + OPA627 (Gain -1) + battery PS +/- 15V (20 x R6). The preamp sounded "fluffy", like I had wrapped the loudspeakers in a large felt pad. I don't know if it actually was the PGA's fault, but now I'm using the same OP-Amps and two 10k potentiometers as a temporary preamp and the sound is back to "normal". To point out the good stuff about the PGA2310: The level changes were absolutely click-free, also the chip was easy to interface (digitally, that is).
I've thought about trying the Wolfson WM8816, but I don't know yet if they'll send me samples...
Good luck with your project, hopefully you'll get better results than I had.
My current PCB has the PGA2310 AGND connection going to my analogue power ground. I started having second thoughts about this and now I'm going with Beanz suggestion of connecting it to signal ground on my redesigned board. I don't want the voltage spikes caused by the electrolytic charging currents from putting 100Hz hum into the volume chips so signal ground it is. I'll let you know the result!
I was quite impressed with the sound of the Mark 1 version (except for the hum of course!). It's only a PC sound system so not proper hifi: the speakers are Audax 4" aerogel + 10mm tweet for the fronts, Audax 3" full range for the rears and Peerless 8" DVC 831858 for the sub. I like Audax.
I managed to get the serial interface working reliably without too many hassles. All control is all done with a Xilinx 9536XL CPLD as I have acess to the tools and programmer at work. Much easier than a uP for such a simple task!
I've already got the Burr-Brown parts so I'll stick with them. I've seen quite a lot of discussion about the Wolfson parts but nobody seems to have actually used them, probably because of their poor availability.
Thanks for the help guys.
|All times are GMT. The time now is 01:38 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2013 DragonByte Technologies Ltd. (Resources saved on this page: MySQL 30.00%)
Copyright ©1999-2013 diyAudio