Is there a better way,5V I2S to 3.3V CMOS

Bill Fuss · 27th March 2010, 01:43 PM

I'm mating an SAA7210 output to the input of the CS8421. It seems to work fine with simple resistive dividers but have not done any critical listening yet.

Just asking if there is a better way of doing this, it will be all in the player chassis, very short wire runs.

Thanks in advance, Bill

Bill Fuss · 29th March 2010, 12:11 PM

No one has ever done this? That's hard to believe.

I could use a 74VHC125, but would it be an actual improvement?

prairiemystic · 31st March 2010, 09:00 PM

Hi Bill, I don't use resistive dividers for high-speed signals, say a few MHz up. It's a tradeoff between loading the driver IC and keeping a low impedance signal that doesn't get rounded off too much due to circuit capacitance. If the CS8421 is close to the SAA7210 and your traces/wiring is symmetrical then resistors should be okay. If you used say 5K and 10K that's too high (slow), 500 and 1k is 3.3mA load and guessing 20ns delay. 50 and 100 ohms is 33mA and too much for the driver.

Being lazy, I use 3.3V powered buffer IC's with 5V-tolerant inputs. Fairchild numbers here: Quad (74LCX125) or Octal (74LCX244) buffers. The TinyLogic parts are good if you only have one or two signals to level translate.

Bill Fuss · 31st March 2010, 11:45 PM

Thanks for the info. I did start with 1k/500 but thinking it might be too much load I changed to 2k/1k which seems to work fine. The leads are pretty long right now but will end up around 4" when I'm done. The scope traces look great except for the ws, the bottom half was blurred. I put a 1k in series with it and it cleaned up nicely. I'm pretty much searching in the dark with this stuff but I'm learning.

I was pretty impressed with the difference, getting rid of all the SPDIF crap, quite a change.

Thanks again, Bill

ashaw · 1st April 2010, 11:33 AM

how about the TXB0108 from TI. $1

27th March 2010, 01:43 PM	#1
Bill Fuss RIP Join Date: Oct 2005 Location: Central Florida	Is there a better way,5V I2S to 3.3V CMOS I'm mating an SAA7210 output to the input of the CS8421. It seems to work fine with simple resistive dividers but have not done any critical listening yet. Just asking if there is a better way of doing this, it will be all in the player chassis, very short wire runs. Thanks in advance, Bill

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Driving 3.3V parts from 5V controller	janneman	Parts	2	26th August 2008 03:23 PM
How to connect CS8414 to SRC4192? ( 5V->3.3V)	dieringe	Digital Source	10	12th October 2006 05:44 PM
3.3v to 5v CMOS interfacing	Zodiac	Digital Source	9	15th May 2006 10:11 PM
Need help with power supply, ~12v source: 9v, 5v, and 3.3v ouputs	mbcouple	Power Supplies	3	2nd September 2005 06:19 PM
6.3V Rectifier Tube Lit by 5V	snoopyma	Tubes / Valves	8	10th August 2005 06:15 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

29th March 2010, 12:11 PM	#2
Bill Fuss RIP Join Date: Oct 2005 Location: Central Florida	No one has ever done this? That's hard to believe. I could use a 74VHC125, but would it be an actual improvement?

31st March 2010, 09:00 PM	#3
prairiemystic diyAudio Member Join Date: Jul 2004 Location: Edmonton, AB Canada	Hi Bill, I don't use resistive dividers for high-speed signals, say a few MHz up. It's a tradeoff between loading the driver IC and keeping a low impedance signal that doesn't get rounded off too much due to circuit capacitance. If the CS8421 is close to the SAA7210 and your traces/wiring is symmetrical then resistors should be okay. If you used say 5K and 10K that's too high (slow), 500 and 1k is 3.3mA load and guessing 20ns delay. 50 and 100 ohms is 33mA and too much for the driver. Being lazy, I use 3.3V powered buffer IC's with 5V-tolerant inputs. Fairchild numbers here: Quad (74LCX125) or Octal (74LCX244) buffers. The TinyLogic parts are good if you only have one or two signals to level translate.

31st March 2010, 11:45 PM	#4
Bill Fuss RIP Join Date: Oct 2005 Location: Central Florida	Thanks for the info. I did start with 1k/500 but thinking it might be too much load I changed to 2k/1k which seems to work fine. The leads are pretty long right now but will end up around 4" when I'm done. The scope traces look great except for the ws, the bottom half was blurred. I put a 1k in series with it and it cleaned up nicely. I'm pretty much searching in the dark with this stuff but I'm learning. I was pretty impressed with the difference, getting rid of all the SPDIF crap, quite a change. Thanks again, Bill

1st April 2010, 11:33 AM	#5
ashaw diyAudio Member Join Date: Jul 2006	how about the TXB0108 from TI. $1

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)