The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> currents or voltages?
https://designers-guide.org/forum/YaBB.pl?num=1123395379

Message started by svensl on Aug 6^th, 2005, 11:16pm

Title: currents or voltages?
Post by svensl on Aug 6^th, 2005, 11:16pm

I was wondering why some people use current-based integrators and some voltage-based integrators for continuous time Sigma Delta A/D converters? What are some advantages of using currents over voltages? Are currents easier to handle, for example in the feed-back DAC? I would appreciate if someone could share their thoughts/experience on this.

Thanks.

Title: Re: currents or voltages?
Post by svensl on Aug 7^th, 2005, 4:25am

After some research online I came across the following abstract.

"It is often said that there is a fundamental difference between current-mode and voltage-mode circuits.
This conjecture is discussed in technical and philosophical terms, and it is shown that there is no such performance
difference to be found, and that it is not possible to make a clear divide between ‘voltage mode’ and ‘current mode.’
And yet performance differences appear in the literature. It is shown that they come from the different design
practices of the current-mode and the voltage-mode research groups. The conclusion of this paper is that the practical knowledge of the current-mode research groups should be re-integrated into main-stream IC design, and that all propaganda of the type ‘current-mode is better than voltage-mode’ should be stopped immediately."

This is from "Why ‘Current Mode’ Does Not Guarantee Good Performance" by HANSPETER SCHMID. I found it quite amusing and valuable as well.

Cheers

Title: Re: currents or voltages?
Post by Paul on Aug 7^th, 2005, 1:12pm

Hello,

I would tend to say that voltage-based designs have an advantage in speed and that there are some differences in arithmetics: it is easier to sum currents, but when calculating differences, you must avoid getting negative values.

The most important thing is that there are to my best knowledge little current-mode designs in commercial production, which IMHO means that they cannot be better than voltage-mode designs. I may of course revise this statement if somebody is able to mention current-mode based commercial designs.

Paul

Title: Re: currents or voltages?
Post by Ken Kundert on Aug 8^th, 2005, 1:33am

I am not aware of today's techniques, but back in the 70's and 80's, many of the high speed oscilloscopes were designed using current-mode bipolar circuits. They tended to be faster than voltage-mode circuits because the impedance levels were low, which greatly reduced the impact of parasitic capacitors.

-Ken

Title: Re: currents or voltages?
Post by Paul on Aug 8^th, 2005, 9:35am

Ken,

thanks for reminding me there are other technologies than CMOS ;) I cannot really comment on bipolar design techniques and focused my statement on CMOS. It seems I must however consider again the impedance issue and its effect on bandwidth and possibly revise my previous statement on speed.

I found two interesting documents from Poulton's group at Agilent (ISSCC '02 and '03):
http://www.labs.agilent.com/Ken_Poulton/papers.public/2002isscc_10_1_tal.pdf
http://www.labs.agilent.com/Ken_Poulton/papers.public/2003isscc_18_1_pg_slides.pdf
From the second document:
"Current-Mode T/H and Gain
Good Linearity: Current mirrors with cascodes are 8 bit linear.
Poor Accuracy: Gain and offset errors."

Paul