Analog Wire - All Comments

Decompensated op amps to the rescue

模拟信号链 — Thu, 23 May 2013 00:44:27 GMT

作者: TI专家Bruce Trump 翻译: TI信号链工程师 Rocky Xiong (熊尧)

大家公认的事实是单位增益稳定放大器比非完全补偿放大器更流行，且取得了压倒性的优势。这说明什么呢

RS-485 - Who says you can't teach an old dog new tricks?

Neel Seshan — Wed, 22 May 2013 02:16:56 GMT

Bonnie, Yes. TI has Isolated RS485: ISO3082 family for 5V RS485 and ISO15/ISO35 for 3.3V RS485.

SN65HVD20 does help when the ground movement is less than 25V and eliminates the need for isolation.

RS-485 - Who says you can't teach an old dog new tricks?

Bonnie Baker - WEBENCH Design Center — Tue, 21 May 2013 16:38:48 GMT

Very exciting product line. Does TI have an isolated RS-485 device? In low power applications, does the SN65HVDHVD20 replace the need for isolation?

DAC Essentials: String Theory

Kevin Duke — Mon, 20 May 2013 18:10:20 GMT

Hi Eric,

Digi-pots could certainly also fit in this application, that's most likely the typical approach, and TI offers a few digi-pots, though this isn't currently an area of focus. It all comes down to exactly what the system level goals are and if you really need to go through all the pains of going outside the box by using a MDAC instead since some of the specifications you would need to go about this aren't included in every MDAC datasheet. A benefit for instance might be that the digi-pots are not going to offer the same number of steps as a 16-bit DAC, but it's up to your application whether you really need that level of granularity or not.

This isn't something I've put a lot of thought into in the past, but this topic has received a lot of attention here on the blog, we'll try to look into this deeper and have a future post about it.

DAC Essentials: The pursuit of perfection

Kevin Duke — Mon, 20 May 2013 14:55:47 GMT

Aaron,

I am glad you have found the series helpful. Be sure to stick around for the remainder of the series and let us know if there is anything you would like to hear more about.

DAC Essentials: String Theory

Eric Fletcher — Sun, 19 May 2013 14:27:33 GMT

Bonnie,

I suggest that you look in the portfolio of your best ennemies at Analg Devices. Their "Digital Potentiometers" are exactly what you need!

(AD5124,AD5204,ADN2850, AD5272_74,AD5291_92, etc). They are available with nominal resistor value from 1kOhm to 100kOhm,

and 128, 256 or 1024 steps. Digital control comes from SPI or I2C buses. I designed with them programmable anti-aliasing filters

based on Sallen-Key architecture, up to 8° order. With behaviour Papoulis-Legendre and Cauer-Elliptic.

Filter for thought

Bonnie Baker - WEBENCH Design Center — Thu, 16 May 2013 21:12:40 GMT

Soufiane,

Very nice article for such a complex topic. I suspect that you along with others and my self have taken the time to calculate active filter's R-C values for all of the type of filters, including Bessel, Butteworth, Chebychev, and Inverse Chebychev, etc. This is a challenge and takes time! There is a new software out there that will help you get beyond the FilterPro that you have highlighted in this article. It is from TI and it is called Fitler Designer. The URL for this site is: www.ti.com/.../webench_filters.shtml . This filter program combines the best from Ti including TI's FilterPro and National's Filter program.

Current feedback amplifier...how do I make it work for me?

AKHILESH KESAVAN — Thu, 16 May 2013 08:38:48 GMT

Hi Xavier,

Interesting Math......Also For Bonnie's question, I think it all comes down to the trans impedance gain you are looking at it/.... As such CFA amplifiers have poor gain accuracy compared with VFA's. Current-feedback amplifiers are rarely used for high gains, particularly when absolute gain accuracy is required.

If interested split the gain stage into two. CFA will be useful where you need high slew rate, excellent settling times etc....

Let’s take this driver out for a spin

Soufiane Bendaoud — Wed, 15 May 2013 20:24:25 GMT

Akhilesh and thank you:

Yes as you stated this is mostly for SAR ADC's. The DC and AC specs of the op amp can have a serious impact on the ADC. In higher resolution converters (>14bits), the noise performance of the op amp becomes critical and can degrade the overall SNR to the point where it just doesn't make sense to use a high resolution ADC.

Let’s take this driver out for a spin

AKHILESH KESAVAN — Wed, 15 May 2013 19:02:34 GMT

Hi Sofiane,

Very good article to start with...I think in a nut shell, we can say that the op amp should not significantly

degrade the overall DC or AC performance of the ADC.

I think here you are talking about SAR ADC driver circuits.. In most cases the selection heavily depends on the input capacitance of the converter (S&H circuit) and appropriately selecting the RC value to help supply the currents.

Selection of these driver op amps will be difficult when we deal with multiplexes data acquisition systems, where one needs to consider the on resistance, On resistance modulation(variation of resistance with applied signal).

Generally we should look for UBW, settling time, ouput impedance, open loop gain, Noise levels.....

DAC Essentials: The pursuit of perfection

Aaron Chen — Tue, 14 May 2013 21:15:49 GMT

These series really help me to know what DAC is :)

Thanks your sharing!

Let’s take this driver out for a spin

Soufiane Bendaoud — Mon, 13 May 2013 16:30:24 GMT

Thanks Bonnie:

That's correct. When I say driving the ADC, I also mean the external RC (C for current absoption) and yes you're right about the output impedance. As far as stabilities, there are several techniques to overcome the problem but I couln't get in the details here. Suffice it to say, it's not just a matter of putting a 50 to 75Ohm resistor in series with the output of the op amp, yes that can work sometimes but the actual value depends on several parameters such as phase margin, open loop gain, noise gain, zo, cap size etc.....perhaps it's time for a refresh on that with laborious mathematics?

Let’s take this driver out for a spin

Bonnie Baker - WEBENCH Design Center — Mon, 13 May 2013 15:57:39 GMT

Soufiane,

Very nice article. This is a nice starter article. I like how you start with the power supply requirements. You have gone through a lot of important specifications, in particular the settling time of the amplifier.

I am curious about what ADC that you are trying to drive. In reality, you are driving the input capacitance of the ADC that requires supplying high speed current spikes to the converter. More than likely, you will need a R/C “lowpass” pair on the output of the amplifier to help supply these currents to the ADC. Therefore, the output impedance of the amplifier is very important as you do have to work through stability issues.

DAC Essentials: String Theory

Kevin Duke — Thu, 09 May 2013 16:47:42 GMT

Hi Bonnie,

Thanks for the comment! Many DACs look a lot like digital potentiometers on the surface but there are a few caveats underneath the surface. The biggest one is exactly what you pointed out - that there is a buffer sitting on the output of a typical DAC. Since we want to deliver as low noise performance as possible and due to some design constraints the output buffer is typically bandwidth limited.

So, to answer your question the string DAC could be used in the way that you have described but only under very specific circumstances. For DACs that are better suited to pretend to be digital potentiometers you're going to want to focus on R-2R or MDACs that don't feature output buffers. Stay tuned for our next blog post on that topic....maybe we'll even make a deep dive on using DACs in this way.

Being negative can still be positive

Loren Siebert 1 — Wed, 08 May 2013 23:03:29 GMT

I had in mind applications where high frequency linearity is a key specification. The LMH6554 has an OIP3 over 40dBm at 250MHz, but to get the rated linearity it is necessary to set the output common mode very close to mid supply.

Sometimes, an inexpensive switching regulator can give that extra flexibility that makes the design come together.