Rail to Rail High Bandwidth and High slew rate opamp

[Akbar Quraishi]

Hello All

I am a student trying to design a rail to rail DC Rectifier with input of 8 MHz. I know that I require an OPAMP with a high GBW and high slew rate but I a not able to find a correct one can someone please look in this matter

Thanks

 

[ronsimpson]

Rail to Rail excludes you from fast.
What slew rate are you looking for?
What GBW?

 

[crutschow]

Note if you are trying to build an ideal rectifier using an op amp, that it's very difficult to do that a 8MHz, even with a high-speed op amp.
What signal exactly are you trying to rectify?

 

[ronsimpson]

THS4221 120/230mHz, 990V/uS TI (R-R output not input)
I can find faster but.....

 

[Akbar Quraishi]

I am trying to sample a sine wave with 5-8 MHz. I tried with LM6142 but it didn't work at higher frequencies. My circuit was simple, a voltage follower followed by an inverting amplifier. i am looking for slew rate of greater than 520V/us to get 10Vpp at 8MHz. Thanks for the positive replies. Hope to get som more guidance form you all

 

[crutschow]

You mean a sample-and-hold circuit?

 

[ronsimpson]

i am looking for slew rate of greater than 520V/us to get 10Vpp at 8MHz.

THS4221 120/230mHz, 990V/uS TI (R-R output not input)
I can find faster but.....

LM6142 9mHz, 25v/uS

 

[AnalogKid]

THS4221 120/230mHz, 990V/uS TI (R-R output not input)
I can find faster but.....

Pretty impressive for a voltage-feedback part.

ak

 

[Akbar Quraishi]

Thank you Ron. do you know any other opamp that can be also looked at

 

[ronsimpson]

THS4541
THS4225
EL8102
OPA836
LMH6611
LT6200
FHP3132

 

[Akbar Quraishi]

Thanks again Ron. I tried the ckt with THS4221-22, but its not giving the result I need. I have attached the ckt diagram. can you please look it for me and tell me whether its correct
Thanks again to all

 

[ronsimpson]

Make R3 and R2 the same value. (10k)

You have other problems but first make the inverting circuit have a gain of -1. not -1.5

 

[Akbar Quraishi]

Yes i did that. What are the other problems it has?

 

[ronsimpson]

I think 10K is too high for high speed. The capacitance (input, output, board, etc) and the 10k makes a RC low pass filter. At 8mhz I would try 1k.

 

[ronsimpson]

What you are doing requires a R-R input and R-R output and I was only looking for R-R output. So some of the op-amps I picked probably will not work.

 

[MrAl]

I am trying to sample a sine wave with 5-8 MHz. I tried with LM6142 but it didn't work at higher frequencies. My circuit was simple, a voltage follower followed by an inverting amplifier. i am looking for slew rate of greater than 520V/us to get 10Vpp at 8MHz. Thanks for the positive replies. Hope to get som more guidance form you all

Hi there and welcome to the forum,

So then why do you need 520v/us?
252v/us should be fast enough at 10vpp and 8MHz, right? (Keep in mind 10vpp is only 5v peak, if that's what you really meant).
Of course you may want to go a little above that for good measure, maybe 300v/us.

Also, what are you going to use for the rectifiers?
Also, what kind of accuracy are you looking for?

My recommendation would be to use a fast ADC instead...perhaps 100MSPS or so. You could then also calculate other things besides peak voltage.

 

[ronsimpson]

Also, what are you going to use for the rectifiers?

Using a R-R input and R-R output amp......you get full wave rectification with out diodes.

See he almost got it. The problem is: the gain(s) are not equal. +1 and -1.5

 

[MrAl]

Hello again,

It is hard to tell what is the output and what is the input in that 'scope' drawing because the waves are superimposed on each other. It would also be a good idea to show what the input is, and what the output voltages are and times.

Also, even if it does appear to work that doesnt mean that will work using real op amps anyway. That's because many op amps do not like to be driven from open loop back into closed loop again without a significant delay. This is one of those op amp aspects that you NEVER see talked about on web sites. What this means is that it may not work at all or it just might introduce some distortion. In any case, real life testing would be mandatory before settling on a particular design.

But assuming it does work or it can be modified slightly to get it to work, how does he intend to filter the output? With the op amp driving the output directly an inductor and capacitor would probably be needed to filter it into DC, rather than just the usual capacitor.

Just for reference, here is some scaled time data assuming the op amps were perfectly compliant:
[Vin is 1 volt peak]
-- Results with first op amp gain adjusted:
-- ...time..........Vo1.........Vout...
-- 0.050000 0.370820 0.463525
-- 0.100000 0.705342 0.881677
-- 0.150000 0.970820 1.213525
-- 0.200000 1.141267 1.426584
-- 0.250000 1.200000 1.500000 (first peak output)
-- 0.300000 1.141268 1.426586
-- 0.350000 0.970822 1.213527
-- 0.400000 0.705344 0.881680
-- 0.450000 0.370823 0.463529
-- 0.500000 0.000003 0.000004 (zero output)
-- 0.550000 0.000000 0.463521
-- 0.600000 0.000000 0.881674
-- 0.650000 0.000000 1.213522
-- 0.700000 0.000000 1.426583
-- 0.750000 0.000000 1.500000 (second peak output)
-- 0.800000 0.000000 1.426587
-- 0.850000 0.000000 1.213529
-- 0.900000 0.000000 0.881684
-- 0.950000 0.000000 0.463533
-- 1.000000 0.000000 0.000008 (zero output)
-- 1.050000 0.370814 0.463518

 

[Akbar Quraishi]

yes I tried the op amps. Didn't worked. I found this : **broken link removed**. but not sure of its working. Can you explain

 

[MrAl]

yes I tried the op amps. Didn't worked. I found this : **broken link removed**. but not sure of its working. Can you explain

Hi again,

They are working at 500kHz in that paper, and may be mentioning the delay i was talking about.
But did you first take Ron's advice and change the gain to fix the basic design flaw first?