The LM359 datasheet states this:
So if I have a 1MHz signal I want to amplify, I could do so by 400 (theoretically). So what does the line I have underlined mean? Is the gain limited somehow to 100?
thanks

 

So what does the line I have underlined mean? Is the gain limited somehow to 100?
thanks

It means that the GBW product of 200-400MHz is only valid for gains of 10-100. For example, at a gain of 1, the datasheet says the GBW is 30MHz.

 

I am even more confused now. So by how much could a 1MHz signal be amplified?
From Wikipedia I had found this for a GBWP of 1MHz: The same device when wired for a gain of 10 will work only up to 100 kHz.
That's why I said I could theoretically amplify it by 400 times. Is that incorrect?
thanks

 

Hello,

The main difference between the LM359 and an other opamp is that the LM359 is a current feedback opamp in stead of a voltage feedback opamp.
The current feedback opamp has some other properties, as you might have seen in the daatsheet.

Bertus

 

I am even more confused now. So by how much could a 1MHz signal be amplified?
From Wikipedia I had found this for a GBWP of 1MHz: The same device when wired for a gain of 10 will work only up to 100 kHz.
That's why I said I could theoretically amplify it by 400 times. Is that incorrect?
thanks

No. The product of gain, which is dimensionless, and bandwidth, which has units of frequency, is going to be a minimum of 200 MHz, with a typical of 400 MHz. Check the heading on the columns in the datasheet. I'm assuming minimum, typical, and maximum. This is valid for gains of 10 to 100. The behavior at unity gain is not specified, in the snippet you've given us. So if you design for a gain of 10 the product will be 10 MHz. which is well below the specified minimum. The circuit will be fine as far as being able to do the amplification but may be overkill for that application. You also need to verify that it will be stable at that gain and frequency. Similarly if you design for a gain of 100, the product will be 100 MHz., and again you are OK with respect to the GBW limitation.

 

The 359 Is a "Norton" or "operational transconductance" amplifier. They are quite different from either conventional op amps or current feedback amps.

The LM3900 and LM359, as far as I know, were/are the only two Norton amps that ever made it into volume production. Neither has been very popular, but you can do things with them that you can't with conventional op amps or CFB amps, such as make a non-inverting integrator.

www.ti.com/lit/an/snoa653/snoa653.pdf describes the LM3900 in some detail and may help you understand some of the oddities like open loop gain (Figure 5). A conventional op amp has a "pole" at a very low frequency and the open-loop gain falls with a slope of "-1", (20 dB per decade of frequency with a negative slope). The 3900 has fairly flat gain up to about 1 kHz, then a slope of -20 dB/decade. This means the notion of a single value for gain-bandwidth doesn't really apply. I think the doc explains why, but I haven't looked at in detail for a very long time. Figure 7 of the datasheet for the LM359 shows its open loop gain, which is fairly flat up to 1 MHz.

Note the ap note for the 3900 calls it "new" - and that the note was published in 1972.

 

In depth look at norton opamp, there is a section as to why BW is
not a constant relationship.

http://www.analog.com/en/analog-dialogue/articles/current-feedback-amplifiers-1.html

Regards, Dana.

 

Ok guys, got it. The links were really helpfull.
thanks again