LM8272: Current Amplifier application

Hi Scott,

You would need to use +/-12 VDC for the supplies, as the maximum supply voltage for this device is 24V. I drew up a quick simulation of the LM8272 in a buffer, with a +/-10V square input. There seems to be a slight settling time on the falling edge of the signal.

The two channels are independent, so I do not anticipate there being any issues between the two inputs.

Can you provide more information on the signal the op amp will be receiving (i.e. frequency, gain)? I should be able to give you a better idea of the circuit characteristics with more information.

lm8272_current_amplifier.TSC

Best regards,
Carrie

Hey Carrie, 

Thanks for the response. The input will be the Analog Output from a NI-9262 cDAQ Module. There are 6 channels, all of which are specified on the datasheet to have independent returns, which I think is incorrect as I believe all returns are common inside the cDAQ module, but I will be using these AO's basically as function generator waveforms, could be a square simulating a clock, could be a sine, a saw, etc... Each channel is specified as +/-10VDC driving 2.6mA / channel. I'll be using 3 of these LM8272's to drive the 6 channels.

Hi Scott,

Thanks for the information. Can you tell me what you anticipate the maximum frequencies of these waveforms will be?

Best,
Carrie

The function generator these are replacing worked to 50MHz

Hi Scott,

If you are wanting to drive 50MHz waveforms, you would need a device that has a higher gain bandwidth than the LM8272, which has a gain bandwidth of 13MHz. 

High Speed Amplifiers, can you help Scott find a suitable op amp for his application?

Best,
Carrie

Hi Scott,

Can you share the input signal you are trying to amplify? Is it a current or a voltage?

Best Regards,

Ignacio

It is a voltage, I will be using 6 Analog Output (AO) channel from an NI-9262 Analog Out cDAQ module to drive up to a +/-10VDC at 2.6mA drive signal operating as function generator signals, which could be simulated clocks, sine waves, sawtooth, pulse, etc... I am using this device to replace an older HP Function Generator capable of generating frequencies up to 50MHz

Here is the technical specs for the AO's

The specifications in the datasheet are correct. The '2mA' on the product page is a simplification. From the datasheet, there are 3 related specs: Per channel = 10mA. This means a single channel can drive +/- 10 mA assuming you're not exceeding the all channel limits

All channels (trip) = +/-16mA. This means that you can drive a total of +/- 16mA combined on all channels before the protection circuitry trips. If you divide that evenly between all channels, that gives you 16/6 = 2.6mA / channel

All channels (hold) = +/-9mA. This means that if you trip the protection, the module remains in the protection limit case until current drops below 9mA combined across all channels. This equates to 9mA/6 = 1.5mA / channel.

So, the 2mA/channel is a bit of a simplification

Hi Scott,

Thanks for the additional information. I would like to ask for your 50MHz requirement. Do you need 50MHz at +/-10V? Any information about your large signal bandwidth and or slew rate will be helpful in finding a device. I was also wondering if you required a dual amplifier or if a single channel device would also work for your application. We do have a somewhat limited dual op amp portfolio especially if the requirement is 50MHz at +/-10V.

Best Regards,

Ignacio

I honestly don't think I will ever need 50MHz, most of the time we will be using these to simulate clocks running up to ~100KHz, what I am trying to do is design to the capabilities that the old function generator provided. Can you tell me what the LM8272MMX/NOPB would be reliable up to in Frequency? If you have a better part that will work better at higher frequency no problem, I'm hoping it is available in the 8-VSSOP package so changes to my design could be minimized as I just completed the layout using the LM8272MMX/NOPB part.

Hi Scott,

Our only drop-in replacement in this package that will work at these outputs would be the OPA2810. I attached the product page below. It is a big jump in terms of bandwidth and slew rate compared to the LM8272. To answer your question, the bandwidth of the LM8272 would depend on the gain configuration you are in. The gain-bandwidth product for this part is 15MHz. The datasheet has additional information regarding -3dB bandwidth performance at different gains, highlighted on page 10.

Best Regards,

Ignacio

OPA2810 data sheet, product information and support | TI.com

Assuming these OPA2810 will work with supply voltages of +/-12VDC to produce the +/-10VDC outputs? if so, I'll update my design. I'm using these as unity gain amplifiers to act as current amplifiers to provide higher drive capability

Hi Scott,

This should not be an issue as the device is rail to rail. The device is on par with the LM8272 in terms of output current as well. We do have a model for this device (attached below) that should be able to capture the behavior of the device. We also have a TINA model for this device.

Best Regards,

Ignacio

sbomam3a.zip