Using THS4303 as antenna amplifier

Eren,

Do you mean the termination resistor (50ohm in most figures)? The reason that 50ohm is used is that it is assumed that the signal source requires a 50ohm terminated load. Sizing this resistor too large in an AC-coupled configuration can result in a large input offset however. For example, if the signal source V_I shown in Figure 44 is going through a DC blocking capacitor, then input bias current can only flow through the 50ohm resistor. If that resistor was instead 10kohm and the input bias current is taken as the 7uA typical, an offset voltage of 70mV will result at the In+ pin.

4V may be appearing at the In+ pin is if you are AC-coupling your input signal and you have replaced the 50ohm resistor with a very large resistance (or left it open). This would create a large offset at the In+ due to the input bias current flowing through the large resistance. Due to feedback action, the op amp will force the inverting terminal to the same voltage, which will appear at the In- pin if it is left floating.

Thank you for your reply Kristoffer,

Yes, my mean was the 50 ohm resistor. You know, the ac coupling capacitor and the Rt (the 50 resistor place)  resistor behave as a RC circuit which create low cutoff frequency point. In my application antenna equivallent circuit as i said before voltage source series with a 8pF capacitor so that to carry the cutoff frequency to lower values, i need high Rt resistor value, I choose THS4303 because it has 1.6Mohm common mode resistor value but I forgot to take in the account the bias current value.

I have a question too,

to put a high resistance value ( for example 10kohm ) to the Rt resistance place can damage the amplifier? I asked this question because when I put 50kohm resistance, the amplifier starting to work normally (Supply current is 34mA) but in five minutes, its supply current is starting to change continuosly (between 28mA - 42mA), then in another fie minutes if i don't close the power supply, the supply current is fixed at 12mA and In+ , In- dc voltages are 4.95V and the amplifier is not working. Close the power supply and open it half an hour later, these steps starting again.

Eren,

I would not expect damage from a high value Rt resistor.

What are you using to generate the 2.5V references, a bench supply? One potential issue is that the 2.5V references you are supplying may not be able to sink current, and there may be some kind of charging effect with increasing offset at the In+ pin over time.

What happens if you turn off the supply and immediately turn it back on? Are the In+ and In- pins still at 4.95V unless you wait longer? Try this - when you turn off the supply, measure the voltage at the 2.5V points (should be 0V). Is there a voltage at those points that is slowly decaying? If so, try adding a 1kohm resistor (or higher) to ground at the 2.5V points so that the 2.5V reference is forced to always source current and there is a path to ground for the THS4303 bias currents.

Kristoffer,

I am using a regulator for 5V for THS4303 supply then to generate 2.5 V using another regulator too for reference points. In the datasheet pointing to make this referance points  low impedance so that I am using a buffer for it.

I don't want to give you any wrong information so that tomorrow, i will start a new test series with a new card and new THS4303 amplifier and note all values and share with you for your advices.

Thanks for your interest

Hi Kristoffer,

I think, i have solved the input offset voltage level problem with your advice.1k ohm resistance made the voltage level stable.Thank you for this solution.

You know, i need high input impedance, i found a workaround solution for it. I want to take your idea about  negative points of this solution.

To set (In-) pin voltage offset to 1V and Rt resistance voltage offset 4V and choose Rt resistance value to make the voltage which will fall on the Rt, 3V with the bias current which flow through Rt and with this way I will have Rt resistance as much as possible and will set the referance voltage to 1V because it is enough for me 1V p-p for output.

Thanks,

Eren;

You did not mention what frequency range you need, but if you want to get good low frequency performance, try an OPA656 FET op amp in a non inverting amplifier configuration. The resistance from the +input to ground can be 1M with the low input bias current FET op amp.

If you were using a loop antenna, a low noise bipolar op amp would be more suitable.

Regards, Neil P. Albaugh   ex-Burr-Brown

Hi Neil,

I want to get 10kHz at low frrequency. At high frequency, as much as possible. The antenna is whip antenna. I will use single supply. Could you suggest any amplifier for this application?

Best regards,

Eren;

Using an op amp in unity gain:

If you use a 100pF capacitor in series with your antenna and the OPA656 + input and connect a 1M resistor between that input and +2.5V, This will give you a -3dB point of about 1.5kHz but your whip antenna is going to be VERY inefficient at low frequency. The amplifier output will be +2.5VDC plus the RF signal; you can block the DC with a big capacitor if you like.

Regards, Neil P. Albaugh   ex-Burr-Brown

Neil;

As you said series capacitor makes low cut off frequency smaller but effeciency is being too low. I have OPA657, could you answer the questions:

1.) Can I use OPA 657 with single supply (Vs:+5V) with full performance

2.) I have used 10M ohm resistance for biasing the input in the simulation and by this way, i get 10kHz low frequency limit., it is working in simulation, is it possible in application

3.) For high frequency application of OPA657 which case (SOT-23 or SOIC8)  is suitable

Best Regards,

Eren;

Yes, an OPA657 is useable if you set its gain to  +7 V/V or higher; it is not stable in unity gain.

A 10M resistor is fine; the choice is arbitrary. An OPA657 has very low input bias current so resistors in this range do not cause large offsets.

Either package should work just fine. Keep your leads very short, use a good bypass capacitor directly on the +5V supply pin, and use a good ground plane.

Regards, Neil P. Albaugh   ex-Burr-Brown 

 Neil;

I need to reach 1GHz at the high end so that I want to use as buffer.

Your mean by saying not stable in unity gain (as using buffer) is the gain peak at the high end of the frequency band or it is not work properly?

Best Regards,

Eren;

The OPA657 will oscillate if used in unity gain. Check the data sheet.

Regards, Neil P. Albaugh   ex-Burr-Brown

Neil,

I have been out for the past couple of days and greatly appreciate your expert insights on this forum!

Eren,

Note that the small-signal (200mVpp output) bandwidth of the OPA657 in G=7 configuration is 350MHz, which is significantly lower than the THS4303's bandwidth of 1.8GHz (200mVrms output) at gain of 10V/V. There are some conflicting pieces information here, as well. In the first post, you refer to the THS4303, which has a fixed gain of 10V/V but the latest discussions I read here involve unity gain. Does this stage need to provide gain? What is the required -3dB bandwidth at this gain?

Hi Kristoffer,

My main purpose is to get 10KHz to 1GHz unity gain, if possible more gain.

I have found first THS4303, it was very very good for my purpose until i learn the input bias current effect. my source signal must be 8pF ac coupled, i can not change the capacitor so that I need nearly 2M ohm input impedance to reach 10KHz but i have experienced that the resistor which is used for biasing is can not be 2M ohm because of the bias current of the THS4303 (7uA) so that I started to search for low input bias current wideband amplifier. For this I have learned from the Neil, need FET input amplifier but I could not find amplifier which reach to 1GHz at the high end of the band  while stable working. So that I choose amplifier in your product list which let to me to reach more gain and highest band.

I want to ask you again some questions which I have asked Neil before to check again:

Can I use OPA657  +5V Single Supply with the performance (G=7  350MHz)?

Best Regards

Hi Kristoffer,

This is not an commercially available antenna which I know. This is my own testing work. I  bought OPA657, waiting for delivery.

I will inform you, when I tried the circuit.

Best Regards,

Hi Kristoffer,

I completed my design, it is working very well.

Thank you for your support

Best regards,