Op-Amps or Transistors that are Functional in a Uniform 8-12T Magnetic Field

Hello Kota,

8-12 Tesla, that is a very strong B field. This is like inside of NMR spectrometer or NMI Imaging system. When you said uniform field, do you have any percentage changes in B field? I have to think about this, and it may not be possible without shield. 

In addition, what RF frequency or ranges are you trying to amplifier? Please be more specific about your applicaiton. I may be able to find alternative way to get you what you want. 

Regards,

Raymond

Hello Raymond, 

Thank you for your reply. Yes this is for NMR purposes. I am not sure exactly what % changes in the B field the circuit has to undergo, however, it won't be fluctuating from 8-12T (I just don't know exactly what value, but it will be in that range) it will be a constant, static field in one direction. 

Unfortunately shielding will not be an option for me since I cannot disturb the uniformity of the field, and shilding tends to change the direction of the magnetic field lines. I found some papers testing transistors' performances under magnetic fields of this strength and it seems like there are consequences in the gain, and noise. I would be okay with having to mitigate these consequences as long as I get confirmation that the parts themselves will not be damaged from the field (ex. if the transistor itself has magnetic metals in it). 

The frequencies are 125-500MHz. Thank you for your help!

Hello Kota,

Unfortunately, an opamps behavior under a magnetic field is not a spec that we test on our high-speed amplifiers. Because of this, it is difficult to give you an answer on how a part will behave or which is more suited for your application. Some of our isolated amplifiers mention being resistant to magnetic interference (http://www.ti.com/isolation/isolated-amplifiers/products.html), however none of these parts seem to have the bandwidth that you need.

Let me know if you need help choosing an amplifier based on your circuit's requirements. 

Best,

Hasan Babiker

Hello Kota,

The forum is redirecting to AMP-HSAMPS for reply. However, I am still able to help you out or provide you with ideas. 

Here is my follow-up questions per the application issues, perhaps HSAMPS group will provide with a different set of questions.

1. If no and insignificant changes in flux density or B field, no significant currents can be induced. But this is not possible, so we need to know the fluctuation of B field, 1% or 7% or x% -- this will may help to estimate the worst case in differential mode signals. I assume an amplifier circuit can handle and reject common mode noises. 

2. An amplifier circuit has to be powered inside of the B field. The operating current will interact with the field, which it will generate motion. And it will not be directional. So you need to keep operating current as low as possible. 

3. What kind of load are you driving at 125-500MHz? what is the driving current? I have a feeling that discrete circuits may be a better way to go, if it works at this bandwidth. You have do input/output impedance matching at this frequency band. 30dB is approximately 32 in gain and it should be able to get from a simple transistor circuit. 

Regards,

Raymond

Hello Hasan, 

I see. Is there a way you can find out what type of material the op-amps are made of (if they have magnetic metals)? Maybe I can find ones that are appropriate for my bandwidth, and get some samples so that I can test their behaviour under the magnetic field myself. 

And yes, I would love some help finding amplifiers for my purpose. Thank you so much for the help!

Hello Kota,

Could you give more details on your circuit requirements? You will be needing a bandwidth of at least 500MHz? What type of signal are you inputting to the circuit and what type of load is being driven? Are there any other design limitations outside of gain and bandwidth?

After we've narrowed down some parts for your design, I can try to see if any of them don't use magnetic metals. If not, then getting some samples may be the best way to go, outside of Raymond's suggestion of building a discrete circuit.   

Best,

Hasan Babiker

Hello Raymond, 

1. I see. I will check what the maximum fluctuation will be. I should be able to have this information sometime next week. However, one of my earlier concerns were even perfectly static magnetic fields (no change in flux) can affect the circuit by affecting any magnetoresistive/ferromagnetic/paramagnetic/magnetic metal materials inside of any components. Let's say, even in an ideal situation where I only get a perfectly stable 10T magnetic field, the circuit's behaviour inside and outside the field will be different due to different materials being affected by the static field. 

2. I see. Yes I will have to keep this in mind when I am designing the circuit

3. What I know at the moment is the fact that I have to impedance match to 50Ohms after the preamplifier since there will be a transmission cable with 50Ohms impedance to connect to another system. Before the preamplifier will be a very small NMR probe, where I will get the signal ~125-500MHz since those will be the resonating frequencies of the particles. And yes, if a transistor circuit is sufficient I am happy to go that direction, however, I will still need confirmation that I can find transistors that will operate in this magnetic field. 

Hello Hasan, 

Yes I will need ~30dB gain for signals 125-500MHz. The preamplifier will connect directly to a NMR probe, where the signals are coming from (the resonance frequencies of the particles I am looking for are in that range). After the amplification, the preamplifier circuit will be connected to a transmission cable with impedance ~50Ohms, so I will require impedance matching there. 

Yes it would be great if you can check if they don't have magnetic materials. Otherwise, would you be able to recommend op-amps that will meet the other requirements and I will test them under the magnetic field myself?

I appreciate your help. 

Hello Kota,

What is the voltage range of your input?

-Hasan

Hello Hasan, 

At the input it will be ~micro volts 

Hello Keto,

If you are building discrete circuits, the type of transistors you will need is in GaAs, GaN or even Germanium high frequency (GHz) Transistors

In addition, you need to tell us what is input and output waveforms, type of load (inductive, capacitive or resistive), and the output current. 

Yes, I agree with film and wiring materials inside of IC is important. If you select diamagnetic (Cu, Au, Ag etc.) or paramagnetic materials (Al, Pt, Cr etc.), you will have better performances. Most high end RF amplifiers or transistors is made of diamagnetic materials. 

Most of uniform magnetic field is constructed in cylindrical shape. Can you drill a hole on a side of tube and insert your probe in this manner, rather than within the field? 

Regards,

Raymond

Hi Kota,

think also of the Lorentz force: Each moving charge is experiencing a force which pushes it in a direction perpendicular to the moving path and the magnetic field lines. Remember how the Hall sensor is working. So, I think it will play a role how the transistors with their pn-junctions are oriented relative to the magnet field.

Kai

Hello Kota,

It may be possible to use a current-feedback amplifier to achieve your high gain at 500MHz (splitting gain into two stages may be a better option). The biggest limiting factor, however, will be your noise. Our low noise parts typically have broadband noise range of 1 to 2 nv/sqrt(Hz) (this is without looking at whether these parts can fit into your design.)  So with a bandwidth of 500MHz, you are looking at an input noise of at least 22 uVrms, which will certainly dominate your input if you are within the uV range.

Best,

Hasan Babiker 

Hello Raymond, 

Thank you for the information. The input is an AC signal from an NMR probe ~micro volts that will last about 100ms, and the output will be connected to a 50ohm transmission cable. As the output current depends on the amplifier design, the output current is unknown. 

Hello Hasan, 

I see. Would I be able to get the part numbers of the low noise parts? Also, do you have a solution/suggestion regarding the noise issue so that my signal is now drowned out by noise? 

Thank you for your help

Hello Kota,

I apologize, my previous calculations were done for a bandwidth of 500MHz and not your true bandwidth of 375MHz. Even then, you would need a part with a broadband noise 0.052 nV/sqrt(Hz) in order to achieve a noise level of 1uVrms at your input. This is still a smaller noise level than any opamp we have available. The only solution would be to narrow the bandwidth of your application. 

Best,

Hasan Babiker 

Hello Hasan, 

I see. Thank you for the clarification. If it is not possible with an op-amp do you have suggestions for transistor models with the lowest noise figures that are made of materials that are less-affected by magnetic fields? It seems like cascaded transistors may be the best solution here. 

I really appreciated your help. 

Sincerely,

Kota Chang

Hi Kota,

what is the equivalent circuit of your NMR probe? Can you show a schematic and give further details?

Kai

Hello Kai, 

I don't have an equivalent circuit of the probe at the moment since I haven't been able to make any measurements on it, but it is a ~3cm long copper trace around a teflon block. the teflon block is about 3mmx30mmx10mm. half of the copper trace (1.5cm) is on the lower half of the 30mmx10mm side, then wraps to the other side

Hello Mike, 

Thank you for the information. I will try to look for components with these in mind and test them myself.

Would you recommend me to request samples? Or should I just try to buy a range of different parts?

Thank you for your help. 

The ATF-54143 has been used in many NMR/MRI front ends. However it was recently discontinued, and I'm not aware of an exact equivalent part. Mini-circuits recommends the SAV-541+.

Hello Mike, 

I see. Thank you for your help!