Output impedance of TX-SDK-V2 and power supply board

Hi Dmitrijs,

Impedances: The pulser (LM96550) output on the TX-SDK-V2 board is "loaded" to simulate a transducer. If you like to attach an external transducer, you'd have to remove the "load" components on the board.

From page 29 of the TX-SDK-V2 user's guide:

"o The output waveforms of the Pulser can be monitored with high impedance oscilloscope probes at PS0 to PS7 or the outputs can be measured at the 8 RF connectors at J3. These 8 Pulser outputs have load components: CB51 to CB58 = 330pF & RB0 to RB8 = 2.49kΩ.
 o The 8 RF connectors at J3 can also directly drive an ultrasound transducer probe. In this case, CB51 to CB58 must be removed."

I don't have the output impedance of the T/R switch (LM96530). However, on the TX-SDK-V2 board, there are external components (e.g. DB0, RA11, CA12, RA41, RA31 for OUT0, etc.) between the LM96530 output pins and the J4 connector where the receiver (e.g. LM96511) would connect which essentially set your output impedance.

Does this answer your question?

Power Supply: Unfortunately we don't have any layout files for the power supply module except the schematic that you've already seen and have noted.

Regards,

Hooman 

Hi Hooman,

Thank you for your response!

Impedances: Since CB51 to CB58 = 330pF & RB0 to RB8 = 2.49kΩ are there only to model the transducer, is there any reason to keep RB0 to RB8 when connecting an external transducer? Could I replace the parallel connection of resistor and capacitor with just a single series 50Ω resistor to make it a source with 50Ω output impedance?

Thank you!

Hi Dmitrijs,

Removing shunt resistive load: All LM96550 Electrical Characteristics are with RL= 2kohm. That's the purpose R80 to R88 serve. Having the resistance or not may affect some of the AC performance numbers (e.g. rise time, enable time, etc.) but I'm not sure if the load "has to be there". I don't think it'd hurt anything if you experiment with removing them while the external probe is connected.

Replacing the combination with a single 50ohm load: I don't believe that the LM96550 can drive a shunt 50ohm load directly without over-heating. What's the purpose of 50ohm loading that you are after because that is not represent a typical probe load?

For probing purposes, use a high impedance probe at PS0 to PS7 instead, as the user's guide instructs.

If I'm not answering your questions correctly, please let me know.

Regards,

Hooman

Hi Hooman,

Thank you for this information!

Replacing the load with a single series 50Ohm resistor: the reasoning for doing this is that it would be much easier for me to electrically match the external transducers to a 50Ohm source. I made my own transducers and without any matching they reflect ~70% of the input power.

I might be confusing the load impedance with the source impedance, but based on your previous answer, I concluded that the source impedance of the LM96550 pulser by default is defined by the parallel connection of capacitor and resistor. Hence, my idea was to swap out this combination to get 50Ohm source that I can easily match to and deliver maximal power to the external transducers.

 The impedance of my transducer is Zt = 12.7 - j12.8 (Ohm). Thus, if I remove all other load, the load on the LM96550 would be much less than 2kOhm.

Regards,

Dmitrijs

Hi Dmitrijs,
I've talked to the LM96550 designer and here is what he said:
"The pulser can drive 50 Ohms. It’s just that the swing is reduce because of the output impedance is about 5 Ohms. So the peak-to-peak swing is about 10% less."

Hope this helps.

Regards,
Hooman

Hi Hooman,

Thank you so much for your response!

Output impedance of LM96550: it is definitely very good to know that LM96550 should be able to handle the low loads on the order of 50 Ohms. The datasheet for LM96550 does not mention the output impedance of 5 Ohms. It does mention however the value of "Output ON Resistance" to be 7...11 Ohms. Is this the output impedance the LM96550 designer is referring to?

Malfunction of LM96550. I am actually not sure anymore if we even will get a chance to test out the performance of LM96550 with our transducers. The chip burned out before we made any changes to the board and before connecting the external transducers. It was the first time we tried to power up the board and I think we followed the power up sequence in the right order. We went through the following steps:

1. We connected the FPGA to PC and the TX board. The SAMTEC connector was also plugged to the TX board, but nothing was connected to the SMA connectors.
2. The ground lead (the black banana cable) was connected to the ground of one of the power supplies
3. -50 V was powered up. We do not have an extra power supply for VSUB, hence J_SUB header was configured in 1-2 position to link -50 V and VSUB together. Thus, VSUB banana plug was left unconnected
4. As soon as -50 V was powered up, the LM96550 sparked up. On a figure below you can see a black mark left in the area where the chip was sparking.

I would be very grateful if you could help us to resolve this issue as quickly as possible. It took us a while to put the setup with all the power supplies together, and in the end, we did not get a chance to evaluate the performance of any of the chips even once.

Thank you!

Dmitrijs

Hi Dmitrijs,

Sorry to hear that you are having problems.

Output impedance of LM96550: The output impedance is shown in the datasheet (Output ON Resistance)  as 7 to 11 ohm.

Malfunction of LM96550: Did you make sure that you followed the instructions circled below?

If you switch to the LM96551, you would be able to use less number of power supplies.

BTW, your attachments / pictures did not come through!

Regards,

Hooman