TX810: un-consistent samples gain

Hi,

I am looking into this and get back to you within 1 to 2 hrs.

Regards,

Shabbir

Hi,

Can you confirm which part of the waveform are you worried about in below? Did you probe the control signal B3/B2/B1 and made sure it is toggling at the same time across burst?

Regards,

Shabbir

you can see that B3/B2/B1 are shorted ,yes, I'm worried about the pulse wave,see the amplitude zoomed in : 

Hi, 

Yes all control pins B3/B2/B1 shorted. What i meant was this shorted signal, you might be toggling from 0 to 1. Since at the start of waveform, I see glitches coming at different instant so I though signal is toggling from 0 to 1 at different time leading to such variation in amplitude later. So can you probe the control signal too?

Regards,

Shabbir

hi,

I have measured the Bs pin toggling time-the variation of the Bs pins is only 12nS,not 4.5uSec as seen in my measurement:

Hi,

- What does blue waveform means? 

- Also can you probe the +/-5V supplies and check if it is stable?

- You have shown the amplitude variation. Across burst do you see amplitude reducing or increasing or random variation? We can check if it matches with supply variation.

Regards,

Shabbir

- What does blue waveform means? -it is one of the samples 

- Also can you probe the +/-5V supplies and check if it is stable?- i see 7mV drop while the T/R is close

you can see while the T/R is always close the amplitude variation is very low over 50 different bursts :

Hi,

Thanks for clarification. I understand that TR switch always ON, there is no issue. 

About power supply,  you mentioned in is 7mV when TR switch is always ON.

For TR switch ON/OFF case also can you probe the +/-5V supply once? Or there also is it within 7mV?

Also just for my understanding, why are you concern about variation during transmit? 

Thanks!

Regards,

Shabbir

I meant while I'm switching on the switch see a 7mV drop,

I'm concern about variation during transmit because the system pulse repetition must be highly accurate and I'm using the AFE to verify it

Hi,

Is it possible to probe the AFE input using oscilloscope and confirm that variation exist at the input too? In continuous turn ON mode, AFE wont see the highlighted part. So I was thinking that there is possibility that issue is disappearing in continuous turn ON mode because AFE is not seeing the glitches highlighted below which may be affecting the amplitude later. Is it possible to confirm about it?

Regards,

Shabbir

the AFE input on C431 TR switch side:

Hi Yonatan,

This is just single waveform. We need to overlap mulitple burst and see. You can keep scope in infinite precision check.

Just for debug, can you replace C28 with 100 ohms resistance and see if it helps in resolving this issue for channel 1? 

Regards,

Shabbir

see scope picture of the wave:

I did changed C28 to 100ohm and measured better result-yet not good as I measured with the switch always close ,

you can see the error over time in percentages from the first pulse to the other  :

with the TR switched before the pulse:

with the TR always close :

Hi,

When you changed the capacitor to 100 ohms resistor, can you confirm what did happen to the below area? Also how much improvement did you get with and without the 100 ohms resistor?

Regards,

Shabbir

error with the resistor:

error with the capacitor: 

wave with the resistor :

Hi,

Thanks. So after adding resistor, there is significant improvement. Variation reduced from 16% to 0.7%. By looking at plot, I think all these measurements are being done at the AFE output data. 

1. Can you confirm after how much time are you repeating the transmit? 

2. Can you now try placing 100 ohms on rest of the channels and see if 0.7% improves to further down? I am thinking that this residual might be coming from adjacent channel as cross talk.

Thanks!

Regards,

Shabbir

looks much better ,yet I see exponential shape to the error ,can you think of a cause to it?

i'm repeating the transmit 100 times with 300ms delay 

Can you confirm what change did you do to get further improvement? Is it after adding resistor to all the 16 channels?

For above issue, I was looking for a large time constant circuit. At the input, initially I didnt find any DC path to ground so suspected low DC discharge. Hence recommended to all 100 ohms resistor. It seems that helped. To improve it further, we can try reducing the INP resistor to 50 ohms but it will have impact on signal chain dynamic range. 

Regards,

Shabbir

Yes,replacing the resistors improved the measurements ,I will try with 50 Ohms,

Thanks

Hi Shabbir,

Yonatan and i are working on the same project, why there is a impact on signal chain dynamic range ?

I did not see in the datasheet resistor on the path to ground. 

what will be the trade off ?