VCA2615: Amplifiers forum

Hi Viktor,

Apologies for delayed response. 

I am supporting the VCA device and I would really appreciated if you could share the background information or thread behind the same?

I am not able to follow the above conversation with Mr. Chen Kung. 

Can you share previous details you had with Kung?

Thanks & regards,
Abhishek

Hi Abrishek

Thank you for your reply.

Below this text you will find initial post and the answer from  Mr. Chen Kung.

Regards,

 Viktor.

=======================================================

Hi all

I use VCA2615 in ultrasonic application.
On the page 18 of Data Sheet
Table 5. shows codes for Feedback Resistor Settings.

I have come across that setting any code other than
'1111' (Open) causes a significant increase in noise.

For example, for code = '1111', input termination Rl=0, amplifier gain = 52dB
(about 16dB for LNA with a single ended input plus 36dB for VCA)
output noise voltage
Uout,rms=1.2mV

For code = '0000'
Uout,rms=3.59mV
For code = '0001'
Uout,rms=3.57mV
..................
For code = '1101'
Uout,rms=2.44mV
For code = '1110'
Uout,rms=2.14mV

Please answer two questions:
1) Why is this happening?
2) What is FEEDBACK RESISTOR for code '1111' (Open) ?

Thank you,

Best regards,
Viktor.

Hi,

for your questions:
1) Why is this happening?

Please refer to the VCA2615 datasheet Figure 61 page 19,
Let's assume you are using active feedback termination,
this shows LNP Gain as 20dB (vs. your setting is ~16dB)
also it shows 8 different kinds of LNP Noise levels,
you can see the curve (it shows Noise Figure vs. source Impedance = Feedback Resistor values)
which is very similar to your test Curve results as well (showing in mV).

2) What is FEEDBACK RESISTOR for code '1111' (Open) ?
Please also refer to the datasheet page 17,
it mentions in the open position, the input resistance of the LNP is about 100kohm.
Thank you!

Hi Viktor,

Thanks for sharing the details. I understood the problem.

So, when you set FB1-4 as '1111', essentially meaning open; the input resistance (Rin) will be 100 kohm as shown in the datasheet.

Also, note that the Rin is the input resistance and with active termination, Rin is a function of feedback resistor and LNP Gain. 

So, for say LNP gain of 18 dB:

Rin (0000) - 26.15 ohm

Rin(0001) - 28.77 ohm

Rin (1110) - 301 ohm and so on.

So, as you see in general, when source impedance is increasing (only talking about active termination case) the noise figure is decreasing from the plots, your data also suggests the same.

As it is a very old device, I am also not much familiar, hence, about the 8 different noise levels, let me go through the datasheet in more detail and align with the design team. I will provide you the correct and precise explanation.

In the meanwhile, i would suggest you to download the simulation files from the product page and play around with that, it may give you an answer you are looking for. In parallel, I will also do the same and share you my findings.

Kindly allow me some time.

Thanks & regards,

Abhishek

Hi Abhishek
Thank you for the quick reply.

1)
>So, for say LNP gain of 18 dB:
>Rin (0000) - 26.15 ohm
>Rin(0001) - 28.77 ohm
>Rin (1110) - 301 ohm and so on.

As I think it is not entirely accurate.

LNP gain = 22 dB (according Table 2. Gain Selection of LNP)
22db ~ A=12.59 - for differential input (and half as much - for single ended input)

So,
Rin (0000) = Rf/(1+(A/2))=130/(1+(12.59/2))=17.9 Ohm (equation 5 on page 18)
Rin (0001) = 19.6 Ohm
........................
Rin (1110) = 205.6 Ohm

>when source impedance is increasing (only talking about active termination case) the noise figure is decreasing from the plots, your data also suggests the same.
yes, of course.
But
for Rin=205.6 Ohm Uout,noise,rms=2.14mV
for Rin=100 KOhm Uout,noise,rms=1.2mV (that is much less !)
I did not find any mention of this noise behavior in the Data Sheet for this amplifier.

So any active impedance matching for
FB4,FB3,FB2,FB1=0000...1110
always give more noise than for
FB4,FB3,FB2,FB1=1111 (Rin=100 KOhm).

Why then do we need active resistance matching?
After all, the chapter "ACTIVE FEEDBACK TERMINATION" talks about active termination as a way to reduce the amount of noise.
But as I think it is impossible to get less noise than for
FB4,FB3,FB2,FB1=1111 (Rin=100 KOhm).

2)
where from 8 curves (Figure 61, Figure 62) of Noise Figure came from and what they correspond to ?

3)
Also I do not understand discrepancy between Figure21 and Figure61.
Figure21 shows NF=2dB (smallest value) for Rs=0.
But Figure61 shows MAXIMUM value of NF for small value of the Rs.
Could you please explain it?

Many thanks.

Best regards,
Viktor.