CMC UCC3808A2 Current sense gain implications

Hello everyone,

I would like to know the implications of the Current Sense Gain value (2.2 V/V) of the UCC3808A2 (https://www.ti.com/lit/ds/symlink/ucc3808a-2.pdf page 6) in the control loop. Particullarty, let's consider the following circuit:



After calculating the current primary peak current (which results to be around 50 A according to this guide https://www.tij.co.jp/jp/lit/an/slua584a/slua584a.pdf), with a 15 mOhm current sense resistor, it would cause 0.75 V at it. That value would be in the the current threshold zone which is to this chip between 0.7 V and 0.8 V.

I want to design the controller but I have the following questions (they are enumerated so that you answer easily):

1) What does the 2.2 V/V Current sense gain value mean/implies?

2) Would I have at the ifb pin the sensed value multiplied by 2.2?

3) if 2) is affirmative, should I change the value of the current sense resistor (15 mOhm)?

As always, thanks for reading, and thank you for your help.

Best regards,

David.

Hi,

1) For every volt the gain is 2.2, so a 1V signal becomes 2.2V.

2) Yes.

3) (50A * 0.15R) * 2.2 = 1.65V

Work the other way round to save time on "iterative calculations": (0.75V/2.2)/50A = 0.00681R.

I'd guess the closest single value current sensing resistor you'll get is 5mOhms. You could add 2*1mOhm to a 5mOhm but I don't recommend that much, maybe my fussiness is unfounded. Maybe paralleling a few e.g. 3*20mOhm or something but I'd use one single resistor, unless your making it/having it made.

An expert here with more experience on current sensing resistors may be able to suggest how to get 6.8mOhms in a useful way.

Does the input have to be between 0.7 - 0.8V or that is the maximum? If the latter then a 5mOhm should be "ideal" = 0.55V at 50A.

Thank you for your reply d123.

According with the cited Resistor sense (Rs) design guide by Texas Instruments (https://www.tij.co.jp/jp/lit/an/slua584a/slua584a.pdf) which uses the UCC28083 (https://www.ti.com/lit/ds/symlink/ucc38085.pdf see page 4) and it has the same Current Sense Gain range (CSG), that CSG was never used in the desing of that Rs. Would that mean that they are wrong? It might be a silly question, but i am surprised because it is supposed to be an Application Report...

Hi,

To be honest, I had a quick look (my day started at 7 a.m. and finished at 22:40 between one thing and another) earlier and now and don't understand the reason for that G 2.2 missing from app note calculations nor have I the time to study the whole thing, I'm afraid.

I'd contact TI to enquire about how to understand the app note versus the datasheet. The TI forum may yield a reasonably speedy reply to help, especially with regard to a TI part.

Hello d123,
Thank you for your help. I did what you suggested and apparently I was mistaken. I share here TI's answer to the question:

There is no gain block associated with the CS signal, the gain described in the datasheet is the effective gain between the COMP signal and the CS signal.

In the controller there is a 0.3V dc offset to give the CS signal a DC bias. This allows the comp signal to command zero duty cycle without have to go all the way to 0V, which is difficult to achieve due to on-state drops and Vbe's in semiconductor devices.

The COMP pin voltage is also then divided down by an resistor network of 80k/60k which reduces the COMP pin voltage to a similar range and the current sense signal or viewed from the current sense input it scales up the CS input voltage with respect to the COMP pin voltage.

Best regards,

David.