Author Topic: Adjusting charge current by adjusting control voltage into LT1248 PFC controller  (Read 1167 times)

0 Members and 1 Guest are viewing this topic.

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Hi,

We have batteries which are 400v +/-5v ( no matter what the state of charge)
We wish to charge them with our “controlled open loop” method…depicted as attached. (LTspice sim and pdf schem)
Its non isolated but we’ll be ok.

We simply vary the voltage on the SS pin of the LT1248 Boost PFC controller  in order to change the charge current.
Do you see any problems with this application of a control voltage into the SS pin of the LT1248?
(LTspice sim works fine, but you never know  with  the real thing)

LT1248 datasheet
https://www.analog.com/media/en/technical-documentation/data-sheets/1248fd.pdf
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
I don't understand why you don't just use the Vsense pin for your control. The SS pin is used for under/overvoltage lockout, you lose that functionality, plus you run the high risk of shorting the output of your opamp via the transistor pull down and doing some damage.
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Thanks, we need to "inflict" a voltage at the VAOUT pin in order to control the charge current into the battery.
(we can add in our own under/over voltage lockout)
I cannot see how we risk shorting the output of the EA opamp.
There are 5 ways to "Inflict" a voltage onto the VAOUT pin...these are as above in #1, and as attached.
Which do you think is the best method?
« Last Edit: June 17, 2020, 09:50:34 am by treez »
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
Nicely drawn, your options. Number 5 seems the most obvious. Internal to the chip LT1248 and connected to the SS pin is an NPN transistor which shorts it to ground in the event of under/over voltage. That is where the opamp output could be shorted to ground. I would leave that pin alone as an input and use it only as designed, as an output. 8)
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Quote
Internal to the chip LT1248 and connected to the SS pin is an NPN transistor which shorts it to ground in the event of under/over voltage. That is where the opamp output could be shorted to ground. I would leave that pin alone as an input and use it only as designed, as an output
Thankyou very much moffy!, yes, very nicely spotted!!...my  circuit number 1 (of post #2) is thus kaput....
...However, circuit 2  uses the SS pin and solves it with a diode....circuits 3 & 4 can also use the SS pin and not be affected by that internal NPN transistor (due again to the diode use).

There are diodes (near the Error amplifier and SS pins) shown in the datasheet block diagram of LT1248 that dont actually seem to be "real" diodes. The datasheet doesnt really explain them...but they cant possibly be there in reality. I've never seen this sort of thing on other PFC controller datasheets.
« Last Edit: June 18, 2020, 09:24:49 am by treez »
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
They can change the internals of an IC without changing pin functions. Or there could be other stuff going on inside you don't know about. It is safest to use the SS pin as an output only. It is only a simplified schematic, don't make assumptions that might bite you in the long run.
 
The following users thanked this post: ocset

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
Please find attached an alternative way to control the VAout pin of the LT1248.
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Thanks Moffy, though i noticed your kind example connects -15V to the Vss of the internal EA opamp of the LT1248. Unfortunately this cant be done.
« Last Edit: June 19, 2020, 06:11:41 pm by treez »
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
It is just an illustration, the supplies can be whatever you like. The -15V is unnecessary with RRIO opamps but I just chose any opamp to illustrate that you can use a single external opamp and make Vout = Vin, by using the Vsense and Vout pins. Isn't that what you are trying to achieve?
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Thanks, yes, sorry i'll take another look.
I must admit, i am foxed as to what those diodes are in the LT1248 datasheet near the VAOUT and SS pins.........i cant see any use for a diode there, i  almost cant believe that they are actually real diodes. The datasheet says nothing of them.
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2220
  • Country: au
The diodes are probably there for protection purposes, stopping +ve high voltage inputs damaging the internals. Or they could just be a by product of the manufacturing process. As you say, who knows why? But at least we know they are there.
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Thanks but surely they cant really be there?...the one on the VSENSE pin would be stopping the feedback signal from getting to the error amplifier?
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2766
  • Country: ca
Hi,
The diodes are there and they are real. The reason that they are not documented  beyond that, is that they are matched.

Consider this model:



The Error Amplifier, EA, has three inputs. There is the reference input, the soft start input and the feedback node, Vsense.

The EA will regulate the lowest of the two voltage, the fixed reference (7.5V) and the voltage on the SS pin.
When the voltage on the SS pin is greater the 7.5V the reference in the controlling input.

D1 and D2 are needed to OR the fixed reference and the soft start inputs.
The voltage at the non-inverting input of the EA is lower of the two inputs plus one diode drop.

D4 add the diode drop to the Vsense input.

The diodes are biased by the input currents of the EA.

Modelling Result



General Comments

You would be much better using an IC that is designed to be an LED driver, like the LT3799, instead of adding work arounds to a PFC chip.

Given the other topics that you have started on the EEVBlog in the last few weeks, I am sure that you are not designing a battery charger. You would get much better help if you told the truth and asked for help for an LED driver.

Regards,
Jay_Diddy_B
 
The following users thanked this post: ocset

Offline ocsetTopic starter

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Thankyou very much Jay_Diddy_B.
Your post was great, and  i see now, amongst your other comments, the bit showing the current sources I1 and I3 really clears it up for me.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf