LM2675 replace with LM76003 ?

Hello
I have an older design in front of me and I'm thinking about modernization and shrinking.
The original design looks like this.
DC/DC Stepdown 3A LM2576HV-Adj make output DC voltage for
RF Power Amplifier with RF power MOSFET amplifier which creates RF power sinus over 10MHz.
Feedback for LM2576 is from output RF Amplifier , which creates from RF DC 1,23V for LM2576HV Adj.
It all works surprisingly well, and it's actually simple.
52kHz converter but requires large electrolytic capacitors, big coil and I need somewhat higher power.

My idea.
use Synchronous Step-Down DC/DC LM76003
this work on 500kHz , suffices with SMD ceramic capacitor, small coil and is in 6x4mm WQFN package,FB is 1.006V.

On the first look, it looks like it's easy to modify FB and it should work,
On second look, 10 times higher frequency, another structure of internal loopback etc.
I can not judge it and so I want to ask more experienced will it work?
with any problems I can count on?
Eventually with any problems I can count on?

Second question.
My device is RF and it is closed and is not cooled by the flowing air
LM76003 have high efficiency, but high switching speed and some conductivity losses I thing need take heat away.
I thing WQFN is on TOP PCB, and button PCB is flexibly pressed against the metal wrapper through which it cool. it is used for such a gray flexible heat-conducting rubber, I do not know how it is named and who sells it?

How about the 150kHz version of the LM2576 ... ?

LM2596 and LM3596 ....

If I were you, a concern would be (since you are feeding an RF amplifier), could there be issues with the higher switching frequency?

You know, intermodulation or beating?

Hi,

@schmitt-trigger
Did you ever encounter such a problem?

I don't know if this can happen.
My thoughts: "switching" means that there are overtones. This overtones need to be suppressed by the output filter.
In simplest case this is just a capacitor.
If there is a switching controller with a higher switching frequency. The energy per pulse is smaller.
My idea is that there isn't much difference in high frequency ripple...assuming that the PCB layout and the capacitor selection is correct.

Usually switching controllers with higher switching frequency need smaller output capacitors.

Klaus

Klaus
Yes it did happen to me, on an analog (NTSC) cable receiver.

An inter modulation product fell very near the chroma subcarrier, generating a very annoying “chroma beat”.

Granted, my particular problem may not apply to the OP’s circuit. But I only wanted to alert him that intermodulation products can interfere higher frequencies than a simple harmonic.

Hi,

Ok.
To avoid nasty async interference one may use a synchronizable SMPS controller.

Klaus

Yes stability of feedback lookp is one of the questions.

Now is 52kHz
LM2576HV + 470uH + 1000uF Low ESR as standard output stage of DC/DC Switch converter
to suppress interference from RF amplifier is added
100nF ceramic C + 330nF PET WIMA -> Wide band Ferrite choke FR3 ->
2x 100nF ceramic + 2x 330nF PET WIMA + 220uF Low ESR elit -> RF toroid in Drain of RF MOSFET.

To have an idea of the size of the solution with the LM76003 I drew it

New 500kHz
LM76003 + 22-47uH + 6x 10uF X7R 100V 1210 + 2x 100nF X7R standard output stage of DC/DC Switch converter
to suppress interference from RF amplifier will be added
100nF ceramic C + 330nF PET WIMA -> Wide band Ferrite choke FR3 ->
2x 100nF ceramic + 2x 330nF PET WIMA + 2x (may be 4x) 10uF X7R 100V 1210 -> RF toroid in Drain of RF MOSFET.

The problem is, what exactly are paramters ceramic capacitor size 1210 10uF 100V from X7R?
I'm talking about ESR , parasitic induction , properties on 500khz, and 10MHz

Very straightforward.
National (Now Texas Instrument) when you use their online Webbench design tool, will produce a bill of materials complete with vendor names and part numbers.

Go to the vendor's website, for instance Murata, and study the particular performance characteristics on the component's datasheet.

My personal experience is that all of the major capacitor vendors (Murata, TDK, Vishay, Kemet, etc) have fully characterized components.

I thing it is no very straightforward.
The TI database does not contain any SMD 10uF 100V,
but above all the TI database contains only one value, for example ESR 2mohm, which is a non-frequency indication :idea:.
A better source of information is THIS
it also does not contain any 10uF 1210 100V X7R, but 63V is already found and has all values frequency dependent

For example GCM32ER71E106KA57 10uF 1210 63V X7R
ESR on 500kHz 2mOhm 10MHz 15mOhm
L on 500kHz any pH 10MHz 400pH

Caution of ceramic C is strongly dependent on the DC component of the voltage
For example if on C will be 40V DC will be C only about 55% nominal value .

Thank you all for the information.
I'll try and see.
Still a question about the use of ceramic condensers in DC / DC converters
Electrolittics capacitor have some (relatively small lifetime) for exmpole 5000h on full nominal load.
As is the case with ceramic condensers (X7R)?