ISOM8110: SMPS reference designs and compatible ICs

Hi Clemens,

The IC i did check was using a complex FB that involved a internal current source and some Pmos and current mirror .

So yea in retrospective it was something very special that didnt meet the optoemulator min current.

It is one of the LLC controllers, the one that ends in UCC256xxx  60 or 61.

I was checking the flyback/forward portfolio and yea never saw something like current sources and all were FB directly in some comparator/opamp.

Yeeeea probably most will be fine but there is that one IC in 100 that cant and according to Murphy's Law that IC will be my IC so i need to be carefull.

Best regards ,

Robert

Hi Robert,

Thanks for reaching out. 

Allow us some time to analyze your inputs and requirements and respond by May 29, Thursday.

Regards
Varun

Hi Kumar,

It is ok i am not in such a rush , as of now i am just analyzing available reference designs and SMPS controllers.

I am myself checking how the FB is treated inside since i will have to deal with a lot of options and that also helps sort them down.

And i think ``opto emulator friendly`` tag /feature in the columns filters would be a very nice thing to have since it would make sense for both your marketing and customers/ engineers that want to avoid optocouplers. 

Also though i am repeating myself : Also are there any SMPS ( isolating ) ICs what work with digital FB? something hysteretic probably?

Thanks  for the help,

Best regards,

Robert

Hi Robert,

Thanks for your detailed inputs. Appreciate it!

I'm still analyzing your inputs and will get back by Friday, May 30.

Robert Alfred Kecs said:

SMPS ( isolating ) ICs what work with digital FB? something hysteretic probably?

You mean instead of an Analog Modulated Loopback control using SOM8110 (Ic=If*CTR) you're looking for an ON/OFF modulation - Under Isolation Portfolio the Answer would be NO here. But if you're looking for a SMPS IC - then I need to to push this request to the Power Switches Team who can guide you the best here. You can also ask it in a separate e2e thread parallelly while I respond on other aspects of your query.

Regards
Varun

Hi Kumar,

Yes i was wandering if there were ICs that have ON/OFF FB and 2-3 pins but if there are none there is no problem.

As for something that i can control with ISOM8110 i am still looking UCC2897A LM5026 / 25 are strong candidates that I am considering as of now.

the LM50xx have FB that goes through a current mirror , there may be problems with IOSM minimal current but theoretically i think it could work ,not sure .

Barely work and work well are VERY different. and yea i am not rushing the IC choice.

Best regards,

Robert

Hi Robert,

We have analyzed your requirements and thanks for trusting on reliability of TI's opto emulators over optocouplers.

Currently, we are working with SMPS IC Team to get a list of devices which will be compatible with ISOM8110.

Also, we're parallelly evaluating the UCC2897A LM5026 / 25 and  UCC256xxx  60 or 61.based solution to work with ISOM8110.

Will keep you posted.

Regards
Varun

Hi Varun,

Any updates? ( I am also posting in order to keep the thread active, i didn't remember after what amount of time a thread is archived and locked but it was over 2 weeks, so writing now just in case it is 2weeks) 

Best regards,

Robert

Hi Robert,

Apologies for delay here. it's taking more time as we're working with cross functional team to get to a solution each controller at a time since we don't have a list ready at our end.

Currently, we're evaluating UCC256612-Q1 and possible solution to make it wok with ISOM8110.

Will Revert back by 18th June once we have inputs from SMPS team.

Also, there is no deadline for a thread to be active. However, we try to visit all the open threads on daily basis and update once we have any solid update.

Regards
Varun

Hi Varun,

Do you have any updates?
Also can i also ask if it is feasible for LM5045 or LM5046 to work with opto emulators? ?

Best regards,

Robert

Hi Robert,

Thanks for your patience. Based on the feedback received from SMPS Team  UCC256612, UCC2897A LM5026 / 25 should be compatible with the ISOM8110.

Team is working on releasing a reference design with UCC256612 and should be applicable to other mentioned controllers as well since they have an inbuilt REF Supply to bias the ISOM8110 properly. I will update on this as soon as I have a sharable information.

May I know what is your thought/criteria (maybe a possible biasing circuit that you might've though of) when you're proposing these ICs for compatibility check with ISOM8110?

Regards
Varun 

Hi Varun,

Well criteria was not so precise for the LM5046/5 i was looking at high power (100-200W) compact reference designs and i saw those 2 and all the other chips in full bridge topology  didnt have integrated gate drivers ( large space saver) so for me in a way i saw this as way better part number then the rest also the self start up to 100 Vdc was also a nice bonus that sort of reinforced my choice. Also 1MHz programable frequency gives me a lot of resign flexibility not to mention that i can get ZVS if all goes well. Yes they are not the newest chips but they are the only ones with this feature set. 

Regarding that UCC256612 i dont remember but since it was LLC it was also for something high power (100-200W). bot all i was considering  increasing the MTBF of the design.

LM5026 / 25 it was for wide range low ish power 10-30W or so applications , why did i ask about them: wide input range integrated gate driver and reasonable capsule for whet we get.  Also 1MHz programable frequency. Since it is active clamps i can see why i can go with higher frequencies and some high frequency transformers. IF it was normal flyback i would think way more if it was worth going at high frequencies, not that i dont want to but EMI is a great concern. Also for this low power application i am very space constraint so efficiency is vey important. That is why i was also considering forward topology. ( and some type of synchronous rectification on the secondary regardless of what i have on the primary. 

Why i was also targeting them was this:

-fairly simple FB input , tho i saw some post that said they dont work with a very similar chip

-i saw some reference design with them or another design that included a dc dc using them

- good price to performant solutions 

- somewhat documented, some are excellently documented while in others what happens in the reference design is sort of hard to figure but enough. 

This is about the best I can say since i askes like a month ago and worked on other things in the meantime , and not i am returning since we sort of need to make some resign decisions in the near future, since u know some stuff can wait since u want to done right the first time but after a time you sort of start needing to deal with it so that you have enough time to properly do that stuff. ( Yes it is multiple things in different projects that is why some of my IC choices may seem random).

Best regards

Robert

Also if you find any flaw in my logic please tell me , i might have large biases due to my design constrains and my experience in the domain.

Hi Varun,

That sort of clears it for me. I picked IC that i sort of expected to be of so i decided well nice to know. Regarding the UCC256612 i suppose the NPN pair Q4 in a current mirror configuration is there to help with the current draw by mirroring the opto current.

On a side note if i remember well for LM5046/5 there was a nice current to duty cycle relation: 1mA was 100% 0ma was 0 duty .

So in that case ill have on the secondary the *normal* FB circuit and on the primary ill have that current mirror and then ill have to play with R24 and R35 / R36 in order to make sure i have that required 0.7mA through the device and then over that the 1mA i need for the FB ( if i understand well) .

I cant thing well of how the current will work but i will use some SPICE and play with resistor values and see what i get.

Best regards,

Robert