LT3045 Paralleling - ILIM circuit pros and cons

[[email protected]]

Quote

For some reason Alexy at ldovr does not believe it. Oh well... 
(c) @Superdad

 

1.       I am not a professional radio engineer

2.       English is not my native language
 

 I don’t agree with @Superdad that ILIM based “unique” way of paralleling LT3045 is better and this is the only reason why I don’t use it. I already explained why the statement about his “way to cut the output impedance nearly in half” is a bit tricky (because it is very low absolute values and a minor part of the full impedance measured at the point of load even with high quality cable/connectors used). And don’t forget - the maximum power transfer from a source to a load is usually achieved when the load impedance matches the source impedance.

Let’s focus on the second part of why I think it is harmful.
 

Let’s check block diagram below, and see how current monitor circuit can impact overall performance.


with a classic design (without ILIM compensation), reference voltage (SET PIN) is a function of the current sourced by VERY HIGH QUALITY CURRENT REFERENCE (100uA) followed by a single resistor Rset and  Cset capacitor in parallel with Rset (Green path).
Vout=Vref = Iref*Rset

And this layout can provide precise reference voltage for the regulator circuit. As you can see there is no gain from the SET pin to the output which means any noise that appears on the reference circuit will not be multiplied (like for the conventional linear regulators with ref voltage and resistor divider).

Another picture (Red trace)   –  if you add ILIM circuit, which is designed for output current limitation/monitoring. It is a bit tricky to use it for the drop compensation, but to simplify let’s assume that ILIM circuit is just adding compensation voltage to the reference point.  

Actually
Vref = Iref *(Rset +Rcdc) +  Iout/500*Rcdc

Points that I am aware of

1. Unlike from the HIGH QUALYTY Current 100uA reference source, ILIM circuit (not a constant but the function of output current) can inject unwanted noise into the reference point, any noise that appears on the reference circuit will hit at the output as well.
2. ILIM feedback (drop compensation voltage) will not appear at the SET PIN immediately, still we need time to charge/discharge Cset Capacitor, with capacity relatively high (few uF) and charging current very low (100uA), and discharge via few KOhm resistor.  All this will cause significant delay in action and in some cases such compensation can make things even worse.

I will appreciate any comment from skilled professionals, especially @Superdad and @JohnSwenson

Best regards
Alex

 

   

[dtossan]

Which is more important? Output impedance? Regulation speed? Frequency response? PSRR/noise? Would be difficult to understand in a mult-variable system, obviously.

Technical arguments are difficult to be conclusive because of the relative weights of the positive and negative factors are difficult to pin down.

Perhaps a comparison (in terms of SQ) of the 2 circuits would give good support as to which approach is better (or same) in a typical digital and/or analog application.

Such a test setup may not be too difficult to arrange.

Of course, SQ as perceived by listeners are subjective - but this is par for the course.

[The Computer Audiophile]

On 10/28/2019 at 8:12 AM, [email protected] said:

 

1.       I am not a professional radio engineer

2.       English is not my native language
 

 I don’t agree with @Superdad that ILIM based “unique” way of paralleling LT3045 is better and this is the only reason why I don’t use it. I already explained why the statement about his “way to cut the output impedance nearly in half” is a bit tricky (because it is very low absolute values and a minor part of the full impedance measured at the point of load even with high quality cable/connectors used). And don’t forget - the maximum power transfer from a source to a load is usually achieved when the load impedance matches the source impedance.

Let’s focus on the second part of why I think it is harmful.
 

Let’s check block diagram below, and see how current monitor circuit can impact overall performance.


with a classic design (without ILIM compensation), reference voltage (SET PIN) is a function of the current sourced by VERY HIGH QUALITY CURRENT REFERENCE (100uA) followed by a single resistor Rset and  Cset capacitor in parallel with Rset (Green path).
Vout=Vref = Iref*Rset

And this layout can provide precise reference voltage for the regulator circuit. As you can see there is no gain from the SET pin to the output which means any noise that appears on the reference circuit will not be multiplied (like for the conventional linear regulators with ref voltage and resistor divider).

Another picture (Red trace)   –  if you add ILIM circuit, which is designed for output current limitation/monitoring. It is a bit tricky to use it for the drop compensation, but to simplify let’s assume that ILIM circuit is just adding compensation voltage to the reference point.  

Actually
Vref = Iref *(Rset +Rcdc) +  Iout/500*Rcdc

Points that I am aware of

1. Unlike from the HIGH QUALYTY Current 100uA reference source, ILIM circuit (not a constant but the function of output current) can inject unwanted noise into the reference point, any noise that appears on the reference circuit will hit at the output as well.
2. ILIM feedback (drop compensation voltage) will not appear at the SET PIN immediately, still we need time to charge/discharge Cset Capacitor, with capacity relatively high (few uF) and charging current very low (100uA), and discharge via few KOhm resistor.  All this will cause significant delay in action and in some cases such compensation can make things even worse.

I will appreciate any comment from skilled professionals, especially @Superdad and @JohnSwenson

Best regards
Alex

 

   

Alex - Are you affiliated with any company in the industry?

[sandyk]

https://www.ebay.com.au/itm/LT3045-78xx-Ultralow-noise-0-8-Vrms-linear-regulator-0V-15V-0-5A-78xx-Layout/252948166542?hash=item3ae4e2b78e:g:jSMAAOSwol5YyCO1

[[email protected]]

3 hours ago, The Computer Audiophile said:

Alex - Are you affiliated with any company in the industry?

Yes, but i am not going to promote it here.

May I ask why did you ask such a question?

[The Computer Audiophile]

5 hours ago, [email protected] said:

Yes, but i am not going to promote it here.

May I ask why did you ask such a question?

I ask because members of the industry must identify themselves in their signature lines. This way everyone reading understands there are possible conflicts of interest. Also, members of the industry don’t have the same freedom to Talk about whatever they want here on the forum. 
 

 

 

[rickca]

1 hour ago, The Computer Audiophile said:

I ask because members of the industry must identify themselves in their signature lines.

See ldovr.com and dxpwr.com.  I see no problem with Alexey's discussion of a technical issue here.  He isn't promoting his products, but the AS rules require him to identify his business.  

[[email protected]]

4 hours ago, The Computer Audiophile said:

I ask because members of the industry must identify themselves in their signature lines. This way everyone reading understands there are possible conflicts of interest. Also, members of the industry don’t have the same freedom to Talk about whatever they want here on the forum. 
 

 

 

Thank you for the information,

I add company details in my profile as requested, but I think most of the CA members can identify me by email.
Is there any place I can read about the rules for industry member's limitations and forbidden subjects

[The Computer Audiophile]

1 minute ago, [email protected] said:

Thank you for the information,

I add company details in my profile as requested, but I think most of the CA members can identify me by email.
Is there any place I can read about the rules for industry member's limitations and forbidden subjects

Here's a link to the signature settings - https://audiophilestyle.com/settings/signature/

[The Computer Audiophile]

1 minute ago, [email protected] said:

Is there any place I can read about the rules for industry member's limitations and forbidden subjects

No, only because written rules, while they sound great, are always used by people to find loopholes and see how close they can come to the edge of what's allowed. Our rule is to not talk about your products unless specifically asked about them, don't talk about competitors or competitor's products, and if you aren't sure, just don't do it. Always ask if needed.

 

Plus, I guarantee you that 100% of the time when a member of the industry tries to break these rules with a sneaky sentence or two, it always backfires. No members of the community like to see one pulled over on them. I'm not suggesting you would ever do this, but I offer it as a nugget of information, based on years of experience. 

[[email protected]]

does my topic violates any of CA rules?

and it is reffered to @Superdad comments about 4xLT3045 chineese board here
(does is looks like "talk about competitors or competitor's products" or "sneaky sentence" in your opinion? )
and few following comments with my name as an opponent
comment one
comment two

 

[Jud]

Hi Alex -

 

Just as a forum participant and not as someone in charge of anything, I think the discussion as you have framed it in the original post works fine. I understood @The Computer Audiophile's comments more as laying out some ground rules than as saying you had violated any of them. As long as this stays on the level of a discussion about best circuit design practice rather than turning into an advertisement, I think we would all be happy to learn something.

 

One item you do need to add though, as @The Computer Audiophile mentioned, is an indication of your commercial affiliation in your *signature*, so we can all see it, rather than in your profile, where we have to go looking for it.

[sandyk]

 

Incidentally, in my previous reply where I gave the link to a product sold by Alexy , I also posted that particular link as it showed a very useful pin for pin replacement for a 78XX voltage regulator that I purchased from Alexy which does a great job in the Input PCB of my DIY DAC, and outperforms the LM3940T in this application.  These are niche products that are not sold by any of our forum advertisers either.