Author Topic: Linear PSU and low-ESR vs normal capacitors.  (Read 2195 times)

0 Members and 2 Guests are viewing this topic.

Offline xKertxTopic starter

  • Contributor
  • Posts: 11
  • Country: fi
Linear PSU and low-ESR vs normal capacitors.
« on: February 28, 2024, 04:46:26 pm »
Hello,
I have a Korad KKG305D bench power supply and while i was cleaning the filthy PCB:s and swapping out the noisy fan, i thought that i could also swap out those cheap chinese capacitors to a good known brand capacitors.
I know it still will be just a ~100€ power supply, but electrolytic capacitors are relatively cheap, so why not. And maybe it will even improve it's reliability and lifetime a little bit.

When i was searching for new capacitors, i read from several sites that unlike SMPS:s, the linear regukators don't like low-ESR capacitors.
So, my question is, should i buy a low-esr caps or "normal" ones? Or, does it make any difference in this case?
For example the output capacitor, or the other capacitors.

Thanks!
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #1 on: February 28, 2024, 05:26:01 pm »
Where?

Context matters.

Where did they say linear supplies don't like them?  Who was speaking?  Are they an authority on the topic, or just repeating n-th hand information?

Where in the circuit did they say linear supplies don't like them?  Input? Output? Compensation? Bypass?  How low is "low ESR"?  Does this include polymer electrolytics or the regular kind?  Ceramic?

What kinds of linear supplies did they discuss?  There are many kinds, many ways to compensate them, whether internally or discrete.  LDOs supporting ceramic caps are a very normal thing these days, but you do have to check that you're actually buying the latest and greatest, not some cheap shovelware from the '90s back when electrolytic and tantalum were all we had.  (Even to this day, '1117s are conspicuously present.)

The reason I ask, is any unqualified, broad, blanket statement, on a subject with even a little nuance, is almost certainly wrong, and at the very least must be verified against other more authoritative sources (and so on in turn), or checked by original research (these are verifiable claims, we can measure them).  Until then, such statements are at best hearsay, let alone fact; be careful repeating, let alone internalizing, them.

Notice I've said almost nothing about the topic in question, just brought up variations of parameters, possibilities.  This is a most valuable life lesson, it generalizes to anything and everything!  particularly valuable as you have the opportunity to learn it without having even modified your power supply (implied: potentially breaking it from acting on bad advice).

As for how to know, how to do the technical evaluation; you need the schematic first, and then datasheets of candidate parts.  Also since your concern is reliability/lifetime, likely many other glaring issues can be noted -- searching for schematics or improvements on the unit in question (also check similar part numbers under clone brands; nonsense/random branding proliferates along cheap products such as this (not necessarily this product in particular, but products of this sort, at low prices) may find teardowns and reviews that are most useful.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: soldar, dobsonr741

Offline xKertxTopic starter

  • Contributor
  • Posts: 11
  • Country: fi
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #2 on: February 28, 2024, 06:47:43 pm »
Well there's not much information about this power supply on internet.
Can't find any schematics or anything else.
I guess it's not so popular model. And that's why i asked.
I guess i was hoping if there will be some general rule of thumb whitch kind of capacitors are used in where.
I still think it's better to replace these cheap capacitors, before they go bad and break something else.
« Last Edit: February 28, 2024, 07:52:56 pm by xKertx »
 

Offline radiolistener

  • Super Contributor
  • ***
  • Posts: 3374
  • Country: ua
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #3 on: February 28, 2024, 07:16:21 pm »
Be careful, using ultra low ESR capacitors (such as solid state capacitors) in circuit which is designed for high ESR electrolyte capacitor may lead to circuit damage or unwanted oscillations. This is because circuit may expect high ESR capacitor to dump oscillations or don't exceed some level of current.

In other words, the circuit may expect some ESR of capacitor as a resistor which is a part of circuit. And if you replace it with ultra low ESR it removes that resistor and circuit going to unexpected mode...
« Last Edit: February 28, 2024, 07:19:09 pm by radiolistener »
 

Offline Vovk_Z

  • Super Contributor
  • ***
  • Posts: 1419
  • Country: ua
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #4 on: February 28, 2024, 07:41:40 pm »
It is safe and reliabe to use a general use type 105 C capacitors of a reputable brand.
 

Offline armandine2

  • Frequent Contributor
  • **
  • Posts: 602
  • Country: gb
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #5 on: February 28, 2024, 08:56:24 pm »
Hello,
I have a Korad KKG305D bench power supply and while i was cleaning the filthy PCB:s and swapping out the noisy fan, i thought that i could also swap out those cheap chinese capacitors to a good known brand capacitors.
I know it still will be just a ~100€ power supply, but electrolytic capacitors are relatively cheap, so why not. And maybe it will even improve it's reliability and lifetime a little bit.


So, my question is, ...

that question aside ... upgrading an inexpensive Chinese built, if not designed, piece of test equipment for something better raises the question to exactly where the value of the product is?  ::)
« Last Edit: February 28, 2024, 09:03:39 pm by armandine2 »
Funny, the things you have the hardest time parting with are the things you need the least - Bob Dylan
 

Offline BrokenYugo

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: us
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #6 on: February 29, 2024, 12:31:00 am »
The whole ESR sensitive regulator thing is very regulator specific, not all linear regulator ICs need it, and something like a lab bench power supply is a whole different animal.

High grade long life parts from the good brands are not cheap.

If it ain't broke don't fix it.
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #7 on: February 29, 2024, 01:14:28 am »
that question aside ... upgrading an inexpensive Chinese built, if not designed, piece of test equipment for something better raises the question to exactly where the value of the product is?  ::)

I mean -- not saying it's good value, but if the value to one is more about developing skills and gaining knowledge, poking around such a unit is an affordable way to do so, while having something that is at least nominally usable at the outset.

Or, think of it as building your own, with fewer steps: the circuit's already designed, you have all the parts necessary right there, it might not be well conceived or put together, but it's most of what you need, there's no enclosure crafting, the displays and buttons are definitely usable, the power stage and controls might even be usable with tweaks; of course one needs to determine how much work is required to get it to a more confident and reliable state.

Or, if you're in a part of the world where good stuff just doesn't exist, it's not imported, there aren't normal sales channels, ordering direct from China isn't simply the convenient drop-ship we have here in the US, or the west in general; it can be a huge PITA to get something even workable, and then to improve it to a reliable state, while spending much less -- and much cheaper -- labor to get there, can be a win.  Particularly for the beginner or hobbyist where labor may be marked down or written off entirely, and, again, learning opportunities.

Whereas if you go the "normal" route of designing (or "shopping" schematics online... blarf!) your own lab PSU, and it's a whole bunch of work just to get anything together at all (transformer, enclosure, controller, power stage, heatsinks, terminals, panel meters..!), and in the mean time you can't work on other more interesting projects without a lab supply handy -- if one has more interest in the latter -- and you can always go back and fix up this unit -- well, that might serve better.

So yeah, maybe not great reasons, but I don't doubt out-of-hand that there is legitimate value to be had in such a unit.  It's not something I'll be buying any time soon (even if I didn't have supplies already), but also not something I'll begrudge someone from considering.  Especially on a meager budget.

Anyway, there's a very real market like this, albeit moreso on the mechanical side of things -- you see trashy iron castings with crap millwork all the time in, like, hobby machine shop level tools -- drill press, vice, X-Y tables, entry-level lathe/mill/combo, bandsaw, etc.  The nature of the market is, if the manufacturer spent any more time on any step of production at all, they would price themselves out of the market -- it's a race to the bottom, remember -- and the low price range always and forever will and must be -- so, that they can offer, frankly anything even resembling the product in question, at all, is remarkable, and given that cost pressure, they can do some pretty good quality despite it.  No, not "quality" in the usual sense of something that's functional, consistent and reliable out of the box -- but something one can get good use out of, give or take reasonable effort put into it, and despite its many shortcomings.  The best explanation of this I've seen is here:



Now, it's not quite the same for electronics, because -- well, give or take your welding capability, and how suitable your substrate materials are and etc. (good luck welding on those cracker-barrel gray iron castings..!), but, with mechanics, you can potentially do quite a lot in terms of repair, re-tread, grind, fit and finish, paint and polish, etc., like the above; the same need not be true of electronics, where MCUs might be locked, or in any case source-unavailable, let alone whatever dev tools would be needed for the no-name chips used, and major geometric errors like creepage and clearance fundamentally affect the safety and reliability of the product in a way that isn't feasible to fix, short of remaking the entire PCB.  So, there is a significantly smaller chance that equipment like this can be improved, say to good western engineering standards, given reasonable effort.  That's unfortunately the nature of something as high-level as electronics; even if the house is good, you can hardly fix the foundation without rebuilding the house atop it.  But still, much of the stuff needed is present, and, if one wanted to make repro boards maybe, with open firmware, and all the right precautions, and, filtering and protection and all that, you could; and doing so would benefit every other user -- if not immediately or directly, then by bits and pieces here and there, overall.

Long story short, if one is interested in (or not discouraged by) doing some fix-ups to their tools, there exist...whole markets for that, basically; I wouldn't begrudge anyone for opting to do so, whether by preference, financial limitations, or other extenuating causes like geopolitics entirely.  Just beware the many ways such a product can fail, and how to familiarize oneself with their working, so as to be able to fix it.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: soldar, dobsonr741, 5U4GB

Offline BradC

  • Super Contributor
  • ***
  • Posts: 2106
  • Country: au
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #8 on: February 29, 2024, 03:42:01 am »
it's a race to the bottom, remember -- and the low price range always and forever will and must be -- so, that they can offer, frankly anything even resembling the product in question, at all, is remarkable, and given that cost pressure, they can do some pretty good quality despite it.  No, not "quality" in the usual sense of something that's functional, consistent and reliable out of the box -- but something one can get good use out of, give or take reasonable effort put into it, and despite its many shortcomings.

I own a pile of "built to a price" Chinese tools, and my last one was a mini-bandsaw I got from Vevor. Sure, it required about half an hours "fettling" to get it precise, but even taking into account my full professional hourly rate I'm about 2 bandsaws ahead on price for a precision I'm happy with.

Like anything (Fast, Good, Cheap; pick two), with tools it's "Now, Cheap, Good" (excepting it's incredibly rare to get Cheap and Good regardless of time). I often pick cheap when I know it's "good enough". I now have several other Vevor purchases on my "stuff to get" list, because I know I can quickly make it better than it has any right to be. Without those bottom of the barrel tools, I'd never be able to justify a lot of purchases.

Heck, my power supplies (linear and SMPS) are the same. Cheap and Now, and a bit of fettling required to get them working to an acceptable level of stability.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7860
  • Country: us
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #9 on: February 29, 2024, 04:15:51 am »
(excepting it's incredibly rare to get Cheap and Good regardless of time)

It used to be easier, but patiently stalking eBay for an older high quality unit is one way to get cheap and good.  And I'm happy to do some "fettliing" with an older high-end product that just needs bit of TLC, not so much with things that are poorly made in the first place.  I suppose I've made exceptions myself, but I can't say I've had many successful long-lasting relationships with bargain products--and I've tried more than a few.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline xKertxTopic starter

  • Contributor
  • Posts: 11
  • Country: fi
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #10 on: February 29, 2024, 04:35:56 am »
Well, i already have this PSU, so i am not going to scrap it and buy a new one (or used) just because it's cheaply made.
It is working and it's good enough fo my needs.
I just don't trust those chinese noname (or fake Rubycon) capacitors enough in the long term, so i would like eliminate this one possible fault point and swap these out before they fail and break something else also.
 
The following users thanked this post: MathWizard

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7860
  • Country: us
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #11 on: February 29, 2024, 04:42:55 am »
Well, i already have this PSU, so i am not going to scrap it and buy a new one (or used) just because it's cheaply made.
It is working and it's good enough fo my needs.
I just don't trust those chinese noname (or fake Rubycon) capacitors enough in the long term, so i would like eliminate this one possible fault point and swap these out before they fail and break something else also.

In a linear PSU the capacitors typically aren't overstressed like they are in a marginal SMPS.  They might eventually die of old age and leakage, but they aren't a particularly weak link and something else may very well break anyway.  As far as Low-ESR, there's typically no need for or benefit from special Low-ESR capacitors except where they are really needed.  If you do want to do some cap swapping, just use any reasonable brand like CDE, Kemet, etc and regular general-purpose versions are typically fine.  Use 105C rated ones if you like, but unless your linear PSU gets pretty hot that isn't even all that important.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline BradC

  • Super Contributor
  • ***
  • Posts: 2106
  • Country: au
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #12 on: February 29, 2024, 06:13:26 am »
It used to be easier, but patiently stalking eBay for an older high quality unit is one way to get cheap and good.

More of a US thing. Affordable high quality disposals are few and far between South of the equator.

Well, i already have this PSU, so i am not going to scrap it and buy a new one (or used) just because it's cheaply made.

Nor should you. My experience with SMPS is it's often the capacitors. With linear supplies (and particularly cheap linear supplies) it has almost always been a semiconductor failure (and one melted transformer).

Pick a nice "name brand" cap and change them out if it'll make you feel better. I took one 30V10A supply and "tweaked it up". Replacement caps, trimpots, front panel potentiometers, voltage reference and associated resistors (0.1% with a +/- 15ppm thermal). I spent more than I paid for it to tickle it up and turn it into my "lifetime PSU". It made no noticeable difference, and a 100% load for 5 hours shorted a secondary on the transformer. I binned it, bought a new and better one and haven't touched it.
 

Offline Seb71

  • Newbie
  • Posts: 4
  • Country: ro
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #13 on: February 29, 2024, 09:04:04 am »
Affordable high quality disposals are few and far between South of the equator.
Those all go to a certain dumpster room in Australia.
 

Online Haenk

  • Super Contributor
  • ***
  • Posts: 1090
  • Country: de
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #14 on: February 29, 2024, 10:31:27 am »
I'd suggest not to change the caps. Even cheap chinese ones are "good enough" for the low stressing and temperature of a linear supply. Eventually they will run dry, but that will take years or decades to happen.
 

Offline radiolistener

  • Super Contributor
  • ***
  • Posts: 3374
  • Country: ua
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #15 on: February 29, 2024, 12:43:21 pm »
Nor should you. My experience with SMPS is it's often the capacitors. With linear supplies (and particularly cheap linear supplies) it has almost always been a semiconductor failure (and one melted transformer).

I have old linear PSU from Aiwa PC speakers. It worked very long time, after 30 years I found that this PSU has dead capacitors, but semiconductors still alive...  :)
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16620
  • Country: us
  • DavidH
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #16 on: February 29, 2024, 02:51:24 pm »
When i was searching for new capacitors, i read from several sites that unlike SMPS:s, the linear regukators don't like low-ESR capacitors.
So, my question is, should i buy a low-esr caps or "normal" ones? Or, does it make any difference in this case?
For example the output capacitor, or the other capacitors.

"Low ESR capacitors" in connection to electrolytic capacitors refers to two different types of parts.

Common aluminum electrolytic and tantalum capacitors have low ESR variations which are not that much lower, and this applies at higher frequencies for switching power supplies.  Using these parts will not be a problem.  They may also be called "low impedance".

There are also polymer aluminum electrolytic (sometimes called OSCON capacitors, maybe because Sanyo was the first to make them available?), and polymer tantalum capacitors which are very different from the above.  They have very low ESR and may cause problems in circuits not designed for them.
 

Online MrAl

  • Super Contributor
  • ***
  • Posts: 1440
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #17 on: February 29, 2024, 06:06:07 pm »
Hello,
I have a Korad KKG305D bench power supply and while i was cleaning the filthy PCB:s and swapping out the noisy fan, i thought that i could also swap out those cheap chinese capacitors to a good known brand capacitors.
I know it still will be just a ~100€ power supply, but electrolytic capacitors are relatively cheap, so why not. And maybe it will even improve it's reliability and lifetime a little bit.

When i was searching for new capacitors, i read from several sites that unlike SMPS:s, the linear regukators don't like low-ESR capacitors.
So, my question is, should i buy a low-esr caps or "normal" ones? Or, does it make any difference in this case?
For example the output capacitor, or the other capacitors.

Thanks!

Hi,

The answer is simple but requires some research because the choice of electrolytic cap depends on the type of regulator being used in the circuit.  For some regulators a low ESR cap is great, but for others the ESR should be within a certain range or else it can oscillate.
You'd have to look inside the power supply and see what chips they are using and what the caps are for.  You can then look up the type of regulator(s) being used and look up the data sheet and see what it says about the capacitors used for filtering and stability.
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21688
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #18 on: February 29, 2024, 07:11:19 pm »
When i was searching for new capacitors, i read from several sites that unlike SMPS:s, the linear regukators don't like low-ESR capacitors.
So, my question is, should i buy a low-esr caps or "normal" ones? Or, does it make any difference in this case?
For example the output capacitor, or the other capacitors.

"Low ESR capacitors" in connection to electrolytic capacitors refers to two different types of parts.

Common aluminum electrolytic and tantalum capacitors have low ESR variations which are not that much lower, and this applies at higher frequencies for switching power supplies.  Using these parts will not be a problem.  They may also be called "low impedance".

There are also polymer aluminum electrolytic (sometimes called OSCON capacitors, maybe because Sanyo was the first to make them available?), and polymer tantalum capacitors which are very different from the above.  They have very low ESR and may cause problems in circuits not designed for them.

To be clear, I've understood the topic as low-ESR electrolytics; good point to be cautious of other types.

Incidentally, polymer and tantalum are available in overlapping ranges of ESR, so you can still find good damping value in these types; it's just that average tants skew higher in ESR, while average polymers skew lower.  Both skew lower than electrolytics.  The lowest-ESR tail is dominated by polymer, while, I'm not sure offhand actually whether dry tant or electrolytic beats out the other; I have a feeling they are comparable, but I'd have to check to be sure.  The main thing is you wouldn't be using dry tants in large values (expensive, risky with respect to surge), so the smaller chip parts tend to have higher ESR, but so do comparable electrolytics, but polymers are fine for surge use.

As for electrolytic ESR, I wouldn't say quite as generally; ESRs are impressively low these days, at least in certain values/ranges where it's needed.  Which is most of them nowadays, heh.

Low ESR isn't really new, it was just specialized before.  Consider Sprague 672D/673D:
https://www.vishay.com/docs/42054/672d.pdf
https://www.westfloridacomponents.com/mm5/graphics/M04/673D107F250.pdf (did they inherit just this exact product line, or 2nd source because MIL use..?!)
(The original actual-Sprague brand catalog, consider http://www.bitsavers.org/components/sprague/Sprague_Electronic_Components_1980.pdf for example, claims "low ESR", but doesn't actually say, and the engineering bulletins are long gone.  Somewhere between then and now, the ESR was published in the datasheet/catalog, but who knows if the formulation or design (other than the outer cans maybe) has changed.)

These values (assuming they're historically representative) are regularly met today, the difference is everyone's doing them, because everything has SMPS; and they're doing it in a tiny fraction of the volume, lol.  These were special, and mandatory back in the days of minicomputers; they're featured prominently in vintage DEC equipment for example.

ESRs definitely get low enough that a control loop can become unstable, if it was designed marginally, or dependent on high ESR, whether that was the style at the time or because they were designed around modern "general purpose" types (high ESRs are still available, I think more often in axials, go figure?).  Unlikely, maybe, to be a problem, replacing anything with anything, but the most confident advice that can be given is "try it, and verify response" (load step response for example, shouldn't have overshoot or ringing) (mind, it most likely will have overshoot as it transitions from CV to CC mode, so ensure settings to avoid that during the step; and test both modes, CC and CV will have different responses).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Wallace Gasiewicz

  • Super Contributor
  • ***
  • Posts: 1184
  • Country: us
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #19 on: February 29, 2024, 11:20:17 pm »
ESR is a big deal now that the Chinese have made ESR meters good and inexpensive. 
NOT that important in a Linear PSU that has to even out the supply at  120 Hz.   
Most of the big caps in a linear filter the output of the rectifiers,

The charging and discharging rate of a capacitor is governed by resistance, You can discharge one slowly with a resistor, Try it.At the low freq that the cap is used, a little ESR is probably good. 

Measuring ESR all the time is a current fad. It is only important is limited applications. 
Leaking caps in your application would be far more important. 

Also, lots of crap out there: 

https://www.eevblog.com/forum/chat/cheap-ebay-caps-went-bad-(on-the-shelf!)/
« Last Edit: February 29, 2024, 11:29:36 pm by Wallace Gasiewicz »
 

Offline xKertxTopic starter

  • Contributor
  • Posts: 11
  • Country: fi
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #20 on: March 01, 2024, 04:06:53 am »
 
Leaking caps in your application would be far more important. 

Also, lots of crap out there: 

https://www.eevblog.com/forum/chat/cheap-ebay-caps-went-bad-(on-the-shelf!)/

Yes, that is why i'm swapping them. Just to be safe. :)
« Last Edit: March 01, 2024, 04:10:47 am by xKertx »
 

Online MrAl

  • Super Contributor
  • ***
  • Posts: 1440
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #21 on: March 01, 2024, 07:39:11 am »
When i was searching for new capacitors, i read from several sites that unlike SMPS:s, the linear regukators don't like low-ESR capacitors.
So, my question is, should i buy a low-esr caps or "normal" ones? Or, does it make any difference in this case?
For example the output capacitor, or the other capacitors.

"Low ESR capacitors" in connection to electrolytic capacitors refers to two different types of parts.

Common aluminum electrolytic and tantalum capacitors have low ESR variations which are not that much lower, and this applies at higher frequencies for switching power supplies.  Using these parts will not be a problem.  They may also be called "low impedance".

There are also polymer aluminum electrolytic (sometimes called OSCON capacitors, maybe because Sanyo was the first to make them available?), and polymer tantalum capacitors which are very different from the above.  They have very low ESR and may cause problems in circuits not designed for them.

To be clear, I've understood the topic as low-ESR electrolytics; good point to be cautious of other types.

Incidentally, polymer and tantalum are available in overlapping ranges of ESR, so you can still find good damping value in these types; it's just that average tants skew higher in ESR, while average polymers skew lower.  Both skew lower than electrolytics.  The lowest-ESR tail is dominated by polymer, while, I'm not sure offhand actually whether dry tant or electrolytic beats out the other; I have a feeling they are comparable, but I'd have to check to be sure.  The main thing is you wouldn't be using dry tants in large values (expensive, risky with respect to surge), so the smaller chip parts tend to have higher ESR, but so do comparable electrolytics, but polymers are fine for surge use.

As for electrolytic ESR, I wouldn't say quite as generally; ESRs are impressively low these days, at least in certain values/ranges where it's needed.  Which is most of them nowadays, heh.

Low ESR isn't really new, it was just specialized before.  Consider Sprague 672D/673D:
https://www.vishay.com/docs/42054/672d.pdf
https://www.westfloridacomponents.com/mm5/graphics/M04/673D107F250.pdf (did they inherit just this exact product line, or 2nd source because MIL use..?!)
(The original actual-Sprague brand catalog, consider http://www.bitsavers.org/components/sprague/Sprague_Electronic_Components_1980.pdf for example, claims "low ESR", but doesn't actually say, and the engineering bulletins are long gone.  Somewhere between then and now, the ESR was published in the datasheet/catalog, but who knows if the formulation or design (other than the outer cans maybe) has changed.)

These values (assuming they're historically representative) are regularly met today, the difference is everyone's doing them, because everything has SMPS; and they're doing it in a tiny fraction of the volume, lol.  These were special, and mandatory back in the days of minicomputers; they're featured prominently in vintage DEC equipment for example.

ESRs definitely get low enough that a control loop can become unstable, if it was designed marginally, or dependent on high ESR, whether that was the style at the time or because they were designed around modern "general purpose" types (high ESRs are still available, I think more often in axials, go figure?).  Unlikely, maybe, to be a problem, replacing anything with anything, but the most confident advice that can be given is "try it, and verify response" (load step response for example, shouldn't have overshoot or ringing) (mind, it most likely will have overshoot as it transitions from CV to CC mode, so ensure settings to avoid that during the step; and test both modes, CC and CV will have different responses).

Tim

Hi,

Funny story.
About 10 years ago I needed a low ESR cap so I called the local parts store we had about 3 miles from here.  They had been around for a LONG time and worked a lot with industry.
I told the guy on the phone I needed a "low ESR capacitor", and he didn't know what I was talking about.  He replied something like, "what is that".
 I was so stunned I didn't know what to say for a good 10 seconds or so.
This was an electronic parts store (not Radio Shack) that was around for as long as I could remember.
 

Offline beatman

  • Regular Contributor
  • *
  • Posts: 56
  • Country: gr
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #22 on: March 01, 2024, 01:52:10 pm »
Go on and replace the crapacitors.I have two cheap AX-1803D lab psu's for easy tasks last two years.I have replace all caps with standard 105c 2000h series capacitors from aishi WH and nippon KMQ series.But prior to this i have measure the crap caps chongx noble or acon brands with LCR keysight U1733C and found out of specs.I check with scope before and after recap  and see the ripple.Big improvement.I have replace the big 4700mf filter cap with nippon kmq 6800mf.No need low esr caps is linear psu.
 

Offline Jwillis

  • Super Contributor
  • ***
  • Posts: 1711
  • Country: ca
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #23 on: March 01, 2024, 03:18:49 pm »
If you intend on replacing questionable caps, you need to consider a few factors. ESR is important as it determines the degree of ripple current that the capacitor must be able to withstand without shortening it's lifespan.I believe the Korad KKG305D has three parallel 2200 63V filter caps. Paralleling identical  caps will decrease the total ESR the same way as parallel resistors so three caps will have a third of the ESR of a single capacitor. The inrush current is 3 times that of a single cap but the ripple current is 1/3.
Increasing the capacitance can reduce the voltage ripple, but don't get to high as that can also increase the inrush current. Excessive inrush current puts extra strain on other components and the transformer there by shortening their lifespan defeating the purpose of installing better quality capacitors. Increasing the values of the capacitors could also change the form factor. Will they fit in the space provided and do the pins have the same pitch. This may reduce your choices.
 

Offline 5U4GB

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: au
Re: Linear PSU and low-ESR vs normal capacitors.
« Reply #24 on: March 02, 2024, 10:01:07 am »
I have a Korad KKG305D bench power supply

Would you consider posting photos of the internals to one of the KA3005D/P threads?  It's been mentioned there but no-one knows much about it or how it compares to the more mainstream KA30xx series.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf