Author Topic: EEVBlog #1116 - The Capacitance Multiplier  (Read 44413 times)

0 Members and 1 Guest are viewing this topic.

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #25 on: August 29, 2018, 04:52:03 pm »
If you are asking why does C422 not go to ground, but rather to the output of the transistor, my own little research suggests that this is a bit more than a simple RC filter. It looks to me like the topology of a Sellen-Key filter like this: https://en.wikipedia.org/wiki/Sallen–Key_topology

Do I have it right? I can't explain exactly how it works though. Need to study this.

Quite right, it doesn't have the unity gain of an op-amp so the stop-band attenuation won't be perfect, but the transition band can be peaked for whatever filter characteristic you need, and the asymptotic slope can be -40dB/dec (until leakage takes over).

Note the feed-forward path through the first resistor and capacitor, so that at frequencies where the transistor gain is dropping off, this path dominates, limiting stop-band attenuation.  Admittedly, this is still a higher roll-off than for a typical op-amp, so it might not be too bad.

Here's an example I made, with vacuum tubes for audio cred:



The 6V6 model is actually pretty close (when the model's bugs have been avoided..) to what was actually used, a 5702 submini pentode.  The simulated response:



You can see it's much sharper and tighter than anything you can do with resistors and capacitors alone, and smaller than most things you'd do with inductors (I know, you can get pretty small ~1H chokes, though the Q is pretty bad).  Plus you'd still need the buffer to drive an LC filter, so might as well do it in one. :)

I don't have a picture of the actual filter, but it went on this radio, just after the detector.  The construction is similar, just imagine it with another submini and more capacitors and resistors around. :P

Also, incidentally, I finished up that radio by powering it with two SMPSs, in a hybrid configuration.  One AC to 12V 2A supply (Meanwell) just to get things going, then a discrete 12 to 6.3 and 100V supply (a slight variant of this) for the tubes.  I added shielding around, and an extra stage of LC filtering outside the discrete module, and that was all that was needed to keep the thing dead quiet on the 20m (~14MHz) band.

Tim
« Last Edit: August 29, 2018, 04:59:09 pm by T3sl4co1l »
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: AviZiv

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #26 on: August 29, 2018, 05:18:05 pm »
Not being a native speaker of English, one thing I would appreciate very much is if you would wrote down some of the more unusual or not as easy to understand terminology. For instance, it took me a bit of googling until I found out that the one configuration you mentioned was spelled "Sziklai pair".

Yep, noted, I'll write down such thing in the future.

Having watched, with occasional amusement, your regular mailbag frustrations and wrestling with mittel-europa place names may I offer a little pronunciation help here. Sziklai is a Hungarian name and like all Hungarian words with 'sz' or 'cz' phonemes is unpronounceable by anybody who isn't a native Hungarian. The nearest I've managed to get transliterates as 'shik-laii' with the 'sh' sound being something between the English 'sh' and 'z' sounds - shape your mouth to say 'z' but instead say 'sh'. I got this pronunciation years ago from a Hungarian descended German colleague.
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6927
  • Country: ca
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #27 on: August 29, 2018, 06:08:53 pm »
Thanks, here I thought it was a Japanese name  :-[  Let's get our Sidney Darlington and George Clifford Sziklai right.

But Sziklai pair still won't work in a capacitance multiplier.
 

Offline ggchab

  • Frequent Contributor
  • **
  • Posts: 276
  • Country: be
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #28 on: August 30, 2018, 11:03:45 am »
Thank you for this very interesting video  :-+
 

Offline IanMacdonald

  • Frequent Contributor
  • **
  • Posts: 943
  • Country: gb
    • IWR Consultancy
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #29 on: August 30, 2018, 11:10:14 am »
But Sziklai pair still won't work in a capacitance multiplier.

BS. PNP series pass, NPN driver.

Though, I wonder if we could achieve time travel this way with less flux.

Incidentally, the way this circuit works is that the cap will try to charge to Vin, and thanks to the light loading of the emitter follower it will mostly succeed. However, when Vout gets too close to Vin the transistor becomes unable to draw collector current because the difference is too small, and the transistor then effectively becomes a diode between the cap and Vout. In which case the cap voltage is rapidly dragged down until the collector can draw current again. Hence Vout tends to stabilise at the lowest instantaneous ripple voltage.
« Last Edit: August 30, 2018, 11:29:06 am by IanMacdonald »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6927
  • Country: ca
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #30 on: August 30, 2018, 06:13:06 pm »
With all due respect, you are mistaken. Unless you have a sketch to clear up misunderstanding on how the NPN driver works at multiplying capacitance.

Once you have high ripple on either transistor's emitter, you are modulating conduction at the E-B junction.
This breaks the basic principle of the an emitter follower here in the capacitance multiplier that relies on pure DC at the base and the resulting near constant collector current (in both transistors) to filter ripple.

People arbitrarily change from Darlington to Sziklai, NPN/PNP etc. and let's not be sloppy.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #31 on: August 30, 2018, 06:27:24 pm »
With all due respect, you are mistaken. Unless you have a sketch to clear up misunderstanding on how the NPN driver works at multiplying capacitance.

Once you have high ripple on either transistor's emitter, you are modulating conduction at the E-B junction.
This breaks the basic principle of the an emitter follower here in the capacitance multiplier that relies on pure DC at the base and the resulting near constant collector current (in both transistors) to filter ripple.

People arbitrarily change from Darlington to Sziklai, NPN/PNP etc. and let's not be sloppy.

I'd be interested to see a simulation illustrating this.

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

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6927
  • Country: ca
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #32 on: August 30, 2018, 08:21:35 pm »
With Szikilai the pass transistor appears to not be contributing, its (emitter) in-phase with the ripple so the result is slightly worse? Only the driver is doing the work.

For discussion I did a basic run in LtSpice and also in Paul Falstad's Circuit Simulator of the capacitance multiplier using a Darlington and Szikilai.

It's kind of a nothing burger circuit because it wastes a fair bit of voltage as heat, and adding a zener turns it into a voltage regulator anyhow.
I've rarely encountered capacitance multipliers. Only seen in esoteric audio gear.
 
The following users thanked this post: thm_w

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #33 on: August 31, 2018, 06:43:03 am »
Thanx for coming back to fundamental, bummer it wasn't a friday :D

Someone please run a SIM on the shizaki pair and end our missery, that and a darlington to compare. My money is lower ripple for the darlington, lower drop out for the other.

JS

If I don't know how it works, I prefer not to turn it on.
 

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 3452
  • Country: it
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #34 on: August 31, 2018, 08:19:50 am »
I just saw this that seems kinda related :)

MIC94305
500mA Switch with Ripple Blocker Technology
http://ww1.microchip.com/downloads/en/DeviceDoc/MIC94305-500mA-Switch-with-Ripple-Blocker-Technology-DS20006029A.pdf
 

Online EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #35 on: August 31, 2018, 08:43:30 am »
I think a better name for this circuit topology would be resistance multiplier instead of capacitance multiplier.

I tend to agree, but capacitance multiplier is the industry term like it or not.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16561
  • Country: 00
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #36 on: August 31, 2018, 01:50:50 pm »
I just noticed that "InfiniVision" is spelled InfiniiVision.  :popcorn:

 

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #37 on: August 31, 2018, 04:32:59 pm »
@radioactive: thanks, I guessed one and missed the other...

JS

If I don't know how it works, I prefer not to turn it on.
 

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #38 on: August 31, 2018, 05:21:57 pm »
Looks like Sziklai wins this time. Same rejection but with lower dropout. One thing that might worth checking is high freq to see how it will work in a SMPS.

Excelent you are running this! Thanks so much, shacky days here or I would running them myself...

JS

If I don't know how it works, I prefer not to turn it on.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #39 on: September 01, 2018, 06:33:58 pm »
Datasheet not very useful -- you're assuming the datasheet is even descriptive enough (it's not), or that the models are anything like an average part (who knows). :-//

What's the library entry look like for those parts, anyway?

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

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6927
  • Country: ca
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #40 on: September 01, 2018, 08:14:21 pm »
I tried your simulation in LTSpice, much different than what I'd drummed up.
2N3055 Bf=73 with 1.5A load and 10kHz input ripple compared to my 2N3904 Bf=300 with 42mA load and 120Hz input ripple.
I could not get the same results in favour of the Sziklai for ripple reduction.

In LTSpice, I noticed the 2N3055 Darlington has a higher voltage drop, I'm getting 8.5V out, compared to your 10.4955V so the base current is way different. 0.85mA across the 5.4k
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16547
  • Country: us
  • DavidH
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #41 on: September 01, 2018, 11:24:34 pm »
The difference in performance between the Darlington and Sziklai pair comes up in audio power amplifier design so there might be something to learn there.  Local feedback designs with a third transistor are much better than either.

Another thing to learn from audio power amplifier design is Ft and hfe droop at high currents which are going to be present in a high current capacitance multiplier.  There will be considerable advantage to using RETs (ring emitter transistors) or equivalent like the D44/D45 series instead of common power transistors.

I myself have only used capacitance multipliers like this at 10s to 100s of milliamps for sort of super power supply decoupling and surge protection.

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #42 on: September 02, 2018, 12:17:29 am »
Wow, are all the parameters identical for those models, except for type (NPN/PNP)? Whose idea was that? :-DD

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

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6927
  • Country: ca
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #43 on: September 02, 2018, 12:45:04 am »
My intuition squawked because of the difference in the two configurations, reasoning the PNP transistor does not contribute with pulsing DC on its emitter, so the Sziklai is using beta of only the NPN for multiplying capacitance.

I went down to 1kHz to see if speed is an issue knowing that both transistor configurations are relatively slow. Very rare to see a Darlington used as a SMPS switch for example.

Perhaps small signal AC analysis is better? If we assume the driver's base is an AC ground, it's acting as a low-pass filter and attenuation would show differences.

No need to apologize, I don't think electronics is about being correct or perfect.
For me the forum is about sharing and learning, I don't care if I'm wrong aside from the troll attack.

Similar transistor models would help in the comparison, I was just lazy and using the ST 2N3055 that comes in the library.

Rod Elliot from Elliot Sound compares the two configurations in audio and switching.
 
The following users thanked this post: radioactive

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #44 on: September 02, 2018, 12:56:09 am »
@T3sl4co1l,

I just did a diff on all the parameters (after removing NPN/PNP) and yes they are!  I normally don't use Qucs for these type of simulations (utilizing 3rd party contributed component models.  Guilty parties provided e-mail address).  After this embarrassment,  I probably won't again without checking much closer.  It is great software for many things, especially s-parameter simulations.  Again, sorry for wasting everyones time in this thread.  No good deed (attempted) goes unpunished... why is that almost always true?

No worries man, I appreciate taking the time to try. :)

Me, I'm just incessantly paranoid about everything I have to work with: models, simulator quirks, datasheets, tolerances...

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

Offline digsys

  • Supporter
  • ****
  • Posts: 2209
  • Country: au
    • DIGSYS
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #45 on: September 02, 2018, 04:03:31 am »
Quote from: radioactive
  ... Any requests from anyone for specific transistors to use? ....
Well, since you're taking orders :-) When I ever get a chance, I want to test these Power FETS AUIRFS-8409-7P at 10A (or something similar).
I can send you a couple (+) if you're interested. I use them in LT4320 bridges, up to 30A + and would love a simpler option for minimizing ripple.
Hello <tap> <tap> .. is this thing on?
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1773
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #46 on: September 03, 2018, 05:31:34 pm »
Hi Dave,

I do aknowledge that the cap multiplier is a common and useful circuit, but to avoid frustration it is OK to show its limitations, too:

- This circuit is strictly for low frequency ripple removal only. Supression of fast switching spikes will be poor
- For the sake of efficiency, the transistors operate at rather low Vce voltages where there collector capacitance is very large.
  This capacitance feeds the unfiltered input right to the transistor base.
- As a safety precaution, the input of the darlington base should have a current limiting resistor attached. When your output is shorted
  and you have no such resistor, the filter cap will discharge thru the base of the first darlington transistor without a current limit,
  possibly killing it.
- The input resistor is a bit tricky to choose, because a proper value is dependent on transistor beta(s). If its too small, transistors
  almost go into saturation becoming excessively slow, if its too large the voltage drop increases too much. Transistor betas
  can vary a lot from device to device and over temperature, so a resistive divider is probably a better idea (or even a string
  of two diodes from the input, with a resistor going to ground, and then a series resistor plus the cap).


 

Offline jasonhanjk

  • Newbie
  • Posts: 3
  • Country: my
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #47 on: September 07, 2018, 01:07:43 am »
I'm using this circuit actually. When my signal into the op amp increases, the noise from the output of the transistor goes up too. Otherwise it's actually a good few mV ripple supply.
 

Offline Leo Bodnar

  • Frequent Contributor
  • **
  • Posts: 801
  • Country: gb
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #48 on: September 09, 2018, 09:39:56 am »
This is barely scraping into the topic of the discussion but I think it might be useful for practiotioners of ripple removal.

Instead of blocking the high frequency ripple you can shunt it out. Wenzel has an apponote here.

Cheers
Leo

 
The following users thanked this post: BravoV, thm_w

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: EEVBlog #1116 - The Capacitance Multiplier
« Reply #49 on: September 09, 2018, 01:20:03 pm »
While some of the circuits in Wenzel's appnote might have fair to good performance the one illustrated above looks pretty mediocre. Surely the highest effective attenuation (with the shunt transistor fully on and theoretically a short circuit) would be that of the divider formed by the 15 ohm resistor and the ~1k to ground presented by the rest of the circuit - i.e. a gain of 0.985 or -0.13 dB. With that relatively huge 1k in the way of the shunt current I don't see how it could be effective under any circumstances. There must be some error in the presented resistor values.
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf