Never used one, Jan, looks interesting, nice way of reducing the charge pulses and thereby the intermodulation with audio signal.
Anyone else had any experience with the EC Designs circuit?
Hugh
Anyone else had any experience with the EC Designs circuit?
Hugh
Of course the effect is similar with a CLC but it may be more compact & cheaper than that coil.
An even simpler version would be a C-CCS-C where the CCS is a floating current source of say 3A.
Maybe someone whould put this is a sim, lf sims are generally reasonable reliable.
jd
Sorry Shaan for the OT...
or not quite because this PS implementation is good!
Hi Jan and Hugh,
Honoured to answer to you both 😎
Since its publication I 've only used "charge-transfered supplies" on my projects, first on low power circuits like DACs and preamps:
Picasa Web Albums - mauricio
(photo shows -ECdesigns'- D1M USB DAC with "floating charge transfer supply" at the extreme right, and two "SuperTeddyReg" low noise regulators, the ones with LEDS)
...and then in power amps, like this single ended classA amp, the ZCA:
Picasa Web Albums - mauricio
(this is the non-floating supply (also published on my picasa next to the former); note only two 10000uf caps for each amp; current is +/-1.2amp through a "PowerReg" gyrator based regulator that uses 33uFtantalum//X7R SMD 100nf as main cap; ripple amplitude is about 50-70 mV last time I checked, under load).
...or DestroyerX' DHRII:
Picasa Web Albums - mauricio
What I understood about John's explanation of his charge transfer supplies is that the active element (Mosfet or Darlington) blocks or passes current from the first bank of caps depending if the Schottky rectifiers are conducting or not, this way practically disconnecting mains-TX-rectifiers complex from the second bank of caps which must provide current for the main part of the cycle. 😉
The floating supply ends with double the V at the end: the first // caps see one secondary voltage each and the second banks see the sum of both V so the caps have to be selected accordingly (see my DAC it has 10000uf *2 parallel and then 10000uF for the floating supply).
These combination of charge-transfer supply plus TeddyReg low and high power regulators make really transparent supplies and I would love to find a simpler and clearer solution... 😉
I hope this helps,
M
or not quite because this PS implementation is good!
Hi Jan and Hugh,
Honoured to answer to you both 😎
Since its publication I 've only used "charge-transfered supplies" on my projects, first on low power circuits like DACs and preamps:
Picasa Web Albums - mauricio
(photo shows -ECdesigns'- D1M USB DAC with "floating charge transfer supply" at the extreme right, and two "SuperTeddyReg" low noise regulators, the ones with LEDS)
...and then in power amps, like this single ended classA amp, the ZCA:
Picasa Web Albums - mauricio
(this is the non-floating supply (also published on my picasa next to the former); note only two 10000uf caps for each amp; current is +/-1.2amp through a "PowerReg" gyrator based regulator that uses 33uFtantalum//X7R SMD 100nf as main cap; ripple amplitude is about 50-70 mV last time I checked, under load).
...or DestroyerX' DHRII:
Picasa Web Albums - mauricio
What I understood about John's explanation of his charge transfer supplies is that the active element (Mosfet or Darlington) blocks or passes current from the first bank of caps depending if the Schottky rectifiers are conducting or not, this way practically disconnecting mains-TX-rectifiers complex from the second bank of caps which must provide current for the main part of the cycle. 😉
The floating supply ends with double the V at the end: the first // caps see one secondary voltage each and the second banks see the sum of both V so the caps have to be selected accordingly (see my DAC it has 10000uf *2 parallel and then 10000uF for the floating supply).
These combination of charge-transfer supply plus TeddyReg low and high power regulators make really transparent supplies and I would love to find a simpler and clearer solution... 😉
I hope this helps,
M
Yes that's clear and that's the same a 'normal' supply with rectifiers works; there also most of the time the rectifiers/xformer are disconnected from the reservoir caps also.
What's different here is that the charging current into the reservoir caps is limited. That means charging takes longer but at lower current levels which keeps the harmonics on the supply lines lower and at lower order. Same thing as with CLC configs. So, not sure what the benefit would be other than smaller/cheaper than a coil.
jd
Member
Joined 2009
Paid Member
I see a lot of Vbe drops in this design. OK if you start off with plenty of voltage from your trafo, but the simplicity of a CLC appeals more and sacrifices very little voltage.
I think the voltage sacrifice would be similar. It results mostly from the fact that the current limiting component 'averages' the voltage on the cap. CLC's do sacrifice substantial voltage compared to a straight C supply.
jd
Thanks Jan,
I forgot to add:
Hi Bigun,
I suppose comparing a CLC over a gyrator based regulator, one saves $$$ on capacitance and (expensive) choke, apart from space economy.
Picasa Web Albums - mauricio
(look under the speaker binding posts of this chipamp is the small PCB that contains the (+) and (-) PowerRegs, for example).
I don't have any choke to compare but I will try to make some DIY...of course John Brown also has something to say about chokes: he winds his own (apart of winding his own resistors) with a "Honeycomb" pattern that reduces interwinding capacitance. 😉
Regards,
M
I forgot to add:
Hi Bigun,
I suppose comparing a CLC over a gyrator based regulator, one saves $$$ on capacitance and (expensive) choke, apart from space economy.
Picasa Web Albums - mauricio
(look under the speaker binding posts of this chipamp is the small PCB that contains the (+) and (-) PowerRegs, for example).
I don't have any choke to compare but I will try to make some DIY...of course John Brown also has something to say about chokes: he winds his own (apart of winding his own resistors) with a "Honeycomb" pattern that reduces interwinding capacitance. 😉
Regards,
M
Gyrator unlike a real L does not store an energy. It wastes it. The same components may be used for a source follower with shunt voltage reference, for better stability and lower dynamic resistance. No space economy: heatsink is needed.
Member
Joined 2009
Paid Member
Why ? - I don't see why this should be the case unless the internal resistance of the choke is high.
Amicus Wavebourn,
🙂
You know I am electronically challenged...
We love your pencil drawn schematics...
Since we already must have a big heatsink...
Your signature reminds me to study Latin...though those gentlemen spoke Greek 😀 and maybe considered Latin too rude a language...in veritas...
Cheers,
M
Happy with his truth.
🙂
You know I am electronically challenged...
We love your pencil drawn schematics...
Since we already must have a big heatsink...
Your signature reminds me to study Latin...though those gentlemen spoke Greek 😀 and maybe considered Latin too rude a language...in veritas...
Cheers,
M
Happy with his truth.
The determining factor is L - the AC impedance. The higher the L, the better it works but the more voltage you lose. The AC impedance limits the current changes, so as soon as the rectifier opens up (Vac > Vdc) it takes time for the current to grow, and soon again the rectifiers close (Vac < Vdc).
jd
Noob question.
Are the C-L-C, C-R-C, Gyrator(?) better than the C-CapMultiplier-C filter shown in the ESP pages?
I mean, I have been using plain C first; result- I died several times due to hum. Rested in peace for some time. Used C-R-C then; result- lower hum(still audible though), does not cause acute problem but there arose chronic symptoms, like FATIGUE. Afraid, I jumped to C-CapMultiplier-C; result- simply NO hum; ears on speaker. Now I can feel the real warmth of my SE Class-A follower, hour after hour, day after day(a month later it will be week after week, hooray!).
Is the gyrator better, electrically/sonically???
>>> Capacitance Multiplier Power Supply Filter
Are the C-L-C, C-R-C, Gyrator(?) better than the C-CapMultiplier-C filter shown in the ESP pages?
I mean, I have been using plain C first; result- I died several times due to hum. Rested in peace for some time. Used C-R-C then; result- lower hum(still audible though), does not cause acute problem but there arose chronic symptoms, like FATIGUE. Afraid, I jumped to C-CapMultiplier-C; result- simply NO hum; ears on speaker. Now I can feel the real warmth of my SE Class-A follower, hour after hour, day after day(a month later it will be week after week, hooray!).
Is the gyrator better, electrically/sonically???
>>> Capacitance Multiplier Power Supply Filter
No.
It is what I was saying in my previous post. Your capacitor multiplier has low dynamic resistance in respect to ground, it is what you actually need. Gyrator would have high AC resistance in respect to unfiltered DC, it is not the end result you need.
Let me stress again very significant point: Gyrator mimics inductive resistance electronically, trading off some energy, converting it into heat. While real coil stores an energy and has an intrinsic inductive resistance.
Last edited:
Hello. I'm confused... because I thought that a Gyrator is a capacitance multiplier. I thought that it is two names for the same thing?
J.
No. Gyrator converts capacitive resistance into inductive resistance. What people call "Capacitance Multiplier" is actually an emitter follower. Or a source follower, with RC in base (or in gate).
Thanks for the fast reply Mr. Wavebourn.
Yes, that's what I thought a Capacitance Multiplier is.
OK, so now I need to learn what a Gyrator really is! Are there any links please?
Member
Joined 2009
Paid Member
I still don't understand.
My reasoning is this:
The inductor has an impedance proportional to the frequency. At dc there is very little impedance. Therefore, the ripple (and other contaminants) will experience a higher impedance and will therefore have reduced voltage. This is desirable. The dc experiences no voltage drop in an ideal inductor. Therefore, the inductor has a clear advantage over schemes with Vbe voltage drops IF minimizing dropout voltage is important for your application. The rectifier diodes see a capacitor at their output, so charging pulses are not limited by the inductor.
Gyrator - is a term not restricted to the circuit shown above. A capacitance multiplier is also a Gyrator (according to the internet) in the strict sense of being an impedance converter. Now that's what I call confusing 🙄
"A gyrator is a four terminal or a two port device, that is designed to transform a load impedance into an input impedance where the input impedance is proportional to the inverse of the load impedance." -- a definition from Wiki. We learned gyrators when learned to design ICs avoiding usage of L there.
Speaking of LC against C filters, in the first case output will be an average voltage, in the second case it will be an amplitude voltage. If to add a C before L, it will be something in-between.
Speaking of LC against C filters, in the first case output will be an average voltage, in the second case it will be an amplitude voltage. If to add a C before L, it will be something in-between.
Member
Joined 2009
Paid Member
I know some people (maybe you aren't saying this) claim that Gyrators don't store energy - but as far as I can see this isn't the whole story. Gyrators do store energy. They store it in the capacitor. They store less energy than the equivalent inductor, but they can still store a lot of energy.
As I understand, a capacitor stores the energy in the electric field between it's plates, the inductor in the magnetic field. The key difference is the way in which the energy is stored and then released.
An inductor stores energy when current flow generates a magnetic field. It then releases the energy when the magnetic field collapses back down and in doing so it induces a voltage in the coil. The voltage is mostly limited by the load impedance, so it can rise to a higher voltage than the supply rails. This fact is exploited in tube amps. So, energy goes in via a current flow, comes back out as a voltage that drives a current through an impedance.
A capacitor stores energy when a voltage is held across it's plates, which drives a current to accumulate charge on the plates. It then releases energy when the charge flows out as a current through an external load. The key difference here is that the discharge current is voltage driven so the voltage can not ever exceed the supply rail that was used to charge the capacitor in the first place. So the gyrator can never properly mimic an inductor even if it can store a good deal of energy.
Or maybe I've got this wrong 😕
Last edited:
- Status
- Not open for further replies.