Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

A commodity-based HV regulator: FlexHV
A commodity-based HV regulator: FlexHV
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 8th November 2019, 10:27 PM   #11
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by carlmart View Post
Am I wrong or this is basically a Jung/Didden super regulator adapted for high voltages?
No, certainly not: the PSRR might qualify it for a super-reg, but the output impedance is several hundreds of milliohms, and with a 358, the noise performance is going to be average at best.
It is a Funk1980/LV regulator, optimized for component availability, low cost and tolerance to the outside world.....

Quote:
Originally Posted by jan.didden View Post
No chance! With so many parts you can probably build three J/D super-regs ;-)
Indeed, and at the cost of one J/D super-reg, you can probably build three (or more) of this one

Quote:
Originally Posted by Ketje View Post
Just for fun, a simplex version, no opamps
Mona
Yes, it is a nice and faithful translation in discrete components of this regulator.
I will probably give it a try in sim one of these days.
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 9th November 2019, 03:47 AM   #12
sgrossklass is offline sgrossklass  Germany
diyAudio Member
 
Join Date: Nov 2004
Location: Germany
Quote:
Originally Posted by Elvee View Post
To bias U3 in class A would require additional current, and since the regulator is fully floating and fed by the voltage-setting resistor, any additional consumption would need to pass through the Rset.
Need to? Why?

Quote:
Originally Posted by Elvee View Post
Remember that the current through R4, R5 and possibly R1 also has to be sunk through Rset, and that the regulator is meant to be universal, even for insane input or output voltages within the 20V to input and 12.5V to output
Ah, I see what you're getting at. My little pull-down resistor is obviously not universal - if that's required, it would have to be replaced by a full-fledged current source.
Quote:
Originally Posted by Elvee View Post
Yes, I mentionned this possibility in the AutoShunt Thread, but the gain will be modest, a bit more than 6dB with a 2.5V reference.
With a less-than-ideally quiet reference, even 8 dB is quite welcome if it means 80 µV rather than 200 µV.
Quote:
Originally Posted by Elvee View Post
A larger C8 would benefit the VLF noise, not the 100Hz ripple nor the audio noise.
Besides, startup seems to take the better part of 10 seconds as-is.
Quote:
Originally Posted by Elvee View Post
I opted for 470nF because it does not imply an E-cap, but if VLF noise or output impedance is particularly important, it could be increased (with a LM358, I estimate the 100Hz PSSR to ~120dB. with a LT1013 it is ~140dB)
And that's not bad for a bunch of jellybean parts.

That said, it still is mighty complex for my tastes. At least I'm slowly beginning to understand how the voltage setting business works - U3 compares (Vout-6.25V) to the midpoint between Vout and the lower end of R13, so that I(R13) = 6.25 V / R13, and then I(Rset) = I(R13) + 2*6.25V/(R12+R15+R16). It's ~= 2.017... mA in sim, as you said (BTW: a value of 2.008 mA is obtained with R15/16 = 62k).

In all its glory, the full formula for output voltage then becomes:
Vout ~= 2 * 6.25 V + 2.0 mA * Rset.
which simplifies to
Vout ~= 2 * (6.25 V + Rset/kOhm * 1 V)

Vout = 2(6.25+Rset) was making me feel really uneasy, I can tell you that much! (Dimensions! That would never have passed in physics lab.)
Attached Images
File Type: png FlexHV-mod3.png (15.9 KB, 114 views)
Attached Files
File Type: asc FlexHV-mod3.asc (8.2 KB, 2 views)
__________________
When you have excluded the impossible, whatever remains, however improbable, must be the truth. ~ Sherlock Holmes
  Reply With Quote
Old 9th November 2019, 09:06 PM   #13
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by sgrossklass View Post
Need to? Why?
It does not actually need to in fact: it is more a self-imposed constraint than an actual necessity.

If the total current drain of the floating regulator + various other currents, like the foldback bias exceeds what Rset can sink, the voltage regulation will be lost under open-load conditions, and the output voltage will creep up, eventually towards the raw input voltage.

You could rightly say that even a small tube preamp is always going to draw enough current to prevent this situation from happening, thus for a system power supply, the open-output case can be dismissed (as opposed to a bench supply, which needs to behave in all circumstances, including no-load), but that would be ignoring some realities: tube gear needs some time before the heater allows cathode current to pass, and a standby switch interrupting the HV supply is not infrequent.
If you use such a supply to derive, say a 200V supply for a preamp from the main 600V of the output stage, the preamp supply could reach twice its nominal 200V until the tubes heat up properly.
This could have consequences for some components, like capacitors.
I think you saw the reason:
Quote:
Ah, I see what you're getting at. My little pull-down resistor is obviously not universal - if that's required, it would have to be replaced by a full-fledged current source.
Note that the regulation properties under no-load conditions are going to be pretty poor: not only is the gm of the MOSfet low, but the output configuration is highly asymetrical (~1 quadrant), requiring a peculiar compensation and a blunting of normal-load regulation performances, but I think it is price worth paying.

Quote:
With a less-than-ideally quiet reference, even 8 dB is quite welcome if it means 80 µV rather than 200 µV.
I completely agree: if you are after an optimum noise perf, even the odd dB shouldn't be scorned at

Quote:
Besides, startup seems to take the better part of 10 seconds as-is.
Yes, could act as a soft-start



Quote:
That said, it still is mighty complex for my tastes.
The (relative: look at the breadboard implementation) complexity results from an evolution starting several years ago, when Funk1980 asked me to have a look at a HV regulator.
I proposed a first circuit, and it was implemented with a LM358, because of cheapness and universality.
Only half of the package was needed (there is a single version of the 324/358, but it is confidential and probably costlier than a 358), and I found a use for the second operator as a shunt regulator.

I then had the idea of converting this working, single-purpose design into a more universal one, but it required an additional transistor and a more elaborate compensation scheme.

That's where it ended now, but you can rewind, and undo most of the mods whilst retaining the universality: use a single opamp, even one that does not include GND in its CM range thanks to Mona's mod, ditch the AutoShunt and use a Lo-Po TL431-like reference (the regular TL431 requires more current than the LT1004 + 1/2 LM358).

If the opamp is half-decent, it will not need an additional buffer, meaning further simplifications.

The present circuit has the advantage of being completely universal (it can use almost anything vaguely non-linear component as a reference), completely commoditized (the LM358 is available everywhere), and has remarkable performances thanks to the complete stasis allowed by the autoshunt, but if you opt for more conventional solutions, there can be significant simplifications


Quote:
Vout = 2(6.25+Rset) was making me feel really uneasy, I can tell you that much! (Dimensions! That would never have passed in physics lab.)
Well, it's like all easy-to-use formula's: for example the 1/Gm one for semi junctions: 26/Ie gives the value in ohm with Ie in mA.
We know that this is not proper and homogeneous, but it is easy and convenient, and as long as we know what it actually stands for, it is a convenient mental shortcut.
Life is too short too worry about such details.... if you know what you are doing
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 10th November 2019, 10:25 PM   #14
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by sgrossklass View Post
was making me feel really uneasy, I can tell you that much! (Dimensions! That would never have passed in physics lab.)
Another nice one to make you really cringe: do you know that the VA rating of an ordinary, silicon steel transformer can be estimated with this formula: P(VA) = A², A being the core area expressed in cm² (for a 50Hz system).
This would mean that 1VA (having the dimension of W) = 1cm4 (1e-8m).
Of course, there is a hidden constant in the formula, having complicated dimensions which happens to be exactly 1 for the very peculiar conditions and units specified.

If you try to draw general conclusions from this example, without really understanding all the implications, you will be going straight and and fast into a wall
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 14th November 2019, 09:17 PM   #15
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by Elvee View Post


That's where it ended now, but you can rewind, and undo most of the mods whilst retaining the universality: use a single opamp, even one that does not include GND in its CM range thanks to Mona's mod, ditch the AutoShunt and use a Lo-Po TL431-like reference (the regular TL431 requires more current than the LT1004 + 1/2 LM358).

If the opamp is half-decent, it will not need an additional buffer, meaning further simplifications.
Here is how would such a reg look like (only an example, has not been tested in reality or refined in any way).

I used a standard TL431 and it works, because of the low consumption of the LT1637 (a randomly selected opamp) and the built-in margins, but that's not something I recommend: use a low-power version like the TLV431.
It should decrease the no-load consumption to under 1mA, allowing you to increase Rset to the point of being a 1/2 or 1/4W type (having a sufficient voltage rating though)

Using asymetric bridge resistors, in Ketje/Mona's fashion would allow you to use an opamp not having GND in its input CM range.

The complexity is significantly decreased, yet there are still optional bells and whistles, like the current-limiter with its indication LED, and the filter cell R2 C2
Attached Images
File Type: png FlexHV4.png (77.2 KB, 70 views)
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 14th November 2019, 10:58 PM   #16
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
If you are after the most pruned-out version, here it is:

A commodity-based HV regulator: FlexHV-flexhv5-png

It still works, but has a 4mV residue for a 40V input ripple: clearly unacceptable
Attached Images
File Type: png FlexHV5.png (68.4 KB, 74 views)
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote
Old 19th November 2019, 04:08 PM   #17
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
An important note about the components selection:

The MOSFET must include the coordinates (Vin_raw, Iout_max) inside its DC safe operating area, even if the foldback option is implemented.
The foldback feature can be used to reduce thermal requirements, but not the fundamental V x I capability.

The zener (D4) needs to have voltage allowing the opamp's output to sit at 3V above the local -6.25V under median conditions.

Since the G-S voltage under those conditions is going to be close to Vto, this would be 3.25v + 0.3V + Vto, but at relatively low current: a few hundreds of µA's.

The Vto depends on the exact MOS type used: in general, legacy types are around 4V, and more recent devices have 3V, sometimes less.

If the zener voltage is too high, the circuit will lose the regulation at low currents, and if it is too low, it will run out of steam at higher currents.
Click the image to open in full size.
__________________
. .Circlophone your life !!!! . . ♫♪ My little cheap Circlophone© ♫♪
Electrify your daily life!!! ....⚡⚡ ELEKTRIA ⚡⚡
  Reply With Quote

Reply


A commodity-based HV regulator: FlexHVHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Big Tube based voltage regulator Sparky95 Tubes / Valves 147 24th March 2018 08:24 PM
APERO : Voltage regulator based on an idea of G. PERROT KARMA38 Power Supplies 12 12th April 2017 03:19 PM
DIY Upgrade of "commodity" S/PDIF sourround processor thrillscience Digital Source 4 17th December 2013 04:30 PM
Chipamp-based Regulator? gootee Power Supplies 12 22nd January 2007 06:05 PM
Yet another Opamp based regulator banana Power Supplies 4 22nd January 2007 03:46 AM


New To Site? Need Help?

All times are GMT. The time now is 09:07 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki