Canuck Audio Mart Hifi and Audio ForumCanuck Audio Mart Hifi and Audio Forum
It is currently Tue Dec 11, 2018 4:13 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 76 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
PostPosted: Tue Oct 09, 2018 5:01 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
Nakamichel wrote:
Center tap transfo ? Floating ground ?

The transformer I'm testing with has dual 18 Vac secondaries.

The scope is old school - with a chassis that is earthed.
So the circuit is connected to earth, when the probe's ground strap is attached.


Going right back to basics,

The 18 Vac secondary was rectified by a diode bridge, to create Vp = 24 Vdc with T = 120 Hz

The photo shows this 120 Hz rectified wave - which was then clipped by a 1 K in series with a 15 V zener.

The clipped DC looks fabulous

However, look closely - the electric drill was on when then this photo was taken.
The positive and negative slopes are dancing around like Justin.

The issue is, the zener is basically the "clock" telling the transistor switch
when to turn on and off - but there is jitter.
There was nothing else connected to this circuit (no transistor, no load resistor or caps).

From what I can figure out, after the 120 Hz is switched and filtered
the jitter is showing up as spiking noise.


Attachments:
Zener.jpg
Zener.jpg [ 118.81 KiB | Viewed 301 times ]
Rectified and Zener.jpg
Rectified and Zener.jpg [ 85.8 KiB | Viewed 306 times ]


Last edited by Uunderhill on Tue Oct 09, 2018 5:20 am, edited 1 time in total.
Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 5:14 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 787
Location: London, ON, CA
So, why use this circuit at all when a standard voltage regulator has none of these issues?

Just askin'........ :|


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 5:39 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
jst_canuck wrote:
So, why use this circuit at all when a standard voltage regulator has none of these issues?

This is developing a circuit at low voltage - before moving on to +/- 100 Vdc rails.

There are LR8 and LR12 high voltage regulators, but these would be operating right at their extreme limits.
Components should be selected to operate well under their ratings.

However, even at that, chip regulators are not perfect.
If possible, using this clipper circuit, before a chip regular, would improve the output even more.

Also, this is the ideal application for Nakamichel's clipper circuit.
High voltage rails are required with a low load current.

.


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 7:48 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 787
Location: London, ON, CA
Uunderhill wrote:
jst_canuck wrote:
So, why use this circuit at all when a standard voltage regulator has none of these issues?

This is developing a circuit at low voltage - before moving on to +/- 100 Vdc rails.

There are LR8 and LR12 high voltage regulators, but these would be operating right at their extreme limits.
Components should be selected to operate well under their ratings.

However, even at that, chip regulators are not perfect.
If possible, using this clipper circuit, before a chip regular, would improve the output even more.

Also, this is the ideal application for Nakamichel's clipper circuit.
High voltage rails are required with a low load current.

.

How about something like this instead? http://www.ti.com/product/TL783

Nothing is "perfect" but there are very common engineering solutions in linear power supply design.

Have you considered ranking these targets for your new power supply design:

- Lowest possible line frequency ripple?

- High rejection of AC line noise?

- "Stiff" high voltage, low current DC?

- Component cost?

Always good to review where one is going regularly to be sure that you have not gone off target :idea:

My experience is that it is not a good idea to use unproven design ideas when common, low cost solutions are on the shelf.
You seem to want a "Voltage Regulator" :?

Re-inventing the wheel is a fun classroom project but otherwise a wasted exercise.
Buy some wheels and get on with your cool car design 8)


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 8:26 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
jst_canuck wrote:
How about something like this instead? http://www.ti.com/product/TL783

Bingo ! Thanks for pointing this high voltage regulator out.

Have a look at Nichamichel 's schematic on the middle of page 2.
There is the zener / transister clipper circuit combined with a voltage regular.
Looks like the ti TL783 will work well in this application ( +100 Vdc @ 20 mA ).

jst_canuck wrote:
- High rejection of AC line noise?

Filtering out Noisy Hydro is key

jst_canuck wrote:
- Component cost?

Sonics have priority over cost.

.


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 8:30 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 787
Location: London, ON, CA
Uunderhill wrote:
jst_canuck wrote:
How about something like this instead? http://www.ti.com/product/TL783

Bingo ! Thanks for pointing this high voltage regulator out.

Have a look at Nichamichel 's schematic on the middle of page 2.
There is the zener / transister clipper circuit combined with a voltage regular.
Looks like the ti TL783 will work well in this application ( +100 Vdc @ 20 mA ).

jst_canuck wrote:
- High rejection of AC line noise?

Filtering out Noisy Hydro is key

jst_canuck wrote:
- Component cost?

Sonics have priority over cost.

.

Not clear at all that this "clipper" has any value at all in your design :?:

Any commercial products in any field of electronics use such an idea?


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 8:51 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
jst_canuck wrote:
Not clear at all that this "clipper" has any value at all in your design :?:


From what I've measured with 78XX series of voltage regulators, noise still gets through them.

So clipping and filtering before a regulator should help - combined with filtering after - should give a high quality supply

which again is 1/2 the sound.


Top
 Profile  
Reply with quote  
PostPosted: Tue Oct 09, 2018 5:46 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 837
Location: Montréal, QC, CA
jst_canuck wrote:
So, why use this circuit at all when a standard voltage regulator has none of these issues?

...My experience is that it is not a good idea to use unproven design ideas when common, low cost solutions are on the shelf.

...Re-inventing the wheel is a fun classroom project but otherwise a wasted exercise.
Buy some wheels and get on with your cool car design 8)

...Not clear at all that this "clipper" has any value at all in your design :?:
Any commercial products in any field of electronics use such an idea?
...Thanks !

But seriously JST,
from my point of view, the clipper is nothing since I have 2 better technologies that bury the clipper.
My top tech produces (from AC) 0 ripple. Zero. The most flat DC possible.
But against most of the existing power supply, the clipper is performing. It will improve any supply.

When Uunderhill will finally understand and control at will the clipper, he will be surprised to see how
well the clipper deals with the drill noise. The drill test is highly exaggerated (no one will plug a drill with their hi-end system)
but it looks like Uunderhill decided to test my clipper against the worst condition. I'm not worry, not at all.

When measuring AC voltage, meters read the RMS voltage. Why ?
Because it's where the DC line would be, higher than that there's nothing constant.
If that proof is not enough for you to understand that higher than RMS it is not possible
to have consistent voltage at any instant T, then it's your choice and I respect it.

I'm aware that my words are just words.
I'm about to change that with my ACA power supply which I'll share the schematic here.
It won't be efficient, yes I know it. But...

Uunderhill wrote:
The clipped DC looks fabulous
Of course.
There's no filtering cap yet but we can already see DC. The clipper is marvelous.

Uunderhill wrote:
However, look closely - the electric drill was on when then this photo was taken.
The positive and negative slopes are dancing around like Justin.

The issue is, the zener is basically the "clock" telling the transistor switch
when to turn on and off - but there is jitter.
There was nothing else connected to this circuit (no transistor, no load resistor or caps).
With only the - side grounded, you are creating a resonant circuit. You need a load, this load must be at least 10x (lower) the clipper resistor.
Or, use floating ground, the valse will disappear. And anything's happening lower than DC line won't appear at the output.


You know, I see you are making progress with the clipper. You took the right road going with experimenting
with a lower voltage for starter. Congrats, you are on the right way. I'm encouraging you to continue that way.

Here tips to continue to learn about the clipper:
Attachment:
clipper w. RVs.jpg
clipper w. RVs.jpg [ 43.29 KiB | Viewed 276 times ]
R will protect the Zener.
RV1 sets the current.

Then, RV2 sets the voltage, the clipped voltage. Thus, you could adjust RV2 and see what happens.
With R to protect the Zener, you can use a lower voltage Zener to go even lower than RMS and see
that lower you go, better is the DC.

FWIW, the LM317T reg is better than the TL783. And would be good to use with your +/-100v DC.
I hope you won't reply that it is just good for 40 volt max...

Mik

_________________
Keep things simple and even simpler

Balance MC-1010 phono stage... in progress (week 2)


Top
 Profile  
Reply with quote  
PostPosted: Wed Oct 10, 2018 3:51 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
jst_canuck wrote:
So, why use this circuit at all when a standard voltage regulator has none of these issues?

Going back to this question : noisy hydro still gets through almost every piece of audio equipment I have.
Regardless of how many regulators are in the power supply.
In the end, I use a hydro regenerator to solve the issue.

Even though this particular cathode follower circuit has a very good PSRR
poor hydro still affects my first preamp version.
The object here is to develop a power supply that blocks even the worst hydro noise.

Think you need to see the rectified 120 Hz waveform coming off the diode bridge.
The waveform is awful - even without the electric drill, a DC furnace motor affects this 120 Hz waveform.

However, as the photo of the scope trace shows - the top of the clipped waveform is poetry - its beautiful.

Nakamichel wrote:
You know, I see you are making progress with the clipper. You took the right road going with experimenting
with a lower voltage for starter. Congrats, you are on the right way. I'm encouraging you to continue that way.

Yes - its important to test the circuit one step at a time and not use the prayer method.
The prayer method occurs when the whole thing is built and the persons prays that it works.

Nakamichel wrote:
-1 No, ringing is not higher because of the clipper. And ringing comes from diodes behavior,
not from circuit inductance. Inductance brings ringing but at a lot higher frequency.

You are right. It seems that the "ringing" was not caused by the transistor switch turning on and off.
but rather the bridge diodes switching off the secondaries.
(v) = - L di/dt
This was showing up on the filtered output.
47 nF accross the secondary - before the zener - solved the ringing issue.
Hopefully, those fancy soft recovery diodes you recommended should minimize ringing.

Nakamichel wrote:
A drill with brushes takes the AC current by pulse. The rising and falling edge are then DC current, meaning,
not a sinewave anymore. So, many small DC pulses are added to the AC line. Any noise is easy to remove
when we know exactly which kind of noise it is.

The pulses from the electric drill mostly show up on the peaks of the 120 Hz rectified waveform.
As the photo of the scope trace shows, the clipper completely removes them.

.


Last edited by Uunderhill on Wed Oct 10, 2018 6:19 am, edited 3 times in total.

Top
 Profile  
Reply with quote  
PostPosted: Wed Oct 10, 2018 4:06 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
Nakamichel wrote:
With only the - side grounded, you are creating a resonant circuit. You need a load, this load must be at least 10x (lower) the clipper resistor.
Or, use floating ground, the valse will disappear. And anything's happening lower than DC line won't appear at the output.


The load for this power supply will effectively be set by the two CCS's in the preamp circuitry.
The total current for both the L and R channels will be 20 mA.

Yes the ground strap on the scope probe is connected to hydro earth.
Once the probe's ground strap is connected to floating ground of the power supply, it connects the ground to hydro earth.

However, the final preamp will have its audio ground "Lifted" from the hydro earth.

Actually, I've already tried a variable resistor on the +ve side of the Zener.
Setting Iz to anything more than 9mA does not reduce the noise of Vout.

OK - I've just tried a variable resistor between the Zener and ground rail.
Adding resistance here, does nothing but remove part of the clipping.

A significant improvement was adding a 47nF cap across the secondaries
which stops ringing.
.


Top
 Profile  
Reply with quote  
PostPosted: Thu Oct 11, 2018 6:30 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
Between mixing concrete for sonotubes, and the turbulence in stocks,
yesterday was not a good day.

OK - I've started looking around for a transformer.
Transformer design is an entire topic within itself, however

The specs are

- dual secondaries
- 100 Vac to 120 Vac secondaries
- low capacitive coupling between the primaries and secondaries
- low EMR (high impedance tube circuits are known to be suspable to EMR
- power rating somewhere are 50 VA
- No Hum (potted ?)

I've found

Hammond 1182 G 117 - but I'm concerned about the capacitive coupling.
https://www.hammfg.com/part/1182G117

Actually, a small isolation transformer, with dual secondaries maybe a solution.

Hammond 185 D 230
http://www.hammondmfg.com/185.htm
.

Any other recommendations ?

.


Top
 Profile  
Reply with quote  
PostPosted: Thu Oct 11, 2018 10:07 am 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 837
Location: Montréal, QC, CA
Uunderhill wrote:
OK - I've just tried a variable resistor between the Zener and ground rail.
Adding resistance here, does nothing but remove part of the clipping.
Does nothing ?!
It's the voltage out setting. You'd need it if you couldn't find a transformer or a Zener with all the specs you want.

Speaking of the transformer, a R-core would be great. Windings are separated.

As suggestion, 1/1 transformer gives 170v peak. With a VDC of 100v, you cut 70v and are under RMS for an ultra good DC.
The supply has to be 10x the load for outstanding stability.

Since your load is 20ma, you need a transformer of 10 x 20ma = 200ma.
0.2 x 120 = 24va

But we have to add to that the clipped part.

The clipper will dissipate: 70v x 100ma per channel. 7 watt/channel, very reasonable for a flat and stable DC.
Let say a transformer 50va per channel and you are in the ultra safe area...

It's the way I'd do it, no doubt.

_________________
Keep things simple and even simpler

Balance MC-1010 phono stage... in progress (week 2)


Top
 Profile  
Reply with quote  
PostPosted: Thu Oct 11, 2018 12:13 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
Nakamichel wrote:
Uunderhill wrote:
OK - I've just tried a variable resistor between the Zener and ground rail.
Adding resistance here, does nothing but remove part of the clipping.
Does nothing ?!
It's the voltage out setting. You'd need it if you couldn't find a transformer or a Zener with all the specs you want.

The rectified 120 Hz changes in voltage, so the current flowing through the Zener and any series resistors changes.
So if a resistor is connected between the Zener and Ground, the voltage drop across this resistor will change,
and the beautiful clipping is lost.

Surprisingly there are 100 V Zeners made 1n4764A
but I see the datasheet says Iz = 2.5mA for the test current - which seems quite low.

https://www.vishay.com/docs/85816/1n4728a.pdf

But the datasheet also says this series of diodes can dissipate 1300 mW
if the leads are 4 mm off the case - so I think Iz could be safely set to 6.5mA.
OR a number of Zeners connected in series could be used.

Nakamichel wrote:
Speaking of the transformer, a R-core would be great. Windings are separated.

OK I'll try to find one.
Having low capacitive coupling seems to be important in terms of filtering out Hydro noise.

Nakamichel wrote:
As suggestion, 1/1 transformer gives 170v peak. With a VDC of 100v, you cut 70v and are under RMS for an ultra good DC.
The supply has to be 10x the load for outstanding stability.

Thanks for pointing this out.

Nakamichel wrote:
Since your load is 20ma, you need a transformer of 10 x 20ma = 200ma.
0.2 x 120 = 24va

But we have to add to that the clipped part.

The clipper will dissipate: 70v x 100ma per channel. 7 watt/channel, very reasonable for a flat and stable DC.
Let say a transformer 50va per channel and you are in the ultra safe area...

As I realized this morning, a 1:1 isolation transformer looks like it will be ideal for your clipper
and this tube circuit.

OK the hunt for this transformer is on.

Also, if I recall correctly from another discussion, you mentioned the importance
of having a transformer with low Zout secondaries.
Looking at the datasheets for Hammond the higher the VA rating - the lower the Zout
of the secondaries.

.


Top
 Profile  
Reply with quote  
PostPosted: Thu Oct 11, 2018 1:37 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 837
Location: Montréal, QC, CA
Uunderhill wrote:
The rectified 120 Hz changes in voltage, so the current flowing through the Zener and any series resistors changes.
So if a resistor is connected between the Zener and Ground, the voltage drop across this resistor will change,
and the beautiful clipping is lost.
Well, I supposed you knew that if you add or change R2, you would have to lower R1 by the same value to keep the overall same current which can't be higher than the Zener max current.
R1 & R2 form a voltage divider, inserting a Zener in between creates a reference point (vs. ground). You can place that point whereever you want inside the voltage divider.

- Zener effect happens at 5.1v. For higher voltage it's avalanche effect. There are "Zener" diodes at over 500v.

- Do not care about the 100v Zener at 2.5ma. Use only 1ma and multiply it with a transistor.

- Lower Z is better. The voltage variation is less when the current is varying.

My monoblocks use 3200va each. But only ~300w is used. Then a 400va transfo should be ok.
However, you have no idea how 3200va is better.

It provides the darkest background without making a clinical sound.
When your system is top, adding a darker background is immediately perceived... and appreciated.

_________________
Keep things simple and even simpler

Balance MC-1010 phono stage... in progress (week 2)


Top
 Profile  
Reply with quote  
PostPosted: Fri Oct 12, 2018 6:44 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4105
Location: Ottawa, ON, CA
A few days ago, it suddenly dawned on me :

What if the jitter from the positive clipper could be canceled with the jitter from the negative clipper ?

The variable resistor is adjusted until the clipping waveform on the negative becomes the mirror image
of the positive clipping waveform.
Using the add function on a scope shows when this occurs.

So far the results look very promising.


Attachments:
Balanced Clipping.jpg
Balanced Clipping.jpg [ 96.31 KiB | Viewed 202 times ]
Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 76 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: BinkyTheCat, kenship and 5 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group