PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Benches > Itsu (Moderator: Itsu) > Itsu's workbench / placeholder.
2024-11-26, 05:50:58
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 ... 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 [32] 33 34 35 36 37 38 39
Author Topic: Itsu's workbench / placeholder.  (Read 137480 times)
Topic Options img
img  AlienGrey
Re: Itsu's workbench / placeholder. « Reply #775 on: 2022-01-03, 11:22:53 »
Group: Guest
Quote from: verpies on 2022-01-03, 00:08:56
So if your XY mode can use that integrating math channel as a source for its Y axis, then you can delete the integrating RC network and connect the scope directly to the secondary winding.
See this video.
Why not match the -4Ω amplifier output impedance with the +4Ω impedance of the DUT @ e.g 400 Hz ?
The sine shape does not matter in XY mode.
Personally, I like the symmetrical triangle current waveform flowing through the primary because it creates constant ±dFm/dt in the primary and constant induced ±EMF in the secondary when the core is not saturating.
For me, too.  Maybe Peter is finally working on the SMF code and updating the DB.
Question; it wasn't me who brought S M F up, so what is it?  :( :o
« Last Edit: 2022-01-04, 10:15:48 by AlienGrey »
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #776 on: 2022-01-03, 11:46:26 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: AlienGrey on 2022-01-03, 11:22:53
what is S M F sesin manager function ? single mode function ? :o
Talk to Peterae about it.  Not here.
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #777 on: 2022-01-03, 11:47:25 »

Group: Professor
Hero Member
*****

Posts: 3498
Currently this thread really is not about rotating magnetic fields.
It is about measuring the BH curve.

Quote from: AlienGrey on 2022-01-03, 10:59:22
No on the yoke you already have 2 windings driven by the Mos-Fets you eather need to advance/delay one phase 90 deg
This "advance/delay" is what I meant by driving "at 90º phase".
Also, not all windings are always wound orthogonally to each other, so this needs to be explicitly stated.

Quote from: AlienGrey on 2022-01-03, 10:59:22
D smith does this on his neon inverter the novice always ignores this trick,
Please make a thread about this trick. I will gladly reply to it there.

Quote from: AlienGrey on 2022-01-03, 10:59:22
...or use a four phase clock to make a rotory vortex a TL494 is an inveter driver at 180 deg,
Itsu and I really do not need to build any custom circuits to generate two waveforms which are 90º out of phase, because we have signal generators that do it out of the box.

Quote from: AlienGrey on 2022-01-03, 10:59:22
Also re  the nano pulser has a coax cable to set that up as a telecom eng you first need terminate the cable and tune its length then you need to open circuit it or short circuit it to produce standing waves.
Standing waves in coax cables are generated by CW and imperfect terminations.  Perfect termination will prevent reflections from the end of the cable. No reflections = no standing waves.
Nanopulsers will not generate standing waves even in imperfectly terminated coaxial cables because their Pulse Repetition Frequency (PRF) is too low.

Again, nanopulsers are yet another topic which has nothing to do with measuring the BH curve here.
   
img  picowatt
Re: Itsu's workbench / placeholder. « Reply #778 on: 2022-01-03, 15:24:55 »
Sr. Member
****

Posts: 453
Quote from: Itsu on 2022-01-02, 16:36:12
Yes, that amplifier has 4 Ohm output impedance, and i use a 4.7 Ohm resistor in series with the dut

Itsu,

Are you confusing the amplifier's "output rating" with its "output impedance"?

Most solid state audio amplifiers specified to drive a 4R load have an output impedance 20 to 400 times less than that (which is the amplifier's "damping factor"). 

When used as you are, a resistor between the amp output and load is often used to protect the amplifier and ensure the amplifier output never sees a load less than the 4R the amp is rated to drive.  Although you can eliminate this resistor and drive your load directly from the amp output to take advantage of the very low output impedance of the amplifier, you must ensure you stay well within the amp's max power rating and reactive load drive capability (SOA).  Loads with a complex impedance, particularly capacitive loads, can cause some amp's to oscillate/destroy themselves.  The resistor between amp out and the load is used to limit the load impedance and prevent the amp from having a bad day.   

PW
   
img  AlienGrey
Re: Itsu's workbench / placeholder. « Reply #779 on: 2022-01-03, 15:48:11 »
Group: Guest
Itsu Hi, and thank you Verpies detailing coax and it’s use in comms and function, Therefore; Itsu could you please delete my recent posts of this particular subject or thread as I don’t want to disrupt your sequence of posting and I’m aware and note Verpies comments as I already know this as telecom TX lines was one of the subjects covered in my youth, but thanks for the refresh !
 ^-^ O0

sil



« Last Edit: 2022-01-04, 10:34:44 by AlienGrey »
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #780 on: 2022-01-03, 15:52:16 »

Group: Moderator
Hero Member
*****

Posts: 4159

verpies,

Quote
So if your XY mode can use that integrating math channel as a source for its Y axis, then you can delete the integrating RC network and connect the scope directly to the secondary winding.
See this video.

Nice video, which cleared up some things for me.
I will try the integrating math setup to see if it works on my scope as well, but my main goal is to "see" any difference in BH curve in my presently used "yoke U" (which is not working for me), and any of the other yokes.
Looks like the "Yoke U" and "yoke glued" are similar and the "yoke R" gives a smaller BH curve, so i think i will use that yoke as new yoke in my Ruslan setup. 


Quote
Why not match the -4Ω amplifier output impedance with the +4Ω impedance of the DUT @ e.g 400 Hz ?

Why did i not think of that, ok,  i can try upping the frequency to 400Hz for the "Yoke U"
For the "Yoke R" which measures 633uH i need 1KHz to get a 4 Ohm impedance

Quote
The sine shape does not matter in XY mode.
Personally, I like the symmetrical triangle current waveform flowing through the primary because it creates constant ±dFm/dt in the primary and constant induced ±EMF in the secondary when the core is not saturating.

Good to know, so i can push some more current in the yokes.


Itsu
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #781 on: 2022-01-03, 16:00:09 »

Group: Moderator
Hero Member
*****

Posts: 4159
Quote from: picowatt on 2022-01-03, 15:24:55
Itsu,

Are you confusing the amplifier's "output rating" with its "output impedance"?

Most solid state audio amplifiers specified to drive a 4R load have an output impedance 20 to 400 times less than that (which is the amplifier's "damping factor"). 

When used as you are, a resistor between the amp output and load is often used to protect the amplifier and ensure the amplifier output never sees a load less than the 4R the amp is rated to drive.  Although you can eliminate this resistor and drive your load directly from the amp output to take advantage of the very low output impedance of the amplifier, you must ensure you stay well within the amp's max power rating and reactive load drive capability (SOA).  Loads with a complex impedance, particularly capacitive loads, can cause some amp's to oscillate/destroy themselves.  The resistor between amp out and the load is used to limit the load impedance and prevent the amp from having a bad day.   

PW

Hi Picowatt,

I don't think i confuse the amplifier's "output rating" with its "output impedance".

The amp has 2 channels capable of driving a 4 Ohm load, so i guess it will have 4 Ohm output impedance  (https://www.caraudio.com/threads/us-amps-xterminator-xt800-2.560707/)

Anyway, i followed the setup's as mentioned in my post #765 above link where this was mentioned:

https://meettechniek.info/passive/magnetic-hysteresis.html

specially cores with a low permeability need a large number of turns.
To reduce this large number of turns the drive signal can be boosted with an audio amplifier.
So a larger drive current is available.
Connect in series with the amplifier a power resistor with a value of 4 Ω to protect the amplifier.

 
So yes, "The resistor between amp out and the load is used to limit the load impedance and prevent the amp from having a bad day".


Itsu
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #782 on: 2022-01-03, 16:02:43 »

Group: Moderator
Hero Member
*****

Posts: 4159
Quote from: AlienGrey on 2022-01-03, 15:48:11
Itsu Hi, can you please delete my posts of this thread as i don’t want to disrupt your sequence of posting  And i'm aware of your comments already I trained on telecom tx lines in my youth thanks  O0

sil

AG,    you should be able to remove your own posts, see lower right part of that post,  thanks.

Itsu
   
img  picowatt
Re: Itsu's workbench / placeholder. « Reply #783 on: 2022-01-03, 16:17:37 »
Sr. Member
****

Posts: 453
Quote from: Itsu on 2022-01-03, 16:00:09
I don't think i confuse the amplifier's "output rating" with its "output impedance".

The amp has 2 channels capable of driving a 4 Ohm load, so i guess it will have 4 Ohm output impedance  (rated for 800W each).

Itsu,

It does indeed sound like you are confusing the amplifier's "output rating" with its "output impedance".  If the amplifier's spec sheet provides the damping factor while driving a 4R load, you can divide the 4R by the damping factor to get an approximation of the amp's output impedance. 

An amplifier rated for 800 watts into 4 ohms is very likely going to have an output impedance of around .05 to .01 ohms.

PW
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #784 on: 2022-01-03, 16:30:20 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: picowatt on 2022-01-03, 16:17:37
If the amplifier's spec sheet provides the damping factor while driving a 4R load, you can divide the 4R by the damping factor to get an approximation of the amp's output impedance. 
Good point!  I did not think about that.
That's actually good news because the lower the amp's output impedance the better.
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #785 on: 2022-01-03, 16:34:33 »

Group: Moderator
Hero Member
*****

Posts: 4159

Damping factor seems to be 250, see below (from manual pdf attached)

So does that mean 4R / 250 =  0.016 Ohm?
   
img  picowatt
Re: Itsu's workbench / placeholder. « Reply #786 on: 2022-01-03, 16:53:56 »
Sr. Member
****

Posts: 453
Quote from: Itsu on 2022-01-03, 16:34:33
So does that mean 4R / 250 =  0.016 Ohm?

Yes...

Just be aware that some amps become unstable when driving loads/complex loads with an impedance lower than the amp is specified to drive (particularly capacitive loads).

The data sheet says your amp can drive 100 watts per channel into 4R and 200 watts per channel into 2R.  The data sheet also says the amp can drive 1R, but does not state the max power level at that load impedance.

You can probably get away with using a .25R to 1R resistor in series with your load to protect the amp, although that .25R to 1R resistance effectively becomes the output impedance as seen by the load.

PW
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #787 on: 2022-01-03, 18:21:19 »

Group: Professor
Hero Member
*****

Posts: 3498
Also, a digital amplifier (half bridge or full-H bridge) is easily realizable with good modern MOSFETS and can cheaply be made more powerful than an analog amplifier.

Such amplifiers are digital voltage sources, meaning that they apply a rectangular voltage waveform to the load.
This is quite a good match when the load is purely inductive because it results in triangular current flowing through the inductor when it is not saturating.

The rising edge of this current is depicted on the diagram below:



The downward curving of the blue current waveform is an indication that the driving frequency is too low (this should be avoided as it affects the measurement accuracy).

The upward curving of the red current waveform is an indication of core saturation but it does not need to be avoided in the XY mode as it does not affect the shape of the BH curve.  It only presents overcurrent damage danger to the driving MOSFETs.  It can be mitigated with the typical overcurrent protection techniques.
« Last Edit: 2022-01-03, 19:30:35 by verpies »
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #788 on: 2022-01-03, 20:29:22 »

Group: Moderator
Hero Member
*****

Posts: 4159
Quote from: picowatt on 2022-01-03, 16:53:56
Yes...

Just be aware that some amps become unstable when driving loads/complex loads with an impedance lower than the amp is specified to drive (particularly capacitive loads).

The data sheet says your amp can drive 100 watts per channel into 4R and 200 watts per channel into 2R.  The data sheet also says the amp can drive 1R, but does not state the max power level at that load impedance.

You can probably get away with using a .25R to 1R resistor in series with your load to protect the amp, although that .25R to 1R resistance effectively becomes the output impedance as seen by the load.

PW


Thanks PW,

i think i will keep 0.25 Ohm or so resistor in series just to be sure.

Itsu
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #789 on: 2022-01-03, 20:37:44 »

Group: Moderator
Hero Member
*****

Posts: 4159
Quote from: verpies on 2022-01-03, 18:21:19
Also, a digital amplifier (half bridge or full-H bridge) is easily realizable with good modern MOSFETS and can cheaply be made more powerful than an analog amplifier.

Such amplifiers are digital voltage sources, meaning that they apply a rectangular voltage waveform to the load.
This is quite a good match when the load is purely inductive because it results in triangular current flowing through the inductor when it is not saturating.

The rising edge of this current is depicted on the diagram below:



The downward curving of the blue current waveform is an indication that the driving frequency is too low (this should be avoided as it affects the measurement accuracy).

The upward curving of the red current waveform is an indication of core saturation but it does not need to be avoided in the XY mode as it does not affect the shape of the BH curve.  It only presents overcurrent damage danger to the driving MOSFETs.  It can be mitigated with the typical overcurrent protection techniques.



thanks,

lots of digital D-class amplifiers with H-bridge designs on the web indeed.

Itsu
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #790 on: 2022-01-03, 21:03:41 »

Group: Moderator
Hero Member
*****

Posts: 4159

I tried to use that integrating math channel "Intg(CH3 + CH4)" as a source for the Y axis but by scope won't accept that in XY mode  :(

Itsu
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #791 on: 2022-01-03, 21:13:05 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Itsu on 2022-01-03, 21:03:41
I tried to use that integrating math channel "Intg(CH3 + CH4)" as a source for the Y axis but by scope won't accept that in XY mode  :(
It is a scope limitation. If it is caused by shortsighted firmware then it still could be fixed.
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #792 on: 2022-01-03, 21:44:57 »

Group: Moderator
Hero Member
*****

Posts: 4159
I am on the latest FW for this scope, so i guess its a limitation of the scope.



Using the "yoke R" (633uH @ 200Hz is 0.795 ohms) without any series resistor at 200Hz shows the below BH curve.
Input into the audio amp was 12.3V @ 3.2A.

This BH curve is the smallest of my 3 yokes and way smaller then the BH curve of the till now used "yoke U".
This means to me that the coercive force is small pointing to a softer ferromagnetic material then the "yoke U" ferrite, and thus lower losses.

Itsu
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #793 on: 2022-01-03, 23:47:29 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Itsu on 2022-01-03, 21:44:57
This means to me that the coercive force is small pointing to a softer ferromagnetic material then the "yoke U" ferrite, and thus lower losses.
Yes, at least at this frequency.
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #794 on: 2022-01-04, 10:20:16 »

Group: Moderator
Hero Member
*****

Posts: 4159

I will see if i can increase the frequency to working (24KHz) frequency level.

Itsu
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #795 on: 2022-01-04, 15:00:40 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Itsu on 2022-01-04, 10:20:16
I will see if i can increase the frequency to working (24KHz) frequency level.
Since it takes time for the current in an inductor to rise, at 120x higher frequency, the current will reach only 1/120 of the previous current level, which most likely will not be enough time to reach the ferrite saturation level ...and the BH curve will appear incomplete. 
Of course this lack of time can be compensated with higher voltage at the amplifier's output but I don't think you can alter that in your analog amp.

It is useful to have a HV digital amplifier in a lab for this and many other purposes.
Yes, a digital amp can output only rectangular voltage waveforms but it can do this at high frequency, voltage and current, when it is based on the modern SiC MOSFETs (like your NVH4L160N120SC1) and gate drivers. 
I find the Full Bridge topology most convenient to use (albeit most difficult to build) because it requires only one power supply rail (@ ½V of HB) and has only two output terminals (not requiring center-tapped windings in the load).
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #796 on: 2022-01-04, 15:56:02 »

Group: Moderator
Hero Member
*****

Posts: 4159

Quote
Since it takes time for the current in an inductor to rise, at 120x higher frequency, the current will reach only 1/120 of the previous current level, which most likely will not be enough time to reach the ferrite saturation level ...and the BH curve will appear incomplete.
Of course this lack of time can be compensated with higher voltage at the amplifier's output but I don't think you can alter that in your analog amp.

Right, at about 6KHz i loose the ability to show a decent BH curve.
The reactance of the 633uH coil is then already 24 Ohm.
At 24Khz that would be 96 Ohm.

So i guess i have to go blind on the 24KHz with this "Yoke R".

Quote
It is useful to have a HV digital amplifier in a lab for this and many other purposes.
Yes, a digital amp can output only rectangular voltage waveforms but it can do this at high frequency, voltage and current, when it is based on the modern SiC MOSFETs (like your NVH4L160N120SC1) and gate drivers.
I find the Full Bridge topology most convenient to use (albeit most difficult to build) because it requires only one power supply rail (@ ½V of HB) and has only two output terminals (not requiring center-tapped windings in the load).

I was working on a full H-bridge design here:  https://www.overunityresearch.com/index.php?topic=3319.msg57288#msg57288 and found it really challenging.
But perhaps i have to revisit that and start looking for a circuit for a HV digital amplifier as, as you said, its useful to have.

Itsu
   
img  picowatt
Re: Itsu's workbench / placeholder. « Reply #797 on: 2022-01-04, 16:23:19 »
Sr. Member
****

Posts: 453
Quote from: Itsu on 2022-01-04, 10:20:16
I will see if i can increase the frequency to working (24KHz) frequency level.

Itsu

Itsu,

At higher frequencies, to protect the amplifier, I suggest increasing the value of the resistor in series between amp and load.  Monitor the output to ensure the amplifier is content driving your test load.  Avoid any signs of unwanted oscillations, clipping, slew rate limiting or tripping any current limiting.  Some audio amps do not like operating near full power at half their maximum bandwidth.   

Also note that published spec's can often be a bit optimistic...

PW     
   
img  verpies
Re: Itsu's workbench / placeholder. « Reply #798 on: 2022-01-04, 16:28:38 »

Group: Professor
Hero Member
*****

Posts: 3498
Quote from: Itsu on 2022-01-04, 15:56:02
So i guess i have to go blind on the 24KHz with this "Yoke R".
Yes, without increased output voltage to push the current into the inductor very quickly you are limited to low frequencies for the BH curve.

Quote from: Itsu on 2022-01-04, 15:56:02
I was working on a full H-bridge design here:  https://www.overunityresearch.com/index.php?topic=3319.msg57288#msg57288 and found it really challenging.
But perhaps i have to revisit that and start looking for a circuit for a HV digital amplifier as, as you said, its useful to have.
Because you took the hard way.

The easy way is to buy several pre-made isolated ±15V DC-DC converters to power the gates and use two isolated gate drivers (...or digital isolators + normal gate drivers) for the high side MOSFETs. 
They are available now from Mouser, Digikey, RSonline, etc... but were not 8 years ago.

https://artesyncom-prod.scdn8.secure.raxcdn.com/assets/w1000/2dbc5050b6.png
Itsu's workbench / placeholder.
   
img  Itsu
Re: Itsu's workbench / placeholder. « Reply #799 on: 2022-01-04, 16:50:35 »

Group: Moderator
Hero Member
*****

Posts: 4159
Quote from: picowatt on 2022-01-04, 16:23:19
Itsu,

At higher frequencies, to protect the amplifier, I suggest increasing the value of the resistor in series between amp and load.  Monitor the output to ensure the amplifier is content driving your test load.  Avoid any signs of unwanted oscillations, clipping, slew rate limiting or tripping any current limiting.  Some audio amps do not like operating near full power at half their maximum bandwidth.   

Also note that published spec's can often be a bit optimistic...

PW     

PW,

I did notice some oscillations and current limiting effects at 24KHz, so indeed, its not happy with that load at that frequency.


Itsu
   
Pages: 1 ... 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 [32] 33 34 35 36 37 38 39
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 05:50:58