What gives with this square wave? - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

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 10th January 2010, 07:01 PM   #11
Yvesm is offline Yvesm  France
diyAudio Member
 
Yvesm's Avatar
 
Join Date: Jul 2004
Location: Ardeche
Quote:
Originally Posted by martyh View Post
Thanks for all of the replies

Just for clarification, I was testing the transformer itself not a whole amp. There was only 10vrms on the primary from a 600 ohm source. My main goal was to find the turns ratio with a sine wave which was easy enough. I was more or less messing around when I switched to square waves and saw the weird spikes at high frequencies.
Too low source impedance !
Use stated load at secondary and a source impedance equal to the reflected load at primary.

So, use a 8 ohms load and a 6K resistor in serie with your .6K source.

As already stated, you may observe different results according the end of the primary you will ground.

Yves.
  Reply With Quote
Old 10th January 2010, 07:44 PM   #12
martyh is offline martyh  United States
diyAudio Member
 
Join Date: Feb 2004
Location: Wisconsin
Quote:
Originally Posted by Yvesm View Post
Too low source impedance !
Use stated load at secondary and a source impedance equal to the reflected load at primary.

So, use a 8 ohms load and a 6K resistor in serie with your .6K source.

As already stated, you may observe different results according the end of the primary you will ground.

Yves.
Yes, I think you are right. From the little I know I would expect the low source impedance to cause problems. I donít know how much I could tell about the transformer testing it at such a low level anyway. I was more curious about the spikes which I had not seen before. I have never tested high level transformers only small signal stuff and I am careful to use driving and loading impedances that match my intended application.
  Reply With Quote
Old 10th January 2010, 09:34 PM   #13
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by rascal101 View Post
I believe that the primary cause is the leakage inductance for this transformer which maybe be a bit high. This means that the coupling in between the windings is not that good or the turns distribution in between layers has not been maximized. To measure the leakage inductance, you will need to short all other windings then with an LCR meter, measure the inductance of the concerned winding.
I did a little work on getting transformer spice models from simple measurements of real transformers, a while back. I am not really sure how well the leakage inductance is modeled, but at least an LCR meter was not required, which was a good thing for those of us who do not have one. However, I was only working with mains power transformers that had either a single primary and secondary or two indentical primaries and two indentical secondaries. So I don't know how applicable the method might be, for output transformers.

It is based on a paper about transformer modelling that is available from Onsemi:

http://www.onsemi.com/pub/Collateral/AN1679-D.PDF

Here is a link where I have the simplified measurement steps and calculations (and the downloadable LTspice transformer-modelling file) posted:

Spice Component and Circuit Modeling and Simulation

Comments on shortcomings or possible improvements would be welcomed.

Cheers,

Tom

Last edited by gootee; 10th January 2010 at 09:46 PM.
  Reply With Quote
Old 10th January 2010, 10:06 PM   #14
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by martyh View Post
Yes, I think you are right. From the little I know I would expect the low source impedance to cause problems. I donít know how much I could tell about the transformer testing it at such a low level anyway. I was more curious about the spikes which I had not seen before. I have never tested high level transformers only small signal stuff and I am careful to use driving and loading impedances that match my intended application.
Actually, the spikes would also be there even if you tried a square wave with a plain inductor/coil/choke, instead of with an inductor that's part of a transformer.

For any inductor, V = L (di/dt), where di/dt is just the rate-of-change of the current. The rising and falling edges of the square waves make the current through the inductor change at a very high rate. So L times that high rate gives a large V, during those times, which shows up as voltage spikes.

Tom
  Reply With Quote
Old 10th January 2010, 10:22 PM   #15
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
can we see what the square wave unloaded looks like?
can we see what the square wave loaded with somethinglike the primary impedance of the xfrmr (an equivalent resistor) looks like?

talking about the generator, not connected to the xfrmr?

fwiw, afaik, driving a higher Z primary with a lower Z source is not an issue, and should not cause the overshoots shown.

_-_-bear
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote
Old 12th January 2010, 01:31 AM   #16
martyh is offline martyh  United States
diyAudio Member
 
Join Date: Feb 2004
Location: Wisconsin
Quote:
Originally Posted by gootee View Post
Actually, the spikes would also be there even if you tried a square wave with a plain inductor/coil/choke, instead of with an inductor that's part of a transformer.

For any inductor, V = L (di/dt), where di/dt is just the rate-of-change of the current. The rising and falling edges of the square waves make the current through the inductor change at a very high rate. So L times that high rate gives a large V, during those times, which shows up as voltage spikes.

Tom
Thanks Tom,

Bear,

The square wave from the generator looks fine none of the spikes or ringing.
  Reply With Quote
Old 12th January 2010, 02:04 AM   #17
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Well, it's not that it's changing the current fast, in fact if you connect a small resistor in series and measure it, you'll see the magnetizing current is an excellent triangle wave. This is because EMF is constrained by the source: as you say, V = L * dI/dt. The triangle wave is the resulting dI/dt slope. Those spikes are most certainly not the result of rapid current change.

A positive spike on the falling edge is some sort of feedforward or nonminimum phase reaction. I wonder if there is parasitic capacitance coupling the falling edge from the inverting side of the transformer through to the output. It could also be a faithful reproduction of crud on the source, which is a good reason why some have asked for the source waveform.

As for meaning, if you want to get uselessly technical, it's simply the convolution of your input signal with the frequency transfer function of the transformer. On a simpler level, there are L's and C's inside and excitation makes them ring. Underdamped resonances ring like a bell, while overdamped resonances thud more like a lead brick.

Of interest to feedback, you want the closed loop combination of amplifier (with its limited frequency response, at HF and LF, both of which are easy to calculate) and transformer (which has many, many poles and zeros, but can be approximated by a few simple poles) to come out stable (no oscillation, little ringing), and better yet, come out better than stable (i.e., overdamped, no overshoot), so it will also tolerate bad loads.

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote
Old 12th January 2010, 06:36 AM   #18
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Quote:
Originally Posted by Sch3mat1c View Post
Well, it's not that it's changing the current fast, in fact if you connect a small resistor in series and measure it, you'll see the magnetizing current is an excellent triangle wave. This is because EMF is constrained by the source: as you say, V = L * dI/dt. The triangle wave is the resulting dI/dt slope. Those spikes are most certainly not the result of rapid current change.
It depends on the resistance of the source, relative to the inductance: The current is only a triangle wave if the resistance is (relatively) very small. And in that case, the voltage becomes a square wave. For a large-enough resistance, the current becomes a square wave and the voltage is spikes.

Quote:
A positive spike on the falling edge is some sort of feedforward or nonminimum phase reaction. I wonder if there is parasitic capacitance coupling the falling edge from the inverting side of the transformer through to the output. It could also be a faithful reproduction of crud on the source, which is a good reason why some have asked for the source waveform.
Yeah, I didn't notice that the spikes on the falling trailing edges were the wrong sign for V=L(di/dt) of the primary's inductance. It seems like something more interesting is going on, there.

Quote:
As for meaning, if you want to get uselessly technical, it's simply the convolution of your input signal with the frequency transfer function of the transformer. On a simpler level, there are L's and C's inside and excitation makes them ring. Underdamped resonances ring like a bell, while overdamped resonances thud more like a lead brick.
Actually, it would be the convolution of the time-domain input signal with the time-domain impulse response of the transformer, not with the frequency-domain transfer function. (I've been out of school for a very long time and had to look it up. But mixing the time and frequency domains just didn't sound right. Plus I remembered the convolution integral mainly being used in the time domain.)

Quote:
Of interest to feedback, you want the closed loop combination of amplifier (with its limited frequency response, at HF and LF, both of which are easy to calculate) and transformer (which has many, many poles and zeros, but can be approximated by a few simple poles) to come out stable (no oscillation, little ringing), and better yet, come out better than stable (i.e., overdamped, no overshoot), so it will also tolerate bad loads.

Tim
I sort-of specialized in Automatic Control Theory. But that was decades ago and it's all somewhat rusted, and is creaky and half-frozen, now. I remember Critical Damping (damping factor = 0.7); fastest response with no overshoot. But that's just Classical control theory. With Modern or Optimal control theory, or whatever the latest nomenclature is, we'd want the control system to "turn it up to 11" until the last possible instant and then stomp really hard, once, on the perfectly-compensated "brakes", as it were, and be spot on without a quiver (in some n-dimensional space, of course).

- Tom

Last edited by gootee; 12th January 2010 at 06:48 AM.
  Reply With Quote
Old 12th January 2010, 11:15 PM   #19
diyAudio Member
 
boywonder's Avatar
 
Join Date: Jan 2008
Location: So.Cal.
Some related questions......

I've noticed that in my own testing (and in Tubelab's post #5 above), that the square wave response is not symmetrical.......why is that? The wiggles on the top of the wave are different than the bottom.

What input/output level should one use for square wave testing?
  Reply With Quote
Old 13th January 2010, 01:18 AM   #20
martyh is offline martyh  United States
diyAudio Member
 
Join Date: Feb 2004
Location: Wisconsin
More shots for the curious. This time I added 820 ohm resistors in series with both legs of the primary and an 8 ohm resistor across the secondary. Top trace is the primary and bottom trace is the secondary. The hf shot is at 20K this time and the lf trace is 100 Hz. The phase shift you see in the lf trace was not there when the primary was driven directly from the 600 ohm source.
Attached Images
File Type: jpg 20kHz.JPG (633.3 KB, 273 views)
File Type: jpg 100Hz.JPG (668.3 KB, 221 views)
  Reply With Quote

Reply


Hide 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
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Sine wave - Square & Triangle wave generator using Transistors / OP-Amps lineup Solid State 20 9th October 2006 01:15 AM
Microphone turns triangle wave into square wave? Circlotron Analogue Source 18 25th January 2003 07:44 AM


New To Site? Need Help?

All times are GMT. The time now is 06:22 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2