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

Software Tools SPICE, PCB CAD, speaker design and measurement software, calculators

LTSpice model for a commercial toroidal output transformer
LTSpice model for a commercial toroidal output transformer
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 12th July 2018, 02:33 PM   #71
dch53 is offline dch53  Australia
diyAudio Member
 
Join Date: Jun 2015
Location: Perth, Western Australia
Thanks very much for this. More I've learned!

However, I am unable to plot impedance. Expression is correct but still showing level and phase. I must have missed something.

Referring to your plot, it appears that Z decreases as Lcfb increases. Is that correct?
Attached Images
File Type: png Plot.png (45.6 KB, 56 views)
  Reply With Quote
Old 12th July 2018, 02:52 PM   #72
Koonw is offline Koonw
diyAudio Member
 
Join Date: Apr 2013
Yes, as Lcfb increased, Nfb increased so the Ra decreased. You can insert the current source in between P-1 and P-2 to get full Ra-a. In this case there is no voltage difference between P-1 and P-2 so the block cap is not needed and gives you more accurate result.

To display impedance level in ohms, on the waveform window, right click to bring up popup manual, click on the "Manual Limits", on "Left Vertical Axis", select "Linear"

To turn off the phase plot, left click on right axis (phase) to bring up popup manual, click on "Don't Plot Phase".
Attached Images
File Type: png P-P xformer effective Z test-2.png (106.7 KB, 54 views)

Last edited by Koonw; 12th July 2018 at 02:58 PM.
  Reply With Quote
Old 12th July 2018, 03:00 PM   #73
dch53 is offline dch53  Australia
diyAudio Member
 
Join Date: Jun 2015
Location: Perth, Western Australia
Got it. Thanks again! That is really clever. I even worked out how to add the labels.

Quote:
Originally Posted by Koonw View Post
Yes, as Lcfb increased, Nfb increased so the Ra decreased. You can insert the current source in between P-1 and P-2 to get full Ra-a. In this case there is no voltage difference between P-1 and P-2 so the block cap is not needed and gives you more accurate result.
Plot of full Ra-a attached.

Earlier you said "Since only the primary Z is measured, the speaker load 4/8 ohms should disconnected or connected for overall result."

I thought the primary Z is supposed to be 4K (in this case) WITH the 8Ω load attached. I see you have R2 set to 80K. I'm obviously misunderstanding something here.

Quote:
Originally Posted by 45 View Post
So if you want to see the effect of cathode feedback and you want to use the Polish 4K transformer with CFb then you can't optmize for 4K without CBf and assume that it will be fine with CFb on! So when you run the simulation without CBf you have to use ANOTHER transformer that has got all those cathode winding turns at the plate so that the primary load is 7K. Otherwise you are comparing two amplifiers where very likely one is optimised in the way you want (for example max power with minimum distortion) and the other won't be! Instead if you compare 7K standard transformer with 4K+24% CBf the operative conditions of the output stage stay the same. The point is that 24% CFb is a lot.
Thanks. I had to read this a few times but it makes good sense.
Attached Images
File Type: png Full Ra-a.png (39.6 KB, 43 views)
  Reply With Quote
Old 13th July 2018, 03:37 AM   #74
Koonw is offline Koonw
diyAudio Member
 
Join Date: Apr 2013
Quote:
Originally Posted by dch53 View Post
Plot of full Ra-a attached.

Earlier you said "Since only the primary Z is measured, the speaker load 4/8 ohms should disconnected or connected for overall result."

I thought the primary Z is supposed to be 4K (in this case) WITH the 8Ω load attached. I see you have R2 set to 80K. I'm obviously misunderstanding something here.
If the load is connected you can see overall result which the is result of both current source and load in parallel. So to measure the output impedance of secondary the load is also set to very high or disconnected.

If connected it's the overall result of both in parallel (sorry for not clarify earlier).

But the thing is that the secondary output Z measured is about 16 Ohms instead of 8, with 0H Lcfb, what gives?

PS: So if we connect 16 Ohms to secondary, and then pri Z measured again between P-1 and P-2 is 2k (half of 4k measured earlier), that is result in parallel of current source and (reflected) load.
Attached Images
File Type: png P-P xformer output Z test-3.png (109.3 KB, 39 views)
File Type: png P-P xformer output Z test-4.png (108.6 KB, 12 views)

Last edited by Koonw; 13th July 2018 at 03:59 AM.
  Reply With Quote
Old 13th July 2018, 03:52 AM   #75
dch53 is offline dch53  Australia
diyAudio Member
 
Join Date: Jun 2015
Location: Perth, Western Australia
Quote:
Originally Posted by Koonw View Post
If the load is connected you can see overall result which the is result of both current source and load in parallel. So to measure the output impedance of secondary the load is also set to very high or disconnected.

If connected it's the overall result of both in parallel (sorry for not clarify earlier).

But the thing is that the secondary output Z measured is about 16 Ohms instead of 8, with 0H Lcfb, what gives?
Ahh. It makes sense now.

Quote:
Originally Posted by Koonw View Post
But the thing is that the secondary output Z measured is about 16 Ohms instead of 8, with 0H Lcfb, what gives?
What indeed! I hope you're not asking me! I'll be interested to know the answer to that.

So, you could also plot damping factor.
  Reply With Quote
Old 13th July 2018, 07:32 AM   #76
Koonw is offline Koonw
diyAudio Member
 
Join Date: Apr 2013
Well, there is nothing wrong as it per design for now, but you can easy bring the output Z down with additional Nfb using Lcfb or Gnfb, just to be aware of the actual output Z to use for accurate measurement. I'll post the damping factor plot later.
  Reply With Quote
Old 15th July 2018, 08:20 AM   #77
dch53 is offline dch53  Australia
diyAudio Member
 
Join Date: Jun 2015
Location: Perth, Western Australia
Default Simulation with the VDV4070CFB Transformer

Schematic is shown in the first attachment. THD vs Power is shown in the second attachment.

THD at 50W is estimated as 3%. The MOSFET screen and output anode traces in the third attachment look pretty awful!

koonw’s output stage simulation suggests 0.5% THD at 50W. Maybe this is a result of the perfect signal sources in that simulation.

THD at the MOSFET sources deteriorates rapidly after 50W output due to clipping of the negative cycles (3rd attachment). I can fix that by changing the MOSFET negative supply from -150V to -200V but that doesn't have much impact on the output stage because the 6550s are turned off then anyway.

Menno van der Veen reports 50W for this configuration but doesn’t give any THD data.

Any suggestions how I can improve the performance of the sim or the circuit?
Attached Images
File Type: png full vdv sim.png (86.3 KB, 13 views)
File Type: png THD vs Power.png (57.8 KB, 15 views)
File Type: png MOSFET Source and anode at 50W output.png (41.8 KB, 14 views)
  Reply With Quote
Old 15th July 2018, 11:49 AM   #78
Koonw is offline Koonw
diyAudio Member
 
Join Date: Apr 2013
Try to increase K factor from 0.9995 to 0.99995, as you probably have too much leakage inductance as shown by the ringing top waveform, esp. at higher output.
  Reply With Quote
Old 15th July 2018, 02:18 PM   #79
dch53 is offline dch53  Australia
diyAudio Member
 
Join Date: Jun 2015
Location: Perth, Western Australia
Default Leakage inductance

Thanks. That made a huge difference! THD < 1.2% up to 80W. the simulation even shows it dropping to 0.9% at 90W. Output tube dissipation at that point is still only 25W.

I'll read up on modelling the leakage inductance. I've read that toroidal transformers have lower leakage inductance than traditional EI cores. I'll also do deeper into Norman Koren's The Emperor's New Amplifier.
Attached Images
File Type: png THD vs Power K 0.99995.png (64.3 KB, 28 views)
  Reply With Quote
Old 15th July 2018, 04:57 PM   #80
euro21 is offline euro21  Hungary
diyAudio Member
 
Join Date: Sep 2004
Location: Budapest
LTSpice model for a commercial toroidal output transformer
Quote:
Originally Posted by dch53 View Post
THD at the MOSFET sources deteriorates rapidly after 50W output due to clipping of the negative cycles (3rd attachment).
Before you changing anything, look at output tube G1 waveform in the simulator.
About +30V.... -150V swing.

Then look in 6550 datasheet the UL characteristic: 0V... -70V swing.
No comment.
  Reply With Quote

Reply


LTSpice model for a commercial toroidal output transformerHide 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
Homebrew Toroidal Output Transformer Osvaldo de Banfield Tubes / Valves 77 12th March 2017 04:00 PM
Toroidal Output Transformer FullRangeMan Tubes / Valves 11 2nd March 2011 05:10 AM
LTspice model for Dynaco A470 Output Transfomer Ray Waters Tubes / Valves 5 17th February 2011 03:33 PM
Output Tube biasing for Toroidal Output Transformer gingertube Tubes / Valves 7 17th April 2005 04:40 PM
toroidal SE output transformer noyan Parts 10 11th July 2004 10:41 PM


New To Site? Need Help?

All times are GMT. The time now is 06:43 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki