Frequency Response of Output Tubes - 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 12th March 2012, 09:21 PM   #11
diyAudio Member
 
richwalters's Avatar
 
Join Date: Sep 2003
Location: Alps:Tube amp designs over 150W, SMPS guru.
Quote:
Originally Posted by Kay Pirinha View Post
Or vice versa ?

I don't think that output tubes, or any tube at all, will produce harmonic distortion in the audio frequency range. But, as told by the others before, non-harmonic distortion is different between the tube types. And this means different sonic perception by the listener or the musician.

Best regards!
Good one.... with respect to SY's comment,"1953-vintage Achtungsbieremesse NOS EL84".
Hav e you tried it ? same flag of convenience.

richy
  Reply With Quote
Old 12th March 2012, 09:30 PM   #12
Account disabled at member's request
 
Join Date: Jul 2009
Quote:
Originally Posted by tubelab.com View Post

I have been working on a monster sized breadboard that uses Plitron toroidal OPT's in P-P. The OPT's are rated for 400 watts at 20 Hz.
So in the end you are going down the large tube - enormous OPT road yourself?

I probably will have to go for a 2000va core OPT. Reduced frequency response but should still adequately cover the 20 hz 20 khz range.
  Reply With Quote
Old 12th March 2012, 10:19 PM   #13
diyAudio Member
 
Join Date: Oct 2011
Location: Tauberbischofsheim, Germany
Quote:
Originally Posted by DF96 View Post
I suspect this is not what you intended to say.
You're right, thank you. I meant linear distortion vs. non-linear distortion.

Best regards!
  Reply With Quote
Old 12th March 2012, 10:22 PM   #14
diyAudio Member
 
richwalters's Avatar
 
Join Date: Sep 2003
Location: Alps:Tube amp designs over 150W, SMPS guru.
My recent design; I got around that single core problem by using 2 o/p trannies in bridge mode, currently evaluating 500W using 8x KT90's at 600VB+...on paper could have gone higher to hit 750W but that's pushing it above 10Khz. Expect considerable F3 harmonic at this frequency. Alot of energy is being sapped up by snubbers as the transformer leakage parasitics make hay and also the PSU which has to find 1.6A ar 600V and be stabilised. I use two coupling caps each p-p side....no problem with instability.
Big drawback....with mine.. grunt weight 100Lbs. I can handle this but others need a fork lift truck.
Next time I do this I will use triple p-p-p...far easier.
GL'k with your design.

richy
  Reply With Quote
Old 13th March 2012, 02:58 AM   #15
diyAudio Member
 
tubelab.com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
Quote:
So in the end you are going down the large tube - enormous OPT road yourself?
I have been following your monster build, but reality has dictated that I will likely never go down the big transmitting tube road.

Several years ago I paid a well known transformer winder to make me a pair of OPT's for a big SET amp. I got one transformer for testing and it was not suitable for HiFi. I had the power supply from a big transmitter (1500 volts at 1/2 amp) and several 833A's so I rigged up a big SET amp that took up the whole workbench. It made an absolutely wicked guitar amp cranking out about 200 Watts from 50 Hz to about 12 KHz. I annoyed my neighbors for about a week then took it all apart. Many of the parts have been sold or traded away withn the rest going soon.

http://www.tubelab.com/833SE.htm

I haven't caught up on email and PM's yet, but I never had any suitable anode caps for the big triodes. Look carefully at the picture and see that vice grips are used for the filament, and hose clamps for the plate and grid.

There was a thread several years ago dealing with a bunch of paralleled 807's and someone mentioned some big Plitron OPT's for sale on Plitrons surplus page, so I immediately bought a pair even though I didn't have a use in mind. These OPT's are rated for 400 watts at 20 Hz and I have personally cranked 525 watts through them without issue. I have no need for a 400+ WPC amp, but sooner or later I will build something that fits the OPT's.

In the mean time they have been used for testing and learning. They are rated for 1250 ohms plate to plate with 2, 4, and 8 ohm secondary taps. They also work quite well at 2500 and 5000 ohms. To make big power you need a lot of plate voltage and a high impedance load OR you need a lot of plate current and a low impedance load. It is usually easier to make a good OPT at a lower impedance ratio (and I already have them) so I choose to explore the low impedance route.

Experiments so far with conventional G1 drive and TV sweep tubes have been limited by my 650 volt 1.7 amp power supply. That will still make 250 WPC from a pair of 35LR6's into a 2500 ohm load. The new breadboard will let me explore screen drive, combined screen and control grid drive, and allow the use of multiple output tubes per channel. Eventually I will build a big amp, but I am not in a big hurry, and I am still learning new things with these OPT's.

I made the statement in an earlier post that with these monster OPT's the tube choice didn't influence the frequency response much. This is true, but there is a flip side to this equation. If the OPT's are junk the tubes are a big influence on the frequency response, the distortion, and the power output.

About 15 years ago I purchased a large quantity of OPT's that were designed for guitar amp use. They were rated for 80VA from 80 Hz to 4 KHz, 6600 ohms to 4, 8, and 16 ohms. Dissection revealed no interleaving, a single half primary winding, the entire secondary, followed by the second half primary wound with thicker wire to help equalize the DC resistances. The laminations look like those from a power transformer. I used them for their intended purpose making guitar amps from 15 to 50 watts using 6L6GC, EL34, KT88 and EL84. I noticed that even though the power amp circuits were very similar the bigger tubes had better frequency response at both ends of the spectrum.

Fast forward a few years to a dumb blonde experiment. I made an amp using these OPT's and 300B's. I dubbed it the 300Beast. It was a fully differential design using zero feedback. It remains in class A up to about 15 watts and starts to clip at 25 watts. Zero saturation or OPT related distortions other than reduced power output capability above 10 KHz. Surprize, this was the most dynamic sounding amp that I have ever built. I have tried to upgrade it several times including trying some expensive OPT's including a pair of UTC LS57's but the junk still rules. That amp died a few years ago when one of the electrolytics shorted, but it will be rebuilt and live again someday.

I have since used the same OPT's in several HiFi amps wired for 3300 or 6600 ohms. I find that the performance is directly related to the driving tube. In a Simple P-P with 6BQ5 / EL84's on 6600 ohms and 350 volts I get about 20 watts with saturation effects seen below 40 Hz. The same amp with 6CW5 / EL86 on 3300 ohms and 235 volts can also do about 20 watts but saturation is not seen until 20 Hz. 6 db of GNFB is used in both cases.

Pete Millett has designed a sweep tube amp that makes 18 WPC. I made a few modifications to increase the power a bit. I used the same OPT's at 3300 ohms with some 24 watt sweep tubes (6HJ5's) and 600 volts to extract 125 WPC at 1KHz. No GNFB was used, but Schade feedback was applied from the output plates to the driver plates. This amp can crank 80 WPC at at 40 Hz without saturation effects. At 10 watts the frequency response is 18 Hz to 21 KHz +/- 3db. It is also a very dynamic good sounding amp, even when driven to ear splitting volume through 15 inch coaxials with 96 db sensitivity. The little EL84 amp will exhibit audible saturation on strong bass (techno) before it clips.
__________________
Too much power is almost enough! Turn it up till it explodes - then back up just a little.
  Reply With Quote
Old 13th March 2012, 03:12 AM   #16
Account disabled at member's request
 
Join Date: Jul 2009
very informative post!

I just have a question. Frequency response is one of the parameters on which OPT suitability is assessed. I have read from several sources that regardless of performance per frequency, the true HIFI transformers range well above the 20khz mark and infact a specific target of 200khz is considered the mark of HIFI.

I have never understood this: human earing is theoretically limited to an absolute maximum of 20khz with other senses perceiving higher frequencies. Given identical khz to khz performance...what use is there for higher frequency response?

I know it is a dumb question but somehow I think it is related to the issue of frequency and frequency response
  Reply With Quote
Old 13th March 2012, 03:12 AM   #17
diyAudio Member
 
astouffer's Avatar
 
Join Date: Jun 2004
Location: Pittsburgh, crumbling wasteland
Quote:
Originally Posted by tomchr View Post
I actually put my OPT's on an HP 4194A impedance analyzer and had it calculate the parasitic components for me. Using those values in LTspice, the simulation predicted the LF and HF poles to within a few percent of the values I measured on the finished amp.

~Tom
Can you elaborate a little more on how you did this? I have a HP 4192A analyzer at work and have run sweeps on a few transformers. One transformer from an old hifi that used a quad of 6V6s looked pretty bad, only had midrange.
  Reply With Quote
Old 13th March 2012, 03:27 AM   #18
6J6 is offline 6J6  Australia
diyAudio Member
 
Join Date: Feb 2012
Tubelab, that 833 amp is totally bonkers! Love it.
  Reply With Quote
Old 13th March 2012, 03:51 AM   #19
diyAudio Member
 
tubelab.com's Avatar
 
Join Date: Jul 2005
Location: West Virginia panhandle
Quote:
the true HIFI transformers range well above the 20khz mark and infact a specific target of 200khz is considered the mark of HIFI
There is one thing that I am certain of. You wont get a monster OPT that goes to 200KHz. There are very few small OPT's that go to 200KHz. As the OPT gets bigger the leakage inductance and the interwinding capacitance will increase. Together these can create a notch that moves closer to the audio range as the OPT gets bigger. The big OPT in the 833SE amp has a severe notch just off the right side of the frequency response plot at about 55 KHz. It also has a visible smaller notch at 30 KHz. together these two team up to cause a 6 db drop in response at 20 KHz.

There are some extremists that believe you need a 200KHz OPT in order to hear a 20 KHz square wave. This assumes that you have hearing at 200 KHz. I have really bad hearing with constant tinitus. At high volume I can hear a difference between a sine wave, triangle wave, and a square wave at 15 KHz.

In reality you do need response beyond 20 KHz in order to guarantee flat response to 20 KHz without significant phase shift. You also need good low frequency response for the same reason and these requirements become more severe if you intend to apply feedback from the OPT secondary.

If feedback from the secondary is a possibility, you need to minimize the number of frequency response affecting elements inside the feedback loop. This includes the OPT, all coupling caps, and some bypass caps.

At big power levels you are going to have to work around whatever your OPT can give you. You may need to apply some feedback to improve it's response.

As richwalters has indicated sometimes multiple small OPT's each being fed by its own set of tubes makes building a big amp easier. I have wired as many as 4 of the guitar amp OPT's together with 8 tubes.
__________________
Too much power is almost enough! Turn it up till it explodes - then back up just a little.
  Reply With Quote
Old 13th March 2012, 07:41 PM   #20
tomchr is offline tomchr  United States
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Greater Seattle Area
Quote:
Originally Posted by astouffer View Post
Can you elaborate a little more on how you did this? I have a HP 4192A analyzer at work and have run sweeps on a few transformers.
Any transformer can be modeled as shown in attached image. All components have been transformed to the primary. From a modeling perspective, it doesn't really matter if the parasitics are on the primary or secondary side, and when measured, they're all lumped together on the primary anyway.

Rs is the DC resistance of the windings.
Ls is the leakage inductance - basically, the inductance of the primary, secondary windings had the core been an air core.
Lm is the magnetizing inductance
Cp is the inter-winding parasitic cap.
The transformer on the right of the model is an ideal transformer. I.e. K = 1.0 and 1:n turns ratio.

n can be measured by counting the turns (if you're building your own tranny) or by applying an AC voltage to the primary and measuring the secondary voltage (CAREFUL!!). The turns ratio, n, is then Vsecondary/Vprimary.

Measure the impedance of the primary circuit with the secondary shorted. This yields Rs and Ls (Cp, Lm are shorted out by the ideal transformer and the short on the secondary).

Measure the impedance of the primary circuit with the secondary open (or loaded by a very high impedance - 10 MOhm for example). This gives you Cp and Lm (in series with Rs, Ls but as Z(Lm) >> Z(Ls+Rs) you mostly measure Lm||Cp).

On the 4194A, I use the Equivalent Circuits function (MORE MENUS -> EQV CKT). Circuit B ((Ls+Rs)||Cp) seems to work pretty well for these measurements.

~Tom
Attached Images
File Type: png TransformerModel.PNG (114.3 KB, 81 views)
__________________
Modulus-86: Composite amp achieving 0.00018 % THD. Damn Good 300B, Novar Spud, 21st Century Maida Reg., Filament Reg., etc.
Neurochrome : : Audio - http://www.neurochrome.com/audio - Engineering : : Done : : Right
  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
in-room frequency response & transient response MCPete Multi-Way 1 8th October 2011 07:31 AM
Influence of power amplifier output coil to frequency response PMA Solid State 22 6th May 2011 08:17 AM
Frequency response vs. output power MikeW Pass Labs 14 31st January 2007 01:21 AM
Output transformers and frequency response dsavitsk Tubes / Valves 21 3rd December 2006 07:48 PM


New To Site? Need Help?

All times are GMT. The time now is 10:52 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