4xKT88 into Hammond 1650T, what am I missing? - 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 8th March 2013, 11:07 PM   #1
diyAudio Member
 
funk1980's Avatar
 
Join Date: Sep 2011
Location: Groningen
Default 4xKT88 into Hammond 1650T, what am I missing?

After a lot of reading and studying schematics (and tickering with pre amps), I thought I had the basiscs down for my first class AB1 ultra linear 4x KT88 into a Hammond 1650T OPT build. My goal is between 100 and 120 watts of power, which I assume is realistic.

For some reason, after drawing all the load lines (1/2 OPT impedance for class A part, 1/4 OPT impedance for class B part, shifting up class A line for bias point), my calculations show I've only got about 81 watts of power. HT will be 500V, fixed bias around -58V, which should, according to other builds I've found, be plenty for the 100W+ goal.

In my noob-ness, I offcourse might have made some errors. Any insight is welcome.

edit: I've attached the curves with loadlines. Tubes are paralleled, so Ia should be doubled.
4xkt88hammond1650T.jpg

Last edited by funk1980; 8th March 2013 at 11:22 PM.
  Reply With Quote
Old 8th March 2013, 11:26 PM   #2
diyAudio Member
 
funk1980's Avatar
 
Join Date: Sep 2011
Location: Groningen
And while writing the 'edit' I might have answered my own question. For power calculations, I used around 430mA instead of the doubled 860mA. Although RMS power would then be 172W... I'm lost.
  Reply With Quote
Old 9th March 2013, 12:28 AM   #3
GoatGuy is offline GoatGuy  United States
diyAudio Member
 
Join Date: Dec 2012
Location: SF Bay Area
Lol... and you're worried about the difference?

The acoustic difference between 120W and 81W is only 10log(120/81) = 1.7 decibels ... in other words, almost inconsequential. I know you're trying to get all the design stuff right (laudable, I might add!) through use of curves and load lines, but really what you're going to end up doing is putting a variable bias on the tubes (though exotic... one for each tube, with separate grid-feed caps to isolate the relatively different DC bias voltages), and using a well measured cathode resistor of low value (on each KT88), measure the voltage across that cathode resistor, use ohm's law, and back-calculate the current flow. Nominal without a signal.

Adjusting the bias to get each KT88 to somewhere between the 1/10th and 1/2 nominal conduction level (follow your curves!). Then, start with some nice clean music source (or you could go signal-generator and sine waves + oscilloscope), and measure the CATHODE RESISTOR voltage (again). Why? Well because its going to be close to ground (safe), and because you know the value of each tube's resistor, you'll be able to balance the current quite nicely. And set it to the nominal value afforded by your load curves.

NOTE that in doing this, you need a 4 ohm or 8 (or whatever!) resistive load on the output transformer. This is to burn up the power that otherwise would reflect back as reverse EMF, driving the high voltage at the KT88 anodes much higher than your specification. You don't want much of that, for the life of the tubes, and your safety.

In the end, graphs and load curves get you "to the ballpark". Real tubes, with real differences in manufacturing, different days of make, different diffusions of helium through the glass-metal electrode seals, different batches of oxides for the cathode, and so on ... simply will need you to balance things by potentiometer. Carefully. Very carefully. And the load curves tell you "kind of where we should be".

Never had an amplifier that EVER "hit it" on the dot when everything was spec'd in advance, plugged in, and turned on the first time. Not even very close.

GoatGuy
  Reply With Quote
Old 9th March 2013, 12:35 AM   #4
GoatGuy is offline GoatGuy  United States
diyAudio Member
 
Join Date: Dec 2012
Location: SF Bay Area
PS: these "low value cathode resistors" can be something that'll give enough DC voltage with the expected current, to give you 3 digits of precision on your digital meter. Most meters of any quality today are happy measuring down to sub-1-volt and giving 3 digits. Therefore ... if the current is expected to be (say) 100 milliamps nominal (per tube, though -58Vg1 and +500Va yields like 50ma for Svetlanas) then you might want to use a 2 ohm resistor. Bypass it (or not, it won't really make much difference) with a 2000 uF 25 volt electrolytic. actually, don't. To finally "hear the full sound", just short out the cathode resistors (in place, with a bit of bypass copper wire!), to see the difference, once the biases are set.

You're welcome.

GoatGuy
  Reply With Quote
Old 9th March 2013, 01:13 AM   #5
diyAudio Member
 
funk1980's Avatar
 
Join Date: Sep 2011
Location: Groningen
Thanks for the extensive reply! That's a lot of useful info that will be very helpfull in the next design stages. And I will take it to heart all this theory is a guide and not an exact real world outcome .

A few questions rise though; In my second post, I concluded the power ouput wasn't 80-something watt, but actualy 172. Enough to seriously saturate the xformer I assume. I forgot the use the double plate current in my calculation because of the paralle tubes, but please correct me if I'm wrong.

If my calculation of 172 watt is about right, then I wonder why there are so many schematics with similar tubes and xformers ratings. Is the excess of power used the provide negative feedback or is there another reason for this 'mismatch'?

Last edited by funk1980; 9th March 2013 at 01:16 AM.
  Reply With Quote
Old 9th March 2013, 04:18 AM   #6
45 is offline 45  Italy
diyAudio Member
 
45's Avatar
 
Join Date: Dec 2008
Location: UK
Quote:
Originally Posted by funk1980 View Post
After a lot of reading and studying schematics (and tickering with pre amps), I thought I had the basiscs down for my first class AB1 ultra linear 4x KT88 into a Hammond 1650T OPT build. My goal is between 100 and 120 watts of power, which I assume is realistic.

For some reason, after drawing all the load lines (1/2 OPT impedance for class A part, 1/4 OPT impedance for class B part, shifting up class A line for bias point), my calculations show I've only got about 81 watts of power. HT will be 500V, fixed bias around -58V, which should, according to other builds I've found, be plenty for the 100W+ goal.

In my noob-ness, I offcourse might have made some errors. Any insight is welcome.

edit: I've attached the curves with loadlines. Tubes are paralleled, so Ia should be doubled.
Attachment 334811
You don't get 120W because you didn't build the composite plate curves of the push-pull amp and thus the right 1/4 load-line. Have a look here for instance:

Part 3 - Of Loadlines, Power Output, and Distortion
  Reply With Quote
Old 9th March 2013, 02:02 PM   #7
diyAudio Member
 
funk1980's Avatar
 
Join Date: Sep 2011
Location: Groningen
Quote:
Originally Posted by 45 View Post
You don't get 120W because you didn't build the composite plate curves of the push-pull amp and thus the right 1/4 load-line. Have a look here for instance:

Part 3 - Of Loadlines, Power Output, and Distortion
Thanks, but I have to disagree. Merlin Blencowe and Morgan Jones, two authors I hold in very high regard, both agree composite curves are somewhat 'overkill' for drawing the power stage loadlines. Reasons being a speaker is not pure resistance and valves are not perfect. Treating half the PP output as single ended and then converting the outcome to full PP requirements is enough to create a basic picture. And this is supported when drawing up the loadlines with the given values on e.g. the Genalex KT88 datasheet for the AB1 ultra linear fixed bias configuration (98W vs 100W given in the sheet).

So the question still stands, why are there so many PPP designs with 1650T (Z=1900 a-a) type configuration and 4 KT88's while the power outputs from the valves will far exceed the xformers ratings when pushed to full signal swing?? Headroom? Keeping away from 0V on G1? Margin for NFB??
  Reply With Quote
Old 5th July 2014, 04:57 PM   #8
diyAudio Member
 
claudiomas's Avatar
 
Join Date: Jul 2006
Send a message via MSN to claudiomas
[QUOTE=So the question still stands, why are there so many PPP designs with 1650T (Z=1900 a-a) type configuration and 4 KT88's while the power outputs from the valves will far exceed the xformers ratings when pushed to full signal swing?? Headroom? Keeping away from 0V on G1? Margin for NFB??[/QUOTE]
4xKT88 into Hammond 1650T, what am I missing?
1650W also Z=1900ohm a-a, best whit this transformer???
I have both transformers Hammond 1650T and 1650W , so i send 608VAC 50HZ from mains power transformer to plate to plate , secondary 8ohm on 8ohm 250W resistors, by oscilloscope and voltmeter same result, 37,5 Vac for 1650T and 38,1 Vac for 1650W, waveform exactly the same ,no saturation at all!
__________________
overbiased

Last edited by claudiomas; 5th July 2014 at 05:16 PM.
  Reply With Quote
Old 5th July 2014, 07:35 PM   #9
diyAudio Member
 
funk1980's Avatar
 
Join Date: Sep 2011
Location: Groningen
Are your values RMS, peak or peak to peak??
  Reply With Quote
Old 5th July 2014, 08:38 PM   #10
diyAudio Member
 
claudiomas's Avatar
 
Join Date: Jul 2006
Send a message via MSN to claudiomas
Quote:
Originally Posted by funk1980 View Post
Are your values RMS, peak or peak to peak??
rms
__________________
overbiased
  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
Am I Missing Something? amandarae Tubelab 7 16th November 2012 10:33 PM
4XKT88 drive Salas Tubes / Valves 44 29th June 2009 01:10 PM
What am I missing? jeff mai Parts 3 19th June 2005 09:41 PM
Wanted 2 hammond 1650T output transformers NotInTheMafia Swap Meet 0 4th April 2005 05:27 PM
am i missing something theChris Analogue Source 1 22nd June 2003 12:25 PM


New To Site? Need Help?

All times are GMT. The time now is 09:55 AM.


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