Testing amplifiers?

arcane secrets and handshakes revealed!

Testing amplifiers?

Postby Blair » Wed Jan 20, 2010 8:54 am

OK,

I see there is another post just below this where someone is asking about testing amps also. Here goes a few questions of mine.

I recently purchased an old Tek scope T922. I'm not sure how great it is, because I have never used a scope. I can get waves to come close to being stationary on the screen, but not anything I would consider stable enough to analyze. I think there is a simple issue with the screen adjustments, or the triggering circuit. I'll get that schecked out later though.

Anyway, the scope does work well in all other aspects.

I recently avoided buying a function generator by building one based on the 8308 chip with an opamp buffer on the output of the generator chip. I could not justify spending the money on a function generator that I may or may not use that much. The generator cost about $18 to build using high tolerance resistors, and it does square, sine, and triangle.

The buffered output also is set up for gain, thus eliminating another preamp in the signal chain.

Anyway, without further boring everyone with these projects, I tried the methods described to me in several of my other posts throughout the bulletins.

I hooked my generator up to my amps with one scope channel connected to the input to see the signal wave, and the other probe up to my OPT with a 150w dummy load connected to my 8 ohm tap. This, to my surprise makes the OPTs sing quite loudly when enough input voltage is applied.

The amps that should easily make over 100w began to distort the tops of the sine waves at 1K @ 23.8v.

If you do the math: 23.8 * 23.8 / 8 = 70.805w.

This was measured using my RMS meter connected to the dummy load.

Now for the questions.

A) These are my big amps. There are three pairs of U/L KT88. I would think they would make a bit more power than that before clipping.

B) These amps are biased around 58mA per tube, so on par with Shannon's reccomendations for 60W with a single pair of KT88s. I'm using 1,900ohm primary OPTs, so they are not perfect, but take 4,200/3=1,400ohms. I could not find a 1,400 ohm primary OPT.

C) What do you think would be the most likely candidate for this clipping? The 6SL7, or the phase inverter 6SN7 not being able to deliver enough current, or maybe needs to be biased a bit harder? The circuit is essentially Shannon's front end with 6SN7/SL7 tubes.

D) should I leave the dummy load connected and run the same testes clipping on to the signal through the circuit to see at which point I see the clipping happen?

BTW, before I continue with any further testing, I need to get a few more resistors for my dummy load. It gets very HOT!!

Thanks,

Blair
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby TomMcNally » Wed Jan 20, 2010 12:35 pm

Here's the dummy load I built a few years ago:

Image

You could always put a fan on them for extended testing.

It's normal for the transformers to sing, you just don't normally
hear it because the music is LOUD !

You should hear a 50,000 watt plate modulated AM transmitter -
the audio amp is 25,000 watts ... they really sing.
User avatar
TomMcNally
Darling du Jour
 
Posts: 2729
Joined: Sat Nov 19, 2005 2:19 pm
Location: Northfield, NJ

Postby battradio » Wed Jan 20, 2010 8:14 pm

Hi Blair ,

What does the B+ measure when the amp starts cliping .Also use the scope to see if the driver stage is cliping before the output stage .

Mark
Mark
Image
User avatar
battradio
KT88
 
Posts: 927
Joined: Tue Mar 10, 2009 12:58 am
Location: near ST.Louis MO.

Postby EWBrown » Wed Jan 20, 2010 8:48 pm

"testes clipping " ??? ouch.... ;-) ;-) ;-)

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby Blair » Wed Jan 20, 2010 10:16 pm

battradio wrote:Hi Blair ,

What does the B+ measure when the amp starts cliping .Also use the scope to see if the driver stage is cliping before the output stage .

Mark


Hey Mark,

The B+ goes from 480v to 440v at clipping.

Today, I got slightly different results. I am getting roughly 30.5v before clipping onset from 20hz to 18K and about 28.2v at 20K before clipping. This is much better being at 117W RMS. Still not exactly what I would expect from three pairs of KT88s, but I suppose with PS sag like it is showing, the only other solution is to stack another toroid on these. That or my 500mA choke is not sufficient and is literally providing too much resistance to the PS.

The toroid is an Antek 4T400, so it should do the job I would think. One last thing I would like to try is to pull a pair of KT88s and test the amp.

Blair
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby Blair » Wed Jan 20, 2010 10:17 pm

EWBrown wrote:"testes clipping " ??? ouch.... ;-) ;-) ;-)

/ed B


Nice Ed!
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby SDS-PAGE » Thu Jan 21, 2010 1:29 am

Those are cool looking, Tom! Are the resistors 8 ohms? What's the power rating? Min
User avatar
SDS-PAGE
KT88
 
Posts: 865
Joined: Thu Feb 22, 2007 4:41 pm
Location: Brandon, SD

Postby EWBrown » Thu Jan 21, 2010 10:35 am

To me those look like 50 watt Dale power resistors. If they are relatively new production, probably Dale-Vishay. I have some vintage 50W Dales in 1, 10 and 50 ohms.

What say you, Tom?

I have a couple of big honkin' Japanese made 8.5 ohm, 150 Watt power resistors, wirewound, in aluminum "heat sink" bodies. These came from a surplus joint by the name of Eli Hefron and Sons in Cambridge Mass, about 36 years ago. These also have a slider contact, so I can trim them to exactly 8 ohms, or just use the tap to take off a small sample signal for analysis. They were originally used in a DC power supply dummy load.

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby TomMcNally » Thu Jan 21, 2010 10:58 am

Yep - Dale 8 ohm 50 watt ... they get a little warm but not hot.

My testing usually consists of running Shannon's Automated
Boonton software, so it only takes a few minutes for a run.

I typically do a run at 1 watt (except on a Darling!) and then
the software finds the maximum clean power level and does
a run at that level. Cool stuff. It's much more impressive
on a push-pull design than on single-ended. The SE amps
make you wish you hadn't looked at the distortion!

... tom
User avatar
TomMcNally
Darling du Jour
 
Posts: 2729
Joined: Sat Nov 19, 2005 2:19 pm
Location: Northfield, NJ

Postby battradio » Thu Jan 21, 2010 11:56 am

Blair wrote:
battradio wrote:Hi Blair ,

What does the B+ measure when the amp starts cliping .Also use the scope to see if the driver stage is cliping before the output stage .

Mark


Hey Mark,

The B+ goes from 480v to 440v at clipping.

Today, I got slightly different results. I am getting roughly 30.5v before clipping onset from 20hz to 18K and about 28.2v at 20K before clipping. This is much better being at 117W RMS. Still not exactly what I would expect from three pairs of KT88s, but I suppose with PS sag like it is showing, the only other solution is to stack another toroid on these. That or my 500mA choke is not sufficient and is literally providing too much resistance to the PS.

The toroid is an Antek 4T400, so it should do the job I would think. One last thing I would like to try is to pull a pair of KT88s and test the amp.

Blair


Hi Blair ,

Measure the voltage drop across the the choke at full power out and use the specified resistance of the choke to calculate the current draw through the choke to see if it needs to be larger .

Mark
Mark
Image
User avatar
battradio
KT88
 
Posts: 927
Joined: Tue Mar 10, 2009 12:58 am
Location: near ST.Louis MO.

Postby Blair » Thu Jan 21, 2010 12:07 pm

Hey Mark,

I'm going to sit down with a pen this evening or maybe tomorrow and take measurements from the PT all the way to the OPTs including all the bias voltages in the driver stage. The signal appears to clip on the output of the 6SN7. I clipped on the sg (pin 5) of a KT88 just before the 1K grid resistor, and the signal clips exactly in the same place on my preamp volume control.

I s a single 6SN7 enough to drive 6 KT88s?

Thanks,

Blair
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby battradio » Thu Jan 21, 2010 12:18 pm

Hi Blair ,

If the preamp is cliping you need to start there , add another gain stage to the preamp or amp .Are you using a passive or active preamp .

Mark
Mark
Image
User avatar
battradio
KT88
 
Posts: 927
Joined: Tue Mar 10, 2009 12:58 am
Location: near ST.Louis MO.

Postby Blair » Thu Jan 21, 2010 12:22 pm

I should have been more specific. I'm using my SS preamp and it will drive to full power without clipping at all. I have tested this with my scope.

I connected the preamp with a sine generator as the source to my big amps. I slowly turned the volume up until It is about 3/4 volume where the amplifier output is clipping, and I see the same clipping occuring on the phase inverter side of the power tubes. I clipped from ground to the cap side of the screen grid just before the 1K resistor. I'll try to get better figures for you this evening. I can clip my RMS meter to the same points and tell you what the voltages are there driving the power tubes.

Blair
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby Blair » Thu Jan 21, 2010 10:19 pm

OK guys,

I took a few measurements this evening.

Here we go:

B+ KT88s :479 idle / 439 clipping biased at 60mA rising to .98mA at clipping.

At the PS: before choke 490v idle / 479v clipping
At the PS: after choke 479v idle / 445v clippping

I have a steady -52v on pin 5 of my KT88s throughout the power band.

from ground to my coupling caps on the 6SN7 side, I have 151v at clipping from 190v at idle.

I'm using CCS LM334Z bias for the 6SN7 with 6.7v on the cathode at idle

Here is a picture of what the 6SN7 is doing just before clipping. Taken on the KT88 side of the coupling caps.

The top wave is the 6SN7, and the bottom wave is the secondary side of the OPT.

My preamp makes perfect sines all the way to full throttle. I tested it again this evening.

Image

Thanks,

Blair
Blair
KT88
 
Posts: 1272
Joined: Thu Nov 09, 2006 10:22 am

Postby battradio » Fri Jan 22, 2010 2:55 pm

Hi Blair ,

The 40 volts that the 6SN7 is losing needs to be addressed .A regulated suppy with a tube or transistor would be one way to do it or a completly
seperate supply with its own bridge retifier connected to the torroid winding .That way the current draw from the output wont affect the the 190 volts .Check the AC ripple on the B+ before the choke at full output to make sure you have large enough filter capacitor .

Mark
Mark
Image
User avatar
battradio
KT88
 
Posts: 927
Joined: Tue Mar 10, 2009 12:58 am
Location: near ST.Louis MO.

Next

Return to test & measurement

Who is online

Users browsing this forum: No registered users and 7 guests

cron