Will a broken output transformer effect the screen voltages on the EL84's?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Sammas

Well-known member
Joined
Sep 30, 2004
Messages
547
Location
Sydney, Australia.
I have a matchless lightning clone that is giving me some trouble.

Right in the middle of the most blistering rendition of stairway to heaven the output volume dropped off almost complete - During the rehearsal, I couldn't actually hear anything coming out of the amp, but when I brought it home and tried it again there is a very faint signal still present when the volumes are cranked.

The preamp gain and tone still work. I am almost positive this is a problem related to the output stage (2x EL84 for 15watt...).

I am measuring 370v (+/- 4 volts) out of the GZ34 rectifier. I am getting about the same on the primary center tap of the transformer, with a few volts lower on the plates of  the EL84's. The screens are really low though. About 215v with a set of Electro Harmonix valves, and about 235v with a set of JJ's.

The EH tubes were purchased new to figure out what the heck is going on.

I replaced all the 22k resistors in the power section. Both of the 100ohm resistors that split the B+ voltage to each EL84 screen measure spot on....

I am baffled. Without the tubes in, all the voltages jump... with the screens sitting about 30v below the plates.

What the heck is going on? Is there a chance the new tubes and the old tubes are both duds?

Or has something more sinister happened? Like perhaps the old tubes were on their way out, caused some parasitic oscillation outside the audible range and cooked the output transformer (it doesn't look or smell cooked)... or perhaps a shorted turn on the secondary? Would alterations to the load on the plates of the output valves adversely affect the screen voltages?



 
Trace the path from first B+ to screens. It does not go through any 22K resistors. It may go through some hundreds-ohms resistors, though usualy more than 100. Screens are often fed from a second R-C filter from the first B+. (I don't know just what "clone" plan yours is built from.)
 
PRR said:
Trace the path from first B+ to screens. It does not go through any 22K resistors. It may go through some hundreds-ohms resistors, though usualy more than 100. Screens are often fed from a second R-C filter from the first B+. (I don't know just what "clone" plan yours is built from.)

Its a Ceriatone Matchless, originally built by Nik @ Ceriatone probably 3 or 4 years ago...
The current layout on the ceriatone website is slightly different to my amp as Nik made some modifications a few years ago.

It looks like the layout is based very closely off this schematic: http://www.prowessamplifiers.com/schematics/post/matchless_lightning.pdf

I have traced the path, the screens (pin 9, right?) are fed the B+ through the choke via a 22k resistor. At the node where the output of the choke splits to the 4x 22k resistors I measure 360v (with 361v directly out of the rectifier). On the screens side of the 22k resistor for the EL84's I get 208v... that is much lower than what I measure on the other side of the other 3 22k resistors...

Could it be a problem associated with the 22UF filter cap?




 
Sammas said:
PRR said:
Trace the path from first B+ to screens. It does not go through any 22K resistors. It may go through some hundreds-ohms resistors, though usualy more than 100. Screens are often fed from a second R-C filter from the first B+. (I don't know just what "clone" plan yours is built from.)

Its a Ceriatone Matchless, originally built by Nik @ Ceriatone probably 3 or 4 years ago...
The current layout on the ceriatone website is slightly different to my amp as Nik made some modifications a few years ago.

It looks like the layout is based very closely off this schematic: http://www.prowessamplifiers.com/schematics/post/matchless_lightning.pdf

I have traced the path, the screens (pin 9, right?) are fed the B+ through the choke via a 22k resistor. At the node where the output of the choke splits to the 4x 22k resistors I measure 360v (with 361v directly out of the rectifier). On the screens side of the 22k resistor for the EL84's I get 208v... that is much lower than what I measure on the other side of the other 3 22k resistors (from 270v to 340v)...

Could it be a problem associated with the 22UF filter cap?
 
> much lower than what I measure on the other side of the other 3 22k resistors...

They do different things. X and Y and another feed small 12AX7 triodes, a few mA each. The one to the screens feeds EL84 screens. Even though the EL84 is not much bigger than 12AX7, it is a POWER tube and eats more juice, even the screens.

> the screens (pin 9, right?) are fed the B+ through the choke via a 22k resistor.

Ha!! That IS what that plan shows.

However skim the EL84 data, page 4. G2 (screens) want to suck 4mA to 11mA each, 8mA to 22mA total. If they try that with a 22K resistor in series, we have 176V to 484V drop. From 360V, that's absurd, because they are supposed to be near 300V (or more?), not 184 to nothing.

> On the screens side of the 22k resistor for the EL84's I get 208v

At idle, because your voltages are 20% higher, that's about what you expect if it has a 22K resistor. Twang it, and screen voltage will drop even lower, much too low for the EL84s to deliver any serious loudness.

I am quite sure the raw B+ to screens feed must be much lower than 22K to get "spitfire" type power output.

A re-draw of a Spitfire (original?) shows screens connected to the choke with individual 100 ohm resistors, no other resistance. Modern blistering renditions of Stairway To Heaven do suggest more resistance in there, but not 22K. 2.2K (2K2, 2200) is the most likely mis-read, and will work with authority. It needs to be 1 Watt, and perhaps should be 2W. My gut says 1K 1W is really a sweeter spot.

> built by Nik @ Ceriatone ... Nik made some modifications a few years ago.

I wish you would consult with Nik. He knows a lot more about these amps than I do. _If_ there are any weaknesses, he's the expert. If there are wrong plans floating around, he may want to know.
 
PRR said:
> much lower than what I measure on the other side of the other 3 22k resistors...

They do different things. X and Y and another feed small 12AX7 triodes, a few mA each. The one to the screens feeds EL84 screens. Even though the EL84 is not much bigger than 12AX7, it is a POWER tube and eats more juice, even the screens.

> the screens (pin 9, right?) are fed the B+ through the choke via a 22k resistor.

Ha!! That IS what that plan shows.

However skim the EL84 data, page 4. G2 (screens) want to suck 4mA to 11mA each, 8mA to 22mA total. If they try that with a 22K resistor in series, we have 176V to 484V drop. From 360V, that's absurd, because they are supposed to be near 300V (or more?), not 184 to nothing.

> On the screens side of the 22k resistor for the EL84's I get 208v

At idle, because your voltages are 20% higher, that's about what you expect if it has a 22K resistor. Twang it, and screen voltage will drop even lower, much too low for the EL84s to deliver any serious loudness.

I am quite sure the raw B+ to screens feed must be much lower than 22K to get "spitfire" type power output.

A re-draw of a Spitfire (original?) shows screens connected to the choke with individual 100 ohm resistors, no other resistance. Modern blistering renditions of Stairway To Heaven do suggest more resistance in there, but not 22K. 2.2K (2K2, 2200) is the most likely mis-read, and will work with authority. It needs to be 1 Watt, and perhaps should be 2W. My gut says 1K 1W is really a sweeter spot.

> built by Nik @ Ceriatone ... Nik made some modifications a few years ago.

I wish you would consult with Nik. He knows a lot more about these amps than I do. _If_ there are any weaknesses, he's the expert. If there are wrong plans floating around, he may want to know.

Thanks for the post PRR!

I have bounced a few emails back and forth with Nik. He echoed your sentiments and stated that the screen voltages are way too low, and to try another (another) set of tubes. I guess I'll trump off to the music shop and buy my second set of EL84's... He did also mention that while the layout is considered "out of date" in terms of production, he did mention that there is nothing wrong with it. Which appeared to be true. I played about 50 gigs with the amp and it was loud enough to cut through even the most difficult of rooms with a full (and loud band).

On an interesting note. The updated layout feeds the EL84's straight from the choke - with no resistor at all. It does also have one less filter capacitor: http://ceriatone.com/images/layoutPic/matchlessLayout/Lightningceriatone.jpg

Its official. I'm fuggin confused! hahahaha


 
> updated layout feeds the EL84's straight from the choke - with no resistor

Still the two 100r resistors, but overall way under 1K of resistance from rectifier to screens.

The 22K in series with the screens on the other plan just must be a mistake. Someone started drawing 22K resistors and went too far. A pair of EL84 can't work well with 22K in the screens.

(On another bigger amp using EL34s, I _wanted_ to tame it down a bit, and 3K in the screens knocked the power down 20+% and also made it more abuse-proof because determined abuse just sagged the screens and thus the total tube-abuse. The '84 is half the tube of the '34, so seven times more screen resistance would really make it sad.)

Yet you replaced "something" at that point?? As I say, I could believe one or two K ohms, not dozens of K. So 1K 1W or 2K 2W are plausable. And for testing, one small 100-300 ohm resistor should bring the screens and grunt right up, yet fail cheaply if the tubes are way-sick (or maybe in extended string-bashing).
 
PRR said:
> updated layout feeds the EL84's straight from the choke - with no resistor

Still the two 100r resistors, but overall way under 1K of resistance from rectifier to screens.

The 22K in series with the screens on the other plan just must be a mistake. Someone started drawing 22K resistors and went too far. A pair of EL84 can't work well with 22K in the screens.

(On another bigger amp using EL34s, I _wanted_ to tame it down a bit, and 3K in the screens knocked the power down 20+% and also made it more abuse-proof because determined abuse just sagged the screens and thus the total tube-abuse. The '84 is half the tube of the '34, so seven times more screen resistance would really make it sad.)

Yet you replaced "something" at that point?? As I say, I could believe one or two K ohms, not dozens of K. So 1K 1W or 2K 2W are plausable. And for testing, one small 100-300 ohm resistor should bring the screens and grunt right up, yet fail cheaply if the tubes are way-sick (or maybe in extended string-bashing).

The only thing I can think of is the resistor to the screens being a lower value, but opening under strain and increasing the resistance to 22k... it would take a whole lot of luck for it to land at 22k though!

The original resistors (all 4) measured 22k. I replaced them all with new 22k's that measure to within 100 ohms of the originals. I still have the originals at home, and even made a glance comparison between the new and old colour codes and they all looked identical to me... I'll have another look.

The screen voltages measure the same before and after the replacement too. I have some 180ohm 2w resistors floating around. I'll replace the 22k and see what happens.

It is jumping the gun a little, but regarding the output transformer... would a fault in it effect the screen voltages?

Thanks a lot for your help PRR! I really appreciate it!

 
I'm sure the 22K+extra cap is a drawing mistake.
Those screens are fed straight from the choke, with some small value series resistor.
100r minimum, adjust to taste but not too far.
Just google the EL84 datasheet and go by it.
 
> take a whole lot of luck for it to land at 22k though!

Indeed. 

> colour codes

In the 3-stripe code, orange (22K) looks a lot like red (2.2K)..... I've made that mistake many times. 2.2K is larger than usual but would be playable. I'm quite shocked it ever played LOUD with 22K.

> the output transformer... would a fault in it effect the screen voltages?

Affect. If the OT primary were open, all the cathode-plate current becomes cathode-screen current. But you would also find that plate voltage at _zero_, and the cathode resistor voltage lower than normal. Yet you report the usual few-volt drop from raw B+ to plates.
 

Latest posts

Back
Top