Sunn Head Blew Another Power Tube

[Hellawatt]

Hi,

So I blew a KT88 in my Sunn Spectrum II. This is the second time. The first time it happened I hadn't really biased it, I just set the bias pot to -55V. The second time it happened, I had biased it with a bias probe to about 552V plate voltage X 0.0478A cathode current = ~26W. Does that sound right?

Both times I blew a tube I was using a silicon diode rectifier, and playing a very high output fuzz/distortion pedal with amp volume cranked up to around 4. Could a super high input signal have anything to do with it?

I suspect maybe the voltages in the amp are getting too high? Would that kill a KT88? I've tried tube rectifiers but I wasn't happy with the sag. Maybe I need to make a bucking transformer. My wall voltage is 120V. If I brought it down to 117V would that be enough to safely use silicon rectifiers?

I asked about this amp in an earlier thread here: http://el34world.com/Forum/index.php?topic=20434.msg215317#msg215317

Thanks

[HotBluePlates]

... So I blew a KT88 in my Sunn Spectrum II. ...

What does "blew a tube" mean in this case? Heater burned open? Plate melted? Screen melted? Plate-to-heater short circuit? Silvery getter turned to milky white?

Looking at your other thread, you note that the amp blows fuses when the "blown tube" is in the socket. But you'll still need to know some things about the tube's failure mode to understand if it is an amp problem or a tube defect.

-  Plate-to-heater short circuits often manifest as continuity from pin 3 to pin 2.
-  You probably shouldn't have continuity between any pins except pin 2 to pin 7, and possibly pin 1 to pin 8.
-  Ideally, you'd have the amp on a lightbulb limiter and measure voltages for the bad tube to determine what is shorting (very low plate or screen voltages while limited might indicate a short from those elements to something else; may require measuring voltages with a known good tube to understand what typical readings should look like while current-limited).
-  Small tube pins, or sprung socket contacts, which result in a loose fit in the socket can cause all manner of strange results; does the tube fit snugly (but not over-tight) in the socket?

[PRR]

> Could a super high input signal have anything to do with it?

I had a 1979 T-bird. Boringly reliable until the day I put the PEDAL TO THE METAL screaming >3,900 RPM for a minute up a steep hill (had to pass a jerk in a BMW). Lost all power.

This was a little funny. The car was smog-tuned, not made for high RPM. The core engine was good for more, but the car was far past 130K miles at this point. And the actual failure was the tiny weights inside the distributor. Flew off their pivots and jammed the sparker, sheared the gear, no spark, no go, call a tow.

Another time I had an old Mustang up to near top speed, 88MPH, it made horrible valve noises, but limped home for a valve job.

In both cases some "minor" part failed when pushed HARD. Failed not from design flaw but old age (wear, wobble, slop).

Agree you need to define "blew". WHAT blew?

Bright light on the socket contacts and tube pins, look for ugly spots. If even one pin makes poor contact, you will keep having trouble.

Bias makes no difference when playing "super high"; bias is like the idle speed on the engine, is over-ridden when you put pedal to metal (loud strumming).

[Ritchie200]

What brand of KT88 are you using?  We need some voltages. If I remember correctly sunn was always conservative with screen voltage. I'd be interested in seeing what that is right now.

Jim

[drgonzonm]

The data sheet I have for the KT 88 shows an absolute voltage of 800v on the plates.  I do not believe it is the tubes, but a component problem. 

I wonder if you are not having a cap failure on the b-rail.  The schematic found on this web site shows the caps are rated for 525v, you are running some what above that. 
Do you put a 5 or 10w resistor in series with the choke or up stream of the first cap?  It still does not solve the problem with unrated cap just after the standby switch. 

Another choice would be to replace the first cap with a higher voltage rated cap. 

Sluckey has a vintage amp voltage adapter in his scrap book, you might want to consider one of those.   

[Ritchie200]

The data sheet I have for the KT 88 shows an absolute voltage of 800v on the plates.

In the 70's from GEC, yes.  Now, no.  Those recycled data sheets from yore are meaningless - at least for this tube.  There are only a couple of brands that will live in the realm north of 600v.  I've had several Chinese and euro brands die a quick death with no signal.  If they are that delicate, I can see problems at 550-560 in the Sunn.  Unfortunately I have receipts to prove out that experiment.

We need some more voltage information.....

Jim

[drgonzonm]

with the sunn schematic showing 480v to the CT on the O/T, a vintage voltage adapter is probably the cheapest way to get the voltage down.  If you use 120/6-12, bucking transformer, you will probably want to use the full windings i.e., the 12v drop to 108 volts. 

I still think you may have compromised that 30microfarad cap. 

[Hellawatt]

... So I blew a KT88 in my Sunn Spectrum II. ...

What does "blew a tube" mean in this case? Heater burned open? Plate melted? Screen melted? Plate-to-heater short circuit? Silvery getter turned to milky white?

Looking at your other thread, you note that the amp blows fuses when the "blown tube" is in the socket. But you'll still need to know some things about the tube's failure mode to understand if it is an amp problem or a tube defect.

-  Plate-to-heater short circuits often manifest as continuity from pin 3 to pin 2.
-  You probably shouldn't have continuity between any pins except pin 2 to pin 7, and possibly pin 1 to pin 8.
-  Ideally, you'd have the amp on a lightbulb limiter and measure voltages for the bad tube to determine what is shorting (very low plate or screen voltages while limited might indicate a short from those elements to something else; may require measuring voltages with a known good tube to understand what typical readings should look like while current-limited).
-  Small tube pins, or sprung socket contacts, which result in a loose fit in the socket can cause all manner of strange results; does the tube fit snugly (but not over-tight) in the socket?

There is no visible difference in the bad tubes.

When the fuse blew there was a bright white flash from one of the power tubes.

I checked both bad tubes for continuity and only pins 2 and 7 are continuous.

I went out and made a light bulb limiter yesterday, and when I turn on the amp with a bad tube installed, the bulb pulses, and the voltage jumps around. I get about 30-45V at the standby switch, and the bad power tube is reading around 30V/30V/-5V at pins 3/4/5 respectively, and the good tube is reading about 20V/20V/-3V at pins 3/4/5 respectively, but the reading were jumping around quite a bit.

> Could a super high input signal have anything to do with it?

In both cases some "minor" part failed when pushed HARD. Failed not from design flaw but old age (wear, wobble, slop).

Agree you need to define "blew". WHAT blew?

Bright light on the socket contacts and tube pins, look for ugly spots. If even one pin makes poor contact, you will keep having trouble.

Bias makes no difference when playing "super high"; bias is like the idle speed on the engine, is over-ridden when you put pedal to metal (loud strumming).

The tubes are a little loose in the socket, but the contacts seem tight. No visible damage to any sockets or pins.

So if my input signal was extremely hot, how would that kill a power tube? Wouldn't the preamp stages act as a gate? I've used the same pedal  into other amps without killing tubes. I would like to make this amp as bulletproof as possible in case I ever use it live. I'm thinking about splitting my guitar signal and sending an octave down to the Sunn.

What brand of KT88 are you using?  We need some voltages. If I remember correctly Sunn was always conservative with screen voltage. I'd be interested in seeing what that is right now.

Jim

I'm using JJs. I have a pair of old RCAs that came with the amp but I am afraid to use them until I am sure that the amp is good.

The voltages I was getting with 120V at the wall and silicon rectifier was B+: 548V/537V/443V/324V at Standby/A/B/C and on the power tubes  pin 3/4/5 were about 531V/534V/-55V.

I wonder if you are not having a cap failure on the b-rail.  The schematic found on this web site shows the caps are rated for 525v, you are running some what above that. 
Do you put a 5 or 10w resistor in series with the choke or up stream of the first cap?  It still does not solve the problem with unrated cap just after the standby switch. 

Another choice would be to replace the first cap with a higher voltage rated cap. 

I recapped this amp with a cap board I made that is a copy of the triode electronics dynaco MK3 board. It should be able to handle 800V.


Thanks for all the help everybody. I really appreciate it. I am learning a lot.

[drgonzonm]

Is the tube you lost, the same tube?  If so, check your O/T, Smell the transformer.  use the 6.3 v filament on the transformer secondary.  Also check DRC on all leads on the O/T.  I just want to make sure you don't have a bad O/T. 

[Colas LeGrippa]

Hi buddy,

I would check the grid leak resistors ( 100K , 220K or so ). Seems to me that a bad contact ( solder, bad resistor ) is responsible for that. A power tube will redplate and destroy itself if the grid leak resistor is faulty.
Check all the bias circuit, especially the connection at the grid leak of the defective power tube. I don't think it has something to do with the tube quality or make.
Colas