CAP just exploded, what now?

[Jonas]

Hi,

I have a SE KT88 working amp (no issues). This morning I removed the KT88 to test some voltages; placed the amp from standby to on - after about 10 seconds at full voltage the heater elevator smoothing cap exploded.

I'm guessing that since the KT88 was not loading down the amplifier voltage, the voltage rating of the cap was exceeded? The cap is rated at 160V and the divider provides about 65V DC for the elevation.

My questions: Is this cap really needed? After I clean up the mess, will there be additional things to look for, check before I restart the amp?

 

[pdf64]

A low impedance path to circuit common for the heater circuit is usually beneficial in reducing hum etc transfer between the heater and cathode.
So yes, a decoupling cap there is probably a good idea.
But feel free to test the hypothesis above by trying it without one.
If you do, please report back 

[Willabe]

I'm guessing that since the KT88 was not loading down the amplifier voltage, the voltage rating of the cap was exceeded? The cap is rated at 160V and the divider provides about 65V DC for the elevation.

So it was working? Meaning you installed the cap so the polarity was correct? Clean it up, then fire it up still without the power tube and measure for the dcv where the cap's + lead was hooked up. What's that dcv?

My questions: Is this cap really needed?

I say yes. That cap is for smoothing the dcv stand off/elevation voltage. It should be there. If the dc stand off voltage is not true dc, smooth, it defeats the purpose. It can modulate the heater causing noise.

When I use a dc stand off voltage on the heaters I take it from the very last B+ filter cap, because the dcv is cleanest/smoothest there.

[shooter]

After I clean up the mess, will there be additional things to look for

WITHOUT power, double-check the R's in the V-divider just to make sure sneaky pete didn't do some damage

[Jonas]

Yes the circuit is known good - been running great on the bench for a few months now with zero issues.

I'm guessing it was the sketchy tube I had just installed prior to the incident.

I cleaned up the amp; inspected with the inspection mirror, pulled all tubes and plugged in the current limiter (no shorts) then power up with no tubes for voltage checks.
 
I have the amp back up and running.
 
The incident also took down my artificial ground connection (no replaced as well) and was quite forceful as it slightly repositioned a filter cap that was close by.

Willabe I will move my elevation supply to the last node

[Willabe]

.... then fire it up still without the power tube and measure for the dcv where the cap's + lead was hooked up. What's that dcv?

Should have taken this measurement.

[pdf64]

... That cap is for smoothing the dcv stand off/elevation voltage. It should be there. If the dc stand off voltage is not true dc, smooth, it defeats the purpose. It can modulate the heater causing noise.
 ...

My understanding differs to the above.
Some valves have imperfect heater to cathode insulation, such that a current path exists which allows the AC heater voltage to appear at the cathode, causing a buzzy hum. Fully bypassing the cathode usually gets rid of that, but sometimes is insufficient, and if the design doesn't allow full bypass, then other mitigation must be used.

The heater to cathode current path is easily saturated by a few micro amps of DC, eg by biasing AKA elevating the heater circuit such that its VDC is a few volts higher than the cathode voltage. Once saturated, no modulation of the current path is possible, because it's saturated.
So pure DC is no better than ripply DC.
By the same token, the circuit common (ground) point used can make no difference.
Provided any ripple etc on the DC is not so great that it brings the instantaneous heater + elevation voltage down too close to the cathode voltage, saturation of the h-k current path is maintained.

Rather than smoothing, the purpose of the cap is to decouple the elevation voltage, ie make it low impedance (eg a high impedance heater circuit could facilitate an unintended feedback path between circuit stages), and to ensure that the heaters are balanced and referenced to 0V AC.

[Willabe]

Merlin calls it a smoothing cap.

https://www.valvewizard.co.uk/heater.html

From Akin's web site, Q&A section;

Q: How can I get rid of the hum if it is caused by the filaments?
 A: The best way to get rid of the hum is to generate a DC filament supply, but this isn't always practical, because of the high dissipation caused by the large current draw of the filaments.  First, make sure the filament wires are run as twisted pairs, then make sure they aren't routed too close to sensitive grid wires or next to coupling capacitors.  If that doesn't do it, try "elevating" the center tap of the filament winding (or junction of the two filament reference resistors) to a potential above the cathode voltage of the tubes, typically between 10V - 50V or so.  You can generate this voltage with a couple of resistors arranged as a voltage divider from the plate supply.  Be sure to bypass the junction of the resistors to ground  with a suitable filter capacitor, or you may inject some buzz or noise into the amplifier from the power supply.

Merlin says use it, Akin say's use it, KOC (Kevin O'Connor) in his books say's use it, and although saying a different reason, your saying why we would use it.

For what ever the reason, because of this, I would use the bypass cap.

[Willabe]

Here's how I did my Princeton 5F2a build, I just posted this in another thread.

[Merlin]

I'm guessing that since the KT88 was not loading down the amplifier voltage, the voltage rating of the cap was exceeded? The cap is rated at 160V and the divider provides about 65V DC for the elevation.

Even unloaded it's not like the voltages are going to more than double. Sounds like the cap was installed backwards, or the elevation voltage is not what you think it is.

[pdf64]

...
Merlin says use it, Akin say's use it, KOC (Kevin O'Connor) in his books say's use it, and although saying a different reason, your saying why we would use it.
...

Apologies for not being clear, but I wasn't suggesting the cap has no useful role.
Rather I put forward my understanding, that its role is largely unrelated to reducing HT ripple.

The cap's decoupling action (which I suppose might alternatively be described as 'smoothing') being to prevent the heater circuit becoming a carrier for parasitics and other induced nasties.

It's a different scenario to the traditional Fender / Marshall arrangement, in which the heater circuit balance point is connected to 0V common, where different 0V reference points (eg at the mains transformer, at the input sockets) can make a difference in the degree of hum contamination of the signal path.

[stratomaster]

Here's how I did my Princeton 5F2a build, I just posted this in another thread.

What a timely post.

Just last night I added a 330k/68k divider off the screen node of my 68 Custom PR. Bypassed the 68k with a MOD 8μF-150v.  This resulted in an elevation of about 75v. And fits into a small heat shrink package similar to a trem roach.  This can conveniently piggyback on axial caps.

Haven't done a full volume test yet, but seems to have helped a bit with overall hum.

However today I came across a post that cited a need for 35v or more above the maximum cathode voltage for effective hum reduction in that stage.  In my amp this is about 64v on the cathodyne triode (Stokes Mod to PI). I'm not too worried about it given the unity-ish gain of the cathodyne, but in theory I should be targeting ~100v elevation.

This seems high (half or more most tubes allowable heater-cathode voltage). 

pdf64 is suggesting saturation can occur at very low elevation voltages, whereas Willabe has mentioned 70-80v before full noise benefits are realized.  I see no harm in targeting higher voltages and benefits to DCCF stages with reduced Vh-k even if they aren't benefitting from reduced hum given the 35v above cathode voltage target specified above.

I'm not sure I have a direct question here, but rather a series of observations that seem to be at odds.  None of which point to DC elevation as a harmful practice, however--but rather seem to indicate two different understandings of the mechanism of noise reduction and/or which mechanism is the dominant one.  The latter may be very much situational and responsible for the discrepancies.