Overfiltering?

[Enzo]

Bob, I am sure you can design a warm up circuit, but to me it sounds like a solution looking for a problem. I mean how many bad 5U4s do you see in a year, and of those, how many are bad from damaged cathodes in the manner suggested? These poor bedraggled tubes can work for decades. Cathode trouble and slow starts to protect them is way way down my list of concerns. If rectifiers were failing left and right, it might be different.

[bob p]

The only way cathode stripping and rectifiers got linked was that there was a contingent of people who said that using solid state rectifiers was bad because the B+ came up instantly, before the heaters got hot, and that you ran the warmup time with cathode stripping going on each power-on. Near as I can find out, that contingent was wrong because of the low voltages in guitar amps (i.e. under 700+ V).

if i understand this correctly, it would appear that when we've been talking about cathode stripping, everyone has been thinking about cathode stripping in the output tubes, caused by instantaneous B+ application with SS rectification. i'm happy to accept the consensus opinion on that, but that isn't really what i was referring to. i was referring to heater (cathode) failure in the rectifier tube that is caused by cold starting them into a huge capacitance reservoir.

in this scenario the cold tube is subjected to a huge voltage surge that is way beyond the tube's warm current limit. in the case i mentioned earlier, the Heath HiFi amp was modded by a self-professed eBay tube guru, and fitted with several millifarads of input capacitance. this caused the 5R4 to experience premature heater (cathode) failures, just as the lifespan vs. current charts would predict. if that kind of abuse won't strip a cathode (as a prelude to cathode vaporization), its hard for me to imagine what will.

You're right, Steve. That's why tube rectifier data sheets list maximum values for the first filter cap. The bigger the cap, the smaller the time the rectifiers conduct at peaks of the AC wave, and the higher the current peaks must be to get enough DC in during the short conduction time. Eats the cathode coating.

that's what i was referring to -- cathode stripping of the rectifier.

[bob p]

These poor bedraggled tubes can work for decades. Cathode trouble and slow starts to protect them is way way down my list of concerns. If rectifiers were failing left and right, it might be different.

Enzo, the only reason that I'm participating in this discussion is because I had to fix a real world problem -- its not because I'm trying to push a theoretical point on everyone. I mentioned that the amp was a genuine rectifier tube-eater in one of my earlier posts. maybe you missed it.

i agree, rectifiers should work for decades if they're operated within their design limits. its when they're operated way outside of the maximum ratings that tube failure is significantly hastened. taking a look at the exponential plots on any of the current or voltage burnout curves, its easy enough to understand why.

maybe i'm looking at this from the wrong perspective, but when i thought about the problem, i came up with a short list of differential diagnoses:

Causes of Short Rectifier Life Span:

1. defective tubes
2. excessive heater voltage
3. excessive heater current

the amp did have a bit of overvoltage on the 5V tap, being a 1950s amp that was plugged into a 2007 outlet. from what i've read, a 10% heater over-voltage will cause premature failure, though i think that current was more of the culprit in this case.

maybe my list is wrong and something else should be at the top. if i'm barking up the wrong tree and i'm not thinking about something that i should have been thinking about, please let me know!

[Steve Conner]

Hi Bob, all,

I doubt the warm-up circuit you designed for that overfiltered amp would help much. The excessive peak currents due to small conduction angle are known to eat the rectifier tubes quickly. Even with the heaters at full temperature, there is not enough emission to cope with them. This was a known fact ever since people started designing tube rectifiers.

If I was "fixing" this, I'd add resistors in series with the PT HV winding to get the rectifier peak cathode current back down into spec, or gut the inverter grade caps and put more modestly sized caps inside the cans, or sneak a pair of 1N5408s under the socket

[R.G.]

if i understand this correctly, it would appear that when we've been talking about cathode stripping, everyone has been thinking about cathode stripping in the output tubes, caused by instantaneous B+ application with SS rectification. i'm happy to accept the consensus opinion on that, but that isn't really what i was referring to. i was referring to heater (cathode) failure in the rectifier tube that is caused by cold starting them into a huge capacitance reservoir.

OK. Difference in nomenclature. I've only ever seen cathode stripping referring to the process which happens to cold cathodes with high voltages. I've heard the diminishment of cathode emission from high currents as cathode exhaustion, but that's just how I've seen it referred to.

in this scenario the cold tube is subjected to a huge voltage surge that is way beyond the tube's warm current limit. in the case i mentioned earlier, the Heath HiFi amp was modded by a self-professed eBay tube guru, and fitted with several millifarads of input capacitance. this caused the 5R4 to experience premature heater (cathode) failures, just as the lifespan vs. current charts would predict. if that kind of abuse won't strip a cathode (as a prelude to cathode vaporization), its hard for me to imagine what will.
that's what i was referring to -- cathode stripping of the rectifier.

I understand. I've just seen it called something else.

E
maybe i'm looking at this from the wrong perspective, but when i thought about the problem, i came up with a short list of differential diagnoses:

Causes of Short Rectifier Life Span:

1. defective tubes
2. excessive heater voltage
3. excessive heater current

the amp did have a bit of overvoltage on the 5V tap, being a 1950s amp that was plugged into a 2007 outlet. from what i've read, a 10% heater over-voltage will cause premature failure, though i think that current was more of the culprit in this case.

maybe my list is wrong and something else should be at the top. if i'm barking up the wrong tree and i'm not thinking about something that i should have been thinking about, please let me know!

While defective tubes is always a problem, I would say that the causes of early rectifier tube failures among good tubes would be the following, probability unknown:

- Rectifier temperature, as caused by internal dissipation and external ambient temperature
- Excessive heater dissipation, caused by high heater voltage in amps not designed for today's 125Vac lines. High heater current and voltage are inextribably linked.
- Excessive current peaks caused by
(a) too-big first filter caps
(b) excessive B+ current caused by sub-acute conditions like:
(i) output tube bias problems
(ii) leaky filter caps
(iii) other component failures

If you want rectifier tubes to last a long time,
- run the heaters at the right voltage
- keep the peak current pulses below the specified maximums
- keep the glass bulb temperature as low as possible.

But those maxims are true for any tube.

[R.G.]

so what do you do when someone is adamant about keeping tube rectification and doesn't want to throw away the expensive inverter grade caps that have already been paid for?

I personally would put two SS rectifiers in there feeding a first filter cap of maybe 20uF, and a small resistor from there to the plates of the tube rectifier. That would keep the pulse current stress off the tube rectifier and let it live a long, productive life, and also keep the tube rectifier in there so the owner can see it glowing merrily and notice that the amp only comes on when the tube rectifier warms up. The owner also gets to keep the expensive inverter grade caps which they have paid for.

Maybe pot up the SS diodes and first filter cap into a "protective phase shift network" block of epoxy and explain that by synchronizing the phases of the AC line harmonics with the main phase, that you can achieve subatomic alignment of the conductivity electrons within the conductance valence bands of the conductors. This will obviously both give longer rectifier life and also increase the coherence of the raw DC power, resulting in purer and cleaner pauses between the notes, increasing the subjective S/N ratio. Of course, you may have to be in optimum listening conditions to hear the difference in an already well refined listening rig.

God, sometimes I'm ashamed of myself. Those poor tweakos.

[dai h.]

lol. don't forget to (grow and) stroke your beard (and look very serious) while saying that.

[Rob Mercure]

R.G. - you're too bad man <grin>. I'd love to see you take a slice out of the gurus on the "audio asylum" BBS - t'would really be funny. But don't forget to sandwich your "protective phase shift network" between blocks of ivory, Seborgium, and Bose-Einstein condensate using only platinum plated dead soft gold wiring with a widdershins helix - the opposite in the southern hemisphere.

Rob

[Steve Conner]

Bose-Einstein condensate

What? Everyone knows Bose aren't an audiophile brand

[bob p]

you guys are both funny and evil.

putting the very funny jokes aside, i had thought about using a current limiting resistor, as that's essentially what i've built into my in-situ cap reforming rig:

i built a "special" plug-in SS rectifier + current limiting resistor for reforming caps in-situ whenever i resurrect an old amp that's been offline for awhile. (i salvaged the octal base from a dead rectifier, and wired two SS diodes and a suitable current limiting power resistor into the octal base.) i just remove the tube rectifier, plug in the current-limiting replacement, and let her rip until the voltage stops dropping across the resistor.

one adverse effect of using a current limiting resistor in a hifi application is that its going to result in poor voltage regulation under load (aka "Sag"). good for guitar amps, not so good for HiFi. that's the only reason i didn't try the current limiting resistor already.

RG and Rob bring up some good points about the Guru HiFi Mojo myths: some HiFi people are firmly entrenched in their beliefs, which makes it hard to convince them that change may be a good thing. although I had thought about doing some of the "evil" tricks like sneaking a pair of SS diodes into the amp without telling the owner, I didn't want that to come back and bite me. it wouldn't be at all good to have someone going around telling people that i intentionally deceived them when working on their amp. instead, i just told them that the fix that they wanted wouldn't be a good fix, and that they'd be coming back with the same problem. hopefully letting them find out on their own that i was both honest with them, and right in the advice i gave will be better in the end than deceiving them and selling them a load of Guru HiFi Mojo BS.

[bob p]

I personally would put two SS rectifiers in there feeding a first filter cap of maybe 20uF, and a small resistor from there to the plates of the tube rectifier. That would keep the pulse current stress off the tube rectifier and let it live a long, productive life, and also keep the tube rectifier in there so the owner can see it glowing merrily and notice that the amp only comes on when the tube rectifier warms up. The owner also gets to keep the expensive inverter grade caps which they have paid for.

so you're recommending leaving the tube recto in parallel with the SS recto, essentially using it as an ambience lamp?

[R.G.]

OK, I'll try to be serious.

one adverse effect of using a current limiting resistor in a hifi application is that its going to result in poor voltage regulation under load (aka "Sag"). good for guitar amps, not so good for HiFi. that's the only reason i didn't try the current limiting resistor already.

What makes the hifi tweako beliefs so perverse is that they are harming themselves with them. In this case, the cure I suggested results in marginally less sag and a higher power supply voltage. A tube rectifier can be modelled as a perfect diode, a battery, and a resistor in series. The battery is the average-ish low current voltage loss across the tube and the resistor is the small signal increase in forward drop per unit current through the tube.

In a high-peak-current setup like you get with a big first rectifier, every single current pulse is large, so every AC half cycle has the peak current times the tube equivalent resistance subtracted from it. This means that even though the big capacitance keeps there from being much ripple voltage, the DC voltage level is lower than it would otherwise be because the DC pulses all go through a big resistor.

You can improve this dramatically by doing the rectification in SS diodes which have much lower incremental resistance, then running the resulting ripply DC through the tube rectifier into a second cap. The tube rectifier now carries the average DC current, not a series of peaks, so the average loss through the tube rectifier is much smaller than it would otherwise be.

Effectively, a tube rectifier into the first filter cap causes a higher DC voltage loss than putting the tube rectifier after the first filter cap. The best of both worlds is if there happens to be more than one of those inverter-rated caps in parallel. Using SS diodes into the first one recovers the maximum DC from the input AC, and putting the tube rectifier after the first cap operates the tube in its least-objectionable position. The sag is lower this way. The lowest-sag, highest performance setup using SS diodes, low ESR caps and the rectifier tube is to do the rectification with the SS diodes into the low ESR caps and disconnecting the rectifier tube except for the heaters. Removing the tube altogether lowers the series resistance and improves the supply. If the tube simply has to remain there, paralleling it with SS diodes improves things in a measurable way.

By demanding both a tube rectifier and large-value low ESR caps, the amplifier owner is demanding poorer performance in some way, no matter how you set it up. The tube has a high internal resistance, big offset voltage, and low tolerance for high peak currents. The large-value low-ESR caps force high peak currents. That forces the tube into high voltage drop and short life, and also lowers the available DC voltage to the following circuit. That lowers its power output (if it's a power amp) and induces sag, which presumably the owner does not want.

RG and Rob bring up some good points about the Guru HiFi Mojo myths: some HiFi people are firmly entrenched in their beliefs, which makes it hard to convince them that change may be a good thing. although I had thought about doing some of the "evil" tricks like sneaking a pair of SS diodes into the amp without telling the owner, I didn't want that to come back and bite me. it wouldn't be at all good to have someone going around telling people that i intentionally deceived them when working on their amp. instead, i just told them that the fix that they wanted wouldn't be a good fix, and that they'd be coming back with the same problem. hopefully letting them find out on their own that i was both honest with them, and right in the advice i gave will be better in the end than deceiving them and selling them a load of Guru HiFi Mojo BS.

Your honest is laudable, and is probably what I'd really do if I had to support myself in your position. I would probably try to educate the owner. But I would be ready for him to mark me off as one of those poor benighted souls whose ears simply can't hear good sound.

You can wind up in the same position either way. A dyed-in-the-wool hifi tweako has the same response to being educated that a religious believer has to being told an opposing view - they demonize the educator. A dedicated tweako would hear you say the truth, then take their amp to someone who would give them the "truth" they want to hear, while telling their friends not to patronize you because you're not a believer in subjectivistic hifi. You would get the rep as the repair guy who sadly is deluded into believing that things other than the politically correct Way is good, just because it can be measured and proved logically, and therefore must be wrong in the way that only a theoretically-trained EE can be wrong.

I realize that the last sounds tongue in cheek. I assure you that it's not a joke. Some people can be educated, some can't. The truth --IS-- always better, but some people simply will not see it, and worse are proud of not seeing it.

So maybe the only usable thing to do is to tell the owner the following: that the reality is that they must simply replace rectifier tubes a lot; it's the price they pay for wanting tube rectification and high-value, low-ESR caps at the same time. There are ways around this, but they all involve getting the high current peaks out of the tube rectifier using SS diodes. There are people who will lie to them about that, but you won't.

In any case, Mother Nature has set up these Rules that demand the world works one way, and the owner does not believe that's how Mother demands it work. Getting him and Mother to agree is not an enviable position to be in. Mother CAN'T be pursuaded, and he won't like it.

[R.G.]

so you're recommending leaving the tube recto in parallel with the SS recto, essentially using it as an ambience lamp?

Sorry, I didn't see this one when I responded to your just-earlier note.
Not exactly. In the hookup I suggest, the tube does in fact have some function and effect in the circuit. It actually conducts current and has an effect. Not much of one, I agree, but some. It's a hook to hang your honesty on. The merry glow of the filaments does have a calming effect on hard-core tweakos, though.

But in effect, yes. Having the tube rectifier in the circuit lowers the available DC and increases sag no matter how you hook it up.

Tubes do a very few things in DC power supplies. They have soft turn on and off, so they do not cause radiated RF ringing. This is perhaps the major good, and it may be quite audible in some setups. You may need to use fast, soft-recovery SS diodes or snubbers to suppress this with SS.

The other things are at best benign. They have a higher forward voltage drop. This loses some of the available DC voltage, causing lower power availability. They force you to use lower value filter caps to improve their life, making for higher ripple or short life, pick your poison. They dissipate heat. A lot of it. They delay the turn-on of the high voltage.

Aside from the lack of RF buzz and turn-on delay, there is not much positive there except for the mood effects you note.

[Le Basseur]

What about a tube rectifier's PIV?

Greetings R.G.,
I don't want to bother but I'd like to clarify one of your statements (the second proposition):

You can re-add squishiness by then introducing resistance or a tube rectifier as you like, to taste. The tube rectifier is limited in this case only by its peak average current and dissipation.

If I get it right,you're talking about the HT SS recto followed by a tube rectifier in series,placed between the 1st and the 2nd cap or between the 2nd and the 3rd,depending on the choke/resistor placement.
This arrangement is pretty rare but I've seen a few schematics using an anode-paralleled 5R4WGB or one of those TV damper diodes (anode on top of the glass bulb).
...wich raises a legitimate question:
The only in-series tube rectifier's data to be concerned are the peak average current and dissipation,as you said?...or,should we think about PIV,too?(as I saw in different catalogs,the tube's PIV is marked as "Uf" or otherwise...no standard here).
If we're dealing with a low-to-moderate HT rail,I'd not be worried,but putting a humble EZ rectifier in a +600...700VDC as a series-pass ...I'm not that sure it's a safe thing to do.
Thanks!
Regards,
Le Basseur

[R.G.]

Greetings R.G.,
I don't want to bother but I'd like to clarify one of your statements (the second proposition):

If I get it right,you're talking about the HT SS recto followed by a tube rectifier in series,placed between the 1st and the 2nd cap or between the 2nd and the 3rd,depending on the choke/resistor placement.
This arrangement is pretty rare but I've seen a few schematics using an anode-paralleled 5R4WGB or one of those TV damper diodes (anode on top of the glass bulb).
...wich raises a legitimate question:
The only in-series tube rectifier's data to be concerned are the peak average current and dissipation,as you said?...or,should we think about PIV,too?(as I saw in different catalogs,the tube's PIV is marked as "Uf" or otherwise...no standard here).
If we're dealing with a low-to-moderate HT rail,I'd not be worried,but putting a humble EZ rectifier in a +600...700VDC as a series-pass ...I'm not that sure it's a safe thing to do.
Thanks!
Regards,
Le Basseur

I had to go back to my note to see what I actually said...

That setup was ACvolts -> SS recto -> modest 1st filter cap -> choke ->tube recto -> big succeeding capacitors. The SS rectifiers need to be rated at 2X the AC input voltages in a full wave centertapped circuit. But there's not much PIV load on the tube rectifier in this setup. It only sees the difference between the input cap and the second cap, modified by what the inductor does. If we accept that the second cap is much bigger than the first filter cap, there can't be any huge current discontinuities to make the inductor do flyback pulses. That leaves only the difference between the second cap and the first cap.

That difference will nearly always be in the forward direction, not the reverse direction. The only time when the tube rectifier sees much reverse voltage is when the AC power is turned off. The first filter cap runs down to zero much more quickly than the second filter cap, and the tube rectifier blocks backward current flow. The inductor holds up the tube recto plates for a while until it runs out of energy too. Then the tube recto sees whatever PIV is left in the second cap. This will be less than the normal B+, but maybe not much less if the second cap is truly huge.

Yeah, in those circumstances, the tube rectifier needs to withstand one times the B+ voltage. The humble EZ81/6CA4 is rated for 1000V PIV. So yes, an EZ81 would probably survive in this setup in a 700V DC supply setup. I would be tempted to use a bigger device, because I like to provide bigger reserve safety factors, but it probably survive.