Dynaco Durability, VTA-style

A personal tale of not-quite-woe:
I had been blissfully blasting my 18 month-old VTA ST-120 into my vintage 1990 B&W Matrix 803s for the duration, knowing full well that the crossover caps were way past due for replacement when, BLAM! - left channel turns from bliss to horror, rectifier fails spectacularly, fuses blow, and pandemonium ensues. Knowing enough to know what I don't know, I contacted Audiobill in a panic. Sent unit to Doctor Bill.
This was two weeks ago. Apparently, from the evidence, a crossover component failed, blew my left midrange driver, presented as a virtual short to the ST-120, and cut me down in my prime. Upshot: due to B&W's crappy policy, replacement speakers for this vintage are unobtainium, and there are NONE ANYWHERE. New speaks on the way - the touted Tyler Acoustics budget Halo 2. But that's another tale....
Here's the punch line: there was NO DAMAGE to anything on the ST-120, other than the loss of two 5AR4s. Yes, the KT-88s are fine, the left OPT is fine, all components are fine. This is one durable beast, the ST-120. Hats off, again, to all parties responsible - Bob, Roy, Audiobill, and my boy deepee99, who turned me onto Tyler speakers. Anybody got a tissue???

Glad to read your amp is fine. You lost two 5AR4? That is a typo error or do you have a special build?

BNR_1 wrote:Glad to read your amp is fine. You lost two 5AR4?  That is a typo error or do you have a special build?

Not at the same time - two in succession. Having not yet realized the source of the problem was the speaker itself, I powered up several times, trying different output tubes. One of the 5AR4s was brand new. Total losses: 2 5AR4s and three fuses.

Got it. I hope these were not NOS or healthy OS Mullards.

BNR_1 wrote:Got it. I hope these were not NOS or healthy OS Mullards.

No, Genalex Gold Lion new builds - still, $100 worth.

Kentley wrote:

BNR_1 wrote:Got it. I hope these were not NOS or healthy OS Mullards.

No, Genalex Gold Lion new builds - still, $100 worth.

May I ask what sort of fuses were in place when the rectifier(s) protected them?

Peter W -- Interesting wording!
These are what I had in my arsenal. https://www.amazon.com/gp/product/B008OZ4U7C/ref=oh_aui_search_detailpage?ie=UTF8&psc=1
I'm aware of your distaste for slo-blo's. Perhaps you, and others, can further elucidate this, and provide some concrete alternatives.

OK - Here goes.

RANT WARNING RANT WARNING

Fuses are one of my pet peeves as they are so poorly understood as to be dangerous at least 80% of the time in my direct experience. So, a bit of a lecture follows as well.

There are three types of fuses in very broad brush:
a) A conventional fuse - consisting of a single element designed to 'blow' at a specific load with very little overshoot.

b) A "slow-blow" fuse - one that is designed to take an overload for some period of time being a function of excess load and overall fuse rating. Typically, these fuses are an inexpensive (to the fuse maker) way to protect real-estate whilst not blowing too often. They are, again typically, used for loads that come on strong, then taper down during normal operation. Tube equipment meets this definition. Overloads may occur at any time during operation, as long as there is some level of cooling time between overloads. And, small overloads may last a _VERY_ long time before the fuse actually blows. The cost of these fuses is only fractionally greater than type a).

c) Dual-Element fuses - these are fuses with an element that handles a potentially very large momentary overload, such as a motor starting or contact surge, and then steps aside for a normal fuse element. The overload function happens ONCE per cycle. Any subsequent overload and the fuse will respond as with type a). These fuses are quite costly relative to normal fuses, often 3 to 10 times the cost.

There are variations all over the place, including quick-blow fuses (very fast response), low-resistance fuses (silver or silver bearing elements) and many more.

HOW TO SIZE A FUSE:

The process is as follows:
a) Obtain a metered variac, fine-pitch ammeter, resistance box + VOM, or other device by which you are able to measure actual current drawn by a device at full load and operating voltage.
b) Ideally, you will devise a dummy-load for amplifiers that allows you to run them at/near clipping to measure the true maximum load.
Measure the load that will run across the fuse using this (these) devices.

Round it up/down to the nearest 1/2 amp, but shade low. So, 4.1A = 4.0A. 4.3A =4.0, 4.4 - to 4.8A = 4.5A and so forth. Remember, you are measuring at/close to clipping, which should not be a normal state.

Obtain DUAL-ELEMENT fuses at that rating. Install them.

Notes:
1. Dual-element fuses do not like being short-cycled. The delay element *will* fail if short-cycled, so for testing (and testing only) a slow-blow is acceptable as long as you are right there with your hand on the throttle (switch).
2. Dual-element fuses are wearing parts, but less so than conventional fuses. So, if you get the fuse very close to the maximum operating value, its service-life will be measured in hundreds of hours, not thousands. But, a cheap investment as compared to a 5AR4.
3. The load-to-failure curve is not linear, but logarithmic - the higher the overload the quicker the failure.
4. An amp with a tube rectifier has two surges, very brief large with a very big overshoot (3 x quiescent), one very slow, but no overshoot. The first when the cold filaments heat up, the second when B+ comes on line. You may watch these if you meter the start.

The "why" of it:  Imagine a slow-blow fuse in a cool location with a 10% overload. The fuse will likely tolerate such an overload for many minutes (using my metered variac, I once tested a 3A slow-blow at 3.3A for over 30 full minutes). 3A @ 120V = 360 watts. 3.3A @ 120V = 396 watts. That is 36 watts over, about 5 x a standard 7 watt candelabra lamp. That is a LOT of heat that is going somewhere - somewhere it is not supposed to go. Keep that up for some time and *POOF*.

My standard ST70 draws about 150 watts at full load (rated at 190 watts). I use a 1.25A dual-element fuse. The manual calls for a 3A fuse. I have not had the 1.25A fuse blow on overload yet over a couple of years, but then I baby the amp, and we are running at about 118V generally.

Good luck with it.

p.s.: Slow-blow fuses are created in the nether regions for the bedevilment of their users, and (almost) no other legitimate purpose.

Wow. Thanks for the very detailed answer, Peter W. Lots of useful info.
I made a very cursory search for dual-element fuses. It appears that I would need to switch out the fuse holder as well, as these DEFs seem to be over-sized. Is this the case, anyone? Opinions and facts welcome. (Bob???).

http://www.cooperindustries.com/content/dam/public/bussmann/Electronics/Resources/product-datasheets/Bus_Elx_DS_2044_MDQ_Series.pdf

All the sizes you might need.

Type MDQ are the style you are looking for for standard glass fuses.

Peter W. wrote:http://www.cooperindustries.com/content/dam/public/bussmann/Electronics/Resources/product-datasheets/Bus_Elx_DS_2044_MDQ_Series.pdf  

All the sizes you might need.

Type MDQ are the style you are looking for for standard glass fuses.

That works. Thanks. Anyone else care to add?

Oh, and N.B., Peter W. - my amp blew both the fuse and the rectifier. Hard to say what kind of cause/effect went on there. But I get your point.

Kentley wrote:Oh, and N.B., Peter W. - my amp blew both the fuse and the rectifier. Hard to say what kind of cause/effect went on there. But I get your point.

When the rectifiers slagged, they dead-shorted the transformer secondary, which, finally, was enough to blow the slow-blow fuses. Put simply, you got lucky! Usually excess current shows up (as heat) in either of two places, the output transformers or the power transformers - knowing that _all_ of it comes through the power transformer anyway.

Occasionally in the repair shop, I'd get "customers" looking for a "NO-BLOW FUSE."  

This usually meant they blew up their transistor receiver's output stage, stuck a new fuse in there, and the fuse blew again (the output stage couldn't "blow again," it was already blown!).  Hopefully they did not try a 20 amp automotive fuse.

Thank you, all, for your info. I now have enough to make an informed decision. First , I will only use top-grade slo-blo's (I ordered Littlelfuses). If problems persist, I will move to the dual-cores. I suspect that the complex series of events I encountered will never occur again. Knock on iron.

http://www.littelfuse.com/~/media/files/littelfuse/technical%20resources/documents/application%20notes/littelfuse_application%20note%20slow%20vs%20fast%20auto%20fuses.pdf

This makes worthwhile reading.

Note that most slow-blow (vs. dual-element) fuses may last as long as 1,800 seconds (30 minutes) at a 10% overcurrent, but no less than one (1) full minute. This exactly mirrors my own experience when I was "testing" the concept of slow-blow vs. dual element.

http://www.cooperindustries.com/content/dam/public/bussmann/Electrical/Resources/solution-center/technical_library/BUS_Ele_Tech_Lib_Dual_Element_Fuses.pdf

Also states that the typical fuse must be rated at 300% of surge current (approximate condition experienced by the OP), whereas a dual-element fuse may be as low as 125%.

Further reading will show that whereas both types of fuses fail in approximately equal times by overcurrent, by percent, as dual-element fuses may be sized much more closely to the actual load, they will provide a larger level of protection by virtue of that fact. Note that in the OP's case, a 6A slow-blow fuse was used. This fuse can hold for a minimum of 120 seconds (2 minutes) at 200% of rating - that would be twelve (12) amps - whereupon it would be a close-run thing between the rectifier and any given transformer as to which fails first.

On the other hand, a 2A dual-element fuse would fail similarly at a 4A. In the case of my mostly-factory 70, my 1.25A fuse would fail very quickly at anything over 2.5A.... you get the picture.

I am advocating a fuse that, while it may be relatively costly, will do more than protect the real-estate. It actually stands a chance to protect the equipment to some degree. Nothing is perfect, but I have, far too often, experienced the phenomenon of the $40 tube or the $200 transformer failing while protecting the $0.40 fuse. Guys and gals, use what you want. I promise that I will never write "I told you so".

Back in the day, even as notoriously cheap an individual as David Hafler did, in fact, supply dual-element fuses in much of his equipment. The cost relationship was then, as now. He never would have done so unless there was good reason.

https://www.flickr.com/photos/66282888@N00/10709830353/ shows a blown OEM fuse from a 70. Note the remains of a spring - part of the Dual Element system.

How well would it approximate the benefits of a dual element fuse to just put 2 fuses in series?

Tube Nube wrote:How well would it approximate the benefits of a dual element fuse to just put 2 fuses in series?

Not at all. As each fuse would have its own behavior independent of the other. So, if a fuse would blow on the starting surge, it would blow whether in series or not.

I have used 3-amp slow-blows in the M-125s with no issues, rather than the recommended 5. This is of course with tube rectifiers. Doubt they would survive the surge from a WS-1-type no-delay s/s rectifier.

deepee99 wrote:I have used 3-amp slow-blows in the M-125s with no issues, rather than the recommended 5. This is of course with tube rectifiers. Doubt they would survive the surge from a WS-1-type no-delay s/s rectifier.

You might be shocked (word chosen deliberately) at exactly how much a 3A slow-blow will accept before blowing. I will continue to represent that the entire concept of the wire-wound slow-blow fuse is a creation from the nether regions except in very, very, very few applications.

Peter W makes some excellent points, and although I have seen the dual-element fuses, I believe as Bob and Peter both mentioned that I have actually seen these in very old Dynaco gear!
So maybe 50-60 years ago they were reasonably priced. As Bob mentioned, I also normally use slo-blo fuses (sorry Peter) realizing that it takes a full 2 minutes to blow at twice the rating.
I don't think there are alot of cases where something will draw double current without blowing something else first (like underrated resistors).
If you truly have a full short - then a slow blow WILL blow in less than 5 seconds. This protects the circuit if you do something stupid like wire up the power supply wrong,
or a power transformer shorts internally. The likely protection Peter talks about is something more like an quad cap that is failing due to decreasing internal resistance, which makes it run hot,
which makes it eventually fail. It could take an hour to fail, or it could take a year to fail. Meanwhile, it's running hot, and so extra current is being wasted. This is where the dual-element fuse will be superb.
Reading and understanding all the specs, an ST70 could easily use a 1.5amp or 2amp slo-blo rather than a 3a slow-blo, since it only uses 170-190w which is around 1.5a.
A 1.5 amp slo-blo in an ST70 would still run for an hour with a load of 240watts (2 amps). Hopefully if that were ever the case, you'd notice that either the quad cap or power transformer was running awfully warm!!
On the other hand (here you go Peter) a $12 fuse is super cheap compared to even a $20 rectifier or output tube!! And most likely, will last you 50 years!

So, folks: If I were to try a dual element fuse in the ST-120, what value would be optimal? - I don't have the equipment to do Peter W's testing.

My Hafler DH-200 kit that I built in 1981 came with an odd-looking fuse. I now know that it was a dual-element fuse! So I can confirm that Hafler used them. I found a source for 5 A dual-element fuses for $1.71 a piece which seems pretty reasonable.

So, can a dual-element fuse replace any fast-blow fuse at the same amperage rating for additional safety?