Immortal Mods: additional fuses, where to put the fuse holders?

[catalin gramada]

Why don't to solder on the wire in line and put a varnish over it? It will not be replaceable but bet you no need it for next hundred years.

[trobbins]

I commend your efforts to add appropriate fusing on the secondary side.

One half of the coin is to add the fuse - the other half is to use a fuse rating and type that aligns with your particular amp.

For heater use, that choice for rating may be somewhat obvious in that valve heater currents are generally known, and you would add some tolerance margin to allow for heater in-rush which can be 4-5x and may take a second or so to get to 2x. That inrush normally points to a T (timelag) type of fuse, and that type may be better if a soft fault occurs, but that is when it can get a bit complex to sort out. You really need to identify what spec the fuse is made to - as that could be unknown (ie. automotive), or UL284 or IEC60127 if a 5x20 or similar style.

For B+ use, PSUD2 is I reckon your best friend, along with a measurement or two, to start the selection process. Then it comes down to what fuse model and type and spec you can get. Of course you can wing-it with a value that you have to hand, but its good to appreciate what that means.

Some additional fuse selection discussion and selection design is in link.
https://dalmura.com.au/static/Valve%20amp%20fusing.pdf

[pdf64]

NTC thermistors are a handy thing to remove start up current surges, and avoid the need to use T type fuses.
They can be used in the primary and secondary circuits.
Modern Fenders tend to have them in the primary circuit, and many of those use regular F type fuses.
Their downside is that if the amp has been in operation but then get switched off accidentally, flipping it back on too soon (ie before the thermistor has had chance to cool) can pop an F fuse.

[R.G.]

As a side note, I mentioned solid state current limiters. There is a suitable circuit here:
Tube Amp Current Clamp

This circuit prevents the current from exceeding the specific current setting - ever. Well, at least as long as the limiter circuit doesn't overheat, die, and let the current go to infinity. It is possible to integrate an opto isolator into this that would signal a circuit back down at ground voltage to shut things down if the overcurrents exceed a preset time.

This may sound like a PITA, but it is dramatically more predictable in its protection than a fuse can be, and opens the door to even more sophisticated protection schemes.

[trobbins]

Its worth appreciating how intrinsically fault tolerant a heater supply is nowadays.

Vintage heater supplies were often chassis grounded on one side of the heater voltage - so a single fault to chassis would cause a direct short. Similarly, a CT taken to chassis was also a single fault situation. Wire insulation wasn't as good with higher temperature or age. Although one advantage was that octal bases had lots of clearance between pins.

Valve heaters don't fail internally to a bolted short circuit - which is a major difference to B+ related faults due to internal valve faults, which can also cause collateral damage to an output transformer.

A humdinger, or elevated heater supply provides fault tolerance, as one fault to chassis does not cause a short circuit.

A modern production amp, or good diy effort, is likely to take enough care to make heater short circuit situations very rare.

I should do some heater inrush measurements on common valves, just to check how spread the response is, and how influential the heater winding current rating is, and how limited a heater winding peak current is when fully loaded to its rated continuous level (both at turn-on and if a fault occurs during operation).

[mikepukmel]

Thanks everyone, for the links, and info. Reading...

[trobbins]

I'd place a caution against using automotive blade fuses for heater supply protection.

A check on the typical industry standards, and datasheets from major manufacturers, for the current-time performance characteristic shows a wide range in 'operate' time for current levels 1.6 to 3.5x the fuse current rating. It is likely that a blade fuse could blow during heater in-rush, even if the fuse rating was 2-3x the heater supply current rating. That's not to say that a particular fuse would blow.

The link indicates the range of the peak current during in-rush for a variety of common audio valves. What is not shown is the time response of that in-rush current.
http://www.gammaelectronics.xyz/audi..._filament.html

[R.G.]

And I view that as another vote for active solid state watching for overcurrents.

Fuses are extremely, extremely hard to predict blowing currents and times for. A fuse's rating is, as I've said before, not the current it will blow at. It's the current it's guaranteed NOT to blow at. A 1A fuse is guaranteed to carry 1A, and to open at some current above that. the manuals on choosing fuses for blowing characteristics rapidly get back to talking about the I-squared-T through the filament. That's the time function integral of energy delivered to the filament, and is related to the power dissipated by the filament resistance and the heating time until it gets to a melting temperature. That has to be compared to the details of how heat gets out of the filament to cool it. Estimating melting time rapidly becomes a complex heat flow calculation/modelling in three dimensions over time.

The best mere mortals can get to is to use charts and graphs, estimate safety ranges, and pick one.

The advent of $0.50 microcontrollers with many channels of A-D and easy to apply opto isolators makes this all get simpler. You can measure current in real time, do moving-average RMS calculations (even if they're table lookup) in real time, and make decisions based on several criteria when to stop the incoming AC power line, and when to allow it to come back on.

$0.50 is really close to the ones' cost of a single fuse.

[R.G.]

Just as I hit "submit" on that, I remembered something else.

I did the math on whether you can count on a heater fuse to blow the AC line fuse. Turn out, it may blow it, but you can't count on it. The heater lines can get wildly hot and not blow the AC line fuse, so they cook for a long time, and eventually burn out the heater winding insulation.

But that takes a long time. Heater windings in particular have a high thermal capacitance, so it can take a long time of overcurrent before they're damaged. Precision is less necessary for heater windings than for other, smaller windings.

[J M Fahey]

I don't think that writers like him were actually PAID for their articles but would receive a free stack of magazines that they could sell in their shop.

I have actually written tons of articles in Guitar Magazines so know the deal firsthand: they actually "pay" you, even if a symbolic amount, for copyright/legal reasons: that way they "own" the article , you lose rights so you can´t ask/sue them later for whatever $$$$ you want for it; also if later , say, you write a book including that article or want to republish it in another Magazine, you need their written permission.
You *can* post a link to said article (or a scan) in your own Net page, but mentioning their Magazine.
One usual payment type does not involve cash but they publish your Ad for free, actually in exchange for the Article and this is clearly mentioned somewhere, usually they bill you for the Ad but write: "paid for with xxxxx published in the same issue" or the equivalent.
Don't ask for half or full page ads of course, usual is 1/8 to 1/16 page Ad ... which is actually quite visible.

Magazines which are actually built around free lance writers, say 3 Guitar players explaining Scales or Licks or onstage setups , a couple tech articles (Electronics, instrument setups, Recording or Live sound tips, etc.) plus, say, a singer explaining Vocal exercises or whatever, or a band Manager explaining how to get gigs or transport cost or whatever , will often dedicate a full page, divided in 8/12/16 small "contributor Ads", or if not enough to fill a page, will spread them all over the Magazine, to fill white space.

The real payment of course is that those little ads bring lots of customers and in general keep you in the Public eye

FWIW I often get people calling me and mentioning some Tech article I wrote as far back as the 80´s , go figure.

Similar to what Justin Thomas says about RG´s Premier Guitar articles.
Guys here also keep clippings or photocopies and now and then upload some old scan ... it doesn´t hurt at all

[trobbins]

A 1A fuse is guaranteed to carry 1A, and to open at some current above that.

Just to embellish a hassle with fuses - for the generic 5x20mm or *AG style miniature cartridge fuses, the fuses with IEC60127 standard compliance have that "1A fuse is guaranteed to carry 1A" rating, whereas a UL248 (ie. US centric) compliant fuse is only guaranteed for 0.8A continuous from a 1A fuse. And many people would I suggest have no clue as to what standard their fuses are compliant to (if indeed they are quality made fuses).

[mikepukmel]

I commend your efforts to add appropriate fusing on the secondary side.

One half of the coin is to add the fuse - the other half is to use a fuse rating and type that aligns with your particular amp.

For heater use, that choice for rating may be somewhat obvious in that valve heater currents are generally known, and you would add some tolerance margin to allow for heater in-rush which can be 4-5x and may take a second or so to get to 2x. That inrush normally points to a T (timelag) type of fuse, and that type may be better if a soft fault occurs, but that is when it can get a bit complex to sort out. You really need to identify what spec the fuse is made to - as that could be unknown (ie. automotive), or UL284 or IEC60127 if a 5x20 or similar style.

For B+ use, PSUD2 is I reckon your best friend, along with a measurement or two, to start the selection process. Then it comes down to what fuse model and type and spec you can get. Of course you can wing-it with a value that you have to hand, but its good to appreciate what that means.

Some additional fuse selection discussion and selection design is in link.
https://dalmura.com.au/static/Valve%20amp%20fusing.pdf

I spent quite a bit of time scanning for more detailed info on fuses, but couldn't find much other than the bare bones ratings. Yeah, definitely appreciate the issues. Would be bad to go through all this trouble, and still be able to smoke the transformers before the fuses blow. Again, great article.

[mikepukmel]

Just as I hit "submit" on that, I remembered something else.

I did the math on whether you can count on a heater fuse to blow the AC line fuse. Turn out, it may blow it, but you can't count on it. The heater lines can get wildly hot and not blow the AC line fuse, so they cook for a long time, and eventually burn out the heater winding insulation.

But that takes a long time. Heater windings in particular have a high thermal capacitance, so it can take a long time of overcurrent before they're damaged. Precision is less necessary for heater windings than for other, smaller windings.

From your calculations, how far over the rated N amps (whatever that is) is the draw that takes a long time to burn the winding but not blow the fuse? its sounding more and more that, as others e.g. Juan has said, might not be worth it to fuse the heater windings?

[pdf64]

The higher the current, the shorter the time the fuse will hold for.
Find a manufacturer's fuse info sheet to see.
It will be 'typical characteristic' rather than a spec.

[R.G.]

From your calculations, how far over the rated N amps (whatever that is) is the draw that takes a long time to burn the winding but not blow the fuse?

The calculations were not general-case sorts of things. The context is that a particularly-adverse-to-learning guy in this forum decided that it was flatly impossible for a fault in an amp PT secondary NOT to blow the AC line fuse, and concluded from this seat-of-the-pants assertion of his that secondary fuses were never needed, and an AC mains fuse was always enough. I measured a few PTs and looked up wire resistances for some PTs on line, and then constructed a plausible example.

When you short a transformer secondary, the current is limited by a few things. These include:
- resistance of the short itself, as all "shorts" are not necessarily good, solid, welded-on metal connections; if it's a short to chassis, for instance, the resistance of the chassis itself is added in, and that may not be negligible, especially for an oxidized aluminum surface or a thin steel chassis.
- resistance of the secondary winding and the wiring to the location of the short
- leakage inductance of the PT from primary to secondary
- resistance of the primary winding and the AC mains wires, AC mains fuse, switch contact resistance, all that minutae.
Each of these numbers is unique to the PT and its installation. You can make realistic estimates of the ranges of each of these.

With that as a base, you can calculate a reasonable secondary and primary current for a guessed-at short circuit, and then make a reasonable estimate of the change in primary current for a secondary "short". The heaters are the bad ones because of the extreme turns ratio between primary and secondary. The turns ratio for a 120Vac mains and a 3.15V (half of a 6.3V) heater winding is 38.1. So 38 amps in a heater half-secondary increases the primary current by 1A. So for a heater winding with a grounded CT, you can get a current that is limited to a significant fraction of 100A by the resistances of the short, wiring, and secondary winding that only bumps the primary up a little. Especially if the amp is simply sitting there, not being run at full output, and moreso if the primary AC fuse has been sized about 25% higher than max normal current, the primary current may well be an amp or two below its stated fuse rating, and the extra amp won't even push the primary fuse out of its rating, ever. Or the extra amp may just kick it a bit over the fuse rating, into perhaps 120-150% of its rating.

Primary fuses are universally time delay types because they have to sustain the power-on inrush current. So they're designed with a slow response to sudden heating. As a result, the primary may take a long time to blow. As in hours for just a little over its rating, or it may never blow if the added amp or so has not even pushed it above its nominal rating.

So the half-secondary can simply sit there and cook, heating just the heater winding and whatever's next to it, until the insulation overheats and the wires touch. At that point, smoke comes out and you get a chain failure that eventually opens the AC mains fuse, but by then the PT is effectively dead.

its sounding more and more that, as others e.g. Juan has said, might not be worth it to fuse the heater windings?

It all depends on what level of risk you're willing to take. It is quite true that the majority of guitar amps run for several decades, maybe for a century without sustaining a half-heater short. In concocting the "Immortal Amp" idea, I was struck by the thought that some things in a guitar amp regularly need replacing. Tubes are obvious, but electro caps do so as well, and then other things have semi-predictable failure modes. So I did the thought experiment native to engineers - if you can't make it last forever, make it easy to fix, and provide protection for things that are very expensive to replace.

If you put in heater fuses, you can be reasonably certain that a heater short won't destroy your PT. But then the odds are that if you wire the amp correctly and use good quality tubes, it wouldn't ever short anyway. Shoot, there's even a mechanism that ages fuses; Repeated near-maximum currents flex the filament and can make it simply break. But fuses are cheap, PTs are not. You have to decide what level of risk you want to assume.

If you think you'll never get a half-heater short when you aren't there to notice it and shut things down, fuses are unlikely to help you. If you want to make really sure you won't ever have the problem, fuses are cheaper than PTs. That is "worth it" is a calculation YOU have to make.