Fuse Sizing

[pompeiisneaks]

I'm curious about something. 

Fuse sizing is a bit of a mystery to me, but thinking it over, I may have come across how it all works in general.

The reason I'm asking, is the Vox AC100/2 I've built seems to chew threw 3A fuses, and I ended up putting in a 6A on accident for a while and although I blew it too, it was something I knew about, i had a high power part of the circuit accidentally right next to ground and it arced and burnt the 6A.  Otherwise the 6A really ran a long time.  ( just didn't realize I'd put that in). 

The original Vox AC100cph has a 6A fuse not a 3A.  The '65  Vox AC100/2 circuit calls for a 3A.

I started trying to make sense of 'why' when something dawned on me.  Isn't this just a simple 'ohms' law issue?   In the old Vox amp, the voltages were suposed to be 30 volts lower.  Am I just now, at a higher voltage with the same circuit, guaranteed to draw more amperage?  If I were to think in those terms, Lets just say we have 3A, and 470  volts it should have at A.  (IT actually has about 500 in this amp)  If that's suppose to be less than 3 Amps, then with that math: R = I/V so I'd have 470/3 = or 156.667 ohms is the theoretical 'resistance' that would draw this current.  Then if I were to push that up to 500 at the same 'resistance' : 500/156.667 that would be 3.19 A.  If the circuit is already near that level, and now we're baselined over it, doesn't that mean I need to raise the fuse?  Would a 6A be too much though?  What should I use?  4, 5?

Or am I doing the math all wrong, and this isn't how you do fuse calculations?

~Phil

[PRR]

> Lets just say we have 3A, and 470  volts

What amplifier takes 3A at 470V ??

That's more power than a toaster or room-heater.

Or is it 3A at 120V, transformed to 470V (which would be nearer 0.77A, which is still absurd for any amp under 200 Watts output)?

[eleventeen]

You're talking about several different things. If a fuse holder is mechanically installed so that it arcs or there is a frayed wire or something weird, then all bets are off.


A 3 amp fuse will blow when more than 3 amps flows through it or tries to. Whether that 3 amps is at 10 volts or 350 volts. A lot could be said about the particulars and of course you know there are fuses that blow super fast to protect delicate electronics and there are plain vanilla automotive type fuses and there are slo-blo fuses that are more tolerant of an inrush current when something starts up cold.


A circuit that calls for 3 amps that is frequently blowing 3 amp fuses has something wrong with it; at least you have to start with this thesis. Because that 3 amps already includes a slosh factor. If you install a 4 amp fuse and lower the blow frequency then you're probably just delaying a day of reckoning when whatever is bad now and blowing 3A's is about to get worse and start blowing 4As. In other words, some OTHER part of the circuit shall become the fuse as you escalate fuse amp ratings. Hopefully it is not a $100 transformer. If it's an ecap going bad, you have a date to replace that pretty soon anyway. So my reco is to just do it while it's convenient instead of on a gig.

[pompeiisneaks]

Both seem to make sense, and that's right, its the 120V side that has the 3A fuse... okay.  (I do have a second 1A fuse on the HT side but its never gone out on me).  So then the fuses going out at 3A are definitely a sign somethings wrong.  I'm still trying to figure out where or what that may be. 

Thanks for the input. 

Why then, did Vox use a 6A fuse on the mains power in the AC100CPH that used to house this amp before?  (its one of those IEC types that holds the fuse inside the power connection area).  I'm reusing the chassis, PT, Choke and OT from it, otherwise its all new things that I've put in it.

The input voltages for either amp are the same, but the current drawn over the PT may be different I guess?  Is the fused from the 120 V input side designed based on the PT or?

~Phil

[trobbins]

So then the fuses going out at 3A are definitely a sign somethings wrong.  I'm still trying to figure out where or what that may be. 

Phil, it sounds like you need to do more than change fuses, or muse over fuse ratings.  As it is the mains side fuse that is blowing, you may not have the experience or tools to determine what is happening, as indicated by "i had a high power part of the circuit accidentally right next to ground and it arced and burnt the 6A."  No one can really advise what is an appropriate fuse or faultfinding path until you properly document what you have - which would normally start with a schematic, and parts information.  And you may need tools that can measure current on the AC mains side, although the old incandescent light tool can be informative for in-rush and current level.

[pompeiisneaks]

trobbins, I have a 5 page thread on the build with everything you're asking.  I was more curious about the 3A vs 6A disparity between the original chassis I'm using and the 1965 vox schematic I'm basing the new build off of, and trying to understand how you determine the amp rating of your fuses.  That's the only real point of this thread.   

~Phil

[trobbins]

:-)


I guess the concern is still whether turn-on is the cause of fuse trips or steady-state current.  Have you measured the easier of those two?


If your fuse current rating is adequately covering steady-state, then choosing a fuse for turn on is a lot more complicated, and apart from appreciating the I-t level of turn-on current, it will be the 'type' of fuse you are using that determines performance.


The linked article doesn't really focus on mains side fusing, but could help identify what your clone has on the secondary side compared to original vox circuit loading.
http://dalmura.com.au/projects/Valve%20amp%20fusing.pdf

[pompeiisneaks]

Thanks I'll take a look. 

I'm almost positive the issue that causes the fuse to blow is power on.  I can't recall more than the once that I mentioned where the fuse blew while it was running, every other time, I go to power on and it just doesn't come off, but it had been working previously.  It makes me think the inrush current is what causes it to burn out.  (almost every time I power on, the PT makes an audible whine for like 1 second and then is fine).

~Phil

[trobbins]

I guess you need to firstly confirm if your parts and circuitry are operating nominally, or if there is some fault or poor performing part that is causing a higher current than nominal.


If everything is nominal, then you have some options depending on your measurement capabilities.  if you don't have anything that can measure mains current then perhaps you go straight to fuse selection.


Given that turn-on is blowing fuses, then are you using T rated fuses, also known as slow blow etc.?  If using a T fuse then your only option is really to go up one current rating increment, until you get stable operation. There are more informative ways to progress, but that depends on measurement or experienced circuitry assessment.


The other option is to guess where your issue lies, and try and change the circuitry to alleviate that issue.  Eg. a thermistor may assist in taming the turn-on current profile, or some more complicated turn-on circuitry may provide the same or better benefit, with the outcome being that you can use a lower rated current fuse.

[pompeiisneaks]

I am pretty sure I'm using F rated (fast blow) as the schematic didn't specify in a way I could say for sure it just said 3 amp fuse.  Is it common to use slow blow on the mains input and then a fast one for the 1 A circuit protection one?

I dont have an easy way to measure the inrush current, no, I don't have a clamp meter.  I guess I could desolder a connection and put my fluke in amps mode and read it.

The circuit seems to be just fine during normal operation, so I don't think the problem is anywhere but on power on.

~Phil

[jjasilli]

Always use slow blow fuses in your amps.  PS & filaments draw more current at turn on.  Fuses protect against amperage (current). 


Current is different on either side of a  tranny ( except for isolation 1:1).  What we have is the Law of Conservation of Energy = Watts.  But fuses protect against over-current, hence are rated in Amps.


E.g. Filament tranny serving 2X 12ax7 tubes @ 6V & 300mA ea.  Total = 600mA X 6V = 3.6W.  That calls for, say, a   4A   750 mA slo blo fuse on the secondary side.  But on the primary side, 3.6W = 120V x Current.  Current = 3.6/120 = 30mA.  A much smaller fuse would be used on the primary side.  Also, the primary side winding wire can be hair thin, since it carries tiny current.  The secondary side winding wire must be thick.


Checkout:  http://www.valvewizard.co.uk/fuses.html

[HotBluePlates]

... I guess I could desolder a connection and put my fluke in amps mode and read it. ...

That seems like a very sensible idea.

Since you have a Fluke, I'll venture to guess you can put it in normal/RMS mode, average mode and peak-hold mode.  You might try measuring the average current draw over a period of time not playing, then again while playing (loud). Then try measuring again in peak-hold mode at switch on and when you're playing at max volume.

You'll likely find some average value of current during normal playing, maybe some higher-level spikes when playing as loud as possible, and possibly similar (+/-) spikes at switch-on. That will guide you to use the correct fuse and choice of fast-blow or slo-blo.

[tubeswell]

Apart from measuring the current draw, the way to calculate appropriate mains fuse rating is to sum all the VA ratings of all your secondary windings, and then divide the total by the AC mains voltage to get the current drawn through the primary - then double that answer to get the ideal fuse rating. e.g (to coin an example from the top of my head):


High Tension = 700VaC @ 200mA = 140VA
Heater = 6.3VaC @ 5A = 31.5VA
Rectifier 5Vac @ 3A = 15VA


Total Secondary VA = 186.5


186.6/PrVaC (say 125) = 1.5A (maximum current drawn through Primary)


So this would call for a 3A mains fuse in this case

[pompeiisneaks]

If I did the math right, for this circuit, I have HT of 500V and the datasheets for each tube at the approximate voltages would be:
2x ECC82 at 1.2mA (4 total for two triodes per ) = 5.6mA
2x ECC83 at 1.4mA (same as above) = 4.8mA
per pair of EL34's at voltage, 30mA so 60mA total

that's 70.4 mA times the 500 V = 35 VA
Heaters 6.3 volts times 7.2A = 45.36 VA
no rectifier tube rectifier, just diode bridge. EDIT <<<

total VA = 80.36 / 125 = .64A, X2 = 1.28 A.  This means even a 3A should be totally fine.  I'm guessing then that I just need slow blow fuses to handle the inrush current.  I'll give that a try and see what I get before trying the Ampere testing of the mains power.  (I do have all the features mentioned on my Fluke, but I have already done some damage once trying to measure current, so I'm wary of it lol)

~Phil

[HotBluePlates]

If I did the math right ... the datasheets for each tube ... 

... means even a 3A should be totally fine.  ...

You can know filament current from the data sheet. Plate/screen current will be 100% whatever conditions you're running in your amp, not what a sheet shows.

Which is why I recommended measuring your actual amp's operation to know.

Or try what the original schematic showed... It just seemed like that hasn't worked out so far for you, which again is why I recommended measuring what your actual amp is doing.

[pompeiisneaks]

Gotcha.  I will attempt it, and the pucker factor will be high ;)  I think I can disconnect the wire between the switch and the mains input (its one of those thin metal post types like for speakers) and get the amp measurements. 

~Phil

[HotBluePlates]

Gotcha.  I will attempt it, and the pucker factor will be high ;) 

Does your meter have clips for the probe tips? If not, go buy some as Fluke typically provides them with all their meters.

If it were me, I'd unsolder a wire running right into the fuse holder, clip the meter leads to the now-hanging wire and a fuse post, and use the existing rubber cover on the lead clips to insulate. Maybe slide something between this connection & the chassis if it's hanging too loosely. Turn the meter on (make sure the probes are in the correct holes for this measurement!) then turn the amp on & measure.

Makes sure once you're done measuring that you move the probes back to the correct holes for V-Ω measurements. Otherwise, when you try to measure volts later, you meter is a nice short-circuit for the amp and you'll pop a fuse inside the meter (and maybe burn something in the amp).

[pompeiisneaks]

I can't access the mains fuse, its inside the IEC connector.  The HT fuse I could do this, but I'm trying to do the mains mostly right?  (I could also do HT to see how the 1 amp fuse there will handle it, I'm sure).   

The one I bought only had standard point leads, I've bought some, though that have either the option of aligator clips or small hook clips.  I could use either. 

I'll give it a try.  I've just not had time to test it.

~Phil

[trobbins]

Phil, measuring mains side current directly by a meter requires a full awareness of the risks and use of measures to avoid being shocked.  Have you assessed what can go wrong, and had advise from a competent person on how to do such a measurement?

[pompeiisneaks]

Phil, measuring mains side current directly by a meter requires a full awareness of the risks and use of measures to avoid being shocked.  Have you assessed what can go wrong, and had advise from a competent person on how to do such a measurement?

I've never had official training of any sort, I've watched tons of videos on it, and been counseled on this particular forum that using the amps measurement, due to being a bit more dangerous, is not recommended.  I tried once and did something wrong and blew up a component (I think it was my tube rectifier) on my last amp.  This is why I'm a bit wary of doing it.

I understand the risk the ammeter is put in line with the circuit and if done wrong can send 120 volts AC into the wrong place and damge the amp, the meter, or even me. 

This is part of why I am hesitant to do it. 

If someone can point me to a video/site/etc that goes over safe practices for measuring amps, I'd love to know about it. 

meanwhile I've always used the 'ohms law' techniques.  Get a known resistor, measure the voltage drop, and calculate the current.


~Phil

[trobbins]

Imho we all need to be aware and not take short cuts - many forum threads on this topic - and some forums even ban discussion/tests/circuits related to mains side due to the diy risk. 

No one post can address the topic adequately.  I can but alert that one risk is personal shock, which can be somewhat managed by an appropriate test environment.  Earth leakage protection is seen as almost mandatory nowadays.  Having a low level current limited power source is a good aim.  Setting up the test so that hand probing is not needed is a good aim - that may be achieved by securing the test leads and other connections via fixed terminals - and then applying power without the need to touch amp/probes/meter.

[eleventeen]

Another approach is just to go find an amp with a similar tube complement and use the same sized fuse it does.

[PRR]

> per pair of EL34's at voltage, 30mA so 60mA total

That's idle.

Your car drinks more gas on the freeway than when idling.

A couple EL34 at 500V is probably 50 Watts FULL output, perhaps 80W-100W of DC in.

Total VA could hit 130VA. 130/120 is 1+ Amp. 2A fuse may last many years.

True, your problem is at turn-on, which ought not to be FULL power.

Which makes me wonder- is your grid bias slow to come up? If the EL34s get the +500V before they get the -45V, they will suck HARD. Monitoring the bias voltage at turn-on is less dangerous than splicing into the 125V power system.

[pompeiisneaks]

Ahh true, I was looking at typical operating parameters, not extreme states.  Even with your notes, the 3A should cover that.

Where do I measure the grid bias?  On the bias circuit somewhere?  Or just directly at the grid itself?  Or is this just another name for doing the 1 ohm measurement at the cathode? 

I know the cathode current comes up really slowly.  Not sure if that's even related(seems like it may be, but I'm still learning all the terms), but I've had my probes on one of the four cathodes over the 1 ohm resistors from pin 1/8 and it can take quite a while to come up.  It seems like several seconds before it goes from 0 to about 10 or so mV then slowly keeps going until it hits the expected range within maybe 30 seconds or so. 

Is that what you're talking about?

~Phil

[eleventeen]

It *is* what PRR was talking about. If it takes anything like 30 seconds to come up, that isn't the issue.

[pompeiisneaks]

well the fuse burnout happens at like 10ms to happen ;)  I've read a lot of things about inrush current being high with power on, why?  I think someone said I could put in a thermistor.  If I do, where do I put it?  How do I size it/determine the right kind?  or is a slow blow fuse the best fix?  Right now, it seems like the consensus is that I should use a slow blow, and then do some measurements of current, but PRR seems to agree with the math that in theory the amp is way below the 3 amp for normal operational use. 

~Phil

[eleventeen]

Have we asked the question/gone through the exercise of turning the amp on to a standby state versus "all on"?


The problem with blowing fuse after fuse is, they're not that cheap any more like they used to be; and a blown fuse is frustratingly non-diagnostic. It just blows and there you are. Yes, you know something went overcurrent, but that's pretty vague.

[pompeiisneaks]

I have no standby switch on the amp.  Per what I'd read here: http://www.valvewizard.co.uk/standby.html

It seems he's not the only person that sees a standby as a relic of a misspent fender mistake that won't go away :)  I have also heard others confirm it many times.

~Phil

[Willabe]

Where do I measure the grid bias?  On the bias circuit somewhere?  Or just directly at the grid itself?

Yes and/or yes. 

Or is this just another name for doing the 1 ohm measurement at the cathode?

No.

Measurement across 1 ohm K R is for current.

[pompeiisneaks]

Thanks willabe, I'll give that a try.

~Phil