[Amps] HV Fuse

Sat Sep 21 09:50:11 EDT 2019

Vic,

I think you're suggesting that small differences in a HV fuse's open time, even if within spec, can effectively create a delayed-action or slow-blow fuse type when multiple fuses are paralleled.  That would require looking at the specs for fast, slow, and delayed-response fuses to determine if it's even a factor.   Possibly the delay in paralleled fast-blow fuses doesn't come near the delay of the other two types.

I'm fine with the main HV fuse used ahead of the glitch resistor.  The amp uses a 3CV1500A7 and its power supply pretty much limits power to 1500W max.  It was designed ca. 1970 under the old FCC power rules.   In this case, a single 900 mA/5KV microwave oven HV fuse should work fine.  There's ample room in the RF deck to mount it on glazed porcelain standoffs.    

The bigger issue is the use of a HV fuse between the transformer secondary and C-input filter in an attempt to save the rectifier stacks during a fault event that occurs after the filter cap and before the glitch resistor.  I've not seen it implemented anywhere but was mentioned by VE7RF in this post:

 http://lists.contesting.com/_amps/2017-11/msg00149.html

Even with step-start, I'm skeptical of its effectiveness due to the high initial surge current at start-up.  So, perhaps just relying on the 20A ABC mains fuses is still best. 

The Alpha 70V uses four Semtech SCH5000 HV rectifier blocks.  They're rated by Semtech at 5KV PIV but only 0.5A at 55 degs C, or 1.0 A with continuous forced air at 55 degs C.  This is a 1960s design and barely meets the margins for this amp.  I have spares so I used them in the repair.  But, if it occurs again, I'll either use modern equivalents with better specs or a string of 6A10s on a new, custom PC board.  

Paul, W9AC

-----Original Message-----
From: Vic Rosenthal <k2vco.vic at gmail.com> 
Sent: Saturday, September 21, 2019 12:40 AM
To: Paul Christensen <w9ac at arrl.net>
Cc: amps at contesting.com
Subject: Re: [Amps] HV Fuse

It would seem to me that if the fuses had slightly different resistances then one would blow first, followed by all the others. 
Interesting question: which would go first, the highest or the lowest R?
The current would divide in inverse proportion to the R, so the lower-resistance one would get more current. But the one with the highest resistance would get hotter for a given current.
I think the answer is that the lower one would go first. The heat is proportional to I**2 * R.

Victor 4X6GP 

> On 20 Sep 2019, at 22:38, Paul Christensen <w9ac at arrl.net> wrote:
> 
> I don’t think fuses in parallel are a good idea :-). But probably you 
> weren’t serious
> 
> Vic, I am serious!  Actually, I would use them in parallel to determine the overcurrent point then try and fine a single fuse of adequate size.  That said, I have seen as many as three used in parallel in another amp design.
> 
> Paul, W9AC
> 
> 
> . I don’t know what the present rectifier modules are,  but rather than trying to protect them with a fuse, I would just make up a few strings of 6A10s, which are basically indestructible. Then you size the glitch resistor so it won’t blow up before the primary fuse blows in the event of a flashover. A nice 50 watt vitreous enameled one of 30-50 ohms should work.
> By the way, I don’t know what the tubes are in that amp, but I presume one or more triodes. A fuse in the plate circuit of a tetrode will cause the screen to ignite like someone hit it with the heat ray in War of the Worlds, unless there is a screen trip circuit.
> 
> Victor 4X6GP
> 
>> On 20 Sep 2019, at 22:04, Paul Christensen <w9ac at arrl.net> wrote:
>> 
>> Yesterday, I lost all four HV bridge rectifier modules in my Alpha PA-70V
>> vapor-cooled amp.   After replacement, HV returned but whatever caused the
>> fault event also took out a 2N3053 switching transistor that controls 
>> a 8.2V Zener bias diode from the EBS circuit.  I used the opportunity 
>> to change it to a more robust transistor in a TO-220 package that 
>> effectively doubles Ic and Vceo.  The transistor is already configured as a "discrete Darlington"
>> with a 2N4401 ahead of it so the shock didn't get very far on the EBS PC
>> board.    
>> 
>> Going back through the list archives, I see mention of using a HV 
>> fuse just ahead of an amp's glitch resistor.  VE7RF recommended a 
>> second HV fuse placed between one leg of the HV transformer secondary 
>> at the C-input filter cap.  Sounds like a nice way to potentially save the HV bridge diodes.
>> 
>> I have sized the glitch fuse adequately and will use a HV microwave 
>> oven type as discussed in the archives.  However, because of some 
>> unknown variables, it's a bit tricky to calculate the HV 
>> transformer's secondary fuse.  I can use PSUD II software but primary 
>> and secondary HV transformer ESR are unknown variables.
>> 
>> I'm also a bit skeptical that a fuse on the HV secondary will work as 
>> intended due to the extreme initial C charging current (35 uF/5KV 
>> filter cap), even though step-start will limit the initial inrush and 
>> that too factors into the equation.  I could just keep stacking 900 
>> mA HV fuses in parallel until they don't blow but that isn't an 
>> intelligent approach to a solution.  Any guidance here?
>> 
>> I don't see sand-filled ceramic HV fuses in this category and I dread 
>> the thought of cleaning up glass shrapnel in an amplifier.  I'm 
>> thinking that Teflon heat-shrink tubing placed only over the glass 
>> portion should keep the shattering damage under control.
>> 
>> Paul, W9AC
>> 
>> 
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
> 