Paul, its simply that there is not one solution for all amps.
At one end you have the Heath SB series which are extremely transformer
limited and have a fairly low value of filtering capacity. They can
easily live with a 15-25 Ohm resistor in the 20-25W range which should
survive any number of arcs. The simple fact that the PS survives as well
as the filament circuit components speaks volumes for the rather low
currents involved in an actual internal tube arc. Oft times even the
grid chokes survive undamaged.
The resistor should always survive if it is vitreous enamel and sized
properly. There should be universal agreement on that subject.
At the other extreme is the homebrew amp running a YC-156, 4CX10000, or
similar and with a filter bank of 200-500 mF total output capacity plus
a CCS + rated transformer. They DO exist. I wouldnt want to be in the
same building when an event happens!
One thing missing in all commercial ham and most homebrews is a real HV
fuse. Having both the resistor and a fuse is the best protection IMO.
Having worked on many amps in the USN I became a fan of HV fuses and
stocked up on a suitable selection back when they went begging at
hamfests. These days they are hard to find.
There are other methods to provide protection to be considered on a case
by case basis. Many involve down time to replace. Another is as simple
as a plate current overload relay which was used as far back as the 1962
designed NCL-2000 which also had a glitch resistor. Im referring to ham
amps here.
Carl
KM1H
----- Original Message -----
From: "Paul Whatton" <paul@g4dcv.co.uk>
Cc: <amps@contesting.com>
Sent: Monday, September 08, 2008 6:39 PM
Subject: Re: [Amps] 3-500Z glitch resistor
> Hi Carl
>
> Carl wrote:
>> The bigger the better up to the point of no improvement.
> No disagreement with that.
>>
>> However the purpose of the resistor is not to absorb 100 % of the
>> energy BUT to limit the current to a managable level until the
>> primary
>> breaker trips.
>>
> If I understand what's going on during a flashover correctly the first
> job is to hold the current down to a level that doesn't damage the
> tube
> or any other components in the amplifier until the breakers trip. The
> first problemm is the energy discharged from the capacitors. After the
> glitch resistor has dealt with that it needs to hang on in there until
> the breaker, or fuses, operate to protect the rectifiers and
> transformer.
>
> I guess it's a question of amp design philosophy. Do we want the
> glitch
> resistor to survive a flashover intact? If we don't mind it failing
> and
> choose, or have, to use a component that isn't rated for the job, how
> reliable is the failure mode? I don't want a glitch resistor that
> arcs
> or goes short circuit, so the volts the resistor can survive under
> overload are also important. Does the time needed to get the amp back
> on
> the air matter? It does to me in a contest. I really ought to use a
> G3SEK control board but being a cheapskate I trade off protection &
> cost
> against recovery time.
>> In a SB-220 for instance there is no room for a 50W or larger enamel
>> resistor and a 25W will survive. The fact that it works is that the
>> original PS diodes have survived.
>>
> The comments I made about the size of a glitch resistor were about
> building from scratch rather than retrofitting. I'd stick by my
> comment
> that a 50W+ wirewound is far better than any thick-film. I have just
> the same problem of limited size with my AEA LA30 which I want to fit
> a
> glitch resistor in, I'd love to use a 50W WW but there isn't enough
> room. So I'll use the biggest, best, vitreous wirewound I can fit and
> pray a bit. But in home-made amps there isn't an excuse and we can
> design in the space for a decent well-rated sized wirewound.
>
> Guess we might all agree on the following? Use a vitreous ceramic
> wirewound. If at all possible use one that can sink the energy and HT
> volts under a flashover within its ratings. If not use the highest
> power, highest voltage, ceramic wirewound that will fit in the box.
> Thick-film resistors, even big ones, are pants in this application.
> Wirewound resistors, even big ones, are a lot cheaper than tubes or HT
> transformers.
>
> 73 Paul G4DCV
>
>
>
>
> Carl
> KM1H
>
>
> ----- Original Message ----- From: "Paul Whatton" <paul@g4dcv.co.uk>
> Cc: <amps@contesting.com>
> Sent: Monday, September 08, 2008 2:08 PM
> Subject: Re: [Amps] 3-500Z glitch resistor
>
>
>> For Vic who originally asked this question, I've seen the figure of
>> about 40A suggested as the current limit so for a 3-500Z with 3kV,
>> 75R
>> should do. It isn't the normal power rating of the glitch resistor
>> that
>> is the problem but as David points out, it's the resistor's ability
>> to
>> survive the energy pulse from the power supply during flashover.
>>
>> In my 2m amplifier I use a modest 2kV supply with about 50uF. When I
>> built the amp I used a 47R 250W thick-film as the glitch resistor (5
>> times bigger than the 50W RCH series). It "felt" and certainly looked
>> like it should be more than big enough. But it failed, fortunately
>> open
>> circuit, on a flashover. Hindsight is a wonderful thing and checking
>> the
>> datasheet I'd exceeded the overload ratings. BTW such a thick-film
>> resistor costs about GBP45 new from RS in the UK, that's an expensive
>> fuse. Fortunately I'd only paid a couple of quid at a hamfest.
>>
>> IMHO a 50W+ ceramic wirewound is a much better and cheaper bet than
>> the
>> biggest thick-film in this application.
>>
>> 73 Paul G4DCV
>>
>> David G4FTC wrote:
>>>
>>>
>>>>> Everyone seems to agree that a glitch resistor between the plate
>>>>> choke
>>>>> and the HV supply is a good idea, to limit huge current surges due
>>>>> to
>>>>> arcs caused by gas, etc. But recommendations for the value seem to
>>>>> vary
>>>>> all over the map.
>>>>>
>>>> I like the look of Vishay RCH series or similar (Tyco do an
>>>> equivalent). The 50W version is rated at 5.5W without heatsink and
>>>> is specified to handle 2500Vrms (that's 3500V peak) short term
>>>> overload. If you want to bolt it to a panel for heatsinking, the
>>>> insulation is rated for 3500Vrms (5kV peak).
>>>>
>>>>
>>>>
>>>
>>>
>>> Checking page 4 of the datasheet for the RCH50
>>> http://www.vishay.com/docs/50006/rch.pdf
>>> shows a maximum overload capability of about 50 Joules.
>>>
>>> A capacitor bank for a linear of 50uF charged to 3kV will deliver
>>> about 225 Joules and
>>> under glitch conditions the resistor will be absorbing most of this
>>> power, more if
>>> the power supply doesn't have glitch detection/protection and
>>> remains
>>> connected
>>> to the mains.
>>>
>>> In order words to be safe you'll need about five RCH50 resistors in
>>> series.
>>>
>>> I don't know what energy the tube can absorb under flash over
>>> conditions without
>>> damage and this will also influence the selection of the resistor.
>>>
>>> Regards
>>>
>>> David G4FTC
>>>
>>> __
>>
>> _______________________________________________
>> Amps mailing list
>> Amps@contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
>>
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|