>
>----- Original Message -----
>From: <amps-bounces@contesting.com>
>To: "Michael Tope" <W4EF@dellroy.com>; "AMPS" <amps@contesting.com>
>Sent: Saturday, January 24, 2004 6:05 AM
>Subject: Re: [Amps] Alpha 78 B- Resistor
>
>
>> ** GLOBAR-KANTHAL's Type SP resistors are superior to garden-variety
>> wirewound resistors for glitch service. (phone order # 716 286 7610)
>>
>
>Even if I replace the 25 ohm wirewound with a globar style
>resistor, it is still shunted with R112 and R113 which
>for a voltage divider which drives the plate overcurrent
>relay circuit. These are small carbon composition resistors
>that will likely flash over during an anode glitch.
** AGREED
>On the
>other hand, if I put a 25 ohm globar in series with the
>anode, during a glitch the anode supply will try to divide
>between the anode glitch resistor and R111 in the B-
>return. In this case, R111 would flash over and the anode
>glitch R would take the full anode voltage and ultimately limit the glitch
>current to a safe value, so perhaps this is the way to
>go (leave R111 as a wirewound and put a globar or
>other pulse rated high voltage resistor in series with
>the anode supply).
** I would try putting the over-current sampler R in series with the
glitch resistor - and then protect the sampler R with paralleled
200A-peak 1n5400-series Si diodes.
>
>> >While I was in the process of installing a new bandswitch
>> >in our club's Alpha 78, I noticed that the plate overcurrent
>> >shunt resistor R111 was cracked into two pieces, and
>> >that the G-10 PWB underneath this resistor was burnt
>> >pretty badly. It looks as if the resistor was running too
>> >hot which in turned slowly cooked all the epoxy resin in
>> >the G-10 laminate. When I scrapped away the soot, all
>> >that remained was the criss-cross pattern of the fiberglass
>> >layup weave.
>> >
>> >In any case, I don't think this was a "big-bang" kind of
>> >failure.
>>
>> ** Amplifiers that use tubes with gold-plated grids
>> usually don't go bang because a parasite that might induce a stentorian
>> HV-chassis discharge in air in TL-922s, SB-220s, et cetera is typically
>> replaced by a discharge inside the envelope through the gold meltballs
>> that stuck to the anode-grid insulator during the parasitic oscillation.
>> [ref: http://www.somis.org/8877.gs2.JPEG ]
>
>
>
>> >The resistor was not blackened at all. Instead
>> >the resistor's ceramic form material is "crazed"
>> >indicating that was overheated.
>>
>> ** Crazing is typical of a sudden overload.
>>
>
>Perhaps, but how do you explain the burnt G-10
>underneath the body of the resistor.
** possibly two. different events?
>The body of
>the resistor doesn't touch the burnt area, and there
>is no soot on the resistor body.
** Wirewound vitreous resistors have no organic parts.
>This leads me to
>believe that the resistor ran hot for a long period
>of time and that the radiant heat from the ceramic
>form cooked the nearby G-10. If the resistor body
>was running that hot (e.g. hot enough to cook the
>board), then would it not be plausible that the
>ceramic form was cracked and discolored due
>to prolonged operation at elevated temperature?
** agreed
>If all the crazing came from a very short event,
>then how do explain the baked epoxy resin
>the PWB?
** Beats me, but gold leakage can cause high anode current flow, even
during standby. Two events seems possible.
>
>I think the resistor was sized for 1978 regulations
>(1KW DC input, 2KW PEP input) whereas the
>HD transformer and the combined plate dissipation
>of the tubes are more in line with 3KW input power
>rating.
** As I understand it a 3cx800A7 will do c. 800mA of cathode-I before
stripping occurs.
>
>
>> >Presumeably the hot
>> >resistor proximity baked the resin in the adjacent laminate
>> >material over time. The resistor is a 25 ohm x 25 watt
>> >unit, which in my mind is too small. This resistor carries
>> >the full supply current of the amplifier which is probably
>> >close to 1.4 amps when running full bore!!
>> >
>> >Anyone run into this problem before with Alpha 78?
>> >By the way, this might be related to the problem that
>> >Jorge, EA2LU is having with his Alpha 76.
>> >
>> >73 de Mike, W4EF
>> >
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|
