[Amps] Re: 8877 dissp

mark mark@sandlabs.com
Wed, 13 Nov 2002 17:10:32 -0500


Intersting John ...
It is true that 8877 filament is not burning mutch energy, however the
exercise is on a general note.
For example, the AMP in discution will be part of a Flow Process that calls
for 10-15 minutes
STANDBY every hour. I whish to minimalise the power consumption during this
time as it grows linearly with the number of AMPS deployed. Thus, I must
understand the POWER usage of the tube so i can minimalise where it is
necesary.
One place where I focus now is the filament ... I intend to reduce the power
to the filament to, say 50% durring the imposed Standby. This raises the
folowing questions :

1). Does this 50% periodic filament voltage reduction stresses the tube?
2). What is the recovery time from 50% to 100% caloric output of the
filament (for a step up voltage)?
3). What happens, if by accident, HV and DRIVE is applied while the filament
is still ramping up?

Home work ... Home work... It never ends :)

Mark
PS ... Interesting work in your lab :) it remainds me of the times I was
having fun in Toulouse, F. :)
You would'n be there by any chance ? :)

M.



----- Original Message -----
From: "John T. M. Lyles" <jtml@lanl.gov>
To: <mark@sandlabs.com>; <amps@contesting.com>
Sent: Wednesday, November 13, 2002 10:14 AM
Subject: 8877 dissp


> Yes, as pointed out by others, plate dissipation is only that
> electrode dissipation. And for 70% efficient amp in class B or C for
> FM service, it is only 1100 watts or so. But that leaves little
> margin for error, mistuning, VSWR effects. I would not be running a
> single 8877/3CX1500A7 at 2600 watts output anywhere except maybe
> pulsed or SSB. The transmitter I worked on the design of with Bill
> Harness years ago was 1750 output max, at 88-108.
>
> The 8877 I am running in our lab at this minute is dissipating 600
> watts, and we are sweeping it completely off resonance, about +/- 150
> kHZ around 5 MHz, driving a load with higher Q than the amp. It takes
> the beating no problem, but you can feel a lot of heat in the
> exhaust, about 50 deg C when it is idling and varying as we sweep it.
> I am building an ALC loop right now, to lock incident power at 1 kW
> or whatever we dial up, irregardless of VSWR. I would never attempt
> this with an 8877 running closer to its maxium rating.
>
> BTW, the filament power in 8877 is miniscule compared to anode
> dissipation. So it can be dealt with strictly as a blower issue,
> cooling the base of the tube when lit. When you get to very large
> tubes, like 4CW250,000B tetrodes, you have to do things like leaving
> water running in the anode, even while filament is lit. The 7835
> grounded grid triode that we have in sockets running (4), has 30 kW
> filament power. We do account for this in our efficiency
> measurements. We see it in the calorimeter on the plate water course
> even.
>
> ABSOLUTE MAX is exactly what it says. The 7835 is ABSOLUTE MAX 300 kW
> plate dissipation. I have cracked the tubes open - ceramic rings - at
> 270 kW plate dissipation. Now i interlock them at 250 kW through the
> calorimeter setup. Maybe CPI-Eimac is more conservative with its
> ratings..?
>
> 73
> John
> K5PRO
>
>
> >
> >Subject: [Amps] EIMAC Confusion
> >
> >In the EIMAC Technical Data for the 8877/3cx1500A7
> >is is stated in the ABSOLUTE Ratings the Plate Dissipation
> >is 1500W. I take this as an ABSOLUTE MAX.
> >
> >However in the TYPICAL Operation at 88-108MHz USEFUL POWER OUTPUT
> >is rated at 2600W.
> >
> >Is this meaning that the tube is providing 2600W to the load while plate
> >dissipation
> >is only 1500W? Is this possible ? If so What is then the TOTAL TUBE
> >DISSIPATION
> >and how is the filament power related to it ?
> >
> >Mark
>
> --