Unexpected brightness of RF power tube grid

[Antonio Vernucci]

I am seeking advice from anyone have a solid and hand-on experience in RF power tubes operation.

I have recently built a CW transmitter using a German WWII tube (RS391 by Telefunken) as final amplifier.

The tube operating parameters match very well those shown in the datasheet, namely:
- plate voltage: 1500V
- plate current: 150 mA
- screen voltage: 400V
- screen current: 25mA
- suppressor voltage: 0
- grid voltage: -120V
- grid current 2.5 mA
- output power: 140W

I was very satisfied of the transmitter behaviour until I noted that one of the grids (I cannot determine which one of the three through the glass) surrounding the cathode becomes very bright on transmit (sign of very high temperature).

Please note the following:
- I tried three different tubes, all showing the same behaviour
- everything looks normal, no self-oscillations
- the dissipated powers, calculated on the basis of measured voltages and currents, are:
* control grid: 0.25W against an allowable maximum of 0,4W
* screen grid: 10W against an allowable maximum of 20W
* suppressor grid: 0W
- reducing the control grid drive level causes a reduction in grid brightness, but if I reduce the grid current to almost zero I can still detect some glow.

At this point I am wondering whether having a grid working at high temperature is a normal condition and I am then worrying about a non-existent problem. That supposition comes from the fact that I can hardly believe that the screen grid, which has a very light structure, can dissipate 20W without reaching a very high temperature.

If you have some experience on the issue, please advise.

73

Tony I0JX
Rome, Italy

[Scott Dorsey]

I have recently built a CW transmitter using a German WWII tube (RS391 by T=
elefunken) as final amplifier.
The tube operating parameters match very well those shown in the datasheet,=
- plate voltage: 1500V
- plate current: 150 mA
- screen voltage: 400V
- screen current: 25mA
- suppressor voltage: 0
- grid voltage: -120V
- grid current 2.5 mA
- output power: 140W
I was very satisfied of the transmitter behaviour until I noted that one of=
the grids (I cannot determine which one of the three through the glass) su=
rrounding the cathode becomes very bright on transmit (sign of very high te=
mperature).

I am not familiar with the tube. However, if you have grid stoppers in place
and a plate choke and you've looked on a wideband scope or spectrum analyzer
to see that there's no parasitic oscillation, and you've verified that the
screen current is good, it's likely okay.

Excessive screen current combined with a hot screen is a big big red flag,
but you say your screen current measures okay. What happens if you drop the
screen potential a bit?

At this point I am wondering whether having a grid working at high temperat=
ure is a normal condition and I am then worrying about a non-existent probl=
em. That supposition comes from the fact that I can hardly believe that the=
screen grid, which has a very light structure, can dissipate 20W without r=
eaching a very high temperature.
If you have some experience on the issue, please advise.

Does the tube datasheet mention this at all? This is a tube that you will
find few people familiar with.
--scott

73
Tony I0JX
Rome, Italy

[Antonio Vernucci]

Post by Scott Dorsey

I have recently built a CW transmitter using a German WWII tube (RS391 by T=
elefunken) as final amplifier.
The tube operating parameters match very well those shown in the datasheet,=
- plate voltage: 1500V
- plate current: 150 mA
- screen voltage: 400V
- screen current: 25mA
- suppressor voltage: 0
- grid voltage: -120V
- grid current 2.5 mA
- output power: 140W
I was very satisfied of the transmitter behaviour until I noted that one of=
the grids (I cannot determine which one of the three through the glass) su=
rrounding the cathode becomes very bright on transmit (sign of very high te=
mperature).

I am not familiar with the tube. However, if you have grid stoppers in place
and a plate choke and you've looked on a wideband scope or spectrum analyzer
to see that there's no parasitic oscillation, and you've verified that the
screen current is good, it's likely okay.
Excessive screen current combined with a hot screen is a big big red flag,
but you say your screen current measures okay. What happens if you drop the
screen potential a bit?

At this point I am wondering whether having a grid working at high temperat=
ure is a normal condition and I am then worrying about a non-existent probl=
em. That supposition comes from the fact that I can hardly believe that the=
screen grid, which has a very light structure, can dissipate 20W without r=
eaching a very high temperature.
If you have some experience on the issue, please advise.

Does the tube datasheet mention this at all? This is a tube that you will
find few people familiar with.
--scott

73
Tony I0JX
Rome, Italy

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Hi Scott,
pleased to read from you after a long time.
Just for fun I am building a transmitter using German tubes and components.
The Telefunken RS391 is a plain power pentode optimized for suppressor grid modulation, but is not a special tube. It is similar to an RCA 803, just somewhat less powerful than it (110W plate dissipation against 125W of an 803).

The transmitter works beautifully, no self-oscillation or other instability signs. All currents and voltages, including the screen current and voltage, are as they should be, no overload whatsoever. The only oddity is that I see a glowing grid on transmit (not sure if it is the control or the screen grid). I tried three different tuvbes, all showing the same effect.

As you suggested I tried to decrease the screen voltage, and the grid brightness also decreases. But I do not see why I should decrease the screen voltage if it is well within the specifications limit.

An easy conclusion would be that having a glowing grid is normal. But I have some difficulty to accept that (I have never seen a glowing grid).

The issue I am raising is not specific to the RS391, but it is of more general applicability. For istance the "typical operation" section of the RCA 803 datasheet reports as normal a screen voltage of 500V and a screen current of 33mA, corresponding to a dissipated power of 16.5W

How is it possible - I am asking myself - for a tiny grid immersed in between two high temperature elements (namely the plate and the cathode) to dissipate 16.5W (which is a significant amount of power) without reaching a very high temperature? Does the 803 screen grid also glow?

73

Tony I0JX
Rome, Italy

[Scott Dorsey]

Post by Antonio Vernucci
pleased to read from you after a long time.

Thank you, it is good to see you here!

Post by Antonio Vernucci
Just for fun I am building a transmitter using German tubes and components.
The Telefunken RS391 is a plain power pentode optimized for suppressor grid=
modulation, but is not a special tube. It is similar to an RCA 803, just s=
omewhat less powerful than it (110W plate dissipation against 125W of an 80=
3).

With an 803, I would not expect to see that behaviour... but this is not an
803. The grid structures are likely different.

Post by Antonio Vernucci
The transmitter works beautifully, no self-oscillation or other instability=
signs. All currents and voltages, including the screen current and voltage=
, are as they should be, no overload whatsoever. The only oddity is that I =
see a glowing grid on transmit (not sure if it is the control or the screen=
grid). I tried three different tuvbes, all showing the same effect.=20

If the currents are in the ballpark, and it's not oscillating, and the load
is good, then you have done everything you can do, I think. It is likely
a design issue with the tube and it might shorten the tube lifetime, but
that might just be how that tube is.

Post by Antonio Vernucci
An easy conclusion would be that having a glowing grid is normal. But I hav=
e some difficulty to accept that (I have never seen a glowing grid). =20

Yes, I don't like it.

I have seen glowing grids before, but only when something was wrong. There
are a lot of tubes that will draw excessive screen current and have a very
obvious screen glow... but they won't do that for very long until the screen
melts and sags.

Post by Antonio Vernucci
The issue I am raising is not specific to the RS391, but it is of more gene=
ral applicability. For istance the "typical operation" section of the RCA 8=
03 datasheet reports as normal a screen voltage of 500V and a screen curren=
t of 33mA, corresponding to a dissipated power of 16.5W
How is it possible - I am asking myself - for a tiny grid immersed in betwe=
en two high temperature elements (namely the plate and the cathode) to diss=
ipate 16.5W (which is a significant amount of power) without reaching a ver=
y high temperature? Does the 803 screen grid also glow?=20

The 803 screen grid does not glow, or at least not to the point where it is
obvious and visible. But it does dissipate a considerable amount of screen
power, and the screen is designed for that.

If I were you, I wouldn't worry about it because I don't think there is
anything you can do about it that you haven't done already. But if I were
you, I'd be using an 833. Or maybe a 6AQ5.... I have 90 DXCC countries now
with the 6AQ5 as a final... but those last 10 are proving to be difficult...
--scott