The AM Forum

Hi Guys

Now that I have the RF side cleaned up I am trying to get the audio side right. With just Bias at 100-120 ma (no audio) the two mod tube plates are showing color, one more than the other, pic attached. Not the whole plate, just portions of it. I have replaced R-56, 57, 53, and 54 shown on the attached schematic since the originals were a little off. But made no difference. I have C-39 and C-40 on order, but they do "test" as good. R-55  & 58 test as good.  Interestingly R-73 and R-74 do not exist. Both screen terminals are tied together without the resistors present.  Any opinions? Should I add R-73 and 74 ?

Thanks

Manny
KC4MNE
 

Hi Manny,

First swap the tubes around and see if the color follows each tube or not.  IE, is one tube softer than the other? (This could be the likely problem.)

Then measure each grid and screen at the tube pins (under operation) to be sure each tube shows the same voltages.  Measure the two filaments at the pins under operation to be sure there is no excessive sag.

Also, be sure the HV is the same on each tube using a HV probe.

Update: The 470 ohm screen resistors are about 1.8% of the screen impedance.  (750V / .03 A = 25K screen impedance.     470 ohms / 25K = 1.8%)     So it will add little distortion.  Still, if you are a perfectionist like me, it will add a tiny amount as the screens fluctuate in current.    I think if the tubes are stable as is, then a screen regulator is all that's needed.

It looks like there is zero grid current flow and high screen voltage, AB1, which is FB.    Any grid current flow with that circuit is NG. Your modulator power output is limited to when grid current starts flowing, but is the cleanest config you can run (AB1) other than pure class A.    Normal recommended screen current flow is good.

Let us know what you find and we can proceed from there.


Tom, K1JJ

I had the same problem with an RCA bta1r1. Turned out to be in correct bias setting on the grids of the modulators. Don’t know if that is your problem though.

I think the four series grid/screen resistors are parasite stoppers .... 1/2 watt for the screen ones would act as a fuse under heavy oscillation .... were the old ones open or still had a value ?

old bc xmtr ops would put the used 4-400's in the modulator after rf power dropped below spec .... sometimes this worked and sometimes not

i would recommend measuring your bias voltage at the junction of C-31, R-53, R-54 and see if it is close to the voltages from the voltage charts (-140V?), and also measuring the bias at the control grids.


I would reinstall R-73, R-74. They were put there to balance out the screen currents and to provide some minor isolation.


Phil - AC0OB

After a second look, I must agree.   In the real world of imperfect radio reception, the 470 ohm screen resistors will have only a small effect on distortion and can be added back in if you want.

The 470 ohm screen resistors are about 1.8% of the screen impedance when at full current. Less when at a lower current.  (Only about 13V of drop.    750V / .03 A = 25K screen impedance.     470 ohms / 25K = 1.8%)     It will add a tiny bit of distortion in exchange for better stability.  But I'm a perfectionist and think if the tubes are stable as is, then a screen regulator is all that's needed.  That's how I run my 4X1 modulators.

Run some careful tone tests with an SDR spectrum analyzer using both in and out resistor configs and find out for sure... probably barely measurable.

T

oh yes, back to the original question .... is the bright spot trouble ?    never really dealt with this before but the molybdeium (sp?) plate will melt at bright yellow to white temperatures ... as long as it doesnt get above brite orange or arc I think you shud be ok .... whaddyu think mr Vu ?

I think we need some more beans for supper... :-)

I'm kinda waiting for Manny to swap the tubes around to see if the irregular color follows them in different sockets. And to measure the pin voltages looking for differences between them.

I've seen this color difference before in my own used pulls (flea mkt deals) and sometimes it's OK.  If it does not show up in the above tests then it is usually low emission, a damaged grid or screen structure that does not cover the plate uniformally causing bright spots on the plate - or plate structure deterioration damage itself.   (The brighter tube with the irregular pattern is the damaged one usually)

This can also cause glass dimples or divots that get bigger and bigger until the tube pops a hole in the glass and implodes. I've had a 4X1 do this once after I used it as a winter heater for several months... :-)

Could also be that one tube simply has lower emission than the other.

The tubes also look like they need a good cleaning. Clean tubes = better longevity. Radiation cooling is a large part, in fact a VERY large part cuz I see no cooling chimneys.  The crud will partially block heat radiation and create glass, seals and internal structural problems as discussed.  But I still see original red temperature paint on the left tube's glass indicating no excessive heat damage - which is a good thang.

This is just speculation until we get some more info.

T

Manny,

I thought of a simple way to test your tube plate structures for damage...

Take the 4-400A tubes out. Put in a known good tube (if you have one) and adjust the grid bias so that it is idling (without audio signal) at 100 watts, then 200 watts then 400 watts. Take pictures of the plate at each setting.  Example: 3KV * 135 mA = 405 watts.  (Adjust R-27 for each idling point)

Then repeat with the old tubes into the same socket and test them one at a time the same way. Compare all pictures.

Usually the plate will start to glow with an uneven look to it at lower power but then become uniform as the power approaches nearer full dissipation.  You will then easily see which tubes are funky. Many times at full power the plates "clean up" despite being uneven at lower power.

At a full 400 watts, tubes with bright spots and dull areas looking like sun spots will probably not last too long as holes start opening in the plates.  I would think the IMD would increase too since the electrons are not uniformly collecting at the plate as per design.

T

swapped tubes around and the more colorful tube changed position. so it is a characteristic of the tube not the bauer.  bias voltage on the grids is around -155VDC. was on the air on the sunday morning fl am group net and sounded great. just not going to worry about it. did get rid of the resistors and nichrome wire and swapped out leads for plain 10ga wire.