My 9-PC-41 has four 5U4 tubes and when the time comes I would like to convert using 1N4007s.

What has been your experience? I know a resistor should be added to keep the voltage as close as possible to the 5U4. What would be a good value to start with?

All of the articles I have found so far address audio concerns: cathode stripping, delay B+ until heaters warm up, this, that, and the other. Any similar concerns in old TVs?

Your practical experiences appreciated.
Thanks,
Phil

I'll address the cathode stripping issue. I think there's some merit on the issue as having a standby switch on guitar amps for example does serve some purpose as you are applying ~450Vdc B+ directly through the output transformer primary to the output tube plates and possibly screen controls/grids. Delaying B+ makes sense.

Look at your schematic and I'd bet nothing really starts to provide any high(er) voltage B+ until the B+ Boost starts coming up. And, as we know some tv type tubes will have a controlled warm up time.

Also, look at how many tube type sets were built with solid state rectification and have operated just fine.

Taking a load of four(!) 5U4's off would sure cool down the p/s transformer and slow down your A/C meter spinning!

Taking a load of four(!) 5U4's off would sure cool down the p/s transformer and slow down your A/C meter spinning!

5 volts x 3 amps x 4 tubes = 60 watts heat saved plus the amount of plate heating due to output current load.

In the RCA 1947-1949 projection sets, one 5U4 powered the audio chassis, one 5U4 powered the tuner, sync and video chassis and two 5U4s powered the sweep chassis.

I'd say this makes the slow and even B+ start provided by the tubes more important.

This, plus the fact that the solid state rectifiers produce higher voltage and can carry more current than the tubes, gives a potential for greater damage when anything fails. I intend to retain 5U4s in my sets.

Back in the late 1950s, we routinely replaced selenium rectifiers with silicon ones, and we learned the hard way that while they helped keep a set going during the summer when the line voltage was low, but in the winter when the line was normal or slightly high, they tended to pop electrolytics and we switched from 10 ohm added loads to 100 ohm loads.

Also, sometimes, especially on transformerless sets, the silicon rectifiers generated a traveling wave that crawled up on the screen. One cure was to add a .01 MFd cap across each silicon rectifier. or from the plus side to the ground. I never understood how this worked, but it usually did.

I'll leave cathode stripping question to the theoreticians, but I will comment on one posting about using deoxit type contact cleaner on these sets. Since the extra high voltage on the CRT is very effective in attracting dust, and deoxit type cleaners leave a sticky film on the surface that tends to trap particles, I'd avoid their use and use traditional contact cleaners on any projection set.

James.

To really do it right, you need to make current and voltage measurements for each tube or supply, and figure out exactly how much voltage you need to drop to get down to the specified levels for each supply. Find the required resistance and wattage using Ohm's law. For ballpark guesstimates, figure on the output being 30V or so higher using silicon.

5U4s are directly heated tubes, and produce B+ in a second or 2 from a cold start. A bit slower than solid state diodes, but still not "delayed B+" by any means. Certainly not like a 5AR4 or other indirectly heated tube.

Where you might have an issue is the 30 seconds or so between power on and the rest of the tubes starting to draw current. You need to make sure that the output of the silicon diodes doesn't exceed the voltage rating of the filter caps during this time.

I prefer the 1N5408 to the 1N4007. 3A continuous If is probably overkill, but the charging surge can be pretty tough on 1N400x series parts. Mount a pair of them in an old octal tube base.

The b+ will be way to high with the solid state diodes, if you use resistors they will have to drop so much and get so hot its pointless. Use the 5U4s.

Just my opinion.

Back in the late 1950s, we routinely replaced selenium rectifiers with silicon ones.... sometimes, especially on transformerless sets, the silicon rectifiers generated a traveling wave that crawled up on the screen. One cure was to add a .01 MFd cap across each silicon rectifier. or from the plus side to the ground. I never understood how this worked, but it usually did.
Those caps are generally known as "snubbers" in that application.
There's a fast spike produced when the silicon diodes switch-on near the zero-crossing point of the AC wave. Electrolytics possess a small inductance value which allows the spike to ride unfiltered on the B+ line. The "snubber" caps finish smoothing that spike out.

Thanks for all the good advice. I'm going to stay with the 5U4 set up. Don't want to push my luck with this old of a set.
Phil

I killed once a line transformer of a 1955 AEG TV-set.
I interchanged "only" the selenium rectifier by silicon diodes and didn`t
realized the higher output voltage!:rant:
Stay with the old 5U4 horses.:D

Best regards,

TV-Collector:stupid:

Thanks for all the good advice. I'm going to stay with the 5U4 set up. Don't want to push my luck with this old of a set.
Phil

Good choice! It's too taxing on the filter caps, and on the whole set, to have instant B+ (even if you regulate it somehow). The 5U4s give a B+curve which rises dramatically, but then actually pulls way back down as the other tubes start to emit and start to create a load, and then as the 5U4s heat up the rest of the way, the B+ slowly rises to specs. Probably better than a fancy timed startup and regulators :thmbsp:

Charles

The main thing you save is filament power. A 5U4 wastes 15 Watts just to light it.

I have a few 5U4 solid state replacement units. They just plug in, look like a small metal tube, and should work fine. You still need the dropping resistor to keep the surge within limits and to maintain proper B+. And of course you get instant B+, although a 5U4 will light up very quickly.

Let me know if you want one or more of my devices.

Bob,
Thanks for the offer. I have a lonnnnnnng way to go on my set before I even consider firing it up. I'll keep your offer in mind
Phil

[QUOTE=bob91343;3050710]The main thing you save is filament power. A 5U4 wastes 15 Watts just to light it.

Not to be picayunish, but the filament wattage saved by replacing a 5U4 (or a 5Y3) with silicons is largely offset by the higher wattage that's now dissipated AS HEAT in the HV winding and added dropping resistor(s). This is due to silicon diodes' low forward drop compared to vacuum diodes.. plus the fact that the silicon diode is "full on" during is conduction period. Wattage that was formerly dissipated in heating the 5U4's plates now heats up the HV winding even more, and in some cases can cause the xfmr to run hotter than before the conversion.
It's a good idea to measure 'before-and-after' current draw from the wall when converting to silicons.

I don't follow how the heat from added dropping resistors is any different than the heat produced by losses in the 5U4 ? Why would the current draw be any higher with silicon diodes ?

The dropping resistors will not only drop the voltage but restrict the current draw. For example if the set has a B+ of 300 volts DC @ 100mA it's drawing 30 watts from the transformer regardless of how the AC was rectified.

Oh wait, I think I see what you're getting at. Since the silicon is more efficient, the resistive dropper will have to waste more power than the less efficient 5U4.

I wonder if zener diodes could be used as rectifiers to simulate a larger forward voltage drop ?

Sure you can use zener diodes but they would have to be in series with the high voltage diodes. You'd need a heat sink.

If you save 15 W per 5U4 that could be a very nice improvement, allowing the transformer to run cooler, reducing the heat rise in the unit, and of course drawing less from the power line. That savings is independent of the rectification loss, which you have in either case if you size the resistors properly.

No, the resistor won't waste more than the 5U4. If you do it right, it will be exactly the same as the tube. Just the filament power is no longer needed.

Oh wait, I think I see what you're getting at. Since the silicon is more efficient, the resistive dropper will have to waste more power than the less efficient 5U4.
At the risk of being a fuddyduddy:o, I would amend that slightly to read thus:
"Since silicon is more efficient (ie., fully "switched on" during its entire conduction period), the resistive dropper and the transformer's HV winding will be forced to dissipate more power than the less efficient 5U4."
The Si diodes' "square" conduction period presents longer "fully on" time than a vacuum diode's 'bell' curve. "Higher duty cycle" if you will. Thus the resistive dropper and the HV winding see greater net loading, power dissipation and heating.

It'd be interesting to do a real world before-&-after test with a wattmeter in the AC line, and a thermometer on the power tranny.

Okay, whos gonna call myth busters??? :D

SR

Okay, whos gonna call myth busters??? :D

SR

:boring:

You forgot to add the drum fill :)

SR