Replacing rectifier and VR tubes witih SS

[KA9JLM]

I have used the 5U4 tube for its plug base and never had to break the glass to remove the tube from the plug base.

I have used 4 amp 1000 piv diodes for their replacement with no problems.

Should work as described by the OP.

[N2EY]

There were some regulator tubes made with a small amount of radioactive stuff in them. Supposed to insure more uniform firing or some such. Used in military applications, which means they may show up in surplus. Many (but not all) have the little circular symbol on them.

Making a 7 pin miniature plug is easy. (Miniature VRs are 7 pin, not 9). Just take an old worn-out 7 pin tube socket and solder pieces of stiff wire into the contacts.

73 de Jim, N2EY

[KM1H]

I don't understand what you mean when you say that a 183D is not one of those times when you should use a SS rectifier. The rest of your post describes different ways to get the voltage and heat out of the transformer and keep it going. Since finding a new "used" transformer for a 183D would be difficult at best, isn't this exactly when you should use a SS rectifier? Especially when combined with your recommendations and others that I came up with below?

I have my doubts about that transformer being bad as a short between the B+ and 5V windings would not be lightly tolerated. My ONLY suggestion to get maximum heat out of the transformer was a bucking transformer, the other methods I commented on were partial or ineffectual. In addition you could use a 5R4 to save 5W as well as lower the B+ a little more.

You recommend adding a 12v bucking transformer. I don't quite understand as my line voltage here is about 120v and the specs for the 183 say 110/120v AC. Won't this take me down to 108v? That's not very far off, but still it seems to be out of range. You're saying 108v is ok?

Its more than OK. I run full bench bucking transformers for the vintage stuff for almost 30 years and the end result is anywhere between 108-115 depending upon what secondary voltage transformer Im using (all are odd voltage high current, no CT as found real cheap at a local surplus store or hamfest) and what the actual AC is at any given day/time as Im at the end of a deadend road on top of a hill and its anything but steady when seasonal loads hit.

The only real difference between 120-125V and 105-110 will be speaker volume as the 183D and most other radios have excessive gain from the RF stages on to start with. National knew this as their sets could also be configured to run off a 180VDC vibrator pack if needed. Pull a 6V6 and use headphones. The low backround noise makes signals simply jump out and the SNR is usually better....maybe not on 6M!

I appreciate your recomendation to leave the heaters on, that's what I learned when I was a young ham of 12. I had an SX-99 by my bed that I left on for years. However, a few years ago, I talked with another engineer who had worked in the tube division at RCA back when, and he said that by actual test, there was nothing to be gained by doing it. Normal use on and off would last just as long. You get more hours of heater life by leaving them on, but the hours that you actually use the tube are not any more. It depends on your average use percentage, or duty cycle. He also said best heater life was at 6.3v, and it was worth it to adjust line voltage, to get exactly on 6.3v. He did say that soft start up, with a thermistor or variac would be good for the filaments.

Ive no argument there and Im sorry that what I said sounded like we should always leave things on 24/7. Thats no so for the very intermittent type of use hobbyists are involved with. Running the filaments a few percent below the 6.3V bogey but well within the -5% spec will add even more emission life as well as collectively reduce heat a bit more.

I want this 183D to hang together for as long as possible. What about doing this-- Add the bucking transformer, AND, add a 6.3v transformer for the filaments, taking the filament load out of the big power transformer. If I were to do that, I could reduce the line voltage going into the main transformer by an even greater amount. There's plenty of room left with the SS rectifier/resistor combination to run the B+ back up to where it should be by making the resistor smaller, or taking it out all together. The original 5U4 has a drop across it of 30v or more. I could probably put in an 18v bucking transformer, use a much smaller or no resistor in the SS rectifier, and still get 255 volts B+. What do you think of that? There's loads of room underneath the chassis. That would reduce the voltage on the windings and I wouldn't be wasting the watts to drop the voltage in the rectifier resistor.

Sounds like youre just making unnecessary work for yourself. Just start with a bucker, the R-S one I mentioned was due to its often convenient availability but something maybe down to 8-10V would do some good. Get a good IR gun and make various tests, you might be surprised.

Carl

[KM1H]

However, one other way to help extend tube life by the use of tube heat sink shields, like those found in the R390a. The IERC types. Also light air circulation in the area around the radios helps in lowering ambient temperature allowing heat exchange to happen more readily.

Have you priced those shields since the audiophools discovered them? Most of the 183D heat is on the other side of the chassis by the 6V6 and 5U4 section. The cabinet is well ventilated, at least for this part of the country, and I doubt if removing most of the 7 pin tube shields would affect performance any.
My own 390A has had IERC shields since I got it going on 30 years now, thats a different beast altogether.

However, while not reducing heat the thermistor idea, if placed on the primary of the B+ transformer, will help reduce those peaks for the first couple of cycles. I have not had one fail and some have been in use here for at least a decade and a half. But yes they do produce quite a bit of heat themselves.

When properly sized they have their purpose in some equipment (mostly SS) but I still wont use them because the heat added far outweighs a questionable benefit in an antique radio. The transformer impedance alone offers a slow start at turn on. Also be aware that thermistors have a MINIMUM as well as a MAXIMUM operating current rating and the best efficiency is where the series resistance is lowest which is clearly shown on the spec sheets.

Also I like adding separate filament transformers. Many of the receivers of this era used B+ transformers that had filament taps for 6V as well as the 5V for the rectifiers. Filament supplies are usually the highest current draw, so getting that out of the main transformer will reduce heat for the HV transformer. Your total heat will be the same within the radio, but the transformer for the HV will be a little cooler.

Again, I feel that is overkill as long as the steps are taken to get the AC down. Having a transformer show almost no temperature rise might sound good on paper but at what benefit if its already built to handle a certain temperature? And we now have it below that with the bucker?

Carl

[KM1H]

while we're at it, let's do the math on the rectifiers...

A 5U4 that drops 40 volts at 200 mA is dissipating 8 watts in the plates and 15 watts in the filament. So most of the heat is from the filament, not the forward voltage drop of the diodes.

That math should include the 40V drop is for one plate only and the 5U4 has 2 of them in a FW circuit.

Carl

[N2EY]

That math should include the 40V drop is for one plate only and the 5U4 has 2 of them in a FW circuit.

Nope. Remember that each plate only conducts half the time!

Or look at it this way:

Suppose that with a 5U4 in a given circuit the resulting B+ is 250 volts.

Then the 5U4 is replaced with silicon rectifiers, and the B+ rises to 290 volts.

That means 40 volts was lost in the 5U4 which is no longer lost in the silicon.

If the load current was 200 mA in both cases, the 5U4 plates are dissipating 8 watts total (40 volts times 200 mA).

Of course, to be really accurate we'd include the 0.7 volts forward drop of the silicon, but that's trivial compared to 40 volts.

---

The really important bit of all this is that you can't just toss in some silicon diodes to replace a 5U4 unless some provision is made for the resulting higher B+. The usual solution of a resistor in series with the diodes simply means the resistor is dissipating heat instead of 5U4 plates. All you save is the 5U4 filament power.


73 de Jim, N2EY

[KM1H]

Nope is right, I suggest you study PS theory a bit more.
Each half conducts every half cycle and tube manuals consider that in their ratings. Its no different than using 2 SS diodes.

Carl

[N2EY]

Nope is right, I suggest you study PS theory a bit more.
Each half conducts every half cycle and tube manuals consider that in their ratings. Its no different than using 2 SS diodes.

Of course. But everything I wrote is still correct.

Let's look at a real-life situation.

Suppose there's a power supply which uses a 5U4 in the classic full-wave-center-tap configuration.

Its output is 300 volts at 200 mA when using the 5U4.

Now suppose the 5U4 is replaced by two silicon diodes whose forward voltage drop is essentially zero. As a result, the power supply output voltage rises from 300 to 340 volts at 200 mA.

So, with all else being the same, the difference in output voltage between the 5U4 and the silicon is 40 volts.

OK so far?

The only place that 40 volt drop can come from is the 5U4 rectifier, because it's the only thing that changed. Since the total current is 200 mA, the total power dissipated in the 5U4 due to forward voltage drop is 40 volts times 200 mA = 8 watts. Add the 15 watts of filament power and the end result is 23 watts dissipated in the 5U4.

If we want to get the same output voltage from both rectifiers, we either have to lower the transformer output voltage or introduce additional resistance to get rid of 40 volts at 200 mA. The added resistor will dissipate about 8 watts, too.

Yes, there are secondary effects such as the transformer impedance, the size and ESR of the input filter capacitor if it's a capacitor-input filter, etc.

But for a first approximation, the total power dissipated in the 5U4 in the above example is 15 watts of filament power and 8 watts of total plate dissipation. That's 4 watts per plate, 8 watts total for both.

If I'm wrong, prove it.

73 de Jim, N2EY

[KM1H]

I dont have to prove anything since there is already ample proof available if you know how to find it.

You seem to take pleasure in popping up on various forums looking to start arguments and when proven wrong just keep raggin' on and on. Have you EVER admitted an error? If not its a sign of a disturbed mind.

Carl

[N2EY]

I dont have to prove anything since there is already ample proof available if you know how to find it.

IOW, you can't prove it.

You seem to take pleasure in popping up on various forums looking to start arguments and when proven wrong just keep raggin' on and on. Have you EVER admitted an error? If not its a sign of a disturbed mind.

I'm not trying to start an argument. I've admitted errors when I've been mistaken.

But in the example I gave with the 5U4 and silicon rectifiers, I'm not wrong.

You seem to want me to be wrong, but you can't prove me wrong, so you resort to personal attacks.

Why not just show where my example is wrong?

Here it is again:

Suppose there's a power supply which uses a 5U4 in the classic full-wave-center-tap configuration.

Its output is 300 volts at 200 mA when using the 5U4.

Now suppose the 5U4 is replaced by two silicon diodes whose forward voltage drop is essentially zero. As a result, the power supply output voltage rises from 300 to 340 volts at 200 mA.

So, with all else being the same, the difference in output voltage between the 5U4 and the silicon is 40 volts.

OK so far?

The only place that 40 volt drop can come from is the 5U4 rectifier, because it's the only thing that changed. Since the total current is 200 mA, the total power dissipated in the 5U4 due to forward voltage drop is 40 volts times 200 mA = 8 watts. Add the 15 watts of filament power and the end result is 23 watts dissipated in the 5U4.

If we want to get the same output voltage from both rectifiers, we either have to lower the transformer output voltage or introduce additional resistance to get rid of 40 volts at 200 mA. The added resistor will dissipate about 8 watts, too.

Yes, there are secondary effects such as the transformer impedance, the size and ESR of the input filter capacitor if it's a capacitor-input filter, etc.

But for a first approximation, the total power dissipated in the 5U4 in the above example is 15 watts of filament power and 8 watts of total plate dissipation. That's 4 watts per plate, 8 watts total for both.

If I'm wrong, prove it.

73 de Jim, N2EY