[FullRangeMan]

This is a important notice about 6C33 Bias and why this robust military tube used in jets, radar and nuke bomb missiles lasts the same few 2 or 3 thousand hours as a fragile commercial audio Pentode as KT etc when used in audio tube amps in the realistic and lucid view of Jac Music site, certainly it will surprise you, consider change your 6C33 amp Bias.
https://jacmusic.com/techcorner/ARTICLES/English/Portraits/6C33/index.html
JACMUSIC Company Logo
Since 1993 Copyright Notice
Portrait of a Tube
Last Update: 12/09/2019 07:11:13
Description of the 6C33 tube
6C33 was intended is a series regulator tube power supply of tube equipment. An important use was in the power supply of a jet fighter, the Russian MIG. Due to the separate heaters, one system could serve as a spare in case of a failure, without having to replace the tube immediately.

In the period from 1960 to 1990, the Russians were preparing for a nuclear attack. From technical remains such as the 6C33, you can see how close the world was to a nuclear disaster.

When a nuclear bomb explodes, there will be a Electro Magnetic Pulse (EMP), which is just like a series of Radio waves. The difference with a normal radio transmitter is, the EMP is present at all wavelengths at the same time, and of extreme energy. There are funny videos of people holding a neon near a 5 Watt small transmitter, and it lights up. Mind you that needs 90Volt to ignite a neon lamp. Similar, that 90V is able to do damage to semiconductors, though current is low, and the semiconductors likely survive, protect themself by conductive effects, either in forward direction, or by a Zener (avalanche) effect in reverse direction.

Since an atom bomb is a 'little' bit heavier than a Radio transmitter, voltages and also possible current is unbelievable high. So the EMP at the moment of explosion will electrify all metal objects even quite far away, and destroy semiconductors in a distance of hundreds of kilometers.

Specific EMP bombs (Cobalt type) have the main intention to destroy all communication systems. Any electron tube near by will also stop working during an EMP too, but the tubes itself are not damaged by high voltage, and will recover once the EML is s gone. However, all semi conductor communication systems of the enemy are dead in just a few seconds. Only your tube amplifier will be undamaged.
Here is the story

The Americans were extremely surprised when a Russian MIG had to make an emergency landing in Japan. The pilot applied for Asylum, and it was the first time the allied forces got their hands on a complete, working MIG. They found it was equipped with tubes, and the body of the airplane was of metal. At first they had a good laugh about the old fashioned Russian technology. Then, later on, this incident was a MAJOR element that gave extreme acceleration to the nuclear cold war. The Russians had their own semiconductors which were not bad at all. The fact that all MIGs were fully tube equipped, was an important Russian secret that now leaked, and it shock-waved through political systems. The Russians were indeed preparing a nuclear attack. These MIGs could fly close to where atom Bomb drops, or drop small tactical bombs, specially for damaging communication systems.

After the communism got bankrupt, It took until 1995 until the Russians officially let go of their "cold war" tube stocks. After that (and not before) the 6C33 was available in quantities. Same as several other Russian tubes of the period, like 6H30. The small brother of 6C33 is 6C19, which has teh size of an EL84. These are amazing tubes, made to do a reliable job during those critical EMP moments, and quality and technology of those is the best as ever could be produced. So you won't find that in any datasheet, but yes they are EMP proof.

I took apart a 6C33, to see what's inside. Most of the quality you can't see with the eye, such as chemical things, and special treatment of the materials. What is visible at first sight, is the extreme ruggedness, and all mechanical parts are made from very hard, and also solid metal. That is amazing. The tube was so hard to take apart. The cutter tool I used, I had to throw it away afterwards. The anodes are 1mm (!) thick metal.

THE PINS OF 6C33.
This may be worth some extra text. There is great dissatisfaction with some amplifiers, where 6C33 burns down the socket. So we must see if that is unintended use of the socket, of the tube, or both. The tube bottom is hard glass, which can be seen by the pins developing a white color where the glass melts on them. The hard glass is do deal with the extreme heat, this doesn't crack, but most of all hard glass is used when tubes get extremely hot. Hard glass doesn't become as conductive as soft glass at great heat.

6C33 Users often complain the pins get corroded over time, and cause bad contacts. People try to scrape that off, but it is too hard substance. Sometimes there is a layer on there, which I heard somebody on a website call 'white shit". Nice word for it. Recently a customer of mine complained about broken off pins. I took a defective 6C33, and tried to break off the pins. That worked indeed, but they were quite strong, until they broke off with a glass-like sound. They way they broke off puzzled me, and I suspect these are not made of normal metal, but of so called amorphous material.

What is amorphous material? You can melt glass and steel together, in every desired ratio. The less steel, the more the material begins to behave like glass, and of course, the less conductive it gets. I know from amorphous transformer cores, the ratio is 1.1 there. That looks like metal, and behaves like hard metal, but also it can splitter like glass when are not careful.

Socket of 6C33 tubeTo me, those pins look like amorphous material (That is in fact a mixture of glass and metal) The advantage of amorphous material is, it will not crystallize, like any metal does, and it needs no explanation this attaches much better to glass, than pure metal. Also thermal expansion can be more closer to glass, than metal only. With the extreme heat of this tube bottom this can become important. I do not know for sure if that is amorphous, but it breaks like glass, and gives a crackling sound like glass too. Click on the picture to see a video.

Today's use of 6C33, and why so many 'bad' 6C33 burn down in 'good' amplifiers.

PLEASE READ THE datasheets. You cannot ignore them, just because they are in Russian. Then without reading it, it may go wrong. Consider blaming the amplifier and yourself, when such robust tubes for MIG airplanes were good when you received them, but do not work well in your amplifier.

Actually I have seen such blatant ignorance with some 6C33 amplifier builders, that I find it valid to point this out here. So of course this is not addressed to good electronic engineers. Actually I find many DIY are much more careful not to damage anything.

It is important to accept for any tube, when you do not use it as intended, you need to be sure nothing goes wrong because of that. Like 6C33 us definitely NOT intended for fixed bias.

The most common mistake is to assume some self-made for those boxes of the data sheet, where manufacturers write nothing. With that, you are taking a risk, and thy crying is very loud when things don't work our that way, and tubes burn down.

This risk can be avoided with awareness of what you are doing. It cannot be repeated often enough how important it is to read a datasheet in the first place, and most of all read it when you use a tube not as intended.
In that relation, we clearly must say this:

INTENDED USE = POWER SUPPLY regulator, and bias is by definition then in a closed loop. So drift, and leakage are no issue, as long as the feedback loop can deal with it.

NOT Intended use is audio with fixed bias, and drift and leakage may destroy the tube.

Just to give you a (stupid) example here: Suppose Nigeria Airlines has the idea to use normal car tires in the landing gear of their Tupolev airplane. So they know the maximum weight and speed, and checked the tires data, and find out it is at the maximum, but not above. They just say, well it is written in there, so it is possible. We don't need to speculate how this will go wrong. Or they use original Russian airplane tires, but they cannot read the datasheet in Russian, so they don't bother what's written in there. They just say "we always used those tires, and we never had a problem with that since 25 years". Would you fly with this company?

Particularly this "We have 25 years experience, and never had a problem with that" . Does it sound familiar to you...? It does to me! It is what I always hear from amplifier companies when they do something wrong.

Something similar happens with 6C33. It's intended use is as forward conductor, in a regulated power supply. This means, the bias of the tube is electronically regulated by the feedback loop. Maximum data of the tube is specified with this intended use in mind.

Now here comes the HiFi use, it is usually OTL. The efficiency of an OTL circuit is very low. So what will the designer do? He uses the highest possible dissipation, to get the power out of the tube. To eliminate risk, they use only 90% of maximum dissipation. I don't know who invented this first, but many think if you use a device at 90% of maximum, it will last a very long time.

The circuit designer has the option to choose between auto bias and adjustable bias. Both have risks and problems. An auto bias circuit works so well because the bias is determined by a closed loop, where the gain of the tube is the 'loop gain'. The higher the gain of a tube, the better auto bias is working. So what is the gain of 6C33? It is an incredible low 2...3x depending on the working point. So auto bias is not working well, and the tube can easily drift away. Of course they find that out, and some use hand adjustable bias. This is probably better, but the whole concept works by the merit of stay safely below maximum. And safety means not 10%, and sure not when you use a product in an unintended way.

Conclusion: With the above considerations in mind, perhaps you understand better now, why 6C33 fails so often with bias problems in OTL amplifiers. The problem CAN be the tube indeed. So the Russians did a bad job, and let many bad tubes pass, and nothing is perfect. But saying this, nothing is perfect, the problem can also be the amplifier. So to understand this situation we have two choices:

    Either all MIG air planes were flying around with potentially bad tubes
    Or.... many good 6C33 are potentially destroyed by bad designed amplifiers.

And I think I already know the answer. FORGET ABOUT Option 1.

When dealing with 6C33 tubes we see often, how good tubes destroyed by bad amplifiers. Do not think you can recognise bad internal design by a bad external design. Also, vice versa do no expect a good internal design us guaranteed by a good brand name on the outside. Often the relation is reversed! If a so called "great" brands only run on selected tubes, this means you have problems coming your way. These amplifier seem to need "better" tubes. The selection however is only needed because this amplifier would simply destroy perfectly good tubes otherwise. Of course the amplifier designer would call such tubes "bad" but this is only partially true. These tubes were good before they were put in the amplifier. They would become bad afterwards. So in a way the amplifier designer is right, but only in this order.
   

As you can see in this picture, there was as class crack and this tube has lost it's vacuum.

The three space holders in the top are made to give more stability, but they can obviously also damage the tube glass.

<----- YOU CAN CLICK ON EACH PICTURE, TO ENLARGE IT
   First I removed the top parts. As you can see, the stabilizer bars where also used to connect the getter pans to. These are curious getters. From the tube top, it can be seen these are the getter flash, but also they have some kind of encapsulated mass inside, when I opened them. I am not sure if such a getter exists, but perhaps they are combines Zirconium Barium getters. Anyway, I have been strongly expecting a Zirconium getter inside, but I did not see it. Since there is so much mass inside these getter containers, perhaps indeed they are combined. If someone knows something more about this.... Please Email! THANKS!!!!
   This is the mica top part. Just normal mica, but much thicker than normal. It is not aluminum treated, so it has stayed transparent. This demonstrates what we know: The tube is not intended for very high voltage.
   This is not reactor of Tschernobyl. Click on this picture (and any picture) to enlarge it, you will see nice details, and you can see from every one this tube is made to be strong as a tank.

   Hopefully you like these pictures. It was fun making those.
JACMUSIC. Since 1993. All rights reserved. Copyright note

[E55l2]

I love the 6c33cb. I had it more then 15Y very reliable, @250mA/200V,  in a fixed bias SE interstage coupled amplifier.

[FullRangeMan]

Hi,
How many hours did the 6C33 lasts with fixed Bias in your amp?
Stay safe.

[E55l2]

I changed the tube after 1500 hours. Not because the tube was broke but just a new tube sounded a little bit better and I had plenty tubes at that time

[borism]

Great article! I was also puzzled by reports of "unreliability" or many "bad" tubes. It was intended for fighter jets and when you hold them in your hand they are built like tanks. It makes a lot of sense that it is the use in some audio amplifiers that must be at fault when the tube fails prematurely.
 
I was fortunate to find a used Almarro 318b amplifier recently. To me this SET amplifier using the 6C33 is among the best I have heard.

Also, when I cleaned the pins of the 6C33 before use I did notice that they seemed made of a different kind of metal than the other Russian made tubes that came with the amplifier (6H9C and 6H8C).

[FullRangeMan]

Yes the 6C33 use a grey alloy intended to hi heat/voltage tubes, its the same used in the GM70,GK71,GMI11 etc. Some Ulianov I had sold lasted 3 years in a 15W amp without softstart, but with a slow softstart ramp and two power on switches it will last more.

[rollo]

  Use them in my Amps. Great tube.


charles

[FullRangeMan]

  Use them in my Amps. Great tube.


charles

How is your bias Rollo?

[Tyson]

I have an Almarro and in the instructions it notes that the 6C33 tube can vary pretty wildly during initial use.  They recommend checking the bias every hour the first day you use it.  After the first couple days of use, bias tends to settle down. 

[FullRangeMan]

I have an Almarro and in the instructions it notes that the 6C33 tube can vary pretty wildly during initial use.  They recommend checking the bias every hour the first day you use it.  After the first couple days of use, bias tends to settle down.

It seems correct as experts at DIYaudio said the bias varies greatly in the first 100 hours and after that it will slowly stabilize. On this bias subject the GM70/GK71 are a dream came true as the bias is said to keep steady around a full year even in new tubes.

[E55l2]

i never had a "wildly" change in bias. My setup was to begin with a fixed bias with a negative voltage lower then -110v so i knew that the current was not to high at the start. After that i turned the bias up to the required current value. It changed of course in time but remember that the steapness of this tube is very high (40mA/V) so 1V up or down in grid voltages means bias will up or down 40mA.
So you have to stabilaze all voltages in your amp OR no stabilisation at all. I had no stabilization and with the natural drift of the voltage of my powerline it never was a problem in my amp. (200V /250mA/ 50W Current max +/- 40mA depending on the powerline voltage).

After the bias setup i check every day at the start of a new tube for over a month and after that every 3 month. But i forgot that often... 
One thing i do always with this tube, i start with a heater warm-up for 1-2 minuits before i turn the high voltage on. I also start with the negative voltage on.

[FullRangeMan]

Nice. The Michael Boelle datasheet say military minimum heater warm-up are 120 seconds. I have read some expert saying after 35/40W dissipation she start oscillate, I suspect they mean say losing Bias.

[E55l2]

Well, if  there are oscilation problem the expert is not an expert.... to prevent such thing you need a gridstopper.

[E55l2]

Another thing which i almost forgot: Keep the grid circuit resistor low.
Michael Boele say lower then 200kΩ. I used it with fixed bias and prefer lower then 50KΩ or even much better with an interstage.

[FullRangeMan]

Well, if  there are oscilation problem the expert is not an expert.... to prevent such thing you need a gridstopper.

This not a circuit error or mal function they mean say the 6C33 start vary naturally, the 6C33 dont like works in hi power.

[E55l2]

This not a circuit error or mal function they mean say the 6C33 start vary naturally, the 6C33 dont like works in hi power.

I guess we are lucky that the 6c33 has not a bad behaviour like that but it is possible that the amplifier designer did make mistakes.

Just logically: why should this happen, the plate is heated just like every other tube. Hi power just reduces life time of every tube.

[richidoo]

modern 6C33 don't match original specs.
A friend designed 6C33 amp based on the original tube specs.
He basically had to redesign the whole thing after he started it up with new prod. tubes.
And he needed to add a continuous autobias scheme. major PITA
But it sounded good once finished.

[E55l2]

modern 6C33 don't match original specs.
A friend designed 6C33 amp based on the original tube specs.
He basically had to redesign the whole thing after he started it up with new prod. tubes.
And he needed to add a continuous autobias scheme. major PITA
But it sounded good once finished.

What are modern 6c33?

I bought long time ago at the new sensor corporation and later also on ebay. The last i used are from 1976 and 1988.

Most problems come from bad designed or more friendly not optimale designed amplifiers. If the cridcircuit resistor is to high there will be drift of bias.

[FullRangeMan]

Hi E55l2,
What are the Pri Z in this your amp?
Do you are using 600 or more?
Thanks

[E55l2]

Hi E55l2,
What are the Pri Z in this your amp?
Do you are using 600 or more?
Thanks

 I had 600 Ohm but that was a stepup transformer to my esl speaker. I also have a 800 Ohm to a normal speaker

Maybe the 800-900-1000 Ohm is the best for a 6c33 for a normal speaker in SE.