Tung Sol KT120: Why do some of them overheat and others don't? 

Hi Steph, it's director of the factory-manufacturer, Irusha Bitukova answering you: Yes, we moved the side getters purposely. These are reasons - 1.They used to be placed too close to the radiator edge what formed the plate-getter-radiator loop and, during getter spraying, caused high frequency breakdown between getter and the radiator rib. - 2. The getters were too close to both plate and radiators. This made the getter harder to spray and was leading to overheating of both plate and radiators and caused the plate material to spray onto the glass. - 3. In the previous design the getter was farther from the glass what gave a larger mirror which blocked the radiator rib and bigger part of the plate. THIS caused overheating of the plate and CATHODE of working tube as well. And THAT IS WHY you got higher current!!! - 4. In the current design the mirror surface is smaller and moved very slightly what opened shadowed before plate half and shadowed lesser part of another half. This made plate and cathode cooler and lowered plate current at same plate voltage but still the output power readings show 24W value when the lower limit is 18W. We'd love to get rid of the side getters at all but then we'd get the tube life time below 500 hrs. Anyway, thank you so much for being a fun of our KT120! You are great too! I'm your fun now and subscribed!

We learn the most when things go wrong. Knowing this helps me deal with the disappointment of failure.🤔

Love this content, thank you. Visual review of the tubes was the chef's kiss. If anyone had told me I'd be addicted to KT120 troubleshooting videos this month, I would've asked them what they've been smoking!

Your thought process along with your natural drive to resolve problems is what makes you who you are and a great resource. I am so glad the Upscale Audio tubes worked out. I am a fan of them too.

FYI: The "getter" vaporizes calcium metal, which condenses to form that silvery deposit. Calcium is an Alkaline-Earth (Group II on the Periodic Chart) metal, which is highly electropositive, and the vapor reacts with the oxygen & nitrogen in any air present, to form calcium oxide & nitride, which are very high melting point solids.

Top observation. Just another sign of no quality control systems in the manufacturing process. I guess we need to see more examples to prove the point, but it looks highly likely you are correct. In the mean time something to look out for when buying tubes. Cheers.

Yeah! You rule, Steph! That’s great news! I’m always impressed by your detective work. Yup, the getter grabs up all the extra electrons, but great work figuring out why the tube was defective. I doubt I would’ve figured that out, and I work with tube gear pretty much every day. We get our tubes directly from New Sensor, but had to go to other people for a bit, and luckily haven’t run into too many issues. We do only run most tubes at 70%, as I mentioned to you. That little bit of extra power can make a big difference, especially when used with high sensitivity speakers. Thanks for letting us know about the tubes. Fantastic vid! You’re awesome, Stephe! 🔊😊🎶

Yay! I'm glad you got a working amp for all your effort!

your sense of reasoning is impressive.

What a super detective job! It just shows again that you have a great natural understanding of how these things work! Thanks so much for this!

Great detective work Stephe! I'm sure you're right about why these stoopid 120's cook! As you say - why did they change the position of the getter? Crazy! You're definitely exhonerated now! You've done a great service! 👏👏👏! People will now just need to look at the getter & the position of the chrome patch on the side of a KT120 to know if it's a good or bad tube! You wait Stephe - I bet your discovery will go viral real quick! You should be proud of yourself! Someone needs to get on the phone to Vladimir to tell Dmitri to rotate those side getters 20°.

That is a really intelligent theory. Well done

You are absolutely right. I have 6 KT120 tubes and 2 of them have getter position not as bad but almost like yours. I just tested one at 120 mA bias and it redplates slightly at 150 mA worse and not at 100 mA bias. So your theory is absolutely correct.

your analysis makes perfect sense to me.

I think your on to something makes sense to me good call steph.I knew you would figure it out

Great video. I use older (2020) kt120 for my CJ- Premier 11A amp and they work great. Great base too like you noticed with yours.

Probably welded where it is because the plates are being welded together anyway. Its just one more "thing" to clamp together in the assembly. It would make more sense over the beam plates though, there are a couple of extra radiator fins on either side of that weld joint in a KT120 for heat dissipation, and un-blocking those would make them even better able to handle it.

Thank you so much for this video. I am really playing with the idea of building these monoblocks, as I have 8 KT120s laying around from when I built a pair of Bob Latino M125s (which are too noisy for the very efficient speakers I am currently using). All of my KT120s are from several years ago, and all have small getters that are off center of the plate...so even more reason to try your build!!

Super smart my friend super smart

Thank you so much, I really struggle to understand and retain when I read, but your explanation here made things perfectly clear, I think I'll subscribe and embark in binge learning

I never thought l would be learning new things about vacuum tubes so I’m enjoying being on this journey with you.

Hi Stephe - I just checked the four KT120 tubes (bought new in 2022) in my PP Luxman A3600 amplifier and they all have getter material centred over the seams of the plate! I'm lucky that I decided to bias mine at 75mA at 500V for long life and I guess I'll stick with that! I'm also blown away with how good my amp sounds - I'm reaching for more and more music now because it makes everything sound sooo good!

Congratulations, the kt120se sounds really good

Kudos!!! As a 40+ year EE tube guy I can say yes, most tubes made in the last 5-7 years are junk. Upscale does add loads of value in their knowledge and stock as I have personally recommend them and have always had a good experience (tubes / phono-stage and turntables... not on speakers however, they need serious help here)...

I'm using Tung-Sol 7581A from Tonal Tubes last 3 years and am happy with them.

Fantastic video! I learned a lot about tube design from your video. Thank you for stepping through the tube parts. Fascinating! Your description makes it easier to understand how the tube operates. I’ve been using tubes for just a few years and really dig the elegance of the tube’s function. Good stuff.

This is great information for anyone looking to buy tubes on the used market what to look for. Thanks

I believe you are right about the getters. That coating is reflective on the inside as well as the outside. That means the heat would build instead of dissipate.

Awesome. Excellent deduction abilities you have, and I believe you are correct based on the trials, even down to how far the getter flash is over! This information needs to be shared with the manufacturer, because even though they should know, maybe they don't! The other part of me says, see if there is excessive current draw from the red plating tubes, and let them red plate to see if they do ok. That does however, keep one on toes as to when there may be an actual problem with the tubes. Output iron is not cheap! Ultimately, and IMO, it is time for another solution, possibly figuring out how to make a KT88 perform well in this circuit.

Bought a brand new pair of EL84 from my Mid-century RCA console and almost immediately one tube started red platng! throw them in a box and continue to use the old RCA ones from the 60's ...lol

I assembled a pair of Bob Latino M125 mono blocks and ran 8 KT 120s in them and they sounded absolutely amazing!!!

I'm cheering on the RU-vid algorithm. Let's find 410 more subscribers to make 10K!

Wonderful news! 🎉 Does this mean that you will run some tests with these new tubes? Can you check how much the tubes can be pushed to get more juice out of them but still be within a range where they will last quite long? Like perhaps test how much different settings of milliamps put out in wattage.

That makes alot of sense! I would never have tought of it! Great work as usual. Thanks! I learn from you! :)

Totally great observation. Very good investigation

Not only does the mirror surface block heat that needs to leave the plate, the curved shape of the the mirror surface both bounces the heat back and focuses it to hit exactly to the middle of the plate

Just checked 3 pairs of tungsol 6550, 2 were from ebay and were low use but 5 years old, the other 4 were from a uk based supplier just after the war started. Thankfully they all have the silver film to the left or right of the join. Hopefully they are all good tubes. Guessing you like the TL output trans then.

Can't wait for Western Electric to make EL34 tubes.

I run mine at 80ma, and I'm happy with them. I think you might be on to something with the getter position. I had purchased 4 KT120's 3 years ago, and they have offset getters.

I believe they had someone new on the assembly line when welding things together. They probably had a box of halos with the standoff wire/tab already attached. This would then be positioned and welded in place. It could have been that, based on the orientation of the base or holding fixture, it was more difficult to hold the halo flipped in one direction and easier to hold it in place in the other. It could be that the standoff wires/tabs should have been bent after welding - conversely, they were supplied bent to clear the plate joint but the assembler didn't know they should be bent so straightened them to look better (cussing the halo makers with every tube). Who knows? I just hope I can find a second tube that will bias to 90ma.

Thank you, I just compare 4 -KT120 that I bought 4 years ago on ebay they're the same as your good one but only running at 65mA. Will build a new one that can run up to 85mA.

Someone had a hangover attaching those getters on the production line.

Well spotted!…

That's a very intuitve observation but there must be a reason why Tung Sol put the saucers where they do. Either it's easier or they have to be in that posistion because of the internal construction. Hovever if this is what's affecting the power handling it would be an essential part of the manufacturing process to get right, which is terrible QC. One question I have is if this amp was built for use with KT88's would the sound quality (ignoring power) be as good because people like KR Audio etc make good KT88's.

This might be informative. Try graphing the transfer characteristics of the good vs bad tubes, for various grid bias voltages, and plate & screen voltages & currents. That should tell you what's screwed up, though not why. But you might determine the values to make the "bad" tubes good.

A friend of mine runs his PP amplifier in triode mode, the tubes ( nos 6550A) are biased very hot in class A .............455 volt plate at 100 mA .............Six years later the tubes arestill fine and within specification ... WHY ,.................the plates are relatively and asthoningly small and fine .................those were made with three alloys ( tri-alloy clad structure )...........the different alloys were blasted together with DYNAMITE ! The tube .........6550A GE .........and some other US manufacturers...... Those were the days ...you bought the tubes, you used them without testing and in most cases all went well. Graphite plates as seen in big emmision triodes like the 211/VT4c and the 845 kive even longer when pushed a little into red plating ( seen in the dark ) , BECAUSE the anode was coated with zirkonium as extra getter..............when heated above a ceratin temperature they become very active as a getter ..........and the tube livesd longer............Seen in some amplifiers , more than 20;000 hours with more than 1000 volt on the plate........ So everything is relative besides badly build low quality tubes. To conclude ................as long as the anode gets it's electons out of the electron cloud without pulling them ( = stripping of the cathode coating ) there's no problem of shorter lifespan ....that's another myth in audio circles ............but the temperature of the glassbulb has to stay under the designed max temperature rated by the manufacturer.........

Wow that makes sense why they were red plating. I been using Tung Sol EL34's in my guitar amp never had that happen before. Getting tubes these days our getting harder.

Interesting. My recent Mullard KT-88’s are just off to one side like the tube you show, sometimes the plate seem o just covered up, but the disk will be off to one side as the wire that holds it moves the disk away from the gap. No red platting, but as you said biased pretty low.

The tube delivers enough current at a voltage and bias ...........but the plates redplate way below the rated wattage as announced in the datasheets . Older versions work correct as you experience . They have saved on the materials , quality and quantity of the plates ( and other elements ).........especialy reducing the amount of nickel deposit on the plate and the quality of the metals ! We have seen this before at the end of the tube era production ( example : redplating of the famous GEC KT88 due to the heavy concurence war with chinese shitty productions sold way less expensive ) Those low qualty badges that were refused by manufacturers ( hifi and guitar amps ) were returned to the factory but resurfaced many years later on the market sold as expensive nos tubes . Proof , the same tubes manufactured in the years earlier work fine ........most amplifier manufacturer switched to this family of tubes , where there is no alternative other manufacturer who builds the same type ..............all those customer will encounter the same problem with their expensive tube amplifiers ...........Mac, Audio Research, Jadis and many others , Good news that your amplifier works fine now ! Kind regards

Great explanation !!

Do you think it would be worth experimenting with cooling fans? A well-aimed, quiet cooling fan could help with heat dissipation, and could perhaps be powered from the 6.3 volt heater line...or an external power supply

Maybe somebody made a mistake on the production line at the factory. Maybe a new employee

You just created another parts shortage. 😊

Nice detective work Skunkie. Could you look and post what the KT150s look like?

One more comment... My measurements show that you get lower THD the harder you drive the tube. How does your THD look at 80mA vs 110mA ? I also completely agree with you that the KT120 sounds really really good. I had not tried it before at 500V and more than 90mA bias because i had not a high enough PSu then. I will definitely continue to work with the KT120 as an SE output stage. Thank you for your work.

I would like to ask if you think it is possible to bring the "color" pre amp feature into this power amp by increasing the input signal into the 6EJ7 and reducing the output to maintain the same volume ?

Hi skunkie. I have the kt88z that you suggested so I ordered more. They are 30mha so they will be ok biasing on a Muzishare x7? Thanks. Really enjoy your videos!

I use KT 120 with my Doge 10 mk2b and with my Galion TS120 SE no problems at all ' Auto bias dials them in perfectly ' I have over 7 sets of kt88 and the kt120 bass ATTACk is better than all my Kt88 but the Golden lion kt88 resolution is hard to beat ' but If you have bright Crispy speakers Mullard kt88 is the way to go ' I actually use 2 golden lions kt88 and 2 TAD kt88 for power tubes ' and Wow !! Sound great with my dynes (Dynaudio towers)

Yet the same Tube Brand..... made in Russia.. I have found various tubes made in Russia 30-40 years ago, which I have many of... work far better than Russian tubes made today, especially tubes that never existed even 10 years ago...like KT120's...

First thing to verify when a tube redplates ( when normaly they don't or never did ) is to check the coupling capacitor ( leaky ) ...........in this case with the amplifier build by Stepf this is not the cause ! Second verification when it is with PP amplifiers biased with a shared cathode resistor , when aging one tube will do all the job and the other will be in starvation mode . It acts as a seesaw ........ Third reason when this happens is caused by grid current with high perveabce tubes , fixed bias that doesn't protect and compensates to prevent run away .......also to high a grid resistor value as seen to often ! ( not in your amplifier Steph ) Fourth reason ..............;;bad and low quality build tube ........ass seen with these KT 120/150/170 from recent production . These tubes that now one else builds are some kind of a bussiness prison .... You're lucky to find good ones or to have build up a decent stock in the past .

When are you posting the schematic and parts list for the 120 amp?

There is just too much fettling involved with tube amps. I don't want to buy a brand new amp and still have to tune it up like an old AMC Gremlin. Also too cost prohibitive. I was thinking about a hybrid amp maybe like the Dayton hta-200, but now I don't even know about that. Also, is a fully tube amp considered a class A amp?

Thank you. I'm not sure what pp means?

Would they redplate when running bias at 45ma? Great video btw🙂

You might have been able to "getter" to the bottom of this problem!! 🙂

what the cost to build that amp for the parts

ill stick with my 300 B shoebill amp

How is it that the tube engineers didnt realize this?

That is a really intelligent theory. Well done