Rerurn to Romy the Cat's Site

Melquiades Amplifier
Topic: Be kind to your tubes

Page 1 of 1 (12 items)


Posted by Romy the Cat on 10-02-2011
fiogf49gjkf0d
I do not have a lot of equipment that uses vacuum tubes I have a phonostage and power amp that are tube based. The phonostage used 12AX7 that in there from 1999 and I think they will survive me. The power amp use gas tube 0A2, the output triodes 6C33C/YO186 and drivers 6E5P/6E6P.  Dealing with tubes I kind of always in defense about the state of the tube and I do not know when to change them.

I do have tube testers and can measure anything miserable. In some cases I observed that mutual conductance, cathode emission of pate current were gradually lowering and it was reflecting that the tube get weaker. In some case I have seen that the same type of the tube after a lot of working still hold the specification as it was brand new, just taken out of box. And I am taking about the tubes that have seen a lot of use. My power amps run a lot and the 6E5P/6E6P driver tubes are used in there not in very spearing way – at 7W on plate. I have seen some tubes that looked absolutely fried but worked, sounded and measured perfectly. I always in defense what to do with them. 

 My initial idea was to put time meter in the power amps in order to monitor how long the tube work. Sine I have 6 channels and 11 tubes in each amp with different degradation rate then it does not appear too practical to me. I keep a log where I wrote when I replace each tube. It was useful and I was trying to get a year from a tube. Then I abandoned the log as getting more experience I can very accurately estimate how long the tube worked by evaluating the density and character of darkening glass on the tubes sides.

Still, the initial question still remains not answered. What to do with the tubes that worked own period and still perfectly operational. I was not able to see any sonic differences or any other advantages to change them. What is your take on the situation?

The Cat

Posted by decoud on 07-29-2013
fiogf49gjkf0d
Romy, have you come to a conclusion about how to decide whether or not to change a tube in the Melquiades? How exactly does the glass change? Is it just the getter? What about the gas tubes where there is no getter? 

Posted by Romy the Cat on 07-29-2013
fiogf49gjkf0d
As much I did not answer back then as much I do not have it now. The gas tube I never changed. Well, as I moved out of city in 2010 I did changed them but not because I need but because I was “reviewing” the system sort of speaking. The 12AX7 in phonostage I never changed. In 2010 I opened the phonostage and measured the tubes – the show full emission – whatever it means… Those small signal tubes with 1mA on plate I think never die, not to mention that it is a phonostage – a devise that is not being use all time. The 6E5P/6E6P I change each year. I do not know if it right or wrong. They are Russian tube and they dies whenever they want. I do measure them after a year, some of them do lose some transconductance, some do not but I replace all of them juts to be on save side. The output DHT on MF channel I did not replace and I do not know how to test it. I do not believe in testing the DHT on my tube testers. They do maintain the original plate current, so I presume they are still fine but I do not know how to test the true cathode emission. This YO186 tube is used in very conservative way, so I presume it will be live for a long time. The 6C33C uselessly live for 1-1.5 year. I do not test them but hear what the amp begin to “ping”. This is very characteristic ping of dyeing 6C33C and as I hear it I try to figure out what channel it is coming from (sometime not so easy) and then replace the tube. 

Frankly I do not hear a whole lot of difference what a tube or two are replaced and I do it rather as a preventive measure. I wish I would have a better way to do it, the way that would a tangible sonic feedback that the amp begun to use new tube or tubes. I do not have that sonic conformation. I remember when was changing the second stage on my Lamms power amps or rectifier on my Lamm’s preamp I was truly huge different. Not in my currant setup, unfortunately.

Rgs,
Romy the Cat

Posted by Romy the Cat on 07-30-2013
fiogf49gjkf0d
The more I think about the more I feel that my old idea to install a timer inside the amp that would measure the time the amps is running was not so bad. I never in fact implemented it but it would be convenient to have it.

What I am thinking is that the situation that people like me opens an opportunity to somebody come up with a very useful product that I personally would not mind to buy.

Pretend that power cord of any devise with tubes that you would like to monitor plugs into a series male-female adaptor. The adaptor monitors when current is flown and RF dispatch data let say each 60 seconds over own channel. A usage monitor with let say 8 channels installed somewhere with 30-50 feet and reports time usage across each devise. I find it to be very convenient and I do not mind to have it. If somebody would come up with something like this send one for me for free transparency testing of that series male-female adaptor.

The cost. I think that rescores usage monitor with let say 4 channels (extendable to 6 or 8 channels) might cost somewhere in vicinity of $300. My 2 channels pool thermometer is water proved, has own buttery, broadcasts up to 100 feet and the monitor receiver has a lot of data about the in-poll and outside of pool statistic. The whole pleasure costs $21    and free shipping. Considering the necessity to interface with audio power cables and the high-end hoopla I think the price tag of $300 would be excusable. Well, I would as much for 8 channels devise once in life time.

The Cat

Posted by decoud on 07-30-2013
fiogf49gjkf0d
probably drive the price of something like this...
http://www.nico-tech.com/3-Product-4-PM.html
...way higher...

Posted by rowuk on 07-31-2013
fiogf49gjkf0d
Romy, I think that the only real "solution" is to develop a boot cycle for your amps that consist of warmup, parameter test and after this is test is passed, change to operational mode. I am pretty sure that your tubes do not simply unexpectedly explode. I am pretty sure that some vital parameter (like transconductance) deteriorates until there is no choice except explosion. If you, with a PC monitored the vital parameters during the boot and shutdown cycles (when no music is playing), you probably could run your system to a preset trigger level and get a notification that doom is near. AD measurement cards with multiple channels are available at low cost as are analog sensors to measure just about anything. No digital operation would be necessary during the audio playback mode.

Posted by zztop7 on 07-31-2013
fiogf49gjkf0d
Rowuk, you triggered a thought with your words: "boot cycle" "warmup" "change to operational mode".  This question is to all the people that have this knowledge >>>
    For the benefit of the tubes, and other parts; is it better to have NO signal to the amp during warmup??? or is it better to have a tiny signal [lowest voume setting]???   The possible thought is that it might be better to have a little 'release'.    zz


Posted by Paul S on 07-31-2013
fiogf49gjkf0d
ZZ, I'm thinking, ramp up the properly-sequenced DC first (and perhaps we go on from here to the earlier-discussed matter of heating and shutdown sequences for each tube...). After that, when I send signal, I start low and ease up on the volume until it "feels right" to open it up, when the system/components begin to express themselves properly.  However, I have no idea if/how this protocol affects tube life, per se. I just try to run tubes with the operating points at spec. Like Romy says, some high-spec 6922s, 12AX7s and a few other high-spec small signal tubes might last 10,000 hrs. if they are run at spec.  OTOH, some tubes might only last 3 - 6 months before they sour off, like the 6H23s.

In actual practice, if a tube "sounds bad", or if replacing it improves the sound, then it's bad (or sub-optimal) for that application, no matter how it measures.  I wrote about my own experience with the Russian 6C33C output tubes that I routinely replaced in that position after 1 year, but I always swapped the outgoing output tubes into a voltage regulator application and got another year out of them that way.  I was curious to find out how long some NOS Svetlana 6C33C output tubes would sound good, and I pushed them for 2 years in that position, but I trashed them after that, out of general paranoia.  Being "thrifty", I always clean and re-grease tube pins and sockets before I give up on a tube.

Best regards,
Paul S

Posted by zztop7 on 07-31-2013
fiogf49gjkf0d
HI Paul S,??? Paul you wrote "re-grease tube pins and sockets" ???   That must be some very conductive HIGH temperature grease.  I personally keep my amplifiers' tube pins and sockets dry.   Maybe you know something here that I have missed all these years.   Respectfully,  zz.

Posted by Paul S on 08-01-2013
fiogf49gjkf0d
ZZ, no surprise, I suppose, but this has already been discussed here at some length (God knows where, though...).  It's the same grease used for spark plug wires (at the plug connection) and high voltage connections in general, and I use it very sparingly.  It worked great for the super-hot 6C33C pins, and the crappy sockets, when I used those tubes. And I hope everyone already knows, no Caig (eg., Pro Gold) or similar contact "enhancers" on tube pins or sockets!  Nothing at all is a waaay better choice!

Best regards,
Paul S

Posted by Romy the Cat on 08-01-2013
fiogf49gjkf0d
The concept of self-testing boot cycle is nice one but it for sure could not be done during the amp booting. Some tubes do need to warm up and reach the cruse temperature in order to be tested, some do not.  Some tunes (like 6C33C) in fact is best to test during the warming up: by looking how the parameters change during the first 10 mine of tube operation a lot might be said about the tube. Fromm a certain perspective any tube amp is a tube tester and all it take is to slightly reformat the tube operation and bring the data in an interpretable format. The interpretable format is kind of tricky as it would depend from the type of the tube.

Generally looking into a mutual resistance of the tubes we can have some idea about the tube life. I feel that change in mutual resistance in a way a good indicator of tube life but we can’t presume that it has to be some kind of fix number as all tube testers do. It is not the absolute number what is important but rather the change of the number as the tube ages. A rate of decrease of tube gain, the “mu” with decreasing filament voltage of plate voltage is good indication for some tube. The testing of cathode emission is a good indication for other tube. Some tubes (like 6C33C) might be very easy to be tested by looking how bias changed since the tube was broken in. so, in my view there is no universal way to test all tubes but any given amp for sure can test own tubes in the best way to itself.

I think it is a good idea to make any amp that more or less expensive to have a self-testing capability. However, this direction has own “issue”. The amplifiers make sell the replacement tubes and they have huge mark up on those tubes. Letting them to moderate the tube testing is not advisable in my view. It is like my spa dealer goes over himself by advocating do not use ozonizers in hot tabs. The little dirty secret is if his clients use ozone for water sterilization then they buy much less spa chemicals and spa replacement parts from the dealer. Me, being skeptical and believing that any industry person would screw anything that permitted to be crewed, I personally would not trust build in amp tube tester. At least I would need to recalibrate then to my own sense of what it right and wrong. I know, it is a bit paranoiac but it is what it is. If you even buy from a “trusted souse” the matched tubes that ended up to be grossly  and even laughably mismatched then you know what I mean.

Rgs, the Cat

Posted by rowuk on 08-01-2013
fiogf49gjkf0d
 Romy the Cat wrote:
The concept of self-testing boot cycle is nice one but it for sure could not be done during the amp booting. Some tubes do need to warm up and reach the cruse temperature in order to be tested, some do not.  Some tunes (like 6C33C) in fact is best to test during the warming up: by looking how the parameters change during the first 10 mine of tube operation a lot might be said about the tube. Fromm a certain perspective any tube amp is a tube tester and all it take is to slightly reformat the tube operation and bring the data in an interpretable format. The interpretable format is kind of tricky as it would depend from the type of the tube.


Agreed, BUT we first start to listen when the WHOLE amp is finished. That means the "boot cycle" is over when the "slowest" section of the amp is done.  If the goal is to replace parts well enough BEFORE disaster and before they start to sound "worse" then we need to identify the parameters that lead to havoc or sonic degradation and test for that. Perhaps a specific impulse measured could identify changes in gain, phase and distortion. Perhaps a change in anode voltage outside of a predetermined range after warmup would give us an indication that the end is near.

Assuming zero intrusion in the sonics of the playback system (all testing sensors either as integral analog devices or switched out of the circuit entirely), boot cycle logs would after a period of time offer very useful data about the life, happiness and death of the devices in question. As the Melquiades amp uses a limited amount of tube types, the formats should be relatively "easy" to generate and maintain. I use EL34 and EL84 output tubes (driven directly from a high gain preamp) and have had no trouble with reliability. The sonics decrease in a predictable way, then the tube is switched out, but I have NEVER had a tube self destruct.

Page 1 of 1 (12 items)