Would you buy used tubes?

 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3787
Registered: Feb-07
For the tube guys out there - I see a lot of used tubes on Agon and CAM.

Just wondering what the consensus is. Bad idea to buy used tubes?
 

Silver Member
Username: Hawkbilly

Nova Scotia Canada

Post Number: 901
Registered: Jul-07
Buy from someone who knows how to measure and evaluate a tube. There is a ton of tubes out there being sold by people who know only about what you or I know about them. Find a place you can trust and buy from there. I'd stay clear of ebay auctions for tubes David. I can recommend Vintage Tube Services and Watford Valves as two places that know valves, have been in the business for decades, and have fair prices.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3789
Registered: Feb-07
Thanks Chris. I avoid ebay like the plague.... for everything.
 

Silver Member
Username: Ezntn

Greeneville, TN

Post Number: 181
Registered: Apr-09
Neil recommended Upscale Audio when I was in need.
I'll second that recommendation.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14783
Registered: May-04
.

It depends on the tube and how it will be used. Tubes begin to wear out the first moment they are placed in service. However, you have little to no idea how a tube was used. Has the tube been used continuously or on a duty cycle that allowed for long periods of inactivity and therefore no wear? Either way it must be sold as a "used" tube.

Is the tube a driver tube, power tube, recitfier tube or just where does it fit in the circuit? Used tubes for phono pre amps - particularly mc pre amps - are generally not a good idea. Is it an NOS tube that offers a good value for the money spent? Is it a white box tube that has no specific manufacturing information to support its origins? Are you buying these tubes as spares for testing when problems arise or are you planning on using the tubes in your new $15k pre amp?

As with any other commodity, if the situation suits, saving a few dollars on used goods can be either a good or a very bad decision.


.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3790
Registered: Feb-07
In this particular case it would be for a quad of output tubes (either 6550 or KT88), better quality than the ones I have now, and saving the current ones as backups. Good plan, no?
 

Platinum Member
Username: Nuck

Post Number: 14867
Registered: Dec-04
I would make the deal contingent on having the tubes tested. I would not mail order used tubes at all, unless from a dealer.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14787
Registered: May-04
.

"In this particular case it would be for a quad of output tubes (either 6550 or KT88), better quality than the ones I have now, and saving the current ones as backups. Good plan, no?"



I don't know. I assume you're dealing with a reputable company who has tested the tubes and can guarantee their quality. Do you have any return priviledges with these tubes? What are the differences in the deal to be had vs. buying new tubes? Not dollarwise but, are there any exceptions to the policies afforded new vs. used tubes?


A reputable dealer should be able to give you an idea of what the tubes were tested for. Transconductance, noise, microphonics are all important. Some lesser dealers will simply test the tube for operation and, should it pass without spitting flames, sell the tubes in "tests as new" condition. This might suit your needs or it might not.

Most tube retailers of any quality are willing to spend time discussing which tubes might suit your needs best. If this retailer isn't offering that service, I would pass.




" ... better quality than the ones I have now ... "


How do you know that? By what you've read on the retailer's webpage?


What are your options? Are these NOS tubes? Then possibly they are better quality but not merely by virtue of their being NOS. Don't get sucked into the "they don't make 'em like that anymore." There are numerous newly manufactured tubes that offer excellent service and top quality sound. There were also cheapass tubes made in the old days.


Are you buying a matched quad? Does your amp require tube matching? What if one of these tubes goes out in six months? Will you be forced to then buy a new quad set if no more "used tubes" are available? That could prove to be a costly gamble.


Most anyone who buys tubes before their old tubes are worn to a frazzle will tell you that you never use your "spares" once you've heard something better.

.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3791
Registered: Feb-07
" ... better quality than the ones I have now ... "

"How do you know that? By what you've read on the retailer's webpage? "

Mostly by doing research, and asking for opinions from people such as yourself. Probably a process of trial and error as well.

From what I've read, most amps require matched tubes for the output stage, but not for the input. Is this true?

I would probably be dealing with the The Tube Store (http://thetubestore.com/info.html) since they are based here in Ontario and come recommended from several other people who have dealt with them.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14790
Registered: May-04
.

I've dealt with The Tube Store and consider them to be very fair, knowledgeable and reputable in their ethics.



If your amp uses fixed or auto bias arrangements for the output tubes, then there's no specific requirement for matching tubes. Most shops only charge a few dollars extra for tube matching any way so most people spend the money just to have a more closely paired set of tubes. If the amp employs user adjustable bias, you'll need to match tubes to get the best performance from your amp.


Input or pre amp tubes do not require matching though you would like to have a pair of tubes that closely track one another as far as tube type, microphonics and noise at minimum for circuits in phono stages. Line stages, pre amp and driver tubes in power amplifiers need no particular matching other than to be of similar make and manufacture. Rectifier and phase splitter tubes are typically only one to an amp, so no real need for any sort of "matching" there.



.
 

Platinum Member
Username: Nuck

Post Number: 14872
Registered: Dec-04
The Jolida has auto-bias, as indicated by small LEDs when biased correctly. While I do not consider this as empirical bias balancing, it is an indication that a range has been reached, so , no, a matched (and pricey) set are not required.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3792
Registered: Feb-07
Like we discussed though, a volt meter may be a good idea? So instead of range, I'm hitting the voltage exactly for each tube?
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3793
Registered: Feb-07
Stupid question here (not that I'd attempt this, it's purely hypothetical). My Jolida uses two output tubes per channel. Can you mix and match tubes? Say a 6550 and a KT88 per channel?
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14793
Registered: May-04
.

A volt meter is a good thing to have around but, if your amp has no adjustment for bias other than disassembling the amp to replace a resistor, then you aren't going to be using a meter for bias adjustment. Consider that even the best amplifiers with fixed or auto bias have some degree of tolerance in their components and that no single tube is an perfect match to any other single tube and the need for exacting adjustments becomes far less important. The tubes will produce the specified output power and distortion as long as you have them within their range of settings.


Keep the tubes similar across the board. A KT88 and a 6550 are very likely to have a different and somewhat distinct sonic character and certainly somewhat dissimilar eletrical requirements. Go with a single type of tube and use it in all locations where applicable.


Make certain the tubes you buy are well suited to your specific amplifier. Some output tubes will operate to spec only within a circuit supplying the appropriate bias and plate current. Running a tube which prefers a "hot" bias in a "cold" amp might disappoint you. A good tube dealer knows which tubes suit which amps.


.


.
 

Platinum Member
Username: Nuck

Post Number: 14875
Registered: Dec-04
If you are still poking around by then, Dave, we cn measure and bias the stages next trip.

Don't mix tubes, keep the 6550's in there, I think. Good tubes,if lacking a tad of slam of the KT88's, as you said, you do not lack bass, either.

The beer was flowing just too well to open the amp, just your service panel, LOL!
The panel is much safer than the amp, trust me.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3794
Registered: Feb-07
I'm actually pretty happy with the amount of bass. I'm surprised actually, since everything I'd read about tube rigs (and the Jolida in particular) is lack of bass. No lack of bass on my system - I even disconnected the sub you make fun of all the time ;-)
 

Platinum Member
Username: Nuck

Post Number: 14878
Registered: Dec-04
Not all the time, just now
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14797
Registered: May-04
.

"Tube bass" is often different than solid state bass due to several factors. The most obvious is the output impedance of the tube amp due to transformer coupling of the outputs. This will raise the output impedance of most tube amps to the point where a reactive load type speaker can affect frequency response in direct relation to impedance of the load at various frequencies.

A typical direct coupled transistor amp will have an output impedance well below 1 Ohm at all frequencies which minimizes the interaction (as predicted by Ohm's Law) to barely noticeable in most cases.

As the output impedance to load impedance widens the damping factor of the amplifier drops also. The DF required before diminishing returns set in is disputed but, none the less, as DF decreases due to load impedance the amp is less in control of the speaker which accounts for the "wooly tube bass" often mentioned by those who do not understand how to mate speaker and amplifier. Keep the speaker as non-reactive as possible and the overall impedance as high as possible as they did in the days when tubes ruled audio and you should hear clean, tight, tuneful bass from most well designed and well implemented tube amplifiers.


.
 

Gold Member
Username: Magfan

USA

Post Number: 1066
Registered: Oct-07
I'm not a tube user, but I'd never buy used power tubes no matter the test passed.
low level tubes, yes, I'd be tempted.

As for 'tube bass', isn't there also an interaction with speaker 'Q'?
A speaker which is low 'Q' ....say below the critical .7 value should react better to low damping amps, shouldn't it? Anything over about say.. 1.1 will sound a little flabby even with hi damping SS.
 

Gold Member
Username: Mike3

Wylie, Tx USA

Post Number: 2380
Registered: May-06
There is one more factor to piggy back on JV's and Leo's posts above. The cables or Interfaces. MIT CVT 1's are the best bang for the buck between my MAC tubes and the Gallos.

As an aside to the "wire is wire" arguments abounding on other threads let me add this from the MIT community site;

"Audiophiles are all familiar with the now infamous "cable debate" surrounding the question of the impact that audio cables have in a hi-fi system. Today's audiophiles understand that differently designed cables do impact the overall sound of their systems. We at MIT Cables agree. But these "just wire" designs only serve to manipulate the sound by changing the relationship of capacitance, inductance, and resistance within the cable. They don't eliminate any problems.

This is where Bruce Brisson and MIT Cables separate themselves from other cable companies. Bruce discovered back in the 1970's that audio cables were inherently flawed for their purpose. What he determined was that a high fidelity system needed more than just cables to connect audio components; they needed devices that were engineered to properly interface them.

So, what is the difference between an audio cable and an audio interface? Simply put, an audio cable is a piece of wire that is used to make a random connection between audio components. An interface is an engineered component that is purposefully designed to efficiently transport energy, with a predetermined bandwidth, from one component to another."


How's that for hi-jacking Dave's thread?
 

Gold Member
Username: Magfan

USA

Post Number: 1070
Registered: Oct-07
wow.
cable as component. Let me stew that one over for a while.
 

Silver Member
Username: Hawkbilly

Nova Scotia Canada

Post Number: 909
Registered: Jul-07
An "interface" sounds pricier than a "wire".
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3795
Registered: Feb-07
My interfaces right now are decidedly low-brow.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3796
Registered: Feb-07
What you mention about impedances and load Jan coincides with a conversation I had with Frank a little while ago regarding speakers and tube amps. He mentioned that I should not be too concerned about the sensitivity of a speaker vis-a-vis it's suitability for mating with a tube amp, but instead look more closely at it's impedance, making sure it's as stable of a load as possible.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14798
Registered: May-04
.

If all other values are are equal with a series of speakers all presenting a stable load, then sensivity is somewhat of a big deal. Paul Klipsch's famous quote remains, "What this country needs is a good five watt amplifier."

The amplifier's simplicity of design, a minimum number of output devices and money to burn on execution rather than bling all result from high sensitivity speakers. The amp remains in its most linear range at all volume levels with uncompressed dynamics. None of that should sound bad.

The traditional problem with this remains the trade offs made to gain sensitivity in a speaker system. A few of these trades are rather obvious and intuitive. What many listeners don't realize until they've purchased those 104dB speakers is the amount of quiescent noise they allow through when the amp is at idle. There aren't many less than first class (priced) tube amps that are really designed to mate with the most efficient speakers and the amplifier's inherent noise presents a floor you must reach above just to enjoy music. When the speaker is capable of the upper 90's or higher in sensitivy and you tend to listen in many domestic settings at less than 95dB you will find signals lost in the amp's circuit noise. You must then turn up the SPL's just to hear the quietest parts of the music.


All things equal, a non-reactive load on a tube amp is, IMO, the most important value to consider when selecting speakers. Sensitivity normally equals efficiency, which is always good as less electrical energy is wasted in the conversion to acoustic energy, but it does come in second to the stability of the load. The little 15 Ohm LS3/5a's are still very (tube) amplifier friendly even at 82dB.




"Q" of the system establishes the speaker's damping. Once a speaker strays from being "critically damped" the quality of bass also changes with underdamped systems being loose and flabby and overly damped systems sounding too dry and analytical. Once again the 3/5a presents a trade off that has been exploited by numerous designers since the BBC introduced the design in the mid-70's. In a small box the size of the 3/5a (a speaker designed primarily by well trained ears and slide rules) a slightly underdamped system can be made to give the impression of much deeper bass by allowing a slightly looser (under damped) bass quality, the well known BBC "bump" (folowed by thr even more well known "Grundy dip" in the lower mids) which places a slight rise in response in the midbass region where the harmonic structure of many instruments provides the "feel" of deeper bass. If you look at the frequency response graphs of many speakers of all sizes, this trick of psychoacoustics has been employed in many systems of all sizes ever since. It is a deviation from "accurate" if you judge the term by simple frequency response graphs. No matter, the slightly underdamped system sells well and - all things equal - tends to portray a more transparent sound overall. Given the wide ranging response errors introduced by most rooms, I'll take a more transparent system in any room over a more accurate system only in an anechoic chamber. YMMV


Critical damping is a goal to be observed and then missed by most deisgners today. Speakers with such alignments tend more toward the slightly dry side in many systems, especially those with high damping factor (low output impedance) solid state amps. While DF in the amp is important, the speaker's system "Q" still determines the quality of bass whether the amp driving it is tube or transistor.


http://www.vaughnaudio.com/tech-papers.html


.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3797
Registered: Feb-07
Interesting.

"What many listeners don't realize until they've purchased those 104dB speakers is the amount of quiescent noise they allow through when the amp is at idle."

So do you mean that hyper-efficient speakers can raise the noise floor of a system?

As an aside, I've tried several of my speakers with the tube system, and the Sttafs remain the best sounding.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14799
Registered: May-04
.

No, the speaker cannot amplify the noise. The highest sensitivity speakers system simply make the existing noise more obvious.

Think about it this way - and I'm doing this without the benefit of coffee - the speaker produces, say, 104dB with one watt of input. Another speaker produces 85dB with a similar input voltage. Work out the math to determine just how much power the amplifier must produce to provide a 95db "average" SPL in the room with each speaker hanging on its outputs. Clearly the 104dB speaker is still well under one watt of power.

The higher sensitivty speaker will simply be reproducing the noise content within that one watt of power at all times. The lower sensitivity speaker requires more voltage to rise above the avaerage room noise, therefore, the quiescent noise of the amp remains low to the listener's ear simply due to the fact the amplifier must be producing significantly more power while the quiescent noise remains at the lowest levels of wattage.

The amplifier's idle noise is then lowered in perspective to the music with the less sensitive speaker while the dynamic peaks are also lower in overall level as the amplifier strains to reach the upper limits of its reserves.


.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3798
Registered: Feb-07
"the speaker cannot amplify the noise. The highest sensitivity speakers system simply make the existing noise more obvious. "

That's what I was really trying to say.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14801
Registered: May-04
.

I knew that.
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3799
Registered: Feb-07
I knew that you knew that ;-)
 

Platinum Member
Username: Nuck

Post Number: 14883
Registered: Dec-04
I am enthralled.

104db speakers make quiet stuff loud. Point taken.

82Db speakers need lots of power to be loud. see above.

Tube amps can make noise at idle, we do not want to make that apparent.

SET amps, quiet ones, love HE speakers, still like the olden days.
 

Gold Member
Username: Stryvn

Wisconsin

Post Number: 1220
Registered: Dec-06
Can it be said that the xo in the less sensitive speaker is eating more of the noise?
 

Gold Member
Username: Dmitchell

Ottawa, Ontario Canada

Post Number: 3800
Registered: Feb-07
Mine makes noise at idle, but not overly so.
 

Gold Member
Username: Magfan

USA

Post Number: 1071
Registered: Oct-07
Even with a kilowatt+ my panels are near silent turned all the way up.......
Panamax on the low power stuff and a dedicated circuit to the amp and a PSAudio Soloist outlet for amp and sub. As predicted here, I had problems with the amp plugged into the power conditioner.
 

Platinum Member
Username: Nuck

Post Number: 14884
Registered: Dec-04
I think so Scott.
The crossover is pretty much a constant at low power=no heat.
The XO eats power at a rate roughly the same as the class of XO.
Sub amps are at 4th order, 12DB slope, with a sharp cutoff, the tradeoff is inefficiency. Add bigger amp.

Your Gallos have no crossover, but a 1st order tweeter with one cap. The L-pad is still first order.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14802
Registered: May-04
.

"Can it be said that the xo in the less sensitive speaker is eating more of the noise?"


Not specifically the Xo. The system is less efficient due to several factors, mostly the sensitivity of the drivers. As Leo and I agreed in another thread, parts count is only a minimal aspect of how much energy is lost in the transduction of energy within a speaker; https://www.ecoustics.com/cgi-bin/bbs/show.pl?tpc=1&post=1853736#POST1853736


Here's another way to think of this; quiescent noise is just another low level signal. The speaker system has no way to differentiate between what is a music signal and what is a noise signal, it merely responds to a voltage input. If you are valuing those low sensitivty speakers because of their low level detail retrieval, then you should also be valuing those low sensisitivty speakers for their ability to reproduce queiscent noise as both repesent just another voltage signal. The lower senstivity of the system simply places the signal farther above the quiescent noise floor by requiring the amp to be run at higher output power while the quiescent noise remains at a fixed level (more or less).


The obvious Catch 22 is the high efficiency speaker system which makes circuit noise more apparent vs the low sensitivity system which places the music further above the noise floor but also can place a strain on the amplifier when complex signals or large dynamic swings are encountered. As you've probably figured out, a highly reactive speaker makes that strain all the more obvious while a less reactive speaker allows the amplifier to operate with its highest efficiency. Therefore, the stability of the load is more valuable than the sensitivity.

The trick is to find that balance point where you have the right sensitivity/efficiency in the speaker to reach what you consider satisfactory levels while maintaining sufficient power reserves in the amplifier. Of course, Alice, that bite of mushroom takes us down another series of rabbit holes fraught with even more trade offs.



.
 

Platinum Member
Username: Nuck

Post Number: 14885
Registered: Dec-04
I had that experience with my last Classe setup, Leo.

Just dead into the air.
Pause the transport and it is dead silent.

Nice.

Particularly fun at parties!

All should have dedicated circuits with the stuff we play with, I think.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 14803
Registered: May-04
.

"Sub amps are at 4th order, 12DB slope, with a sharp cutoff, the tradeoff is inefficiency. Add bigger amp."


Well, first of all, not all sub amps use 4th order filters. But it's not essentially true to label a higher order filter as less efficient, particulary in a sub amp, since the signal is falling away at a set rate anyway. IOW, a filter is only removing the portion of the signal you don't want to begin with. Since the signal is rolling away, the sensitivity of the two drivers or speaker systems becomes more of a determining factor in "efficiency" of the overall system IMO.


"Add a bigger amp"? You could, especially if you believe "watts is cheap". That's not a very valid statement with tubes however, if you want to maintain quality.

.
 

Platinum Member
Username: Nuck

Post Number: 14886
Registered: Dec-04
Plus, 4th order is a 24db slope.
I was referring to my sub amp only.

The sober light of day is nasty sometimes.
« Previous Thread Next Thread »



Main Forums

Today's Posts

Forum Help

Follow Us