I have been experimenting with some SET Audio Amps and a couple of guitar amps that use EF86s and noticed that these tubes all seem to be inherently microphonic.
I have rolled NOS Phillips, Telefunken, Mullard EF86/6267 as well as new issue JJs, Mesa, Sovtek and others and they all seem to be microphonic to some degree...a problem that doesn't seem too acute using my little audio SET Amps but a real pain in the #** in applications like guitar amps that rely on the EF86 like the Doctor Z... tube dampers help somewhat, but I'm interested to know if there is something about the actual physical construction of this tube that makes this tube more susceptible to microphony?
Thks Leo
I have rolled NOS Phillips, Telefunken, Mullard EF86/6267 as well as new issue JJs, Mesa, Sovtek and others and they all seem to be microphonic to some degree...a problem that doesn't seem too acute using my little audio SET Amps but a real pain in the #** in applications like guitar amps that rely on the EF86 like the Doctor Z... tube dampers help somewhat, but I'm interested to know if there is something about the actual physical construction of this tube that makes this tube more susceptible to microphony?
Thks Leo
Aren't all tubes microphonic to some degree?
For what it's worth, the 4 April 1956 Philips EF86 datasheet suggests that Philips did pay attention to microphony when designing the EF86, as you would expect for an audio preamplifier tube:
"A sensitivity of 0,5 mV for an output of 50 mW (resp. 5 mV for 5 W output) 1s permissible in those equipments where an output of 50 mW in the loudspeaker does not produce an average acceleration on the tube greater than 0,015 g at any frequency higher than 500 c/s and greater than 0,06 g at any frequency lower than 500 c/s"
For what it's worth, the 4 April 1956 Philips EF86 datasheet suggests that Philips did pay attention to microphony when designing the EF86, as you would expect for an audio preamplifier tube:
"A sensitivity of 0,5 mV for an output of 50 mW (resp. 5 mV for 5 W output) 1s permissible in those equipments where an output of 50 mW in the loudspeaker does not produce an average acceleration on the tube greater than 0,015 g at any frequency higher than 500 c/s and greater than 0,06 g at any frequency lower than 500 c/s"
All components are inherently microphonic. But pentodes are particularly bad because they have more spindly grids to wobble about. The EF86 has an unfairly bad reputation because people try to wring too much gain out of it in low-level preamps. It's simply not a good choice for a guitar amp front end; much more suited to lower-gain or less 'abusive' conditions!
https://www.youtube.com/watch?v=z7yHtoxkdb8
https://www.youtube.com/watch?v=z7yHtoxkdb8
Judging by the experience of those who use them, yes, they are microphonic.
Forget them in high gain monsters, as Merlin states, but Service Techs repairing (usually old, British) regular gain amps, have big problems finding usable replacements.
We are talking amps which worked properly with them since the late 50's, and obviously nobody complained way back then.
Unfortunately going NOS is not the solution, because all the "good ones" have already been used, by definition what remains is the rejects.
Damper rings help, but more important is to shock mount the sockets; making them "float" on a couple rubber grommets.
Forget them in high gain monsters, as Merlin states, but Service Techs repairing (usually old, British) regular gain amps, have big problems finding usable replacements.
We are talking amps which worked properly with them since the late 50's, and obviously nobody complained way back then.
Unfortunately going NOS is not the solution, because all the "good ones" have already been used, by definition what remains is the rejects.
Damper rings help, but more important is to shock mount the sockets; making them "float" on a couple rubber grommets.
As Mr. Fahey said, you adapt to the lie of the land, as you find it.
Don't use a ceramic socket. Mica loaded phenolic conducts fewer vibrations.
Perhaps a damper more capable than silicone rubber "O" rings will help. Herbie's Halos have gotten "good press".
There are 2 good EF86 variants currently being produced by New Sensor: the EH EF86 and the "TungSol" EF806SG. The 50% less costly EH would be my choice, in a guitar amp.
Don't use a ceramic socket. Mica loaded phenolic conducts fewer vibrations.
Perhaps a damper more capable than silicone rubber "O" rings will help. Herbie's Halos have gotten "good press".
There are 2 good EF86 variants currently being produced by New Sensor: the EH EF86 and the "TungSol" EF806SG. The 50% less costly EH would be my choice, in a guitar amp.
Ok. Let's say they are to differing degrees. But it seems that the EF86 was particularly prone to this judging by anecdotal evidence of others and my experiments with the New Stock I've used to compare the problem to the NOS and used EF86s I have on hand.
Perhaps that's where the real problem exists, when the tube is being used to wring too much gain out of a circuit...ie: in guitar amps like the Doctor Z...
But it's interesting that even the New Stock will ring loudly..as loudly as some of my 50 yr old examples...which tells me...at least...that the tube structure seems more susceptible to this symptom than many other tubes...
Which begs the question: Why do boutique amps like the Doctor Z and others seem to insist on using the EF86? Seems like a guitar amp environment with high SPL would only exacerbate its highly microphonic tendencies.
Are EF86 Tubes inherently microphonic?
Only when connected to electricity..
It was one of the design parameters..some are even closer to spec and rattle while in use..
Regards
M. Gregg
OK, as you and Eli point out the type of socket used and implementing some form of vibration resistant shock mounting will help alleviate this problem...are there suitable tubes one could use to substitute for the EF86 with a re-wire of some sort?
I have been repairing valve equipment for over 50 years and in that time I have come to hate the EF86! They seemed to crop up a lot in the old reel to reel tape decks where they gave problems even mounted on a rubber encased holder. It would be my last choice in any audio amplifier, perhaps that sounds cruel and unfair but experience tells me otherwise. I class it in the same league as Selenium Rectifiers, British 'Hunts' capacitors oooh maybe I am pushing it a bit there, no not as bad as Hunts capacitors :-D
Given your experience Eli, is there another tube, less pricey, less prone to microphonics that would be an overall better choice than the EF86?
And it seems that this malady was acknowledged a long ways back, which gets back to my question...why didn't the tube designers of the period come up with something better?
tapping the 6l6 makes louder noise than preamp tubes. interesting that only one piece from pair is so sensitive
but all tubes will ring ifhard
I haven't found that to be the case...yes any tube particularly warmed up for some time will exhibit microphonics if tapped hard enough...but the EF86 IMO whether used, NOS, or New, exhibits audible microphonics even with light tapping....I've now run through 40 tubes, my entire stock of EF86s (new, used, NOS) and they are equally microphonic and become more so after extended run time at load voltages.
Last edited:
The EF86 is typically run with a voltage gain close to 200. Any tube run with that much gain is bound to ring like a bell! Any 'better' tube you find is probably only apparently 'better' because it ends up running with less gain, so its microphonics (which in actuality will be similar) aren't amplified as much.
huh - I'm a pretty big EF86 user and love the tube. Haven't noticed any microphonic issues - at least compared to oh, the C3m or the 71A, both which were horrendous.
Of course I've never used the EF86 in a guitar amp, but I've used 'em for hi-fi: SE amp driver, connected as triode linestage, and the front-end tube of my Eico amps.
Another option is to use the rimlock EF40 - same tube, just a different base (rimlock socket). With old style sockets, it "locks" into place with a centering nub.
Of course I've never used the EF86 in a guitar amp, but I've used 'em for hi-fi: SE amp driver, connected as triode linestage, and the front-end tube of my Eico amps.
Another option is to use the rimlock EF40 - same tube, just a different base (rimlock socket). With old style sockets, it "locks" into place with a centering nub.
In spite of the negative remarks in this thread, the EF86 is highly regarded by quite a few people. It's linear (better than many a triode), has good "tone", and is quiet. AAMOF, it's almost quiet enough to use in the 1st gain block of a phono preamp.
It would seem that the trouble comes from trying to "put 10 lbs. of sh*t into a 2 lb. bag". That always ends in a very messy floor and a big stink. Hold the stage gain at or less than 100X and the EF86 is a winner.
It would seem that the trouble comes from trying to "put 10 lbs. of sh*t into a 2 lb. bag". That always ends in a very messy floor and a big stink. Hold the stage gain at or less than 100X and the EF86 is a winner.
I'm with Kstagger. I have an EF86 amplifying the reverb spring return (25 mv) in front of a 15" woofer in my Hammond H182 organ. No problem. The woofer is driven to 30 watts by PP 7591's down to 60 hz (16' C) and with new B+ capacitors will shake a 12AX7 and a 7199 loose in their sockets. The EF86 does shake loose in its socket and crackle sometimes. But the 1967 UK manufactured EF86 is not picking up any music. Changing those sockets with a hammond grounding ring around it is not simple. I destroyed the grounding ring taking it off, and the project stopped there until I make another. Nobody sells them IMHO.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.