EF86

[johnheath]

Hi…

Anyone tried the EF86 in a preamp? Tube Amp Doctor in Germany claims that they are very quiet and noiseless but on other sites around the internet people tend to say that are noisy? I do not know what to think of this and buying them just for a blind test is… a bit costly =)

Thoughts?

Thanks

/John

 

[emrr]

It depends on the preamp firstly as to how noisy one might seem.

 

[johnheath]

Well it is a preamp that I have designed on my own but I read about the ef86 when looking at Hifi amps and thought that it might work in a preamp for recording?

Noisy? That is my question.

 

[emrr]

It has a long history of being used in a lot of different preamps, of many varieties.  What type of preamp are you talking about? 

 

[mjrippe]

Is a Neumann U67 noisy?  Is an EMI REDD 47 noisy?  A well designed circuit with a good, tested tube will be just fine.

 

[emrr]

A pentode connected EF86 as front end without NFB will be noisy.  As would any other pentode in the same circuit. 

 

[ruffrecords]

A pentode connected EF86 as front end without NFB will be noisy.  As would any other pentode in the same circuit.

On the other hand, using an EF86 with a decent high ratio input transformer you can make a reasonably low noise mic pre like the REDD47.

Cheers

Ian

 

[johnheath]

Thank you guys

My thought was to use it as a pentode... haven't thought so much about the NFB but I guess it can be done.

I have no experience of this specific tube but since I saw it in a Hifi preamp stage I thought that it could work well in a micpre.

Ian... what would you call a decent high ratio for a input transformer? I have a lot of respect for your knowledge so if you say it might work I trust it =)

Could you say anything about the EF86 being more or less noisy than ... a 12AU7? as for a first preamp stage?

Mjrippe: I do not know must be my answer... I am asking without knowing about this tube. But like most other things... if it is good, people tend to use it. So in this case  the EF86 was a bit new to me so maybe not so many people us it? What do I know?

Thanks

/John

 

[Conviction]

I have no experience of this specific tube but since I saw it in a Hifi preamp stage I thought that it could work well in a micpre.

Could you say anything about the EF86 being more or less noisy than ... a 12AU7? as for a first preamp stage?

I'd worry more about microphonics than noise. Having that said, I have at least 6 microphone preamps that uses EF86 as V1, where microphonics ain't a substantial problem.

You could try EF806S's, although they're a little more expensive.

As have already been pointed out, the EF86 was invented with microphone preamp stages in mind. It's a mike pre tube. Period.
I'm a little surprised you haven't encountered them earlier. They're very common in older European studio gear.

 

[ruffrecords]

Ian... what would you call a decent high ratio for a input transformer? I have a lot of respect for your knowledge so if you say it might work I trust it =)

Could you say anything about the EF86 being more or less noisy than ... a 12AU7? as for a first preamp stage?
Thanks

/John

The REDD 47 mic preamp uses a 1:7 step up transformer. In my mic pre designs I use a 1:10 step up ratio input transformer.

In a good mic pre design, the overall noise is determined by the noise generated in the first amplifying stage.

Tubes on their own are too inherently noisy to connect direct to a microphone. To improve noise it is normal to use a step up transformer as the first stage. This provides an almost noiseless gain stage so the noise  of the subsequent tube is almost irrelevant.

As a rule, triodes are less noisy than pentodes. Both types produce shot and 1/f noise but pentodes also suffer from partition noise so they tend to be noisier than triodes.

Despite what others have said, negative feedback makes no difference to noise performance.

Cheers

Ian

 

[johnheath]

Conviction... I have not noticed them until yesterdeay but after reading about them I realize that  thay have been widley used... even in famous guitar amps.

I will surely look into them more to investigate their potential.

Ian... thanks for the simle and very useful answer... as usually.

I will mess with a EF86 and a 12AU7 in a micpre to see if I like them.

Best regards

/John

 

[ricardo]

http://www.r-type.org/exhib/aai0113.htm

EF86 was THE audio Low Noise device for several decades.

While its true that triodes can be quieter than pentodes, this isn't always the case.  The circuit has to be optimised for each type of tube.

For the most critical applications, it was usual to operate EF86 as a triode (screen strapped to anode) .. but then you would have to make up the gain some other way.

One caveat is that it is a small signal device so might not be happy driving certain output transformers (eg if used inside a mike) that were meant to be driven by beefy triodes.

To get the potential performance out of all these modes of operation, you need optimise the circuit to fit the application and the tubes.

Here's a tutorial by Great Guru Baxandall on how to do this properly.
http://www.douglas-self.com/ampins/wwarchive/wwarchive.htm#baxpre55
In dem days, mike input transformers were (relatively) even more expensive and difficult to get than today .. otherwise he would have used them.

(Actually he DOES recommend input transformers in
http://www.douglas-self.com/ampins/wwarchive/Baxandall%20preamp%20Feb55%20p2.jpg)

Some of his later articles have tutorials on how to design & wind your own.  If you see these, copy & link to them cos the knowledge of true masters like GG Baxandall will soon be Unobtainium.

 

[johnheath]

Thanks Ricardo - useful and interesting reading =)

I will use a 1:7 ratio input transformer and a SRPP 12AU7 or 5687 as output tube... the EF86 will be the first tube...

Thanks

/John

 

[emrr]

Despite what others have said, negative feedback makes no difference to noise performance.

Sure it does, if you think further into the statement.  What you say is true for input noise, but does not address total system design. 

If you rely on a pentode for all the gain it's capable of, without negative feedback, and compare it to two triodes arriving at same gain, the pentode will frequently be the noisier path.  If those two triodes are triode connected pentodes, they may be quieter than two triodes.  Apples to apples, same total gain figure. 

If you use a pentode in the way that it is used in most high quality designs, using all its gain to give headroom for large quantities of negative feedback with a subsequent stage, and again arrive at the same gain found above, it will probably be the quietest of all because of the negative gain influence on overall noise/etc. 

So....three ways to use that pentode, all giving totally different overall noise result.

 

[johnheath]

Hi Doug

I find every information about this topic interesting and I guess tahat you were answering Ian (Ruffrecord).

But... since you are fairly convinced that a NFB would make the overall noise less than without... can you give me some hints about how to couple this NFB?

I am not familiar with this pentode (as mentioned) but I know how to create a NFB using normal triodes and power output tube such as those of a guitar amp.

Is it a different way to use the NFB on a EF86 or is it coupled on the cathode as "usual"?

Best regards

/John

 

[emrr]

Yes, responding to Ian responding to me. 

I'm not trying to convince you of one way or another to use it, just pointing out how different paths will have different results with the same tube.  NFB is as much a question of what you want the preamp to sound like as it is related to measurements.  If you grasp it's use elsewhere, it's no different here. 

Dig around the tech documents section, there should be a bunch of different EF86 preamp schematics in there. 

You haven't said if you are planning to use it in triode or in pentode yet, that alone defines a lot about where the rest of the design will go.  How much gain?  etc etc etc.  We don't know what you are trying to do at all, so you'll only get general commentary.

 

[ruffrecords]

Despite what others have said, negative feedback makes no difference to noise performance.

Sure it does, if you think further into the statement.  What you say is true for input noise, but does not address total system design. 

I really hate to disagree but I am afraid I have to. For any given topology, applying or not applying negative feedback makes no difference to the obtained signal to noise ratio. It just alters the gain, for both the signal and the noise.

What will affect noise is the topology used, and the examples you quote are different topologies. Whether NFB is used or not in each one of them makes no difference to their noise performance (S/N ratio) but which topology you use will make a difference.

Cheers

Ian

 

[emrr]

I'm not using the word topology, you're not using the word NFB, they each inform the other.

I may be wrong, but experience tells me if you build up the circuits I describe, on average I will be right in the assessment. 

I will never pick a pentode feeding a cathode follower over two low to medium mu triodes for either noise or sound, both having equal gain. 

The OP hasn't stated a purpose or described in any detail the preamp he wants to build, yet we're raising a speculative cloud of dust over nothing. 

 

[ruffrecords]

I will never pick a pentode feeding a cathode follower over two low to medium mu triodes for either noise or sound, both having equal gain. 

Agreed. They are different topologies and therefore have different noise performances.

The OP hasn't stated a purpose or described in any detail the preamp he wants to build, yet we're raising a speculative cloud of dust over nothing.

I think it is important the OP realises that NFB does not improve noise performance. If it did, we would all use tons of it.

Cheers

Ian

 

[emrr]

Well.....here I go again, sorry....we wouldn't use tons of NFB if we didn't want it's sonic characteristics!  As PRR has said many times, frequently air motion on a microphone is the greatest 'noise' in a preamp, swamping all else, in which case you can pick and choose your characteristics.  Yes, he has to define what sort of tube preamp he wants to build, what he wants it to sound like, how he wants it to clip, etc.  Then commentary can be meaningful rather than theoretical.