Author Topic: Triode-connected EF86  (Read 2802 times)

Offline handsome

Re: Triode-connected EF86
« Reply #15 on: July 11, 2019, 02:14:32 PM »
a good way to eyeball linearity besides the ubiquitous plate curves is to look at the graphs (if offered) of mu versus grid voltage. This basically shows changes of gain versus signal level (far quicker than drawing a loadline which shows the same thing) - in other words does your device amplify more or less depending on signal level? We always assume mu to be a constant. Very linear devices show a pleasingly straight line where they are most linear and thus where they should be operated. Change of gain over signal level is essentially distortion. See graphs of the 6SN7...

Offline El Sid

Re: Triode-connected EF86
« Reply #16 on: July 11, 2019, 05:06:13 PM »
a good way to eyeball linearity besides the ubiquitous plate curves is to look at the graphs (if offered) of mu versus grid voltage. This basically shows changes of gain versus signal level (far quicker than drawing a loadline which shows the same thing) - in other words does your device amplify more or less depending on signal level? We always assume mu to be a constant. Very linear devices show a pleasingly straight line where they are most linear and thus where they should be operated. Change of gain over signal level is essentially distortion. See graphs of the 6SN7...

Ding! A little light just came on...

Offline Ampdog

Re: Triode-connected EF86
« Reply #17 on: July 12, 2019, 05:42:54 AM »
Reluctantly joining, as this can get as diverse as cab-[oops - sorry!],

Several things to keep in mind:

The use of the EF86 as triode-connected input stage of a power amplifier, often stemmed from the conversion of an existing EF86 pentode fed power amplifier to a lower gain instrument; merely the simplest way.

Also, in the application of a particular valve in a line of equipment: Economic factors, not necessarily spectacular performance - in a economy driven world, the latter is often not the prime objective!

And then one may ask: With so many dozens of contenders in any category, why would valve manufacturers not have been able to make optimal examples (triodes) instead of users having to resort to off-application alternatives (triode-connected pentodes)?

Etc. - one can have a long discussion on this (as is in fact often the case on other forums). Often the rule of "horses-for-courses" emerges, and as said the quietly ubiquitous economical factor.

Not even daring to start a discussion, it must also be pointed out that available data, iwhen not scarce, is misinterpreted and vague. (Take many characteristic graphs: E.g. the universal RCA ones are small and not nearly accurate, having probably been hand-drawn.)

Further, in audio work one gets impressed by and Gm, often not noticing that figures are given for ten times or higher anode current than is used in audio small signal applications. Where more complete Data Sheet folders (e.g. Mullard/Philips) are/were available giving variation of characteristics with anode current, one might be in for an unpleasant surprise at the quite different data at say 0,5 - 1,5mA anode current (typical voltage amplifier condition) than perhaps the data sheet 5 - 20mA figures!

Having done what I indicated I did not want to engage in  (:( ), let me point to just one rather forgotten valve very suitable for audio work because of low noise and at a time low cost, the 6GK5. Quoting a relatively high of 70 and similar high Gm of 15mS (noise is a.o. inversely proportional to Gm) depending as said, it is one of the quietest triodes for audio work around - not to be ignored. The 6GK5 is a frame-grid valve - frame-grid valves make for superior stable and consistent performance compared to ordinary valves.)

But there I go again ......  as said, many options; choice at present probably as dependant on price and availability as anything else.

and having said that I reluctantly join in  :), to me there will always be the question: Why would one want to prefer a triode input stage in a power amplifier, with its Miller capacitance-to-unknown-input-driving-conditions being an unknown when designing for stable NFB. Apology if unnecessary technical, but I have been 'bitten' by that factor - so!

And all of the above not necessarily condemning other preferences, just my conclusions after a relatively long association with the orange shining bottles - and not denying some personal bias  :) :)  ....

(Closing thoughts regarding pentodes as inputs: They need not necessarily provide uncomfortable high gain; depends on the design.  Low anode load resistance, high current, high Gm, resulting low noise?  Remember the famous pre-amp of my esteemed colleague, the late John Fuller??  :)  :))
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline handsome

Re: Triode-connected EF86
« Reply #18 on: July 13, 2019, 11:04:58 AM »
Indeed there are but many devices out there and all are interesting....but good engineering requires that devices be fit-for-purpose. Choosing an appropriate device is actually very simple: first determine how much gain you need, then determine what kind of load (resistance and capacitance) your device will be driving. The first requirement brackets the mu of your device, the second it's current capability and plate resistance. As Ampdog points out once you have chosen a device then consult its characteristic curves and be sure it is indeed capable of it's claimed parameters (and within it's dissipation rating) when used in your circuit.

Offline Ampdog

Re: Triode-connected EF86
« Reply #19 on: July 14, 2019, 04:45:50 AM »
As Handsome says . . . .
         and then one gets the unpleasant surprise from Morgan Jones that e.g. an ECC82 (often considered as a miniature 6SN7) is not a very linear valve  :'(  (Under correction I understand that the ECC82 was primarily developed as an r.f. valve.)

Someone possessing the Morgan Jones book(s) might come in here with his (MJ's) views on triode-connected pentodes; hopefully he said something about that?
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline fredeb

Re: Triode-connected EF86
« Reply #20 on: July 19, 2019, 07:38:01 PM »

Pentodes connected as Triodes by Tom Schlangen :

http://www.audiodesignguide.com/New2A3/ETF06TS.pdf
...evolution is the gradual development and stratification of progressive series of wholes, stretching from the inorganic beginnings to the highest level of spiritual creation.
Jan Smuts

Online vinyljan

Re: Triode-connected EF86
« Reply #21 on: July 21, 2019, 08:40:09 PM »
Pentodes connected as Triodes by Tom Schlangen :

http://www.audiodesignguide.com/New2A3/ETF06TS.pdf

Interesting reading. 



But is operation not unstable when Va is LESS  than Vg2   (=500V for tetrode) :thinking:
In a maximum bank turn in a normal category airplane, like a typical Cessna, everything is fine until you turn across your own wake

Offline fredeb

Re: Triode-connected EF86
« Reply #22 on: July 21, 2019, 08:58:54 PM »
Interesting reading. 



But is operation not unstable when Va is LESS  than Vg2   (=500V for tetrode) :thinking:

There is a typo in the text there , because both ways he's saying the same thing , it is meant to be :

Unstable operation when Vg2 > Va
Stable operation when Va > Vg2
...evolution is the gradual development and stratification of progressive series of wholes, stretching from the inorganic beginnings to the highest level of spiritual creation.
Jan Smuts

Online vinyljan

Re: Triode-connected EF86
« Reply #23 on: July 21, 2019, 09:23:22 PM »
Thx.  Looking at a few schematics after I read it I actually understand the necessity for some connections, layout etc better :thumbs:
In a maximum bank turn in a normal category airplane, like a typical Cessna, everything is fine until you turn across your own wake

Offline Ampdog

Re: Triode-connected EF86
« Reply #24 on: July 24, 2019, 09:32:49 AM »
Needs to be said: EF86 was one of the first valves designed to particularly avoid e.g. hum introduced by the heater circuit - spiral twisted heater etc. Also other internal shields fitted; a look 'inside' with a magnifier will be educational.

[Also (O.P. kindly pardon O.T.!), the American 6AU6A also comes into the picture with a moderately high Gm - not specifically designed for audio though; no particular low-noise measures. I used a number of them during my early years.]

and still O.T.: A powerful double triode to be kept in mind is the E182CC.  It has a high Pa = 4W/triode and high Gm. Only problem: Because it was primarily designed for computer work (Va full on-off) the spread between valves is rather high. But a low Va swing (to within 30V of zero) makes for a formidable driver. (I use it in my 100W amplifier. At 500V+ and 8 mA standing anode current, it can produce a wide anode swing of up to 400Vpp. with Rl = 33K.)

But I digress; back to EF86!
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)