diyAudio (
-   Tubes / Valves (
-   -   Input tube recommendations... (

SpreadSpectrum 13th January 2009 04:29 AM

Input tube recommendations...
I'm looking for tube recommendations for the first gain stage in an amp.

Here's how it will be used:
1. DMOS cascode plate load
2. LED bias
3. Input = 6V pk-pk
4. Outside of GNFB loop

Here's what I want:
1. Reasonable price (I consider more than $30 starting to get pricey)
2. Available, preferably in production
3. Very low distortion
4. Lowest noise that I can get

I have been using 6SN7 in this position and was wondering if there is a tube with similarly low distortion, better noise performance, and maybe slightly higher gain.

Any suggestions, or should I just stick with the 6SN7?

dsavitsk 13th January 2009 05:06 AM

what does the next stage look like?

Without knowing, perhaps 5687, 6N6p, 6H30, 6DJ8, 6N1p

SpreadSpectrum 14th January 2009 12:27 AM

The following stage will be a source follower with about 5.5pF input capacitance.

The last four were tubes I was considering. The 5687 is new to me but looks quite interesting.

Any idea which of these has the lowest distortion/low noise?

SY 14th January 2009 12:37 AM

Triode connected D3a. With a CCS load and LED bias, the large signal linearity is astonishing. Superbly low noise. Not in production, but pretty easy to get for about $15.

Eli Duttman 14th January 2009 04:16 AM


The last four were tubes I was considering. The 5687 is new to me but looks quite interesting.

If you are looking at the 5687 and 6Н30П, you should also look at JJ's ECC99. All 3 fall into the high gm/low RP "super triode" category. The ECC99 is quieter than the 5687 and it's being produced currently.

SpreadSpectrum 15th January 2009 03:56 AM


Triode connected D3a.
Looks like a neat tube. Looks like u is a little high, so I'd have to do something to decrease sensitivity.

Quote: should also look at JJ's ECC99.
Thanks for the suggestion. I hadn't looked at that one yet.

I'm actually leaning toward the ECC99 or 5687 at this point, since they look to be a good fit for my requirements. 5687 has the advantage of being on Morgan Jones's distortion chart and scoring fairly well. I wish the ECC99 had been tested.

Has anyone tested ECC99(or any of the other tubes mentioned in this thread) in a similar(to Jones's) configuration? If not, I'll probably just end up buying some of both and doing my own tests.

SY 15th January 2009 12:25 PM

My tests of the D3a were in a similar configuration; there's a story behind that... I had done some 6SN7 measurements with LED bias and a solid state CCS load. I got results very similar to Morgan's, but wanted to try his specific test setup, which used a D3a as an active load. Morgan kindly gave me a couple of D3a to try. I asked him if he had done the same sort of distortion measurement on the D3a, but he hadn't. Just out of curiosity, I set up the D3a for a distortion measurement and thought I made a mistake. I rechecked the jig, looked at the distortion with two different methods and got the same astonishing result. I immediately wrote him that he had put the wrong tube on the bottom! The D3a (in triode) absolutely blew away the 6SN7 for distortion at high swings.

For the record, at 18dBU, 2nd was -70dB, 3rd was lower than -95dB (my measurement limit at the time), nothing else measurable. At 25VRMS (~30dBU), 2nd was -58dB, 3rd and higher were unmeasurable.

Looneytunes 15th January 2009 12:53 PM


Originally posted by SY
Triode connected D3a. ...the large signal linearity is astonishing. Superbly low noise. Not in production, but pretty easy to get for about $15.


Is there a particular brand that you can recommend?


SY 15th January 2009 12:55 PM

I've only tested Siemens.

dsavitsk 15th January 2009 07:41 PM


Originally posted by SY
I've only tested Siemens.
That's about all I've seen easily available. This tube can be a bit all over the place, so buy a few more than you need to find some with similar specs.

Also, vibration dampening is probably useful as these tend to ting when tapped (and I don't think it is oscillation as if it is then these will oscillate in a circuit where 6688's and 7788's won't).

All times are GMT. The time now is 05:27 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2