RH84 and Electron stream question.

[M Gregg]

what has changed? well, the diode will add its imprint as soon as the screen draws current, since the diode has its own current voltage characteristic. and because the screen grid is another control grid for the anode current it will have an influence on the sound.

Which way is conduction? (NB I'm not referring to the diode symbol which is pants under this situation )

I assume you are talking about the VD across the junction..

NB I'm thinking about the Fixed bias idea..

Regards
M. Gregg

 

[M Gregg]

what has changed? well, the diode will add its imprint as soon as the screen draws current, since the diode has its own current voltage characteristic. and because the screen grid is another control grid for the anode current it will have an influence on the sound.

Just thinking about your comment,

Any thoughts about different types of diode..ie UF/soft recovery?
Do you think the diode type will effect the type of sound or just the VD..

Regards
M. Gregg

 

[DF96]

If you use a reverse-biased zener you need to ensure that the screen current is enough to get the zener past the breakdown knee in its characteristics. Otherwise you are just adding more distortion - which will certainly sound different from less distortion. If you use a zener in the forward biased direction then it won't be much different from any other diode.

I really don't understand the interest in making minor bias changes or minor distortion changes. If you hear a small difference how do you know which is more accurate (assuming that is the aim, which it may not be)? If you hear a big difference then you are probably not drawing enough current, and so introducing more distortion - some may prefer this. Forget all the nonsense you may see about a diode passing DC and blocking AC - simply untrue and a sign that whoever says this doesn't understand electronics.

 

[M Gregg]

DF96,

The interest is when you switch between triode / UL/ etc connection and add UL diode and triode diode the U/L diode sounds like the triode selection with more power. If you don't like triode connection your not missing much..

This isn't about the conduction although interesting or the AC/DC aspect..
However I haven't tried different types of diode wondering if the characteristics play a part or just the volt drop. Although I did try UF 4007 and thought the standard 4007 was better but it was slight if any..

Changing the VD using a zener does have an effect but that is because the voltage on the grid is different and you can match it to the triode voltage. Which is what the RH seems to be doing..or should I say trying to do..

Regards
M. Gregg

 

[DF96]

The main effect of a forward biased diode (and a reverse biased zener with enough current) is a shift in bias, combined with a small series resistance. A shift in bias will cause a small change in gain and a small change in distortion characteristics - this may be audible, or may not. The amount of bias shift will depend on the diode - generally the higher the current rating the smaller the bias shift, but it will be in the region of 0.6V. Any change will be well within the range of normal valve characteristics, so you could get a bigger effect by omitting the diode and simply doing 'valve rolling'.

 

[M Gregg]

so you could get a bigger effect by omitting the diode and simply doing 'valve rolling'.

I tried that and as always you do get a change, however its not the same kind of change. If you switch a tube between U/L and triode you can't get that kind of change by tube rolling and its that kind of change that you get with the diode. I am wondering and its only an opinion if the back EMF from the OP Tx could possibly be something to do with it. (on the OP tube)

I have heard similar changes in the past with reversed biased diodes across power tubes.<<just a reflection don't do this!

Regards
M. Gregg

 

[M Gregg]

Still just having fun..

Interesting it is in sort of U/L..


Regards
M. Gregg

 

[M Gregg]

Just some more for a laugh,

Pentode..
I have changed the zener (G2 22V) to a resistor and slowly reduced it bringing Ig2 to spec and measuring the VD across the resistor while listening to the sound change.. and as you would guess it gets louder than the zener and different.
I managed to get down to 300 Ohms. Which I can now switch between ES U/L and pentode. Its just something to try when I get bored..I guess a stiff zener reg supply is on the cards..

Anything for a laugh

But I have an itch to build higher power for a laugh.. So I was looking at the tube cad super charger has anyone tried this?
http://www.tubecad.com/2013/04/blog0259.htm

its probably easier to just build with a bigger tube..

Regards
M. Gregg

 

[M Gregg]

Something else I want to try just for fun,

is the old valve radio idea by dividing the 47K B+ dropper and tap off for a G2 supply..

Regards
M. Gregg

 

[M Gregg]

Something else I want to try just for fun,

is the old valve radio idea by dividing the 47K B+ dropper and tap off for a G2 supply..

Tried this and it works but its rubbish..HF typical pentode problems.

So its back to the resistor pentode set up to give tube current and Ig2 current.

This is how it stands now..

Ignore this its not of interest..
Synergy.

The 300 ohm is a 2 watt Takman MF..but

To those that say you can't voice an amp..
I have tried quite a few different types of resistor in the 300 Ohm position G2. (Has to be measured and set)
And they all sound different the kiwamie sounds a bit compressed and the top end is to controlled.

The Takman MF is very good, however I had to add some wire to the takman MF after cutting it for other experiments..However there was a change in sound linked to the treble. Yes I know it sounds nuts so I changed the lead out wire again and smoothed the treble out..
Its obvious it must be my imagination..but anyway I prefer it.

With the Kiwamie the ES U/L beats pentode hands down. With the Takman in place the pentode beats the E/S U/L.

Years ago the audio DIY got a bit ridiculous when I used to cut leads off carbon film resistors and grind back the ends then silver solder silver leadouts..I don't remember why..however its interesting to note that Audio note seem to do this and even some resistors specify the lead out type it must be all imagination but I could swear there is a difference in the treble anyway.. that's life I guess..Just noticed I didn't change the switch label to U/L pentode.

Its all just for a laugh..
I must get round to a proper project...

Regards
M. Gregg

 

[M Gregg]

Just noticed I had not put the input volume in.
Its just an update I didn't want to repost but..

Its just for fun, the input is 100K log pot (carbon) with magnetic tantalum .5 W 22K and 1M resistors.

I haven't ripped it all to bits yet.. I don't know why.
I tried networks across the primary etc but they just messed up the sound.
I have deliberately left the cap type out..but I have used a Kimber Kap 450V.
I guess others would work and give some play room for type /sound ect.
Kimber kaps have a slight bright edge but the detail is quite good.(each to their own)

Regards
M. Gregg

 

[M Gregg]

Just an update,

looking at the zener in the RH design with no series resistor as in the RH revision 2.
The 1N4752A seems to work well the voltage of the zener is 33V 1Watt so its worth looking at the zener voltage.
The revision 2 RH zener is connected as in post 47#(shown as 22V) but you already know that.

So through this thread is a variety of output connections and driver ideas all with different sounds.
I have used the switch to switch between all the different outputs which has been a good way of comparing each idea.

Regards
M. Gregg

 

[M Gregg]

After a great deal of listening..post 51#

I have decided that the 22K tantalum on the input is overkill so the 22K is now a metal film.
The diodes have been changed just for a trial on the U/L tap to UF just to see if there is any noticeable effect.

Its worth experimenting with the input resistor "type". Perhaps along the lines of Takman MF but even standard types are worth investigating.
The 1meg stays tantalum.

Yes its tedious but that is the nature of components. The Tantalum 22K (input) had an adverse effect on the HF giving a glassy/sharp character and feel.
I couldn't live with it so its out.

Regards
M. Gregg

 

[DAK808]

Hi Greg, kudos for your efforts to determine different sonic attributes to changes in components, etc. I have made 5 or 6 RH amps and my favorite is the "universal" amp so I can try 6l6 types, 6550, kt88, etc. The cost of the output transformers is a lot more than simply using some OPTs from an old console amp, but with the upgrade in OPTs the sound quality is also much better.
Going back to the zener diode on the G2 grid , I also feel that the zener adds a sonic impact and I prefer a dropping resistor instead. I also use another choke for the driver + G2 supply. I feel that the driver supply choke adds more dynamics than the simple voltage ladder supply. Keep up the good work, cheers, Dak

 

[M Gregg]

Hi Greg, kudos for your efforts to determine different sonic attributes to changes in components, etc. I have made 5 or 6 RH amps and my favorite is the "universal" amp so I can try 6l6 types, 6550, kt88, etc. The cost of the output transformers is a lot more than simply using some OPTs from an old console amp, but with the upgrade in OPTs the sound quality is also much better.
cheers, Dak

I modified a SE EL34 to RH universal and tried lots of things in a thread here.
http://www.diyaudio.com/forums/tubes-valves/257466-just-thought.html
Then its got crossed with the PYE Mozart and became a multi switchable ES/UL/Pentode/Triode/ES Triode/ES U/L with the bias set with CCS. It works well and sounds quite good. This went through many component changes/cap changes etc and ended up as a sort of PYE Mozart universal. However the small EL84-EF86 is interesting because its quite compact.

That 22K "type" is bugging me....

Regards
M. Gregg

 

[M Gregg]

LMAO..

Well you guessed it, after fitting the UF 4007 in the U/L Elec Stream the 22K is back as magnetic tantalum..why..why..

Well because I lost detail with the MF..sometimes its just nuts!

I have obtained or should I say waiting for some old "Black Gate" for the el84 cathode bypass just to try for fun. I was going to use Kaisei but then I saw the black gate. So I want to compare them to the silmic 2 I already have in.. just curious its been a while since I used any and I want to know ...

I have seen this so many times now its mad, as you increase clarity in one area its shows up the other problems..this is always the problem with tweaking parts.

What do I mean by clarity<<when you can hear a string vibrate back and forth or the resonance of the metal in a trumpet..the gravel under foot etc..and so the winged serpent flies LOL

Regards
M. Gregg

 

[M Gregg]

Ok,

I have found the stock value of 0.005uF Polypropylene across the primary of the O/P transformer is quite good just direct with no resistor. YMMV.

I have used 630V which has worked OK in the past on SE equipment 1000V is better and a must on PP.

Black Gates are in the cathodes of the OP tubes across the regulators.
Initial impressions are good.

Compared to the Elna silmic at the moment the BG are clearer its as if the bass is more controlled and its the bass control that seems to reveal the details or prevents the masking in the midrange. There is also a little more weight to the sound which is strange because the bass is not so in your face. Perhaps less harsh, but its early days. However its not a massive change.
Its subtle but noticeable.

Regards
M. Gregg

 

[M Gregg]

Reflection on the Black Gate caps,

Compared to the Elna Silmic 2.

The Black gates are not as good!
The Elna has more extended highs and lows period!
I was waiting for this huge break in period but they are typical 70's sounding warm mids with reduced highs and the bass is not as extended.

would I use any more..no! (YMMV)
I guess at the time they would have been a good cap, however if you want to tame the HF of your equipment then...

With regards to the 0.005uF across the output transformer its worth trying it and removing it. It made a difference with the Elna as cathode bypass, but the BG's do the same thing..

So as far as I'm concerned its myth busted..I think things have moved on!

Regards
M. Gregg

 

[M Gregg]

I would say,

its worth playing with different caps in the cathode bypass on the RH.
I will leave the BG's in and listen if I get any changes I'll post.

The BG's are so close to the silmic 2's, however the BG's change over an on period whereas the Silmic have an initial start up period then seem to stay fairly constant. I even thought about taking them back out and putting the silmics back in..perhaps they will open up..

Regards
M. Gregg

 

[M Gregg]

The Back gates are removed,

Because they suck compared to the silmic.
There you go!
The silmic are back in and its like a breath of fresh air.
Myth busted..YMMV..but take the sock's out of your ear's first...

I like Rubycon caps but the Black gate can stay in the past..
The effect is like adding distortion to the music.

Regards
M. Gregg