Tube rolling thread | UltraSonic Studios
Jan 28, 2021 at 1:40 AM Post #16 of 3,643
Now, the datasheet does not have any triode connection curves so it is not very clear for me what grid bias needs.
Your schema seems to be OK, as you can see pin 3 is not used in EL39 and could be used to connect anode cap. in fairly simple EL39/EL34 adapter.
As you can see EL32 is a different beast.

EL38 is almost the same as EL39 just switch pin 1 and 6 (or connect together if p1 is not on the ground on the amplifier to have EL39/EL38 adapter)


About curves, this is mostly the case for many pentode, Tomas will probably get a pair of el39 from DL box of tube even if the concern is for infinity, I'm pretty sure i will be able to extrapolate bias grid voltage for eternity. If not the best would be to measure it once strapped with a curve tracer (Utracer3++ or 6, etracer ...) os ask if someone have it in one of the user forum...
 
Jan 28, 2021 at 1:51 AM Post #17 of 3,643
Your schema seems to be OK, as you can see pin 3 is not used in EL39 and could be used to connect anode cap. in fairly simple EL39/EL34 adapter.
As you can see EL32 is a different beast.

EL38 is almost the same as EL39 just switch pin 1 and 6 (or connect together if p1 is not on the ground on the amplifier to have EL39/EL38 adapter)


About curves, this is mostly the case for many pentode, Tomas will probably get a pair of el39 from DL box of tube even if the concern is for infinity, I'm pretty sure i will be able to extrapolate bias grid voltage for eternity. If not the best would be to measure it once strapped with a curve tracer (Utracer3++ or 6, etracer ...) os ask if someone have it in one of the user forum...

I discussed with Tomas in the context of EL34 in the output socket without adapter and he told me:

If you wanna use EL34 which has g3 on pin one I need to connect pin 1 to pin 8, cathode, which in turn will be grounded.

So Pin 1 is connected to Pin 8 and grounded. But this might be the situation in my amplifier, I do not know how others will be wired but I think a standard would be useful otherwise our advices here might not be accurate for different users of the same model.

For grid bias I found a calculator:
https://www.vtadiy.com/loadline-calculators/power-stage-calculator/

From what I know now (might change, so I am not sure if it is final, input stage will have 180V/4mA and output stage 200V/20mA). If I want to use 6V6 in output stage I set Operating Mode = Triode, Then 200V and 20mA and the Grid Bias voltage is around -13.29V. I'll set it with the Digital Multimeter to -14.5V to be sure current is lower than 20mA then just adjust.
(I never tried that, I just plan how I would do it, so any advice is welcomed)

GridBiasCalculator.jpg
 
Jan 28, 2021 at 4:40 AM Post #18 of 3,643
I think that those calculation are made with spice model of selected tube.. (I suggest to add it on post 3 of the tread)
About strapping a Pentode as Triode, the wiring is the one on page 23 of this reference presentation.
strap.png
pin for EL34WirringTriode
1 + 8K + G3 (it is even done in some tube) K (cathode)
5G1G (grid)
3 + 4 (with resistor)A + G2 using a 100 Ohm ResistorA (anode)
2 - 7heating heating
@SonicTrance
Pin 6 is not used in EL34 but is used as g3 on some tube. Does wiring to pin 1 or 8 make any sense ?
 
Jan 28, 2021 at 4:59 AM Post #19 of 3,643
I will add it after at least one of us will try it and test it. I know about 2 ways of determining the Grid Voltage (this one above and one Tomas showed me, with the Triode Curves).
In this method, we obtain a value of -13.29V for our Grid bias voltage for Vplate = 200V and Aplate = 20mA (as we prezume we will have in the amplifier on output stage).
But using the datasheet and triode curves from them we deduce a value close below -15V (by intersecting the 200V with 20mA we are close to -15V for grid bias).

6v6datasheet.png

Since we have more than 1.5V difference between the calculator and a datasheet, I think we need to decide what tolerance we have, how it is the best way to do and later decide which values are appropriate to our circuit. So I would say we need to test this in the first amplifier when it is build and see the real situation.
More than this, since tube rolling is a hobby for me, I am not every time carefully watching these graphs so I make stupid things like using the pentode connection graphs or even intersecting wrong.

How I see it:
- we can use the calculator
- we can use graphs
- we can use both and choose medium value
- we can put a table on post 3 with tube type, Vplate, Aplate and everyone based on its build will know what to choose
- we can put all three ways there and everyone has to do however he/she wants

Because of these things I have not, yet, put anything there. Another reason is that I am not sure about operating points on the final build.
 

Attachments

  • 6v6datasheet.png
    6v6datasheet.png
    73.4 KB · Views: 0
Jan 28, 2021 at 5:18 AM Post #20 of 3,643
@SonicTrance
Pin 6 is not used in EL34 but is used as g3 on some tube. Does wiring to pin 1 or 8 make any sense ?
It can make sense. But, I usually stay away from wiring unused pins inside the amp. This is because some tubes have internal connections on unused tube pins that can cause issues.
I can do some more digging on this.
 
Jan 28, 2021 at 5:28 AM Post #21 of 3,643
But using the datasheet and triode curves from them we deduce a value close below -15V (by intersecting the 200V with 20mA we are close to -15V for grid bias).

6v6datasheet.png
This is correct. I'd always look at the triode curves as first option. Granted these were drawn by hand and is not super accurate all the time but it's the best we get apart from a curve tracer.
The actual load line will be flat like that and reach all the way to about -28Vg. So the signal can swing from 0Vg to -28Vg. So, it will swing from 60 - 340V on the B+. Eternity will have a B+ of around 360-370V so there's headroom for that kind of swing. B+ in Infinity is 410V.
 
Jan 28, 2021 at 5:38 AM Post #22 of 3,643
Then we can make a procedure like:
- search for Triode curves IF available -> use that value
- if datasheets are not containing graphs (e.g. EL39) or triode connection ones check the calculator, then add 2V to be sure
- use the table as a guide, for what has already been tested on the thread by someone (keeping in mind to carefully choose your operation points)

We can also make this everyone by itself, but then we loose time to calculate something already tested. The power of this thread are the users. I let you guys choose, then I'll just update the 3rd post with what you will consider the best way. I have to admit I do not have enought experience for this, just learning.
 
Jan 29, 2021 at 7:52 AM Post #23 of 3,643
Regarding EL32, I have 2 quads: Marconi and Mullard. The problem is that the Mullard ones have no logo just:

MullardEL32.jpg


And on the back something like "1005-I IMXA". It is not clear to me, what are the date codes for these?

el32.jpeg


PS. I edited post 3 with stuff regarding how to determine grid bias voltage. If something is wrong please let me know to fix it.
 
Last edited:
Jan 29, 2021 at 10:44 AM Post #25 of 3,643
Interesting. I took a look here. I think I know this one, but from what I knew these were sold unassembled and you had to built it for yourself. A kind of IKEA of tube amps. LOL
 
Jan 29, 2021 at 10:53 AM Post #26 of 3,643
That's the one but I bought a used mint unit from someone cheap and he even throw in some current production KT88 / 6L6GC and 12AU7 tubes. I'm intrigued in this amp because it will give me a taste of these power tubes while waiting to see how your and DL amps will turn out. :)

In the process, I will also be buying some additional KT66, KT77, 6550, 807 tubes.... down the rabbit hole I go again....
 
Jan 29, 2021 at 11:17 AM Post #27 of 3,643
Before taking the decision to write Tomas and ask him if he can build me a headphone only amplifier based on 6V6 as outputs and, if possible, 6SN7 as input, because I wanted to use a part of my tube collection on it and see where it goes, I searched several months for tube amplifiers based on these. I looked a lot at WBA Virtus-01, ALO Studio Six, Cayin HA-6A or several Prima Luna models. It seems that even if 6V6 are mostly used in speaker amps, everyone who tried them on headphones had nice words for them. In the end I decided to try a full custom build and Eternity took form. We will see where this will go, I am fully convinced it will be interesting for all of us.
 
Jan 29, 2021 at 11:22 AM Post #28 of 3,643
Well I have owned Studio Six for a year and I know how it sound. I prefer Studio Six over my Glenn OTL amp. Studio Six is a very well build product. Quality of construction is first class. I've several pairs of NOS 6v6gt and 6v6 tubes. Studio Six also use 6sn7 as driver and a GZ34 as rectifier but I did use GEC U52 and Cossor 53ku in it too. Not sure why I sold off Studio Six but it's all part of my audio journey. I move along.
 
Jan 29, 2021 at 3:01 PM Post #30 of 3,643
Lots of interesting gear indeed. Interesting that he talk about Bifrost 2, which I just bought to accompany my Yggdrasil. That Schiit Freya + is something I'm interested in too purely because I've also Mjolnir 2, so a few Schiit gear and it should pair well together.... not that it has to be all Schiit.... pun unintended.

That LTA Ultralinear amp is very interesting.... similar to SE triode amp with negative feedback. That's what he said.
 

Users who are viewing this thread

Back
Top