Thanks Valery,
I may just order the proper device and wait. I was hoping to get them working before the end of the week because I leave on vacation on Sunday and will be away for a week. Probably better to just use the proper part than stacking 2 or 3 zeners to get me to 62V. I always get impatient when I have been waiting a wile for my boards to arrive.
Blessings, Terry
I may just order the proper device and wait. I was hoping to get them working before the end of the week because I leave on vacation on Sunday and will be away for a week. Probably better to just use the proper part than stacking 2 or 3 zeners to get me to 62V. I always get impatient when I have been waiting a wile for my boards to arrive.
Blessings, Terry
Oh yes, I know what you mean (impatience) 😛
Well, stacking will do as a workaround, but the proper ones are always better, so... one week is not that much 😉
Greetings,
Out of curiosity:
a). Have you considered a tube SRPP stage at the input instead ?
b). or better yet, a Differential SRPP stage ?
c). If not the above, then maybe a Current Source on the Depleted mode N Mosfet DN2540, or in case of differential - a pair of those ?
This is a fairly easy trick - you can build a CC Source with the DN2540 and a single resistor. A pair of those on top of those 12AX7 could do some extra wonders ...
d). If you do not like the DN2540, then maybe a regular pentode pair ? Pentodes are WONDERFUL in terms of CC Sources ...
e). A direct coupled differential stage consisting of a set of two tubes - very similar to that as in the good old Williamson Clone input stage ?
Just curious, if you investigated any of those routes.
Cheers,
Ziggy.
two series connected 30V 400mW Zeners give you a 60V 800mW equivalent.
If you select a pair of higher voltage rather than exactly 30V, you can get closer to the 62V specified. eg. 30.6V and 30.5V give 61.1V
This voltage depends very much on the current passing through the Zener. If you don't measure your Zener at the same current as in the circuit, then you won't get the same voltage.
Equally if you change the installed Zener current slightly you can use this to trim the final voltage.
BTW, the Zener voltage drifts a lot with temperature. Give them time to stabilise to the final temperature.
If you select a pair of higher voltage rather than exactly 30V, you can get closer to the 62V specified. eg. 30.6V and 30.5V give 61.1V
This voltage depends very much on the current passing through the Zener. If you don't measure your Zener at the same current as in the circuit, then you won't get the same voltage.
Equally if you change the installed Zener current slightly you can use this to trim the final voltage.
BTW, the Zener voltage drifts a lot with temperature. Give them time to stabilise to the final temperature.
Hi Ziggy - not yet to all of the questions 🙄, however I have already thought about some of these configurations. Differential SRPP looks very promising to me.
I have already obtained a small transformer giving me 6.3V and 160V AC at the same time - good for experiments with tube front-ends at their "habitual" supply voltages (this one will give me around 220V DC).
Thank you for some ideas - I like the sound qualities of the current design with the tube, so I will definitely work further in this direction.
I am thinking about a fully tube-based front-end, combined with solid state (possibly mosfet) power section, utilizing the benefits of both worlds (voltage amplification with tubes and current amplification with SS).
Cheers,
Valery
Why do you use transistors Q6 and Q10 instead of a second set of direct coupled tubes ? The "more tubes" ... the "cleaner sound" .. No ?
Instead of resistors at the tops of the anodes - did you try CC Source and Mirror of some kind ?
Ziggy.
Already trying to simulate exactly the design you mention - tube triodes in the second LTP and current mirror load at the first one 🙂
Member
Joined 2009
Paid Member
Maybe you would be interested in some ideas I posted here for a hybrid with tube front end. Unfortunately it wasn't built as Carlos is too slow 😀
http://www.diyaudio.com/forums/solid-state/194268-dx-hybrid-amplifier-will-produced-soon.html
Careful guys, keep in mind this is the solid state forum. Don't be sneaking in a fox in a hen costume 🙂
Member
Joined 2009
Paid Member
Member
Joined 2009
Paid Member
I think my favourite is the schematic in post #5 of that thread, maybe it would benefit from your input to further improve it. I don't have the time to build this one myself and would be happy to see somebody else pick it up and do something with it. I like the idea of hybrids and hope to have a go at making something myself, perhaps later in the year (I really don't need any more amplifiers !).
Last edited:
Member
Joined 2009
Paid Member
Interesting addition - I have never used a current mirror before, mostly because I didn't want to push the OLG too high but it might be a good thing here where tubes have relatively low transconductance. Some may prefer the sound without the mirror, some with.
12au7 is a rather low transconductance one, so we'll see... another advantage - current mirror improves linearity...
Let me fantasize a bit ... based on the good old dinosaur, the good old Williamson, the input stage of which (L1a and L1b) is "direct DC coupled.
... if you would "cut-copy-and-two-times-paste"
as from :
wzmacniacz przeciwsobny na EL34
the input tubes: L1a and L1b ...
... into L1a, L1a_differential, L1b, L1b_differential,
it could turn out to be a nice differential, two stage, DC coupled amplifier, on the output of which you can choose to use a set of 4 signals with DC offsets at 1/3 and 2/3 of the rail voltage ... obviously some tweaking of the resistors to accommodate for dual rails and CC sink ....
Instead of R3, I would replicate them into R3 and R3_differential,
Instead of R5, I would replicate them into R5 and R5_differential,
better yet, you could throw them out completely and substitute them with respective depleted DN2540, with a single current programming restistor each, as a means of "dynamic resistor/ CC source". This would greatly linearize the tube stages and even them out a bit.
Instead of R2, make it an R2 and R2_differential, into a joint CC Sink (to the minus rail). Program the current as per preference.
Instead of R4, make it an R4 and R4_differential, into a joint CC Sink (to
minus rail). Program the current as per preference.
The rest would be those cascode transistors and voltage shifting, as you did earlier ...
Mniam, Mniam ....
Member
Joined 2009
Paid Member
This is why I specified a 6DJ8 (a have a couple of the russian equivalent) but my goal was not high OLG, rather to leave the tubes with an opportunity to be heard. High OLG will linearize everything in it's path until it simply doesn't matter if you use tubes, BJT or MOSFETs - they all start behaving more like ideal linear elements.
So "ideal linear silicon" sounds exactly the same "ideal linear tube" ?
{{ Ok, ok, ok, ... I did not even ask. I have this "bias" thing. I am "all tubes" straight out of the DAC chip }}
- Status
- Not open for further replies.