Welcome! Need support, you got it. Or share youe ideas and experiences.
Return to K&K Audio / Lundahl Transformers
75.190.166.26
In Reply to: RE: New schematics posted by KevinC on September 11, 2014 at 08:28:51
Hello Kevin:
Curious about the 2-stage PP amplifier schematic and the headphone amp. I have a 3-stage PP amplifier with interstages that I am going to be modding. I have a desire to eliminate the input phase splitter xfmr, as it doesn't have ideal performance. Going fully differential with a CCS to a negative supply is likely my choice for the first stage, similar to your headphone amp.
Removing the first-second stage IT is another possible goal if I can gain a performance increase. This leaves only one IT for the driver to the 3rd stage 300B's. Your headphone amp is a good option to do this. How stable are the DC operating points over the course of a month? Is this a set-it-and-forget-it design, or do you find that periodic tweaking is necessary? I really like the idea of DC coupling the stages to avoid having a capacitor.
On the other side of the thought process, simply forcing the first stage to be fully differential (as opposed to the Lynn Olson design) could allow the input transformer to remain, with its benefits of shielding and galvanic isolation. As long as I don't require phase balance performance from the input xfmr, it performs well. In other words, it does well transforming differential signals with low distortion, but it fails to provide accurate gain and phase relationship between the two secondaries.
Your thoughts ? Currently using LL1692A/PP (2:1.75) for IT, and LL1620AM/PP for OPT. Input transformers are selectable between a Tribute and Cinemag. Thanks!!
Follow Ups:
Are you feeding your amps a balanced or unbalanced signal? Feeding a balanced signal obviates the need for an input transformer. I typically install an input transformer and connect the RCA jack to the primary and the XLR jack to the secondary. The secondaries then act like a CT grid choke for balanced inputs.
Kevin Carter
K&K Audio
www.kandkaudio.com
Feeding amps with balanced output, actually from a LL1689AM/PP preamp. So yes, considering removal of the input transformer entirely (or having provisions to insert or bypass, depending on mood).
If going fully differential with a CCS in the tail, I wouldn't be concerned about balance in the first stage. But then the question is do I IT to the second stage or DC couple. Hence my question about bias stability.
I have built several PP amps where I used direct coupling between the first and second stage. The problem that I encountered was not bias instability, but the consequences of imperfect first stage tube matching. If the second stage tube had low grid bias voltage, a small plate voltage offset in the first stage due to mismatch in the input stage tubes "ate" a fair amount of the voltage swing available to drive the second stage to symmetrical clipping. It's not that I envisioned clipping the second stage in normal use, as that would be bad design, but simply want to have plenty of "headroom" in this transaction. The least fussy solution I found was to use a driver (second) stage tube that had a high enough grid bias voltage that a few volts of plate imbalance wasn't a big issue. I ultimately settled on triode-connected EL34 drivers in those amps. Of course there are other ways to deal with this as well, short of adding another IT.
Kevin Carter
K&K Audio
www.kandkaudio.com
Kevin:
That makes perfect sense. I think that is what I will need to experiment with over the next few months; either DC couple with CCS or keep the IT. I will be using a trioded 6W6, so should have ample headroom at the grid, very similar to the EL34. Formerly using the trioded 46 as a driver.
The more I study the headphone amp schematic on your site, the more I like it. That trick with the input transformer, RCA on one side and XLR on the other, is slick. I will definitely go that route. You effectively force voltage balance through the coil inductance. Technically, one only requires impedance balance to have a balanced line; voltage balance is in theory irrelevant. The differential input handles that. However, by providing voltage balance you permit each triode to amplify similarly scaled signals, which are then passed on to the next stage. This prevents one side of the amp from having to do all the work. At least that's the way I envision it.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: