Welcome! Need support, you got it. Or share youe ideas and experiences.
Return to K&K Audio / Lundahl Transformers
184.23.134.23
I'm interested in building a DHT filament supply using some Lundahl chokes. I may use either a LCL passive filter (as suggested by Thomas Mayer) or I might use the choke filter as a preliminary filter in front of a Coleman Regulator. I'll probably try both to hear the difference.
I've looked at the LL2733 spec sheet and am now wondering which connection alternative to use. If I understand the drawing correctly, I have 3 options:
Serial Connection gives me 400 mH and both chokes filter one line of the output.
Common-Mode Serial gives me 200 mH applied to each leg of the output.
Parallel Connection gives me 100 mH at twice the current rating.
I don't need the high current alternative (for tubes like the #26, 4P1L, 71-A, 801-A, 300B, etc.). So that leaves the two serial connections. Can someone explain the relative advantages of these two options?
I've never actually seen a schematic of Thomas Mayer's passive filter for filaments, but I've seen several of his posts on forums and he was quite clear that it was LCL and not CLC. I understand that this would reduce the large capacitor-charging spikes (and resultant noise) that occur when a diode bridge must feed a capacitor directly.
However, by specifying "LCL" , this suggests that there is no capacitor following the second choke. This seems odd to me. Am I just reading this wrong?
My final question is about 4P1L tubes. Can someone recommend a reliable source for obtaining these? The only sources that I've found appear to be located outside the USA and I'm a little uncertain about which sources are reliable. Also, are there specific tube brands to seek or avoid?
Thanks so much for any assistance!
Follow Ups:
Here is a link to what Dave had to say about Thomas's LCL. If you read the whole thread Thomas says one supply per tube and the last L is placed in the signal box.
(separate power supply chassis)
I like Daves idea of a LC with a CMC on the filament with a hum pot.
Cal
NT
I guess I got ahead of myself. Here is the link.
Cal
I don't know the configuration that Thomas Mayer is using either. You would have to contact him to understand his "LCL" implementation fully.
The common mode series connection offers the same degree of filtering that the simpler series connection does, but adds a degree of common mode filtering, removing "noise" that is common to both power supply leads. It is worth pursuing for the small, but significant sonic improvement that it brings.
My experience with buying tubes from Eastern European or Russian suppliers on eBay has been mostly very good. If the picture shows 4P1Ls wrapped in the original packing, then it's very likely that they are new. Naked individual tubes shown without the packaging visible are to be avoided. I have never seen a 4P1L that was manufactured by any other than the "Svetlana" (winged C) factory, although I suspect that they exist. I've heard really good sound from tubes produced in the 60s, 70s, 80s and early 90s.
Kevin Carter
K&K Audio
www.kandkaudio.com
Thanks, Kevin!
Your advice on sourcing the tubes sounds reasonable. I've had good luck on eBay (following careful, and hard-learned lessons).
After some searching I did find a schematic of Thomas's LCL filter. It does use just one filter bank (with 30k uF) and the last choke connects directly to the DHT. He uses the LL2733 choke.
If I'm calculating this right, a stereo implementation of his topology (4 chokes) would add about $400 to the project just for the filament supplies! The Coleman Regulators seems like a more feasible solution.
Thanks again for your help.
I suspect that the rationale for this is the relatively high audio impedance that the second choke represents.
Iron ain't cheap, but then, if used properly, they also don't sound cheap!
Kevin Carter
K&K Audio
www.kandkaudio.com
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: