diyAudio (
-   Tubes / Valves (
-   -   Advice on SE DHT for Lowther Fidelio DX2 (

Rudolfs1972 30th July 2007 10:23 PM

Advice on SE DHT for Lowther Fidelio DX2
This topic may be old as mountains; still, since I failed to find a concise answer(s), here I go:

I am looking for an advice on good sounding SE DHT for 97 dB/W Lowther DX2 Fidelios.

I use only digital sources (currently Arcam DV-27), rock+ classical+ occasional jazz & pop, likely to serve as the only amp (= also for movies), subwoofer present but not desired, if speakers ever change they will still be Lowthers.

Therefore my goal is not an universal amp to work reasonably well with many sources and speakers, rather a device which will have best synergy with Lowther speakers (and a digital source).

My intention is to use generous amounts of iron - interstage transformer(s) (1:1, possibly1:2 step-up if 2A3 driver), if necessary an input step-up transformer. Aternatively one direct coupling (possibly plate & grid chokes there).

I will use no NFB, would like to avoid PSE, I do not fancy triode connected tetrodes/pentodes unless for a very good reason, any exotic tubes are out of the question... so low to medium mu mainstream triodes, start with Chinese and then roll them.

No comments on PSU are asked for, still they will be appreciated.

My questions:

1. Output -- will 2A3 be sufficient? Is it reasonable to operate it in A2 on peaks?
300B surely would be enough, but then -- are there enough benefits to justify 845 or 211 given HT DC danger and solid state rectification (space and budget is limited after all)?

2. Transformers -- if 2A3 is chosen I ponder over parafeed (Sowther or AE Europe transformer + choke), or air gapped transformer, and certainly air gapped if 300B (or 845/211) will be used -- double C-core Lundahl or AE Europe (I live in Belgium), also for interstages Lundahl or AE Europe, but here main question would be if parafeed sounds considerably better.

3. Voltage amp / Driver -- I believe it would not be very good idea to repeat the same tube, -- like "step-up--2A3--step-up--2A3" -- unless made in different factories. My notion is that 6SN7, which I expect to use somewhere in the circuit, is not strong enough driver for 300B (is it sufficient for 2A3?), but smth like 6SN7--1:1--2A3--1:1--300B actually may work, also in A2.
Any other tubes to use here? 6n30? Alternative approaches, e.g. push pull driver for SE output, like 1:1+1--6SN7p-p--1+1:2--300B/2A3?

Well... thank you in advance then!

arnoldc 31st July 2007 12:45 AM

I have heard these speakers from a friend, driven by an Airtight 300B.


1. We connected a 45 amp on these speakers and they were easily driven to party level (How big is your room?) so a 2A3 will even put a stronger drive. We used a Rotel 10xx (I forgot) HDCD player as the source. I like the 45 better than the 300B when paired with the Fidelio, but YMMV.

2. Since my friend is the distributor of Tamura and Tango, I have heard amps using these transformers. Personally, I have used Tamura, MagneQuest, Hammond, and James. I will make a bold claim that MagneQuest DS-025 and 2A3 are meant for each other. My experience with Lundahl is limited to power and interstage.

3. You may shoot me for this, but I never liked the 6SN7, however good it is and even proven by Morgan Jones in his 3rd edition book. I like 76/56 tone better (but high Rp), and 417A/5842 as drivers. I also tried 6DJ8/6N1P, 12AU7, 12AX7, 5687, 6072. Of that bunch, I like 6072 best, and for some reasons beyond me, I got great synergy with 12AX7 driving an 807 in Tetrode UL mode, but that's not your interest.

Please do not let my post prevent you, or influence you, from pursuing other things. It's the joy of DIY.


Gerrit Boers 31st July 2007 07:09 AM

If you can read Dutch, have a look at DHT Rob's Tyran:
Tyran amplifier

It's an all DHT interstage coupled design.

As you have only digital sources you might also take a look at my approach:

Prototype and schematics

This is a 'no stage' design. Because the Benchmark DAC1 can deliver such a high output voltage (+28dBu) I decided to drop the input stage. A LL7905 input transformer is used to directly drive the grid of the 300B.

Rudolfs1972 31st July 2007 04:00 PM

Push-pull driver and/or parafeed output?
Sorry I ommitted inportant info, -- my room is ~40 sqm / 100 m3, close to square, solid walls, opened to two of other rooms and has a considerable glass area.
So I see that 2A3 may be sufficient for me and 300B enough and to spare.
I forget about 211 and 845 for now. That's a relief as >1kV DC is lethal.
Gerrit's page is very interesting.

I am still pondering about the driver (will use interstage, only which tube) and output arrangement:

I had a look at MagneQuest DS-025, for 15 Ohm it will use the whole secondary which seems fine, but again I came across the parafeed.
Sowther offers a nice parafeed transformer and a choke to go with it, so does MagneQuest and surely AE Europe could come up with smth, still I imagine it may be easier (and cheaper) to get a well designed air gapped transformer?

Driver => output:
I believe main argument for a strong driver is that it does not run out of steam before the output is heavily clipping, say at least 3dB, better 6dB reserve here; one needs power to drive Miller capcitance anyway; just add ability to handle grid current (readily provided by interstage), A2 is no problem?
2A3: WHAT COULD DRIVE it in A2, and what performance could be acieved - round 5-6 W @ 5% THD ?
300B: would that be 10-13 W (med. load, moderate idle dissipation), -- so not really necessary unless easy to drive in A2.

push-pull for the driver still seems interesting: say Lundahl LL7905 input/ splitter, then two halves of 6SN7 (or smth. else - I will make some research there) in push-pull, driving 2A3 (or 300B?) via interstage. That would by large cancel 2nd harmonic (and there would be very little 3rd), give good transformer bandwidth and PSU rejection, --
Is this capable of driving 2A3 / 300B adequately? If not 6SN7, then another dual triode (current production, since matched system is preferable) ?

Now when I saw in Rob Tyran's page THD and P out as function of the load resistance for the 300B, it stroke me how convenient it might be to have a 300B amp with 5k:15 Ohm transformer (parafeed or not; on use of A2 extra research necessary).
That would allow connecting 8 Ohm and even 4 Ohm speakers at a penalty of increased distortion (and power)... if I only had other speakers (Dali Trios do not count). Anyway 5k load for 300B would be fine... but I got carried away on this minor issue.

Gluca 31st July 2007 05:14 PM

Parafeeds amp benefit of the smaller ungapped transformer (better highs and lows) and have a superior PSRR. You need the extra iron the loading choke is calling in. In a very short sentence: I prefer them to series feed amps.

I would not use an interstage transformer but a grid choke or a direct coupled design.

With a 2A3 you can stay with a simple 2 stages design while the 300B will need an extra stage. My take on the subject is to build a 2A3 amp: I think you wont need those 10 watt and anyway you can easily add a 3rd stage with a 300B to the basic 2A3 amp using the same OPT and PSU (you need to take care of the extra current the 300B will draw). So if you will feel the need in a future date with a few mods you will have a new amp.

Consider to drive the 2A3 (if this will be your choice) with a beafy SRPP (6H30 et similia).


Rudolfs1972 31st July 2007 11:18 PM

Reasoning behond question on parafeed
My understanding is that parafeed, by introducing a separate choke to take care for DC, indeed aloows small top qiality output transformer.
Bt as a trade off parafeed:
1) introduces a cap in the signal path, -- what might be considered a mortal sin by some, -- and
2) supposedly (by removing DC from output transformer) changes conditions for the core in such a way that it increases the histeresis loop, resulting in a magnetic analogue of a cross-over distortion --
Parafeed transformers are smaller and have smaller and different cores with different properties (e.g. mumetal or mixed material)...
Somehow my intuition (which usually is good) is strongly pointing towards 1:1 interstage(s) and a series output transformer, but I would be more confident if smb. else also shared a rational argument and/or experience.

All times are GMT. The time now is 09:34 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2