Some of the DIYAudio fans have built this version of the 4P1L with great success. There are several upgrades that can be easily implemented to improve this. I haven’t tried this myself, but my recent experience with the Gen3 and the 01 preamp gen2, I think are worth trying:
Replace the gyrator FET for a cascoded pair (M2 and M4 below) to improve PSR
Replace voltage reference by a cascoded LND150 for better HF and PSR response
Optimise the LF pole of the gyrator load by increasing R4 to 4.7 MΩ and reduce C1 to 220nF
Bias 4P1L to about 30mA. This will reduce distortion
An idea discussed with Rod Coleman, Andy Evans and others in DYIAudio forum which still is evolving. The purpose is to increase the output driving capability of this preamp using a cathode-follower stage. A bonus is to provide the filament bias using the cathode-follower valve filaments:
Albert sent me recently some great pictures and feedback of his 26 pre-amplifier using gyrator as the anode load similar to my design on the 26/01a.
Here are Albert’s notes on his 26 preamp:
“I have my 326 with separated power supply, 5U4G – 8uf PIO – 20H – 50 ASC – 30H – 50 ASC + 18uf pp feeding the Salas’ HV Shunt V2 Regulator, then to 2 ( left and right ) cascoded DN2540 gyrators. Rod’s filament supply coupled with filament bias @ 850mA with anode @135v. I don’t have anything to check the distortion, but this is the most good sound preamp that I have finished to date. It pleases my ears so much that I have the intention to turn it on once I’m in the house. It sounds gorgerous with very good image seperation. Sound stages are wide, deep and absolute quiet. The highs are sparkling smooth with lots of details. Mids are lush and warm with admirable vocals. I’ve found it even better at the lows if compare with 2x 156c. It goes deeper with clear notes.”
A nice pair of CX226 in the new preamp. They are in very good shape, very low distortion measured in the workshop time ago and no hum induced, so great for this preamp.
Sound is definitely warmer, without loosing any of the bass and bright treble from this preamp. A great preamp so far, happy like a bunny!
Today made significant progress on the new version of the 26 pre-amp. Only bits outstanding are the CCS modules part of the HT supply on board. Tested glow valves, output transformers and filament circuits…
Before breadboarding a pre-amp with these nice OT provided by Thomas Mayer, I decided to simulate some options in LT Spice to see what results I got.
CX301a DHT pre-amp
First one is my preferred CX301a thoriated-tungsten DHT. I wired the LL2745 in 5.6:1 step-down configuration. This should provide a low output impedance which is what we want in this configuration where we need to drive the cables to the amplifier with sufficient capability:
CX301a preamp based on LL2745
Looks very promising. Of course gain will be much lower than a gyrator-based pre-amp, the 01a anode load is optimised providing very low distortion: 0.015% based on my SPICE model with curves taken from real CX301a. Gain is low at 3.1dB, but we don’t want loads of gain in this pre-amp. Sound is what we are after…
26 DHT pre-amp
Now it’s time for the revered 26. I used Dmitry’s model based on the RCA manual curves. I’d like to simulate this again using a model based on starved filament curves with a real 26.
Originally planned to bias the 26 in a different operating point based on feedback from Andy Evans, however after playing a bit with the OP I found that a more linear point was around Ia=5mA and Va=114V @ Vgk=-6.85V
26 preamp based on LL2745
26 looks more interesting in principle as the output impedance with this model is lower than the CX301a thanks to having a lower Ra (7KΩ against 11KΩ) so LF response will be slightly better in a side by side comparison.
Will be trying these two with filament bias and Rod Coleman’s filament regulators. The HT will be provided through a Salas HV shunt regulator.
New Lundahl LL2745 OT pair arrived today in the post. Very excited, Thomas Mayer kindly send me this pair for trial. These are specifically designed for the low current and high anode resistance DHT used in pre-amps (e.g. 26, 01a, CX112, etc.). Wiring and primary details are similar to the LL1689. Hope we can get a final datasheet in September 🙂
These are gapped at 8mA. So the primary inductance is about 200H in theory (160-180H in practice according to Thomas). The labels are incorrect. This is 2×2.8:4×1, providing the option of wiring it as 5.6:1 if both primaries windings are in series and all secondary windings are in parallel.
Will rebuild the 26 or 01a preamps and do some test.
I would like to build a quick breadboard with the 01a to do a comparison against the gyrator loaded preamp I’m using at the moment.
