• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

#26 pre amp

Chokes

Outside winding to the lower impedance?
I assume the choke is highly interleaved for reduction of capacitance, so there won't be much difference to the ears. Perhaps it offers another resonant frequency to the tube with wires swapped?

Depending on the choke of course. Many will simply be a single section wound with a start in the middle and the end up outside in the same manner as on a normal spool.

The lower impedance on the outside gives it a bit of screening effect. This may or may not be a good thing of course, but if the hum levels change when one swaps the leads around then at least it gives on a bit of a clue.

Also the capacitance between windings decreases as one goes in towards the centre as the length of wire per turns decreases. Not hugely but perhaps enough to be audible as an interaction with other things.

If one is using EL mains transformers, particularly where the primary is on a separate section with a split bobbin the neutral should be on the outside, and the live to the centre. That way the neutral gives a bit of screening and also the losses could be a fraction lower.

It's all very much a try it and see game of course, as intuition isn't always right.

Best,
Susan.
 
With any of my PCB-based Regulators, reducing the ripple to 80-140mV should be enough, even for line amps. So LC with 10mH or more and 10000uF is easily good enough.

Of course, this all has to do with differential mode ripple and noise.

When you use a high-capacitance trafo (eg Toroid) the noise that gets in is common-mode. The regulator will reject this, but not completely. Hence the recommendation for split-bobbin trafos.

The old 2004 Regulator circuit rejects CM and DM noise much less effectively.

For my understanding, can we install the full rectifier board plus necessary capacitors on the supply unit, then connect through long cable to the main chassis and directly to the regulators without further caps?
 
Made an improvement to my battery grid biasing today. Initially I was injecting the -9V through a grid bias resistor, and using a 0.22uF cap before that to block DC. Since I also have an LL1676 transformer at the input I decided to remove the 0.22uF cap and the grid bias resistor and put the battery at the bottom of the input transformer secondary instead, between the transformer and ground. This gives the advantages of no cap in the signal path, and much lower resistance to ground for the grid (more stable bias).

The bass improved considerably, and the entire presentation became clearer and more balanced. The battery is bypassed with a 10nF Multicap RTX polystyrene. The transformer secondary is loaded with an RC network of 2nF+850ohms, and the transformer is run in a 2:1 stepdown with primaries and secondaries in parallel.

I highly recommend this bias arrangement; the preamp loves it and so do I!
 
Made an improvement to my battery grid biasing today. Initially I was injecting the -9V through a grid bias resistor, and using a 0.22uF cap before that to block DC. Since I also have an LL1676 transformer at the input I decided to remove the 0.22uF cap and the grid bias resistor and put the battery at the bottom of the input transformer secondary instead, between the transformer and ground. This gives the advantages of no cap in the signal path, and much lower resistance to ground for the grid (more stable bias).

The bass improved considerably, and the entire presentation became clearer and more balanced. The battery is bypassed with a 10nF Multicap RTX polystyrene. The transformer secondary is loaded with an RC network of 2nF+850ohms, and the transformer is run in a 2:1 stepdown with primaries and secondaries in parallel.

I highly recommend this bias arrangement; the preamp loves it and so do I!

Attached is a schematic of the new input and biasing arrangement. It sounds far, far better to my ears than the standard arrangement with cap and resistor.
 

Attachments

  • 26 Preamp Input Circuit.jpg
    26 Preamp Input Circuit.jpg
    14.6 KB · Views: 638
Hi Magz,

Thank you for posting your results and schematic. I was thinking about doing the same thing, but I was a bit worried about damaging the transformer if it is spec'd for 0 current with the 26 drawing a very little grid current.

Does anyone have any thoughts of limiting any potential damage to an input transformer in this setup?

Cheers,

Rich
 
Those Lundahls I'm using are amorphous Cobalt core with no air gap, so they'd be very susceptible to core saturation, but I haven't seen it. The DC grid current is very tiny, unless the tubes are very gassy.

I asked Kevin Carter of K&K Audio (Lundahl dealer) about it before I did it, and he said that normal tubes without high leakage will work fine. My results have convinced me that's true.

In any case, the transformer wouldn't be damaged by the DC, the core would saturate, which can be remedied by passing an AC signal at the saturation level...demagnetizing, in other words.
 
In case anyone is interested, I measured the AC impedance of a Duracell Alkaline 9V and an Ultralife Lithium 9V, only one example of each, both new. Both measured 9.4V DC. I measured using my Escort LCR meter with a 1000uF, 25V Panasonic FM cap on the + terminal to block DC. The results at 100, 120, 1000, 10000 Hz for each:

Duracell Alkaline: 12.2, 11.2, 2.8, 1.6 ohms

Ultralife Lithium: 4.3, 4.0, 2.4, 1.2 ohms

I was curious how they compared as I'm using battery grid bias in series with the grid. I prefer to use the lithium because of its longer shelf life, but I wanted to check the AC performance first.

Listening-wise, I don't think I can tell a difference.
 
In terms of core saturation, how would one know that a core had been saturated and needed demagnetising?

Your bass response would drop off noticeably.

I've been back and forth with Kevin C. and Per Lundahl about this several times over the last few years in reference to a couple different projects. It seems the LL1674 and LL1676 are susceptible to developing some core magnetization if they are subject to too much DC current, which will cause the bass response to drop off. Per has implied that simply measuring the DCR with a multimeter is enough to do this.

The cure according to Kevin is: "High level (just below saturation) low frequency sine waves do the job the quickest. You can do that in circuit if you have a source of sine waves and a 'scope to take the primary up to saturation and back it down a bit."

Saturation is said to be ~16VAC rms or so at 20Hz on these. The datasheet shows 22VAC rms at the secondary at 30Hz giving 1% distortion.

All that said, I have now put about 20 hours or so on the LL1676 input transformers with the battery below the secondary, and the bass response is still thunderous. Magnetization from grid current doesn't seem to be an issue, at least with the 26 globes I'm using.
 
Last edited:
Good power supply - Black art?

I've worked some time on my 26 preamp, but have some trouble figuring out the optimal power supply for the job. As of now I have 20H-300uF-5H-100uF, and finally for each channel 5H-20uF-0D3 regulator tube.

This works and sound fine, but I was wondering if there is some kind of theory behind making a REALLY REALLY good power supply. I do not have a scope but uses Duncan Amps PSU designer. Is there something that signifies a good PSU that I should look for when designing or is it just a question about ripple, taste and black art?

Regards,
Vidar
 
The 26 preamp draws a fairly small idle current, and the peak signal current is also small.

There are many valid choices to make, but here's some considerations:

- choke input filter is recommended (best regulation, lower EM radiation)
- Large L, small C in a LCLC or LCLCLC (etc) is perfectly valid, and allows the use of audio-MKP capacitors (or better), and no electrolytics.
- next most important:high quality trasformer. choose EI, not toroid , and make sure it is split bobbin (Hammond call this 'dual bobbin').

If you can't get a good trafo, please use an isolating trafo (eg 230:110V or 230:230v) and treat its secondary as a kind of electrostatic screen (connect one side to ground, connect cap 1uF 440VAC MKP across secondary).

- experiment wit gas tubes and shunt regulators, and find something you like - or stay with passive!

- try some tubular ferrites on the mains lead.

- RC snubber across the secondary: try 47nF 1500V in series with 100 ohm carbon composition.