• 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

Member
Joined 2004
Paid Member
Firstly THD was quite high - 0.1%, but that settled after a while to 0.07% - seems reasonable.

Bad tubes.... or Cunningham CX-326s? :)))
My five pairs #26 tubes (at 6Veff output) has 0.026-0.041% THD... except CX-326s (0.097, 0.124).
Furthermore CX326 has mediocre sounding. :-(
For me the best sounding pair is Majestic G-26 tubes.

10R 50W filament resistor got to about 110 degrees C
I think your heatsink undersized... or no heatsink.
I use passive (HP computer) CPU heatsinks.
 

Attachments

  • Heatsink.jpg
    Heatsink.jpg
    64.6 KB · Views: 822
  • image036.jpg
    image036.jpg
    65 KB · Views: 816
4P1L

We're worse than nuts. I. for one, have gone over to the dark side and am using 4P1L instead of 26. That's just after stocking up on 50 nice 26s!! I blame Ale....

The 4P1L is even more microphonic, but it's quiet - much quieter than the 26.

Andy

With a qick scope around I found a 4P1L Tube Price of a staggering €1.60 each!

If that a real deal, I can see the attraction already :)

How much quieter? Anyone got a real mV or uV comparison?

Not so pretty though :(

Best,
Susan.
 
ps these things are SO microphonic. You guys are nuts! .."

We're worse than nuts. I. for one, have gone over to the dark side and am using 4P1L instead of 26. That's just after stocking up on 50 nice 26s!! I blame Ale....

The 4P1L is even more microphonic, but it's quiet - much quieter than the 26.

Andy

With filament starvation (i.e. 500mA if filaments used in parallel) and proper socket suspension, 4P1L microphony is barely noticeable.
Knowing your building layout don't expect it to be quiet if you bolt it to you 4mm aluminium top plate :)
Ale
 
Bad tubes.... or Cunningham CX-326s? :)))
My five pairs #26 tubes (at 6Veff output) has 0.026-0.041% THD... except CX-326s (0.097, 0.124).
Furthermore CX326 has mediocre sounding. :-(
For me the best sounding pair is Majestic G-26 tubes.


I think your heatsink undersized... or no heatsink.
I use passive (HP computer) CPU heatsinks.

I am not sure what type my valves are, nearly all the markings have gone. But it is good to know that they are normal results.

As for the heatsink, they are mounted straight on the 4mm aluminium chassis, so no purpose heatsink.

Here is a photo I took earlier, it isn't that clear but it shows the large instability issues. y axis shown in 2dB steps. Again, didn't have much time to continue after I fired it up but I am glad that it is almost up and running, it's only taken a year....

Charlie

edit: the frequency response is a multiple tone response for rapid results - so this isn't acurate
 

Attachments

  • unstable.jpg
    unstable.jpg
    693.3 KB · Views: 789
Last edited:
Knowing your building layout don't expect it to be quiet if you bolt it to you 4mm aluminium top plate...>

I'm not getting any problem with microphonics in my solid chassis, but my room is well damped and speakers are about 89db. I even used to use 1G4G and they're horribly microphonic in other rooms and systems.

By quiet I mean they don't hum at all or make funny noises when you put your hand on them!! They also all test pretty much the same - quality control is excellent.

Andy
 
One year on...

OK so worked on the preamp this evening and the end was in sight so I carried on. I have no way of testing properly at home but I was able to make sure I had the right voltages in the right places. And managed to get it to a working state - still a lot of work to be done tweaking etc... making it safe for my cat..

But it is working, and I am currently listening to Joni Mitchell Blue as I type this. In fact it sounds like she is sat right next to me. The tone is superb, very soft yet still retaining all clarity. Unfortunately I just moved and am not yet used to my new listening environment, but I am looking forward to listening more and more.... it is very intriguing!

I will hopefully take it into work tomorrow and get some proper results and eventually .... some day... make it safe! I could not get the full 1.5V across the filaments but I think this is down to a slightly low PS.

Thank you to all for your help. Ale, Rod, Thomas, Andy and DIYers here and in the London Circle - thank you for being so patient with my questions...

.. Andy I hold you fully responsible for getting me into DHTs and may never have money again because of it, but at least i'll be happy!

Cheers

Charlie
 

Attachments

  • 26 pre.jpg
    26 pre.jpg
    781.8 KB · Views: 674
.. Andy I hold you fully responsible for getting me into DHTs and may never have money again because of it, but at least i'll be happy! Cheers Charlie

Well, my cunning plan is to use nothing but 4P1L tubes, which can be bought for around $2. Input, driver, and paralleled for output. Admittedly you have to heat them up, but just think - no more ridiculous prices for Mullard 12AX7 and the like! You can laugh in the face of tube collectors and those ridiculous ebay prices!! Just buy a box of 100 and never use anything else ever again!

Andy
 
There is one situation where the series-connected filaments, using my Regulators, might be a good idea, but beware of the prerequisites:

- Push Pull, cathodes connected together at one end, and Grounded.
- Fixed Bias, with separate bias pot for each side of the PP pair. The bias adjustment will be Vf higher on one side (i.e. a difference of 1.5V between them for a 26 DHT pair, or 2.1V for 4P1L)
- balance resistor might be needed across one filament, since the filament current is often spread 10% or more across samples of the same DHT.

The arrangement would make no sense for cathode bias or filament bias, and battery bias would be difficult to balance the PP pair.

I have not tried this yet, but can prepare the circuit drawing, if there's an interest in it.

The big advantage is that one transformer winding will do for the 2 PP filaments. With PP, there's little risk of crosstalk spoiling the sound.
 
Jamesdb was asking about putting the filaments of two 26 tubes in series (or maybe parallel?) which got a negative response from me and Rod. We've had this discussion before, and I think the general idea was use individual supplies. Anyone who's had success with combining them is welcome to describe their experiences!

Andy


Nobody remember the amazing Kevin Kennedy's 26 Preamp ?
kennedy
 
Hello Rod,

Thank you very much for informative advice.
I now completed the first step of the 26 pre-amp partly modified from my 71A pre-amp. Frankly, I like the sound of 26 pre-amp.
I shall try "Filament bias method" as a next step adopting a new PSU.
 

Attachments

  • 26 pre-amp-1.jpg
    26 pre-amp-1.jpg
    972.1 KB · Views: 1,138
  • 26 Pre-amp.jpg
    26 Pre-amp.jpg
    149.2 KB · Views: 1,129
Hello Hiro, I am pleased to hear that the change to the 26 is a success!

The output transformer, with Hitachi FineMetal core looks very interesting. How does this trans sound?


When I replaced from Tango NP-8 to Finemet core OPT, FM15K(200H)-150x2 in my 71A pre-amp, my first impression was of expansion of bass range and smoother in high range with less distortion.
The sound of 26 pre-amp just suits my taste.

Specification of FM15K(200H)-150x2:

Winding ratio 1 : 0.22
Inductance 212H (5Hz) 15K
216H (100Hz)
256H (1KHz)
Maximum current 15mA/15K
DCR primary 783 ohm x2
Secondary 43 ohm x 2
Frequency response 8Hz ~ 30KHz (-3dB) ( Drive with rp=15K ohm)
 

Attachments

  • 26 & 71A.jpg
    26 & 71A.jpg
    162.7 KB · Views: 1,059
  • 26 pre-amp-2.jpg
    26 pre-amp-2.jpg
    235.8 KB · Views: 848
Well, I finished up my 26 preamp last night and man, does it sound great! I'm running the 26 directly into the AVC in a parafeed arrangement using the mu output of the plate ccs. Square waves look excellent at all volume settings, frequency response extends below 10Hz (the limit of my equipment), and even with only 3 hours running time it sounds clear, punchy and warm with rock-solid bass even at lower volumes. Unfortunately I had fly to Florida for work today, until next weekend, so details, pics, etc. will have to wait.

The AVC is the Bent Audio TAP, with full remote functionality, stacked 3x3 for about 150H inductance. Using LL1676 transformers on input for galvanic isolation, phase inversion and gain cut 2:1.

Stay tuned, more details in a week.