• 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.

First shot at a battery DHT preamp stage

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Its Alive!!!

Using diodes worked a treat, one 1N4701 for the time being (first thing to hand).

Plate volts have risen a small but appreciable amount.

LED next, and again further improvement.

Resistance didn't work, but was easy enough to try (perhaps not enough R)

Now, I wonder.... maybe a neon?

If so that may set the grid at about half B+ and could work a treat.

Or things may go awry, as I'm not certain negative resistance will be a good thing....
 
I did a battery linestage many years ago using CCS loaded STC 3A/109A triodes with a string of 80 x Poundland 800mA/H Nicads for HT and 4 x 2000mA/H NimH for the filaments . If I was to re-do this , I'd use 2P1P or 2P2P valves which are DHP with B7G base . 2P1P works especially well at low voltages , has low Ra , but has a very low current (60mA) filament so probably microphonic . I have briefly tried 2P2P , these are similar to 2P29L and sound pretty good . Food for thought , although it's a right P.I.T.A when you want to listen to music but the batteries are flat as the linestage was left running overnight...

316a
 
One thing to remember; although these valves use exotic technology, their transfer functions are similar to familiar ones. They are not designed to work at particularly low voltage (only a few valves are). The most negative grid voltage on the 1j29b (pentode) characteristics is 2.5 so I would think that much more than this will cut current flow off completely. It would probably be best to treat it as if it were a small EF95 of something similar.
 
Hehe yes batteries draining because of absent mindedness would be annoying - quite often my mind is absent!

I have now got plate to bias nicely about 1/2 HT:

(47V-57V and HT of 90V, using Ra between 17k and 68k, anode current between 0.6mA and 2.2mA, gain between 5.5 and 6.5 V/V, and 100k load)

I still have quite a lot of DC anode current left to play with.

With even lower HT, I wouldnt be surprised if Ra becomes very small indeed, in order to maintain plate volts and around 1/2 HT, and keep current at the level I see now.
 
Hi Piano3,

My little experiments show otherwise. (At least when triode wired)

(Vg = -5V, at 2.2mA quiescent current, plate at 57V, HT of 90)

How it sounds and performs WRT THD, is another matter, and something I havent tested yet.

Of course, higher HT is always going to allow greater headroom for large signal output, at lower THD, than a lower HT will.
 
Last edited:
Oh I see... I missed that point of reference.

I agree, a neon isnt a good idea.

I dont want the grid at half HT, upon engaging my brain; I forgot to step back from that comment and state my error of judgement ;)

Looking at the circuit operation at that stage, felt quite alien, in comparison to IDHT circuits. I believed that I may have to elevate the grid very much more than a DHT circuit.

On that count, I was very very wrong! So not a typo, just me not getting the gist if what you were saying

Initially, using a single diode, the impact on plate voltage was small - perhaps setting plate volts at around 30V.

Now I have things working better but far from optimised, I understand the mechanics of this type of circuit a little better.

It's not actually as far removed from working with IDHTs.

Which pleases me no end!

There are still many question marks about how well these tubes will work at low voltages - I.e. just because a skilled designer can and has made a very nice radio receiver operating on just 9V plate supply, doesnt automatically mean an audio circuit using low voltages, can be made to operate at low THD.
 
Last edited:
I havent done a huge amount of audio things with discrete semiconductors - though more than with valves.

I made Rod Elliot's P3 amplifier, using MJE3055/2955, with MJE340/350 as drivers, rather than BD139/140. Not small signal though :rofl:

Most of anything using discrete components has been simple, and generally instrumentation, front ends for H bridges.

But the datasheet are easier to come by, and read. :rolleyes:
 
A Theremin?
Something like this...?
 

Attachments

  • IMG_1633.jpg
    IMG_1633.jpg
    537.3 KB · Views: 212
Quick lunchtime test and update

I grabbed 10 minutes to set up and run a quick test of the simple DH Triode strapped 1J29B circuit.

HT 90V
Va 50V
Vg -3.3V (2 series LED string)
A ~5 V/V
Rload 100k

Signal input of up to 1 Vrms

A quick glance at the FFT and it seems the circuit is performing OK

Opinion welcome, vis a vis usual OL THD vs Gain)

I see H2 down at -40 to -50dB or more (signal level dependant), H3 down another 10 to 15dB.

Higher harmonics down in the noise floor at or below -60dB.
 
I got some more play time today, so I switched out the LEDs and substituted a diode string using 1N4007.

Due to the nature if these valves, or my circuit, I have found that biasing them so the Anode is slightly above or below 1/2 B+ gives me the best gain, and also THD.

As Piano3 or Lampie mentioned, the valves seem to perform best with about 1.5mA of quiescent current.

Anode load of between 27k and 39k, using between 5 and 7 series connected 1N4007, for Vg -3V to -4.2V respectively; seems to work well.

Also, connecting G3 to K, or A when Triode strapped, makes little difference, certainly nothing I can measure.

Gain is between 5 and 6 V/V depending on Anode load etc, due to the slightly different operating points mentioned above.
 
Quick test in pentode mode.

(Without changing anode load, which I forgot, and should've done)

B+ 90V
Va 50V
Vg3 = 0V
Vg2 68V tapped off B+ using 100k, 4uf bypassed to 0V
Vg1 5 to 7V using LEDs and diodes.

Gain (in the messed up example) about 23V/V
THD similar to triode strapped, at the same dBV output level (input signal to get there is a lot less, obviously)

Both measure about 1% (Open Loop)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.