Hello,
I'd like to build the Mullard 3-3 amp:
http://www.r-type.org/articles/art-003h.htm
I'd like to omit the rectifier tube. But when I use an original power transformer from a EL84 based radio, and the greater voltage drop (from what I saw e.g. 40..50V) vs. silicon 4-diode rectifier is missing, the anode voltages will be too high.
Just adding a higher series resistance is not the same as the effect of a diode - in the former case, the voltage at the tube anode will more directly depend on the current going through it.
How would one do this?
... save from trying to obtain a transformer that gives just the right voltages. I might have another one, where the seller claimed it has curious 259V output, still a bit high vs. 250, but it's a start...
Then there's also the us having 230V today, where the transformers are for 220V, adding to what I need to reduce (if I don't want to age things prematurely).
I'd like to build the Mullard 3-3 amp:
http://www.r-type.org/articles/art-003h.htm
I'd like to omit the rectifier tube. But when I use an original power transformer from a EL84 based radio, and the greater voltage drop (from what I saw e.g. 40..50V) vs. silicon 4-diode rectifier is missing, the anode voltages will be too high.
Just adding a higher series resistance is not the same as the effect of a diode - in the former case, the voltage at the tube anode will more directly depend on the current going through it.
How would one do this?
... save from trying to obtain a transformer that gives just the right voltages. I might have another one, where the seller claimed it has curious 259V output, still a bit high vs. 250, but it's a start...
Then there's also the us having 230V today, where the transformers are for 220V, adding to what I need to reduce (if I don't want to age things prematurely).
I'd download http://www.duncanamps.com/psud2/download.html
Plug in your specs and see what you need to do to correct your voltage, Perhaps a CRC?
Plug in your specs and see what you need to do to correct your voltage, Perhaps a CRC?
After jumping through a number of hoops of presumable downloads, to always be led to a site telling me "it has moved", I quit the silly game.
Where I come from, CRC means cyclic redundancy check. What does it mean in your land? 😀
Where I come from, CRC means cyclic redundancy check. What does it mean in your land? 😀
It means the same thing in my land. On this forum it generally means Capacitor, Resistor, Capacitor filter.
It means "Capacitor Resistor Capacitor" First cap connects to the rectifier's output, then a series resistor from that point connects to the second capacitor. Sometimes, an iron-cored inductor may be used in place of the resistor for better filtering of the hum component, in which case CRC becomes CLC.
Have you seen this modern interpretation of the Mullard 3-3 amp?
Heated Cathode
Note, that user fitted an EZ81 rectifier valve for the power supply
Heated Cathode
Note, that user fitted an EZ81 rectifier valve for the power supply
I have not seen this, helpful alone for the explanation it adds for understanding, thanks!
Mine is really to be mono, though. Basically my aim is to build a tube amplifier with a typical speaker of the time, a very old school DIY cabinet around (steam-bending the wood for the arched top will be "fun" or something... 😉 ), and use it solely to listen to certain recordings that I like to have "tube radio sound", without "using up" my tube radio's radio-specific tubes when I'm not actually listening to radio (it does have an aux in, but I don't want to use it that way).
Whether this circuit will color the sound anywhere near what my EL41 based radio does is another story, but I also liked the sound of other mid 1950's tube radios, so let's see.
Mine is really to be mono, though. Basically my aim is to build a tube amplifier with a typical speaker of the time, a very old school DIY cabinet around (steam-bending the wood for the arched top will be "fun" or something... 😉 ), and use it solely to listen to certain recordings that I like to have "tube radio sound", without "using up" my tube radio's radio-specific tubes when I'm not actually listening to radio (it does have an aux in, but I don't want to use it that way).
Whether this circuit will color the sound anywhere near what my EL41 based radio does is another story, but I also liked the sound of other mid 1950's tube radios, so let's see.
Another transformer I "shot" in the bay arrived, it supposedly has 255V output and a 240V input "tap", so I guess that problem should be solved.
I worry about the missing delay that a rectifier tube would provide, heating up in parallel with the amp tubes and not exposing their cold cathodes to the full voltage.
I worry about the missing delay that a rectifier tube would provide, heating up in parallel with the amp tubes and not exposing their cold cathodes to the full voltage.
Voltage on cold tubes isn’t a big a deal as people make it out to be. Lots of classic gear from the tube era used silicon diodes including McIntosh.
A consideration is that with directly coupled designs, like this one where the driver stage is not decoupled with a capacitor, is that voltages will not necessarily flow at the right current from the outset if there is not the benefit of some delay from a tube rectifier. The Mullard 3-3 looks like a trivial circuit, but it is a very unusual design with that feedback to the screen of the EF86, It can be a bit unforgiving to build, I have read here, and gain can be a bit on the high side too.
I have renovated a few old radios from the 30's, using a bluetooth input on the gramaphone input they all seem to have, and the sound can be quite amazing! Here was a 1936 Paillard 70 TR that I replaced a few capacitors in, that had a 2A3 which only measured about 10% of emission. The video was part of an advert to sell it when I moved.
Paillard 70TR
(purchase price $40, renovation price around $20, sale price $10; not a sound business opportunity!)
I wouldn't worry about the other tubes in the radio so much. For one thing replacements are plentiful and cheap, unless they are very unusual, and in the case of that radio I looked at, they are going to outlive us anyway.
I have renovated a few old radios from the 30's, using a bluetooth input on the gramaphone input they all seem to have, and the sound can be quite amazing! Here was a 1936 Paillard 70 TR that I replaced a few capacitors in, that had a 2A3 which only measured about 10% of emission. The video was part of an advert to sell it when I moved.
Paillard 70TR
(purchase price $40, renovation price around $20, sale price $10; not a sound business opportunity!)
I wouldn't worry about the other tubes in the radio so much. For one thing replacements are plentiful and cheap, unless they are very unusual, and in the case of that radio I looked at, they are going to outlive us anyway.
Last edited:
I toyed with the idea of measuring the (integrated) current going into the filaments, and only when it stabilized at a certain level, I consider the tubes warmed up and enable the HV+ to the anodes with some HV power MOSFET or something.
What you describe might not be helped by that.
Does this direct coupling make the whole thing especially dependant on the exact type of the preamp tube? I don't have an EF86 yet, only one vaguely in the ballpark - though it can do much higher freq, could be another problem, eh? But I am circling some bay offerings like a vulture, lol. Let's see.
I wouldn't put it that way exactly. Alone the "magic eye" (round EM34 / soviet replacement) - which does not contribute to sound, but... I like them - gets eaten up rel. quickly & is not cheap (I already looked at round TFT dispays and considered making a digital emulation 😀 The displays tend to be a bit big, though, and might look adequate only at a distance)
You do not have to have the EM34 on all the time. Put a switch on the back, or see if there is anything that can be done when it is switched to 'phono'.
You have to have an EF86 for that circuit. The circuit was specifically developed to introduce the new Mullard tubes when they were first released. Maybe the Russian 6X32N is close enough, I use that in a Quad II and it works fine.
The circuit is deceptively simple looking. There is direct coupling in two directions, two feedback loops. It is a miracle of synergy.
This might be a better option? Many built and lots of threads on this site: EL84 RH84 Rev.2.
Ah! Also looking at the stereo version someone linked above, it explains that a bit. Very interesting.
I'll save that one for just in case & who knows when/what 😉
But I bought too many specific parts for the Mullard already, like HV capacitors and 3W resistors etc. I'll be looking to get a EF86 then. With some luck my father might even have a used one that's not in too bad a condition.
Either diodes or tubes like 5Y3 and 5U4 that heated up & conducted long before other tubes in the unit did.
The job of those seems "more primitive" than an amplification valve, though. Perhaps they could be more rugged. But either way, since their simpler nature probably also meant they were cheaper, it seems still a more economical prioritization of damage than making more expensive tubes degrade earlier, too.
You could add a EY88 in series between a solid state bridge rect and the filter cap. Pollin has them for 50c.
Cheers,
V4LVE
Cheers,
V4LVE