• 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.

New project 300B

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
When driven over 8 watts in class B, the bias point shift for the 300B, regulated supplies please! One 300B can play 15 watts no problem if properly implemented. All schematics I see are mostly 8 watts designs limited to class A.

At no feedback it produce 1%thd at 1 watt and continue to climb steady. So this design is good. but the power supply not.

Personally I would use a regulated cathode bias and negative bias for the pre-driver stages. All regulated.

Otherwise why build a S.e.t. using the most expensive tube/transfo for no power?
 
I get 10 W (3 % THD) out of my DG300B. Class A with regulated supplies all around.

Using a regulated B+ is also one of the best way to reduce hum on the amp output.

Tom

It is possible to get at 15.5 W at 3% thd, with a wide bandwidth.

I know we are talking low power, which is silly, with 10W 3%, you must get grid current distortion, only avoidable with a cathode follower which can cope with grid current.

To get to the next level you must regulate the cathode bias of the 300B, as well as the grid stabilized with a negative potential. This is more complicated than usual:emoticon:

I didn't built it, I got those figures from a designer.
 
10 W @ 3 % is actually with a cathode follower. You can get a bit higher output power with a source follower, but I liked the sound of the cathode follower better. You're absolutely right about the grid current. That plus the ~200 Vpp swing needed to drive a 300B well is the challenge of a 300B design.

Also, that's with a 5 kΩ primary on the OPT.

Tom
 
tomchr,

There are good reasons to design an amplifier that uses the JJ 40 Watt plate dissipation 2A3.
With a 2.5V filament, there is less field difference across the filament as it affects the multi-electron paths past the uni-potential grid on their way to the plate, than for the same construction that uses a 5V filament.
That JJ 2A3 has a plate structure that is at least as big as a real 300B, and it also has a glass envelope that is bigger than a real 300B. More plate and more glass = 40W dissipation.
If people want to tube-roll on such an amplifier that is designed for 40W on the plate, they should know that they do so at their own risk, when using tubes that "just plug in", because all real 15 Watt plate 2A3 tubes do "fit".
Especially with an amplifier like that, right?

But then I have to ask the question about persons who start with a 6L6 amplifier, and plug in the not-so "compatible" KT88, KT90, KT120 that has the same wiring and same socket.
Or, how about the person who starts with a KT88 amp, and plugs in 6L6 or 5881 tubes; the connect and fit so nicely.
How well might that work?

I use non regulated B+ for 300B and other tube amplifiers, and always get less than 500uV hum. All without global feedback too.
 
Thank you all for the strong interest in this project of mine. But at the moment I'm more confused than before, summarizing, is it better a fixed or automatic bias? In my opinion with a transformer that delivers 320v x 1,414 = 452v-50v of voltage drop we have about 402v and with a fixed bias a - 75v we have a dissipation of about 30w right ?. With two 10hom resistors connected to ground on the cathode of the 300b
 
Which type of calculation you done with voltage?
With 400 vdc the bias of -75 v is too high.
Last, you must have two resistors of 47 ohm and one of 1 ohm 1 w to ground for current monitor


Walter
Why? 47k 10k and 220 ohm is not enough???
Also 10k to ground......

Anyway I can reduce the grid bias from the trasformer from 250v to 220v
 
Last edited:
That JJ 2A3 has a plate structure that is at least as big as a real 300B, and it also has a glass envelope that is bigger than a real 300B. More plate and more glass = 40W dissipation.
If people want to tube-roll on such an amplifier that is designed for 40W on the plate, they should know that they do so at their own risk, when using tubes that "just plug in", because all real 15 Watt plate 2A3 tubes do "fit".

I wholeheartedly agree. However, if someone buys a "2A3 Amplifier" they have a very reasonable expectation that they can plug in a 2A3 and have it work. No instruction manual, warnings, or disclaimers will prevent a customer from plugging in a tube, and if the anode dissipation is higher than what a regular 2A3 can handle, the tube will go *poof*. This will result in a rather unhappy customer as his $2000 pair of NOS super-duper mesh carbon plate blah-blah-blah tube has just turned into a puddle of molten glass. The message of "I told you so" will need careful wording to prevent the customer from having a similar meltdown.

For DIY it's a different story, of course. While most complete their builds just fine, others have a bit harder time with them. It seems well-understood here that you have to pay attention to the tube type before you plug it in. Maybe use an old pair of tubes that you don't care much about as a sacrificial test pair the first time you power up the amp.

I use non regulated B+ for 300B and other tube amplifiers, and always get less than 500uV hum. All without global feedback too.

Yeah... With enough chokes and capacitance, you can certainly get the hum down pretty far. You'll probably notice that a sizeable chunk of your build budget goes to the power supply.

0.5 mV hum is enough to be annoying for me in a smallish living room with even inefficient bookshelf speakers. I seem to recall getting below 100 uV on the DG300B and further improvements are likely possible if the power transformer is moved further away from the OPT than the ~30 mm I allowed. :)
No global negative feedback in the DG300B either.

The mains hum in my solid state circuits sits around -125 dBV usually (~500 nV). That's without any chassis or attention paid to the wiring. Just naked board on a lab bench. :)

Tom
 
Here the tubelibrary.inc, but as a .txt file. Just re-name it to .inc form.

Google "Norman Koren vacuum tube models". He's an accomplished photographer with a vacuum tube hobby. He's made some scripts to model the tubes and has published quite a few SPICE models.

Usually a search for the tube number followed by "spice model" will get you going too.

Beware that there's a 1:1 relationship between the symbol and the model. I.e. the pin numbers in the symbol need to match with the corresponding pins in the model. This becomes an issue if someone has a symbol that uses, say, pin 2 for G1 and pin 3 for G2, but someone else uses pin 3 for G1 and pin 2 for G2. The pin numbers are arbitrary. My point is that if you get really wonky results, I suggest setting up a basic test circuit in the simulator to validate the tube model. If you can generate something that's pretty close to the data sheet curves, you're good. But if you get really weird results, you probably have a pin number mismatch.

Tom
 
tomchr,

When I said my amps have less than 500uV hum, some have less than 100uV (another 14 dB down or more).
But either my hearing is bad, or the frequencies of the hum is less "annoying". I even listen to some amps on medium efficiency speakers that are only 2 feet away.
I test the hum on a low uV DMM, a scope, and an FFT. The nature of the hum is low frequency, 60 and 120 Hz.

I have heard other's amps that have the higher frequency hum resulting from the wiring aperture of impulsive current of cap input supplies coupling into other circuits, power supply ground loops (that also have other circuits involved), and magnetic coupling of power transformers and chokes to the output transformers.

My "expensive" B+ power supplies use one choke, 2 or 3 resistors, and 3 or 4 caps.
I am not building them for mass production and sale to the public.

The 40 Watt 2A3 I have, can play in any amp that is designed for a real 2A3.
But the advantage of using DC on a 2.5V filament versus DC on a 300B 5V filament is the more even spread of effective bias across the tube filament.
Even WE mentions the bias difference for DC and AC filaments, versus plate current.
The 40 Watt 2A3 can be looked at as a super 300B, with only the filament different.

I am sure some persons have plugged in 300B to the real 2A3 amp socket, no way to prevent that. At least it will not destroy the 300B, but the HV supply may rise, and the B+ capacitors may be stressed over their voltage ratings. My B+ caps are voltage rated to work even if no tubes are plugged in (especially important remember the voltages for choke input supplies which have 0.9 x the rms, versus cap input supplies that have 1.4 x the rms). The choke input supply caps should have the same voltage rating as the cap input supplies, because the voltage will rise to the same value, before the tubes warm up and give the choke a load that makes critical inductance take over.

The same goes for 6L6GC amps. Some plug in KT88s before asking the question.
Depending on how robust the amp is, it either gets warmer, or it blows up.
No commercial amp manufacturer can prevent the tube rollers from doing that.
 
Last edited:
Hi folks, I made some changes to the file of the simulator, I corrected the value of the inductance and resistance of the transformer and I lowered the voltage values on the ecl82 a little less watt but less distortion.
what do you think?
 

Attachments

  • 300B_SE_ECL82_1_rev.2.asc
    6.2 KB · Views: 89
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.