never say never - 300B amps...

[Tweeker]

Hi Andrew,

Now thats worth knowing. The power supply for the plates will be based on an LM317K (since I've seen it done and have a copy of the schematic) So adjustable is fine. Whilst we're at it what does reducing the heater power do for tube life? It must be good right? up to a point anyway.

I may have misunderstood what was being supplied here. I assumed it was a Maida regulator supplying 300B plates; if its a heater or filament supply your probably fine, provided its within ratings.

 

[dave dove]

andrew

i think your initial question
re: 300b operating points
is a real good one:
-why don't we operate 300Bs at lower voltage higher current-

i don't see any reason why we shouldn't be
running at close to plate dissipation
-except for the output transformers limitations
or historical conservatism-
i would like some guidance on this also

as to heaters
if you're building on a breadboard
you should start with AC
then try the alternatives

dave

 

[barossi]

Hi,
i am also in the prozess of building a 300B amp. I would like to build the JELabs 300B amp. I will use Lundahl transformers LL1664/3K-80mA.

I am not really sure which operating point i should use?!
I have 900 Ohm /50W cathode resistors in stock.


Regarding the "Typical Operating Conditions - Chart"

Anode Voltage is U-Ugrid, isn't it?

I did some simulations with psud2, but i am not very happy with the results:

An externally hosted image should be here but it was not working when we last tested it.

Kind Regards

Barossi

 

[Sheldon]

You can greatly reduce the overshoot and ringing by adding an RC filter at the end. Try 10-20R with 47uF.

Sheldon

 

[Sheldon]

Anode Voltage is U-Ugrid, isn't it?
Barossi

No, anode voltage is the voltage at the anode minus the voltage at the cathode. The examples in the table are for fixed bias (grounded cathode), so anode voltage would be the voltage measured at the anode (measured voltage, minus zero volts).

With a resistor between cathode and ground (cathode bias) the cathode voltage would be positive and would be subtracted from the voltage measured at the anode.

Sheldon

 

[jlsem]

You can greatly reduce the overshoot and ringing

That overshoot and ringing is nonexistent in practice. It's an anomaly of PSUD.

John

 

[Sheldon]

That overshoot and ringing is nonexistent in practice. It's an anomaly of PSUD.

John

Yes and no. If you use the soft start option (which approximates a vacuum diode start up), you won't see that overshoot on start-up. But if you check the behavior with a stepped load, you will see it more or less in proportion to what you see in that graph. It can be an issue, and it can be reduced.

Sheldon

 

[kmtang]

To check if there's overshoot or ringing, I think it is best to change the current draw with "STEP CHANGE", say started with 80mA to 100mA after 1sec. Then, you will see if there's excessive overshoot or ringing occurs.

I just tried to run the PSDU with the same value of components and noticed that it is better to use 100uF as the last bypass capacitor.

Johnny

 

[Tweeker]

I suspect the ringing really is there, 5H 15R is an extremely high q choke. Its not as horrible in practice as your operating in class A so arent ringing said tank circuit often.

 

[barossi]

Hi folks,

thanks for your help!

I just added "softstart" in PSU2. Here it is:

An externally hosted image should be here but it was not working when we last tested it.

So i think it won't be an issue here!

I use Lundahl LL1673/20H. If you connect the coils in parallel you will have 5H and 15 Ohm.

So i will use the following operating point for the 300B:

U~420 V, Ra= 3K (Lundahl LL1664/80mA - 3K), Rcathode=900 Ohm, Ucathode ~70V

Kind Regards

Barossi

 

[Nordic]

Barosi, that jump is still within the 20% or so you have to accommodate for line fluctuation... provided line is behaveing nicely.... but a softstar is generaly so easy to do, its probably better for when the power co. is naughty...

 

[Sheldon]

Barosi, that jump is still within the 20% or so you have to accommodate for line fluctuation... provided line is behaveing nicely.... but a softstar is generaly so easy to do, its probably better for when the power co. is naughty...

The soft start we are talking about here is not hardware, but just a software option that roughly simulates how a directly heated rectifier would behave in the simulated circuit. An indirectly heated rect. would be slower still. If you were using silicon diodes, the original posted results would be a closer approximation.

Barossi,

Now that you account for the rectifier warm-up, it's time to play with stepped functions. You may choose not to modify the circuit, but it's good practice to push it around a bit at this stage. It's only a simulation, so it costs nothing in parts and little time, but helps for understanding the behavior of the circuit.

Leave your window at three seconds (or two) and set your step for about 1 second. Start with 150ma then step up to, say 250mA. Then try those settings in the reverse order. Have fun.

Sheldon

 

[zarniwoop]

can we go back to the AC heating issue?

I still don't understand how people are able to get low hum amounts when AC heating a 300B. I get about 7mV of (mostly) 120Hz hum at the OPT 2ndary that is due only to the AC heating. Not horrible, but its noticeable on Hi-eff speakers. The 60Hz component from the heating is nulled via a hum pot, but its the 2H that is largely the culprit.

the PSU is very stiff, and I've heard that some people get some cancellation because the 120Hz psu ripple is out of phase with the heater hum. I've also tried injecting some O-O-P 120Hz at the grid of the 300B (see steve bench), but I couldn't get that to work for some reason.

I am tempted to just do DC, but I keep hearing about people with <2mV hum using AC. What else to consider?

 

[jlsem]

Yes and no. If you use the soft start option (which approximates a vacuum diode start up), you won't see that overshoot on start-up. But if you check the behavior with a stepped load, you will see it more or less in proportion to what you see in that graph. It can be an issue, and it can be reduced.

I'm just saying that with all of the power supplies I've built with that configuration, I couldn't get them to behave like PSUD predicts. There may be a slight overshoot, but it happens at a much slower pace than simulated. Computer simulations can only take so much into consideration and usually end up as approximations of approximations. They can be fun to play with doing more exotic circuits, however.

John

 

[Tweeker]

Hi Zarniwoop,

I still don't understand how people are able to get low hum amounts when AC heating a 300B. I get about 7mV of (mostly) 120Hz hum at the OPT 2ndary that is due only to the AC heating. Not horrible, but its noticeable on Hi-eff speakers. The 60Hz component from the heating is nulled via a hum pot, but its the 2H that is largely the culprit.

There really shouldnt be much harmonic content with straight ac power, a poco or household issue? Are you sure its not PSU noise related?

 

[zarniwoop]

There really shouldnt be much harmonic content with straight ac power, a poco or household issue? Are you sure its not PSU noise related?

You always get twice the AC heating frequency as the main noise component with DHTs. And its independant of frequency used. I've heard it explained as a 2H product.

With lower voltage (eg going from 5.3VAC to 4.8VAC) there is a corresponding drop in output hum.

 

[Tweeker]

I suppose cancellation with a prior DHT stage is possible. Itd have to be low mu, or lower voltage filament.

 

[Nature B0y]

The hum pot doesn't fully account for the hum mechanisms which exist in a DHT. One of the mechanisms this is not fully accounted for is the change in emissivity due to the change in voltage across the heater/cathode. Since the relationship of the emissivity of the tube to the heater voltage depends on the magnitude (and not the sign) of the voltage across the heater, an effect similar to full wave rectification is realized, where the fundamental of that hum mechanism is 120 Hz and there is a harmonic component at each 120 Hz upward from there (when the heater is fed from a 60 Hz source).

There really shouldnt be much harmonic content with straight ac power, a poco or household issue?

There's plenty of harmonic content created by AC heaters. Do a simulation of a full wave rectifier. The 5th harmonic (at 600 Hz in this example) is somewhere around 30 db down from the 120 Hz level.

Also, check out Steve Bench's work on DHT hum cancellation here. In addition to the hum pot, notice how he's using solid state rectification to create a waveform that exactly like the voltage across the heater rectified. Then, by feeding a small amount of that rectified waveform into the grid, the wave is effectively inverted, and the hum is considerably reduced. As always, his descriptions are awesome, too.

 

[andrew_whitham]

Ok then here's the DC heater circuit I've come up with. nothing shocking I'll grant you. I've actually had equal performance with the LM338T but a higher input voltage is required.

I can see no noise on the 'scope! (well OK at 5mv a division theres some blurring of the trace, no more than 1mv I guess)

But the problem I have is Heat from the regulator, can anyone say what kind of heatsink is needed? I'm using a small bit of aluminium extrusion but its too hot to touch. It'd be nice to know I'm not far of the mark in requiring a bigger heatsink.

should also point out I'm messing with supply voltages, using the LM338 with 9VAC was best since my 6.3 transformer sunk to 5+ v under load...

Cheers

Andy

edit: that LT7805CT should be a LT1805CT...

 

[Sheldon]

But the problem I have is Heat from the regulator, can anyone say what kind of heatsink is needed? I'm using a small bit of aluminium extrusion but its too hot to touch. It'd be nice to know I'm not far of the mark in requiring a bigger heatsink.

You don't say what the voltage drop across your regulator is, but if it's around 3 volts and your tube draws 1.5 amps (figures just guessed at), then you need to shed 4 - 5 watts. That will be hard to do with just a small piece of aluminum extrusion. However, if you mount that extrusion to your metal chassis with some thermal grease, you should have plenty of area to radiate away the heat.

Sheldon