In your example, f x C x Rl = 0.02 24. The ratio this implies is 0.02 1Vripple : 1 24Vdc. For a 400Vdc supply, that would be 8V 16.7V of ripple (I believe, but not certain that is Vp-p). There are a lot of simplifying assumptions in this, but useful nonetheless. (This is a 4.2% ripple factor.)

(Edit: Thanks for @Phyrgian77 for pointing out I forgot a factor of 2 for full-wave rectification. My numbers work for half-wave, but for full-wave, the ripple would be 8.3V and the ripple factor would be 2.1%.)

@andrewRneumann I see what you're getting at now. Since resistance and current are inversely proportional, and Merlin gives the following formula for finding capacitance (C) from peak to peak (Vpp) ripple voltage, given a DC load current (I) and mains voltage frequency (f):

@andrewRneumann I see what you're getting at now. Since resistance and current are inversely proportional, and Merlin gives the following formula for finding capacitance (C) from peak to peak (Vpp) ripple voltage, given a DC load current (I) and mains voltage frequency (f):

Yes, exactly. I = Vdc/Rl. Substitute that in for I and you will get 2 x f x C x Rl equalling the ratio of Vdc : Vpp (ripple). f in this case is mains frequency, so the factor "2" is in there for full-wave rectification. My above example forgot to take into account the 2, so it is off by a factor of 2. I will fix it... again... *hangs head*.

Yes, exactly. I = Vdc/Rl. Substitute that in for I and you will get 2 x f x C x Rl equalling the ratio of Vdc : Vpp (ripple). f in this case is mains frequency, so the factor "2" is in there for full-wave rectification. My above example forgot to take into account the 2, so it is off by a factor of 2. I will fix it... again... *hangs head*.

Well, I'm so used to seeing the formulas for capacitive reactance, that I was wondering why it excluded angular frequency. My brain was stuck in the "why isn't that 2ĻfC?" box.

Sorry, professors, I got lost, so I'm gonna apply my high-school methods; should be worth at least a laugh.

@joulupukki wants to build a Robrob Blackvibe around an Antek PT? And Rob specs 465V B+, running a 355V PT through a GZ34 into a 4000:8K OT?

Let's try some Antek PT secondary voltages in my favorite B+ calculator. As our brave OP noted in post #32, Antek makes a 100VA 350V PT; that would give nominal B+ 473 (with a choke). Close. (The next smaller Antek 100VA, at 300V, would only get nominal B+ to 405V.)

The calculator estimates DC current draw at ~250mA. And if I look at their load test...

I might guesstimate working voltage at ~340V running .25A. At 340 V, the calculated B+ runs around ~460V. Really close.

One thing I don't know, though. Is 100VA enough to supply this current at this voltage? Professors, if you can stop laughing, throw me a rope.

@King Fan I'm pretty certain that 'calculator' is wildly inaccurate. Merlin's method isn't that complicated.

It seems like you're talking now about a pair of 6L6GCs when the OP was talking about building one with 6V6GTs. Nevertheless, let's do a quick comparison.

Take a look at the Hammond 290UX. That's the Blues Deluxe transformer designed for a bridge rectifier and a pair of 6L6s. Let's compare that to the Antek because comparing center tap transformers designed for a standard full wave (non-bridge) rectification with non-center tap transformers is comparing apple to oranges.

Quick reference of specs from the 290UX datasheet we need to figure this out.

Primary Voltage (Vpri): 117
No load secondary voltage (Vsec): 331.3
Primary DC resistance (Rpri): 4.14
Secondary DC Resistance (Rsec): 44.94

I'll go through this again. First, (Vpri/Vsec)^2 is the impedance ratio.

(117/331.3)^2 = 0.1247179178

Rpri/0.1247179178 gives us the reflected primary impedance.

4.14/0.1247179178 = 33.194909544

Now we add that in series with Rsec.

33.194909544 + 44.94 = 78.134909544

That's the source resistance (Rs).
Rs = 78.135

Now, Antek doesn't give you all of the same specs to do the above calculation. However, they do give you the no load and loaded voltages. So, we can use Ohm's law to figure it out.

AS-1T300
310V @ 0A
281V @ 0.47A

310 - 281 = 29

So, 29 volts dropped at 470mA.

29/.47 = 61.702127659

Rs = 61.7

This is lower than the Hammond 290UX. That automatically tells you it's suitable for a similar load.

Ah, OK, I missed out on the 6V6s. IME, though, that calculator is surprisingly accurate. It predicted in advance the B+ I'd get on my 5G9 and my Bassman Micro within a few volts, and used in retrospect, comes really close on my other builds too.