Calculate minimum load current of choke?

[pchw]

Hi,

In MJ's book, the minimum load current of a choke input supply
is roughly V/L, and let say, I have
- a 275-0-275 transformer.
- 2 chokes, 10H and 15H
- full wave bridge SS rectifier
- EZ81 tube rectifier

Question 1, use the SS bridge with LCLC configuration, with 10H as the 1st choke. Using the V/L formula, the current must draw at least 450/10 = 45ma to operate correctly. What about the 2nd choke?

Question 2, will use EZ81 instead of SS rectifier be any different in the calculation?

Question 3, how about the cap input case as CLC, is the V/L formula still good to use?


Thanks!!

[cbutterworth]

I have had the same issues with my own PSU. You're correct about the calculation it is roughly V/H for a choke input filter. However, you can use a "tuning" cap before the first choke. From what I can gather, a cap of at least 0.68uF will begin to be cap-input like, but it doesn't seem to be an instanteously thing.

My own PSU uses a 275-0-275 into a 5V4; tuning cap; 10H; 60uF; 10H 80uF. My preamp draws around 24mA and B+ sits at between 280 and 250V depending upon the size of the tuning cap. With no tuning cap, I get low B+ (240V) but ripple, lots of it. With the 0.68uF, hum is much reduced, but audible.

With the present configuration of 2.5uF for the "tuning" cap, I get very low hum (still audible) and 348VDC. In other words, the filter begins to change from being an LCLC into a CLCLC filter. When I model differences with PSUDII, changing out the cap values from 60uF to 80uF, then with a final 180uF cap gives extremely little ripple, but I have yet to make those modifications.

Anyway, you should experiment with PSUDII. You can download it from http://www.duncanamps.com. It is free software!

It runs in Windoze, which means that I either have to run it at work, or wait forever for it to run on my Pentium 90 laptop. If you're using full wave recitifcation, you'll enter 275 into the offload voltage calculator, then the current rating of the trafo and the regulation (I use 10%), it gives and offload voltage of somewhere like 302VAC. Estimated values and empirical values differ by a few volts and are well within 5% (a good tolerance for a scientifically based person). I cannot measure ripple empirically, but assume that if the B+ is close then ripple shouldn't be too far off.

Anyway, experiment with PSUDII. I would guess that the 15H should come first as it'll likely have the greatest unfluence on the amount of ripple. Let me know what success you have, as I may change out both my chokes for 15H.

Best of Luck,

Charlie

[pchw]

Hi Charlie,

I have been using psud2 to model the PS, but some models I simply couldn't figure out how to use psud2. For example, start with
a LC input section than split the rail into left and right channels B+,
each channel then has another L'C' section. What puzzles me is, what should be the V' used for calculation with L', the rms V of the transformer secondary, or the voltage after the first LC section. I suspect it would be the later because the first LC section reduces the
ripple significantly. Another mystery to me is whether the same formula holds with tube rectifier. Take your 5V4 as an example, the 5V4 can only deliver 240VDC in choke input, so to calculate the minimum load, should it be 450/L or 240/L or neither. In your latest set up, it is very much a cap input case, something like CLCLC. So, what should be the V used to calculate the load drawn for the first L and the 2nd L? Sorry for being so grumpy, just try to figure out how to make good use of a few old chokes in the parts bin :-) I actually don't have any 15H as in my theoretic example. The highest inductance is only 10H..............

Thanks!!