Double the Inductance

[Lee S]

Hi All.

I am about to begin the build of the Loftin White amplifier. As usual I have to start analysing things and poking around before I have even started. The original circuit diagram shows the PSU as a 800V ct trafo into a GZ34 and then straight into a RCLC arrangement of 47R > 40uF > 10H > 100uF. Now, I have seen other amps and other circuit diagrams with a sort of twin LC arrangement after the first cap. Basically there are two chokes and two second caps. How would I work out what chokes I need to perform this? Obviously if I am splitting the supply, Kirchov's and Ohm's law suggests that I need to half the current capability of each choke in relation to the original choke (so 2 x 100mA chokes instead of 1 x 200mA choke?), but what of the inductance? Isn't inductance in parallel the same as resistance in parallel, meaning that to get the same 10H as the original circuit suggests, I need to get 2 x 20H chokes for each LC section? Does it work like this? Is there actually any proven advantage to having two LC stages like this or would the one work OK?

I have a 100uF + 100uF as second cap, so can accommodate 2 chokes if need be, but I realise that using 2 x 100uF caps in parallel after the chokes equates to an overall second cap size of 200uF. Is this problematic? I have plumbed all of this into PSUDII and it all works and seems OK.

Any thoughts, opinions or advice?

Thanks,

[andrew Ivimey]

My 'original' circuit shows 365-0-365. Perhaps we need to touch base here.

You could argue it doesn't matter as long as you achieve 450volts HT but at least, we ought to agree on first principles and the PSU is first. Because you can tune the PSu any voltage should be more or less easy to get.

I think I used HT windings as above into GZ34 to 20mfd solen to 10Henry choke to 200mfd elna cerafine; it worked for me.

(Philip Ramsey finished things off by monoblocking, by the way)

[Lee S]

My 'original' circuit shows 365-0-365. Perhaps we need to touch base here.

Hmmmmm... My diagram just shows a 800V ct. It's odd you say that though, as David Counter said that Philip at Bluebell recommends a 375-0-375 trafo. The Hammond 374BX. I was going to get the 378X. Like you say Andrew, as long as the 450v is arrived at, then no one is going to complain. I don't really want to sacrifice any more filtering and was guessing that the bigger trafo would have an easier time of things for an extra tenner.

I am not going to monobloc as integration is one of the criteria for this project, but I am surmising that the PSU will be plainly heard on a SE DC amp of this simplicity and would like to get it right. However, I will be breadboarding the first couple of incarnations so hopefully will have chance to play before finalising things.

Cheers,

[david C]

as you say Lee that original schematic does show 800v ct,

all I can think of is that Philip changed his mind and didn't alter the drawing,
my one works fine with the 750v ct tx,
my power supply is 47uf> 10 henries > 100uf b'gate,

I have a feeling that Philip did a new ps, splitting after the first cap into 2 chokes and 2 caps

[andrew Ivimey]

And we all know that Philip's circuit was published a good many years ago in a Japanese journal don't we, but I don't think I have the original.
I have the Hammond transformer you suggest Lee and I thought it was 365-0-365, 'snot important.

More to the point I always thought it was lacking a little in the current supply so if we were to think more power, add another rectifier even, then things would have the right level of welly potential.

I am sure that the last time I spoke to our mutual chum about this he actually did a full monoblock, each mono amp having a 374BX trafo all to itself.

onwards!

[Lee S]

I have a feeling that Philip did a new ps, splitting after the first cap into 2 chokes and 2 caps

Yeah.... You mentioned this before David. I have actually mailed Philp and asked him what his thoughts were. Maybe he will get back to me with some info.

Cheers,

[david C]

Yeah.... You mentioned this before David. I have actually mailed Philp and asked him what his thoughts were. Maybe he will get back to me with some info.

Cheers,

to be honest Philip's much easier to contact on the phone, he's sometimes too busy to answer emails

[andrew Ivimey]

Ha Ha, too true.

And Philip will always phone you back, once he knows you are (or is it, have) bona fides.

One reason, surely, why he doesn't join a BB, is that we would all bombard him with questions. He would never get on with his own life.

[Lee S]

So back to my original query then.... What do I pump into PSUDII if I want to have a twin LC stage after the first cap? Would it be half the inductance and twice the capacitance if I used two 10H chokes, thus giving me an equivalent single LC of 5H and 200uF?

[Nick]

Well, the simple answer is you put in whatever gives you the result ypu want.

Its more complex as the TX and cap will be supplying both stages, but you are only modleing the one. You could try a current tap after the first cap to approximate the other leg.

What I would do is use the same inductance and cap in the two legs as was used in the singlecase, but expect a slightly higher voltage because half the current is going through the same choke DCR.

you could estimate the difference by using ohm's law, the average current and the DCR of the choke.

[andrew Ivimey]

Why, it must be asked, Lee, do you want two stages of LC?

It really isn't necessary!

I once did something like this with a 300B amp (though the type of valve is irrelevant). I choked it to death and yes this could have been related to, though not the single cause of, a very dull sounding amplifier. Then step by step I took the chokes and caps off. The amplifier came alive again and until I took the last choke off the hum was not noticeable. I didn't have a scope in those days and I am sure I would have been able to see the rise of the hash but I couldn't hear it.

Having said that, if your gonna your gonna and Nick's advice is sensible.

[Nick]

In the process of messing about with power supplies, I have come to the conclusion, that there may be hidden downsides to two or more stages of LC smoothing, remember that the energy starts flowing in both directions, and there can be strange phase responses goining on at the LF end. Agreed that it reduces ripple, but its not always a free lunch. Andrews findings would seem to match that.

Of course all that is entirly speculative, and just my musings, so could be total tosh

[Clive]

I believe Lee is thinking about 1 LC per channel, similar to the Bugle PS, not LCLC. Maybe that's more likely to be good?

[andrew Ivimey]

Ha ha, I did Bugle with both (as per schematic) and with one enormous 10Henry350ma choke that I had to hand while Philip sent me the proper chokes.

I am not sure the 350ma elephant wasn't quieter.

But I'm using the right chokes now.

[Lee S]

Hmmm... Consensus of opinion here seems that one choke is as good as two smaller chokes. It just made me wonder why some designers and manufacturers used the two choke solution and thought that it may be beneficial to use one choke per channel. I will just get the single big choke for now and see how it goes.

Cheers,