How can you enhance Sijosae splitter precision without a trimmer?

Dec 14, 2020 49 Replies

If you like using way too many parts. ;)

There's no good reason I can see for doing the supply-pin wraparound trick for a rail splitter. The output is always at mid-supply, so you don't need the RRO capability, and it has serious disadvantages for this use.

Because the output comes from collectors, its output impedance rises rapidly with frequency as the loop gain declines, so that all you've got holding the splitter output still is the capacitor. That rising output impedance looks inductive, so there'll be a nasty resonance with the output cap that you'll have to manage somehow.

Using the complementary follower approach gets you a low output impedance at all frequencies, isn't inductive, and takes fewer parts.

But a TCA0372 and a BFC is the ticket.

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

This is a premature post precipitated by all of Tauno's hard LTSpice work posted in another thread. Your LTSpice circuit in this thread is also appreciated. It makes things easy for me. My ultimate goal is to apply EE theory to analyze the splitter.

Now for some sad news. Although there's a TL084 in my parts bin, there's nary a 2N4401 nor a 2N4403. But, there's a TIP31A and a TIP32A available.

MacGyvering [1] circuits is its own reward. It brings a joy unfulfilled by hurly burly online orders.

The two syllabic second word of "Big Farad Capacitor" captures the rhythm of the BFC better, if you ask me.

Note.

  1. formatting link

Danke,

Don, KB7RPU There was a young lady named Bright Whose speed was far faster than light; She set out one day In a relative way And returned on the previous night.

Plus a lot of rusty bolts and springs, obviously.

Well, if 1/10000 is 'Big', I suppose so. ;)

Cheers

Phil Hobbs

Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

1/100 is /relatively/ big and qualifies as a BFC for me. YMMV.

Anyhow, a Ulyssesian mistake was made when my eyes took a peek at the ballyhooed TCA0372 and saw its 1.0 A output. Its sexy specs siren song got under my skin and made me order it. Regardless, the Sijosae splitter will be built and analyzed while we wait for the arrival of the TCA0372. This thread illustrates how quickly a side project evolves into the main project.

Danke,

Don, KB7RPU There was a young lady named Bright Whose speed was far faster than light; She set out one day In a relative way And returned on the previous night.

One spec to keep an eye on with the tca0372 is the maximum power dissipation. How much current do you need? I made a rail spiltter from the lm675... following the recommended circuit on the data sheet.

formatting link
(fig. 3)

George H. (late to the rail splitting party.)

formatting link

The price is crazy. We pay 44 cents, and we'd gladly pay 10 times that.

You do have to watch power dissipation with the TCA0372, true. On the other hand, you get two amps for half a buck. ;)

The SO16 is about 12 K/W, and a square inch of top-level copper is about

50, so at 25 C you should be able to dump a watt or so, but not much more than that. Of course for the extra five bucks you can buy a couple of 2W zener diodes to put in the supply leads, and still have $4.75 left. ;)

Cheers

Phil Hobbs

Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com

The heat comes out of the 4 big V- pins.

formatting link

Haven't quantified that.

I took a look at the TCA0372 data sheet from On Semi, and I can't quite wrap my brain around Figure 1. It looks to me as if the positive-going side of the output stage has been drawn upside down, resulting in a design which have poor current gain when operating and which would be prone to latch up or (almost) short the rails if the inputs were driving in a hard "go negative" direction.

Am I missing something, or did the schematic's author put in the wrong-polarity Sziklai pair here?

The diagram has been wrong for decades.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required