EMRFD Message Archive 9500

Message Date From Subject
9500 2013-12-03 17:46:50 Jim Miller Interesting Linear tech switcher
9554 2013-12-15 09:08:19 drmail377 Re: Interesting Linear tech switcher

I've looked at this part. It is interesting but the nearly impossible to prototype 20-Lead 3mm × 4mm QFN package (with bottom leads and difficult footprint) is a Major put-off. Also, the reference jig has no external FET or transistor, this is typical for "low EMI" parts. Again, prototyping with this part is very difficult. I would stick to a reference "demo" board design as a basis, then turn a clone board.

This part seems really targeted at around 12-36VDC in to 5-3.3VDC out. Don't be mislead by "Low EMI" claim thinking this part is a one-stop solution for low noise regulation for communications in the HF range - it isn't. But the part is interesting as a pre-regulator followed by a linear LDO specifically designed for noise rejection. For truly low noise communications applications, especially at HF, you really need an LDO post-regulator with good power supply rejection. This will help to keep your jitter/phase-noise in check.

This part Is interesting as a regulator block for the likes of an embedded Linux/Unix/xBSD ARM board, like from Olimex or the Raspberry Pi. However, this part is a stand-alone switcher with no load cable-loss sense and compensation.

In this embedded ARM SoC application, for best efficiency keep the switching frequency below 1MHz and keep the load between 0.5 and 4.0A. At less than 1MHz switching frequency and 4A output, your selection of the output inductor is going to be a bit tough, and physically bigger than you may like. Keeping the inductor smaller helps reduce EMI. Lots of trade-offs here.

The attractive thing in the embedded ARM Linux/Unix 5V regulator application are no external FET or Transistor and more current at 5V to power external devices such as USB rotating/SSD storage and/or back-lit displays that take more than the 05.A-2A required for the embedded board itself (you will typically need a USB hub with DC pass-through as well).

There's a example jig (8614.asc) for LTSpiceIV in your default installation of LTSpiceIV (e.g., in WinXP: Program files > LTC > LTSpiceIV > examples > jigs > 8614.asc). I suggest moving the 8914.asc jig to a separate directory and run it from there (I rename to something like LT8614_Test_01.asc).

If you don't see the 8614.asc jig file in your installation of LTSpiceIV, update your installation from within the application.

The default jig simulation for the LT8614 demonstrates a 36V input with 5V output at 4A and targets startup behavior. The 5V output stabilizes at 0.5ms from my run of the default jig sim. That's all. No further .measure options or .plt files included. Also, compression is on by default, and even with compression on, the .raw file is around 32MB for just a 1ms startup sim.

There are other parts out there that I've looked at like this part (don't ask me for a list, it's been awhile). But TI/National is a good place to start, although I tend to stick with Linear due to decent LTSpice simulation modules and support.

If you can turn a board to physically use this part (due to very difficult mounting), then this part is very attractive because it looks like the LTSpiceIV simulation is pretty good (I beat it up for an hour or so).

Just my quick take

9555 2013-12-15 12:00:36 William Carver Re: Interesting Linear tech switcher
Demo boards are rarely bargains, although TI has some that are........
I checked for LT8614 demo boards and Linear says "contact factory". I'm
not going to bug then, but if anyone does find out the price of the demo
board it would be interesting to know whether its affordable.

9556 2013-12-15 12:56:55 Nick Tsakonas Active Fet Receiver Front End Mixer
Hello group,

I came across a document titled "An Active Fet Receiver Front End Mixer" by Kenneth Voyce, dated back in 1969 and I though I should share it.It is a 75 pages long document describing balanced FET mixers in detail,a very interesting reading to say the least.

You can download the document from this link

(there is a downloadable PDF under the "Files in this item" section)


9557 2013-12-15 16:57:00 Michael Dunn Fast, high-Z buffer?
Might anyone be aware of any parts or designs that implement a
low-distortion, very fast (>1GHz) buffer/amp, but with a high-R input
(to be suitable as a scope input amp). I'm looking at the schematic
for National's old "Damn Fast" buffer, which may provide a suitable
topology for attempting a discrete version.

Michael (VA3SSP)
9558 2013-12-15 23:41:10 Russell Shaw Re: Fast, high-Z buffer?
9559 2013-12-16 06:55:11 ford@highmarks.co... Re: Fast, high-Z buffer?
I found the PDF paper at the second link to be a fascinating study in wideband RF amps. It would be interesting to build a prototype of the sample circuit using a more modern device.

The book examples are now on my Christmas list but at those prices I'm afraid the coal in my sock tells me they are a bit out of my "casual reading" budget. Several are also exceptionally old and likely to include antiquated design techniques that would lend them only to academic discussion rather than yield practical circuits using modern components.

Where there are nuggets there may well be a treasure trove.I wonder how you found that paper? Where did you look for it?


-----Original Message-----
9560 2013-12-16 07:35:40 Russell Shaw Re: Fast, high-Z buffer?