AN: Open Instrumentation Project Progress

Hello Peter,

Just an idea: What mankind really needs is a simple spectrum analyzer for EMC pre-compliance work. Nothing fancy, just something that ranges from 150kHz to 1GHz and can show the ballpark numbers. IOW where people can see how good their chances are before they head off to that expensive lab test. This tool would also be useful when they have come back from the EMC lab with a black eye and must chase those nasty clocks that pushed them over the cert limits.

It doesn't have to be much, maybe a glorified scanner without a front panel but with a USB connection and nifty display software.

Wouldn't a bunch of antenas and an anechoic chamber be necessary too?

Best Regards

Steve sousa

Err... isn't this a job for an old grid dip meter? Speaking of which, why aren't there any NEW grid dip meters? They're as useful a test tool for RF as strobe lights are for mechanics, but all the ones I see are leftovers from 30+ years ago.

Hello Steve,

Not necessarily. When I do field work I mostly won't even have access to a shielded room. There just isn't any on site. Sometimes I rent an antenna and other times I just use my EMCO probe kit.

Much of my EMI diagnosis for clients happens after a failed EMC test. Then I already have a baseline, the plots from the lab. So I know that, for example, 319MHz was 8dB above the line. Now I'll take the analyzer and go after that. Trying to find what generates it and where it comes out.

The first thing to do before any changes to the unit is to log my own baseline. It doesn't matter so much what kind of antenna I use, even if it was a coat hanger from the dry cleaners. It's important to meticulously document positions, distances, angles etc. Also other objects in the room. That's why I take my digital camera if the company allows it.

After finding the root cause(s) and remedies a subsequent scan should show a drastic reduction. Not just 10dB, considering the uncontrolled environment that would too close for comfort, but something well north of 20dB. What I prefer to hear from clients later after the 2nd EMC run is something like "they couldn't see 319MHz anymore".

Nah, those are too flimsy for that. You need something where you can connect a probe or antenna to and where you can narrow the detector BW to 10kHz or so.

Mine is the Heathkit HD-1250. Still works fine and it's probably the most "modern" dipper that was made (came as a kit though). My oldest is a Boonton Megacycle meter. Now that is a true grid dip meter because the detachable head contains a real mini-tube that has a grid. The Heathkit contains a FET.

Why no new ones? Look what comes out of universities these days. Most of those folks can't even solder. With hams there also is a disturbing trend towards not building anything with their own hands.

Looks like you've just volunteered to design one!

You've been conspicuously absent from the "favorite drinks" thread, I must add... :-)

Michael

Hello Michael,

Maybe some day. But I am not a USB expert.

Kind of busy right now and I guess most has already been said. My favorites:

1. Beer (the heavier, darker ales and stouts)

1. More beer.

2. Schnaps (the stuff my wife makes from gin, sugar and blackberries)

Atmel seems to make a series of USB controller chips. I stumbled on them while I was doing research into what microcontroller to use for a first project. I guess they're similar to using a MAX232 for the serial port. (They're supposed to shield you from the messy details of USB.)

You'll master it within minutes, I'm sure.

Take care,

Michael

Yep, seems like a good idea, Joerg maybe this

Warning engineer designed website, may offend many non engineers

martin

Hello Martin,

That is quite a good approach. Nowadays a USB interface would almost be a must as most laptops do not offer a parallel port anymore. Also, I'd probably spend a couple bucks more on the log detector :-)

It's quite good but bogs down older PCs because it is all placed on one long page. Better to split it up and use hyperlinks.

Just wondering...

Having a laptop next to a spectrum analyser will corrupt all/many of it readings, wont it?

martin

Hello Martin,

To some extent, yes. But you'd be surprised how much noise modern and expensive spectrum analyzers let off. Once I was at a client and nobody had a clue where a new forest of noise peaks could possbibly be coming from. There was nothing on the plots from the EMC lab, total silence in that range. So I got some aluminum foil from the cantina and waved it here and there. Surprise: When I held it in front of the $20k+ analyzer's LCD screen the noise disappeared.

One can live with problems like that. Furnace filter mesh, fencing materials, aluminum screen door stock, all that helps as long as you can see the display through it. One of the first questions when arriving for an EMI job at a new client is where the next hardware store is located.

USB and power cables would need some #43 ferrite but almost any EMI job requires to keep a couple of pounds of those at hand anyway.

On Wed, 04 Oct 2006 22:29:07 GMT, Joerg wrote in Msg.

Wouldn't an FT245R from FTDI be an almost drop-in replacement?

robert

What about an ethernet interface..?

Meaning there is actually something interesting to read and no BS ;)

Sort of, I just think that one long scrolling page is a PITA, this one is even worse

martin

Reminds me of a person at uni that didn't know how to use an editor so he wrote all documents in the same file.. until the editor couldn't handle it ;)

Maybe the html author have same problem? =)

Way better..

No

Programming a USB parallel port is not like programming for a hardware parallel port.

Software that uses a hardware parallel port can not just use a USB port.

A total re-write of the I/O in the PC app would be needed.

donald

PS: If anyone has code available to interface to a USB parallel port, please share any links or code.

That would be ok but needs setup. Which means extra work. A well engineered USB device plugs in and the matching software miraculously shows up ;-)

True. But a wee bit of skill application in organizing a document would have been nice.

My site is the other extreme: