RE: Differential voltage probes 3

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>

Hi Steve,

At 11:26 PM 6/28/2003 -0500, you wrote:
>Hey Terry,
>
>I know very little about the technology that we use for measurements, but 
>i can already see how this simple probe will be well worth the $45 to put 
>it together!  Ok, so its super fast and clean, and can take up to what, 10kv?

Right now we are thinking just 1kV RMS.  It would have a "safety" margin to 
5kV peak for spikes and other things, but really a less than 1000 volt thing.

>Is it possible to go for 20 or 30kv ratings by seriesing more of these SMT 
>resistors?

Yes ;-))

>  How much bandwidth would you loose?

That is a big question there.  A LOT depends on how good the new Ohmite 
resistors are...

>Im alreade getting sick of my plain old probes, i want something that will 
>be able to measure into the kv without me blowing any more channels on my 
>scope!  And wow, they put it all in a singe 8(assuming from the diagram) 
>pin chip?  Cant beat that, though you better make sure your soldering 
>skills are up to par before you go and solder that one on.  $9 is a bit steep.

That chip is 1/4 the size of your finger nail...  You could hardly read the 
numbers on it...  Soldering this little surface mount stuff is an issue the 
"simple" version was meant to avoid.  But getting to 150MHz...  Sort of 
like soldering fly's legs...  Probably have to "offer" the board pre-built 
and tested since I see what "unskilled" surface mount soldering can "do" 
every day =:O


>Ok, so excuse my ignorance, but how is this probe going to be packaged?
>Will there be a little box tha plugs into the scope, and then a shielded 
>cable with your 2 tips?  Will this cable mud things up?

Not sure yet...  A marketing thing...  Maybe packed with happy meals, or 
that new Harry Potter book :o)))  Engineers just design this stuff, we 
don't "package" it :-D


>Well, it sounds cool!  When its all done in working, im sure at least a 
>few people would like to buy one.  I bet you could make a bunch of printed 
>boards that size for not too much money.  What do you think a final cost 
>may be?  $60?  i guess you need a box and some cable, the cost of the PCB.

I think it would be real cheap. Right now the board would be very small so 
a ton could be packed on one PCB panel...  Not out to make any profit or 
anything...


>I dont have to tell you to let us know how your first setup goes!

I hope to get the parts ordered up this weekend if there are no other ideas 
that would change things.

Cheers,

         Terry


>Steve Ward
>
>
>>From: "Tesla list" <tesla-at-pupman-dot-com>
>>To: tesla-at-pupman-dot-com
>>Subject: RE: Differential voltage probes 3
>>Date: Sat, 28 Jun 2003 19:46:25 -0600
>>
>>Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>
>>
>>Hi,
>>
>>Ok, using bleeding edge, oh!, I mean leading edge parts...  I got it back 
>>down to one IC.
>>
>>http://hot-streamer-dot-com/temp/DiffProbe21.gif
>>
>>The fabulous Ti THS4151ID apparently does all the power supply offsets 
>>and CMRR fixing internally.  That chip is basically the whole 
>>enchaladia!   So I just added Ohmite's little 5kV resistors, some 1% 1K 
>>resistors, some big NPO caps and it's done.  It would almost fit on a 
>>quarter ;-)
>>
>>There are some relatively expensive parts now, but the total cost is not 
>>that much.  The input resistors are $4 each, the IC is $9, the NPO caps 
>>are $2...  But I think the total cost would be about $45.  Yes, it is all 
>>that little surface mount stuff and you have to be careful since the high 
>>voltage is very near.  Probably a good candidate for HV conformal coating...
>>
>>The input resistors are 5%, but hopefully they are somewhat matched as 
>>they come off the reel.  The PCB around those resistors would have to be 
>>routed out to insure voltage standoff.
>>
>>I have no idea what the frequency response would be, but it "should" be 
>>pretty darn high!  The IC is 150MHz and I am feeding it just like it 
>>wants.  With the input resistors being so small, a LOT of response 
>>problems just vanish.  Caps across the 1k resistors could limit the 
>>frequency response if needed.
>>
>>Unless someone has a better idea, I think I will give this one a try.
>>
>>Cheers,
>>
>>         Terry