cool resistors

formatting link

They will sell us a 22 Meg, 0805, 1%, 25 PPM "fine film" resistor.

We have a silly requirement to increase the input impedance on one of our digitizers, above the normal 1 Meg, to prevent wire loss errors. They are running #16 wire.

Reply to
John Larkin
Loading thread data ...

I looked at that series just lately (50 M for a nanoamp front end) but they were thick film, super expensive even in the low-spec version, and there was no distributor stock for the high-spec version.

I settled on putting 5 of these thin-film ones in series: RES SMD 10M OHM 1% 1/8W 50ppm THIN 0603 MCT06030C1005FP500 Vishay $0.06120 @ qty 100 (Digikey)

as opposed to

RES SMD 50M OHM 1% .2W 100ppm THICK 0805 HVCB0805FDD50M0 Stackpole $3.00 @ qty 100 (Digikey))

The Stackpoles have a very loose linearity spec unless you get the exotic ones.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Sadly, we don't have room for 5 resistors. And 25 PPM is easier on our error budget.

High-ohm precision resistors are a nuisance.

Reply to
John Larkin

Do check the linearity spec, though--that was enough to really make a mess of my error budget.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Plus, at 22M, you could skate by with two 10M 0603s at 12 cents, vs one 0805 at 3 bucks.

A layout would have to be pretty tight to make that a good trade.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

On Fri, 4 Sep 2015 17:27:57 -0700 (PDT), Phil Hobbs Gave us:

Or four sixes or fives. The accuracy actually goes up in such instances, especially if matching and culling techniques are implemented, and high precision segments of circuit assemblies rightfully demand that anyway, so no auto-pick and place from a reel.

Also do not mask between terminations and that area gets cleaned better and leaves a nice air gap to remain clean. We even incorporated grooves cut on the top PCB layer on some of our stuff.. Especially if HV was involved.

Reply to
DecadentLinuxUserNumeroUno

Is that about the point you carve up the PCB with slots and guard traces?

Or carve up your hair into ragged piles... :)

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

#16 wire? 1Meg? Charge them 10X normal for the stupidity.

Reply to
Robert Baer

Stack them? -or- Mount on side? -or- Vertically like shape of "M"?

Reply to
Robert Baer

I dunno, Robert. After all, #16 has about 4 ohms/kft of resistance. So, all it takes is 250,000 feet to cause a .1% error. ;)

Reply to
John S

On Sat, 5 Sep 2015 01:35:41 -0500, "Tim Williams" Gave us:

Only if it is HV. I have plenty of HV designs with slots, and guess where the potting goes? And guess what that provides.

Your only hair loss is due to you lack of understanding about leakage currents, much less the use of slots to abate them.

Reply to
DecadentLinuxUserNumeroUno

On Fri, 04 Sep 2015 23:56:28 -0700, Robert Baer Gave us:

Resistors are not charged.

The wire is to use its skin effect likely.

The circuit only carries the current it carries. There is no 10X normal.

Reply to
DecadentLinuxUserNumeroUno

That's a healthy premium. I normally only charge 20% more for stupidity (mere ignorance is free of charge).

John, have you tried Ohmcraft? They're somewhere in the Rochester, NY area.

formatting link

--sp

--
Best regards,  
Spehro Pefhany 
 Click to see the full signature
Reply to
Spehro Pefhany

Apparently the requirement for hi-Z input comes from the military customer. Nobody wants to mess with the military. Or reason with them, either.

That's interesting; I'll have my guys check them. Thanks

Reply to
John Larkin

If the resistor is just for DC bias, can you drive the other end of the resistor from a nearly-unity-gain buffer? ("Bootstrapping")

What I mean is, instead of putting a 20M resistor from the input to ground, put a 1M resistor from the input pin to the output of a 0.95x buffer that has its input connected to the input pin. That way, the voltage across the 1M resistor is 1/20 of the input voltage, so it passes the same current as a 20M resistor to ground would. You'll see the same DC input resistace, provided you test it with the power to the op-amps turned on.

The above is hardly worth it for 20M input resistance but can be useful if they ask for gigaohms.

Chris

Reply to
Chris Jones

It's the front-end resistor in a voltage divider. And the customer only wants about 20 boards, so we're not inclined to redesign it.

formatting link

The voltage divider is relay-switched in for ranges above +-1 volt, which is all the ADC can handle.

We have bootstrapped the supply rails of low-voltage chopper opamps as followers, to get super high input impedances with high voltage swings. That is sort of a horror.

Reply to
John Larkin

I have been tempted to do that with ADA4528-1 opamps before, as they are very good apart from the low supply voltage limit. Fortunately there is now the ADA4522-2 which runs off +/-27.5V, and from the datasheet it looks nearly as good in other respects.

Chris

Reply to
Chris Jones

...like users are going to connect with cables that long??

Reply to
Robert Baer

You give resistors away for free? No charge? Interesting. Send me a million of them.

Reply to
Robert Baer

You have misunderstood the entire post context and meaning, as usual. So, no surprise why you are labeled as the infamous AlwaysWrong. Can't you find another forum for your stupidity and outbursts?

Reply to
John S

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.