Do they make resistors in values over 10M?

Yes, they make resistors of most any value, even into the Giga Ohms. Resistors beyond 22Megs are not common and are not used in everyday electronics that's why you don't see them in Jameco.

Resistance values beyond 10 Meg are usually constructed differently (long and slender and/or glass encapsulated) than standard resistors to prevent contamination of the package with dirt, moisture, etc from affecting its intended resistance value.

Also keep in mind that circuit boards requiring such resistors must be free of rosin or other materials to keep conductive paths on the board itself from affecting the resistance value required in the circuit.

Dorian

Yes, you can get them from Farnells and RS to name just a couple.

Plan on putting thousands of volts across these do you?

You can get them as small as 0805 SMD size from Farnell, 2M2 to 50G!

High values resistors like this are important for many circuits and are designed in, they aren't just hacks.

Dave :)

I don't think high resistance will be easier to find in higher power ratings, 1W into 10M is more than 3100V and I don't thing many people need that.

The large-valued resistors I have used were up to 100 Meg, in sealed glass tubes. As others have noted, you have to be pretty careful to actually *get* the indicated value in your circuit, due to leakage. I used to mount them up off the board slightly (some folks used teflon standoffs, but I just made the leads a bit longer). Cleaned the board and resistor after assembly using a solvent to remove handling oils, etc.

The circuits I have seen and used these in are current to voltage electrode preamps, and especially photodiode preamps.

Best regards,

Bob Masta dqatechATdaqartaDOTcom D A Q A R T A Data AcQuisition And Real-Time Analysis

Digi-Key has Ohmite brand metal oxide film resistors to 5 gigohm (5000 Megohm). I've used resistors as high as 100 Gigohm in photodiode transimpedance amps. Amazingly, such high values are available at 1% tolerance. You can probably series a bunch of smaller values for many applications.

Steve Noll | The Used Hi Tech Equipment Dealer Directory: |

As a side note a lot of lower end DMM's will read funky at 10 MOHM and above as this is approaching or exceeding their internal resistance.

phaet> Just out of curiosity. Occasionally I run into a schematic on the web

Thanks for the pointers and tips, everyone. Also, thanks to Dorian for the heads up on 'leakage' across the board.

Sometimes I think it'd be wild to throw a 50Megohm+ resistance in the feedback loop of an op-amp in amplification mode. In theory I could get some wild clipping that way, you know, with gains approaching 5000 or more. I suppose you get to a point where it starts looking like an open circuit though, and the opamp refuses to do anything.

Abusing components is fun!

-phaeton

It may depend on the op amp, but you ahve what is open loop and closed loop gain. If you remove the feedback resistor you get the open loop gain. This is the maximum gain and it is usually much greater at low freqeuncies than it is at the higher ones. Feed the opamp with a small ammount of signal and it will generate some nice square waves or similar without the feed back resistor. Capacitors and diodes can be added for other effects.

I've tried that (no feedback loop connection at all) with a few (741, TL082, JRC4558D) but iirc all 3 of them fell completely silent when I disconnected it.

I must have been doing something else wrong. I'm also starting to think one of my breadboards is flaky. :-(

-phaeton

That's because any real op-amp has a finite offset voltage in the range of microvolts to millivolts, and as soon as you go open-loop the gain in the ballpark of millions immediately slams the output into one rail or another.

You can use an op-amp to make a rather louser comparator by running it open loop (it'll just be really slow compared to a proper comparator).

Ahh... that makes sense.

the op-amp won't "refuse" it'll do its best to comply - even without feedback those things don't have infinite gain.

check the data-sheet for the "open loop gain" that's what you'll get with an open circuit.

If you want clipping, a pair of antiparallel diodes in the feedback loop can do it for you without sacrificing as much frequency response as rail-to-rail clipping does.