WTH? NOS CC Resistors don't measure anywhere close to stated values!?!

[Justin Thomas]

My experience with new CCs is that they drift up mightily when soldered and don't recover. Maybe I'm a little ham-fisted in the soldering department, but my CF & MF don't have that problem. And I'm not about to pay for NOS prices - $2/each for 1W? Um, no, CF/MF are JUST fine, for much less. I'll steal CCs from old equipment and if they measure within 20% I'll use them. And even if they drift 25% in a piece of gear that still functions well and sounds good, I leave them in. It's a guitar amp, after all...

Justin

Justin

[R.G.]

And remember: the only place CCs make an audible difference is at inputs, where they add excess hiss and noise, and in places where the audio signal is bigger than 70Vpk-pk. That pretty much means they make a useful difference in the plate resistors of the phase inverter.

And what about current production CCs? Are the problems you all mentioned now solved or they will show again over time?

The production processes are much tighter, so they start out more on the nominal resistance, don't have the middle tested out, and are in general more uniform. I suspect, but do not have data to prove, that the better processes make them less variable. But the problems of resistor change with soldering and lifetime high temps remain, as do the issues with long term drift.

CCs are useful for a different purpose. They have a greater ability to absorb pulse/transient energy without being destroyed, within certain limits. But that's unlikely to be something that's useful in a guitar amp.

[bob p]

I'd like to thank everyone for their ideas on the CC resistor problem, especially El Duderino, whose idea about finger resistance was a very good one. (I was using alligator clips. )

As I went back to taking measurements with my hand-held meters as I watched the Olympics, I learned something interesting...

* One of them started to flake out, by reading tremendously high values, then slowly descending toward the stated value taking over a minute to get there.

* The other started to flake out, by intermittently going back and forth between reading OL and a low resistance value.

With both of these hand-helds misbehaving, I decided to check them on an AC powered bench meter. My Fluke 4-wire benchtop told me that everything read in-spec. WTH?!?

Right after both hand-held meters started really misbehaving, the low-battery indicators came on. So I changed the batteries. Now the readings are normal.

With two different brands of hand-held meters, BOTH meters started measuring high resistances with very poor accuracy when the battery levels started to get low, and the inaccuracy came along HOURS BEFORE the low-battery indicator came on. With both hand-held meters the first sign of a battery problem was flaky high-resistance measurements, not a low-battery warning.

I guess the moral of the story is that hand-held meters suck. I ran into this problem because I wanted to kill two birds with one stone -- I wanted to run through checks on these resistors while I sat in front of the TV watching the Olympics, so I grabbed a hand-held meter and took measurements while I was on the couch. The damned meter kept doing the auto-off thing on me, trying to save it's batteries, and those batteries turned out to be the root of the problem. I never would have had the problem if I had used the bench meter in the first place.

Burned by batteries. Something to think about.

[The Dude]

Glad you got it figured out! I suspected some sort of measurement error- too many resistors- too far off spec. I hadn't thought of batteries, though.

[R.G.]

Burned by batteries. Something to think about.

You don't know the half of it. Go find the video of the Tesla in France on the test drive.

Found a link. https://www.reddit.com/r/teslamotors...customer_test/

During the drive, the car complained about charging irregularities, and the salesman told the prospective customers to get out. It then burst into flames.

Talk about burned by batteries!!

[bob p]

And remember: the only place CCs make an audible difference is at inputs, where they add excess hiss and noise, and in places where the audio signal is bigger than 70Vpk-pk. That pretty much means they make a useful difference in the plate resistors of the phase inverter.

I absolutely avoid CC resistors on inputs, and I only use MF there because they're even quieter than CF.

IMO one of the biggest problems with CC resistors is that today's Chinese CC resistors are exceptionally noisy, and the noise gets worse as the values go up. I haven't met a Chinese CC resistor in the 1M to 5M range that I didn't end up throwing into the garbage. They're really that bad.

RG, are you suggesting that we should be using CC for the screens on our power tubes? If that's the case then guys like me who use MOX and those other guys who use vitreous enamel are doing it all wrong. lol.

[R.G.]

No, not at all. MOX or wirewound VE will be fine. Screen grids don't get much pulse duty. I meant pulse duty like 1000% of rated power for a few milliseconds. For screen grid duty, MOX ought to be fine, especially as MOX is (or was the last time I checked) viewed as the pulse-power replacement for CC.

[Zozobra]

I'd like to thank everyone for their ideas on the CC resistor problem, especially El Duderino, whose idea about finger resistance was a very good one. (I was using alligator clips. )

As I went back to taking measurements with my hand-held meters as I watched the Olympics, I learned something interesting...

* One of them started to flake out, by reading tremendously high values, then slowly descending toward the stated value taking over a minute to get there.

* The other started to flake out, by intermittently going back and forth between reading OL and a low resistance value.

With both of these hand-helds misbehaving, I decided to check them on an AC powered bench meter. My Fluke 4-wire benchtop told me that everything read in-spec. WTH?!?

Right after both hand-held meters started really misbehaving, the low-battery indicators came on. So I changed the batteries. Now the readings are normal.

With two different brands of hand-held meters, BOTH meters started measuring high resistances with very poor accuracy when the battery levels started to get low, and the inaccuracy came along HOURS BEFORE the low-battery indicator came on. With both hand-held meters the first sign of a battery problem was flaky high-resistance measurements, not a low-battery warning.

I guess the moral of the story is that hand-held meters suck. I ran into this problem because I wanted to kill two birds with one stone -- I wanted to run through checks on these resistors while I sat in front of the TV watching the Olympics, so I grabbed a hand-held meter and took measurements while I was on the couch. The damned meter kept doing the auto-off thing on me, trying to save it's batteries, and those batteries turned out to be the root of the problem. I never would have had the problem if I had used the bench meter in the first place.

Burned by batteries. Something to think about.

The battery was low in my meter and before it died I was reading waaaay out of spec line voltages out of my outlet. We're talking 290VAC here....

[Richard]

Is there a useful way to check your meter's accuracy? If I ever get suspect values I will definitely check my meter. As a hobbyist my ways would be to check something of a known resistance or voltage, such as a pickup or a fresh battery. I don't recall having weird readings ever cause problems for me.

[J M Fahey]

Oh, at least with cheap meters, (only kind I use ) , low battery (think 5.6V) does make the meter crazy.

If curious, measure a 10k resistor, even "5%"v ones are usually within 2 or 3%

A fresh 9V carbon battery is around 9.6V unloaded.

But you might as well buy an alkaline and put it inside your meter

Never ever touch the tiny adjustment trimmers inside the meter, you do NOT have an NBS calibrated reference or master meter to compare readings so don´t.

[Dave H]

Is there a useful way to check your meter's accuracy?

I have a few 0.1% resistors I use for that. They are 100, 1k, 10k, 100k and 1M. Sometimes the switches on cheap meters are a problem. I have to rotate the switch a few times before it will read zero with the probes shorted on the ohms ranges.

[Mike Sulzer]

And remember: the only place CCs make an audible difference is ... and in places where the audio signal is bigger than 70Vpk-pk. That pretty much means they make a useful difference in the plate resistors of the phase inverter.

Several of the stages in the preamp chain can be operating at or in excess of that level when the preamp is overdriven. When using a high output humbucker, even the input stage can get close or even over that level. In some circumstances the output stage is not overloaded and the phase inverter plate resistors might only have a few volts swing. So this is a more complicated question.

That said, I do not think it is the CC resistors that make any significant difference in the sound under those conditions. Not when the tubes are overdriven so hard; they are what control the sound.

[Malcolm Irving]

...
Back when the quasi-standard for resistors was 20% and 10% was the premium stuff, manufacturers would try mightily to get a batch of a million resistors to come out with an average value dead on the nose of the nominal resistance. So if you measured a million and plotted their values, you'd get a bell shaped distribution with a peak very close to the nominal. The problem was the width of the bell curve. They could not make all the million come in within even 20%, as the process variations were too wide.

So they tested them all. The ones with values outside the 20% were either tossed into the nominal bucket that they were nearest to (i.e. a 1k at -20% makes a pretty good "810 ohm" resistor - you just have to redo the painted rings) or discarded if no other use could be found for them.
...

I think the idea of the 'preferred value' system was that a resistor that was out of tolerance for one value fitted straight into the adjacent value. E.g. in the E6 range (+or- 20%) a 220 ohm resistor which came out 20% high is 264 ohms, which fits perfectly as a 20% low 330 ohm (which is the next value in the range). That way, in the old days, they could make a resistor, measure its value and then paint the coloured rings on afterwards. That method would imply an 'even distribution' of values within the tolerance.

[dai h.]

re: moisture sensitivity

I've got some old packets of Japanese CC (which come sealed--unlike NOS Allen-Bradley CC), and they have small packs of silicon dessicant inside them.

[Mike Sulzer]

re: moisture sensitivity

I've got some old packets of Japanese CC (which come sealed--unlike NOS Allen-Bradley CC), and they have small packs of silicon dessicant inside them.

I suppose that tube amps using CC resistors would tend to "bake them out", perhaps sounding quite different if left off for an extended time until "re-bake out" was complete. I think of CC resistors as a bad dream, approaching night mare status.