One ohm resistor for measuring bias question

[dbishopbliss]

In the instructions for adding one ohm resistors to measure the bias voltage Doug says, "All you need is a multi-meter that reads milivolts."  My multi-meter is probably 15 years old digital model from Radio Shack.  Simply put, its no Fluke.  I have no idea how accurate it is when measuring milivolts. 

Then it occurred to me... if the current was the same (e.g., 35 miliamps), could I install a 1K resistor with a voltage drop of 35V across the resistor?  I'm sure my multi-meter is accurate in the 10s of volts range and it would be pretty easy to remember that 35 Volts corresponds to 35 Miliamps.

Would this work?

[sluckey]

Then it occurred to me... if the current was the same (e.g., 35 miliamps), could I install a 1K resistor with a voltage drop of 35V across the resistor?  I'm sure my multi-meter is accurate in the 10s of volts range and it would be pretty easy to remember that 35 Volts corresponds to 35 Miliamps.

Would this work?

NO!

[stratele52]

Check some resistor value on your bench and look how is accurate , resitor can have 10% precision . If meter reading is  right , probably same for mili volts and volts too.  You can read a new 9 volts battery , it give you more than 9 volts = 9.3, or very close .

There a lots of people here on the forum doing some work on their amp without knowing what is their meter .

[Willabe]

Check some resistor value on your bench and look how is accurate , resitor can have 10% precision.

Yes, plus 100!

Put the 1ohm R's in anyway. If your meter can't read them they wont hurt anything. Then try reading the bias when your finished with the build and see if your meter will read them.

If your meter can't read them, then when you get a new meter someday they already be there.

You can't put a 1K R in place of the 1ohm R because that would cathode bias the amp and you already have negative grid bias for that amp. Some guys use a 10ohm R instead of a 1ohm R IF their meter won't read the 1ohm bias R.

I think your too worried, try the 1 ohm first and see what happens.

I used a cheap RS digital multi-meter for a good 10 years and had NO problems with measuring across a 1ohm R or anything else for that mater. There's a good number of well experienced builders/repair guys here that use a cheap run of the mill RS DMM (or some other brand) and they have said it works fine for them.

Flukes are great but you don't have to have one to get the job done. I finally killed my RS meter and bought a Klein DMM at Home Depot and it's fine.

Besides lets say you spend the, what $350.00 on a great Fluke. Since your fairly new to building/repairing and you make a mistake and burn up the new Fluke, now what? Cut your teeth on a $50 to $100 DMM, that way if you kill it way less painful.


              Brad        

[G._Hoffman]

Check some resistor value on your bench and look how is accurate , resitor can have 10% precision.

Yes, plus 100!

Put the 1ohm R's in anyway. If your meter can't read them they wont hurt anything. Then try reading the bias when your finished with the build and see if your meter will read them.

If your meter can't read them, then when you get a new meter someday they already be there.

You can't put a 1K R in place of the 1ohm R because that would cathode bias the amp and you already have negative grid bias for that amp. Some guys use a 10ohm R instead of a 1ohm R IF their meter won't read the 1ohm bias R.

I think your too worried, try the 1 ohm first and see what happens.

I used a cheap RS digital multi-meter for a good 10 years and had NO problems with measuring across a 1ohm R or anything else for that mater. There's a good number of well experienced builders/repair guys here that use a cheap run of the mill RS DMM (or some other brand) and they have said it works fine for them.

Flukes are great but you don't have to have one to get the job done. I finally killed my RS meter and bought a Klein DMM at Home Depot and it's fine.

Besides lets say you spend the, what $350.00 on a great Fluke. Since your fairly new to building/repairing and you make a mistake and burn up the new Fluke, now what? Cut your teeth on a $50 to $100 DMM, that way if you kill it way less painful.


              Brad        

You really don't need a $350 Fluke for amps.  You can get one that does everything you need, with real Fluke quality, for $150 or so.  Or at least, you could a couple years ago when I got mine.


Gabriel

[PRR]

Digital milti-meters read MilliVolts! The basic scale is 200mV. When you read higher voltages, you DEcrease accuracy with a voltage-divider of maybe 2% resistors.

This isn't a moon-rocket. If the DMM reads near 1.56V on a fresh flashlight battery, it is way more accurate than tubes need.

My 30 year old RadioShed DMM is up to the job.

> could I install a 1K resistor

Think. You apply, say, -45V to the grid for bias. Now you jack-up the cathode "by 35V" to measure current. The tube now sees 80V of bias! Whatever you measure this way is much less than it will run in normal operation (without the 1K). If you can even get 35mA this way, it will probably smoke when the 1K is removed.

You can't leave 1K in there. Take a 5F6a with 4K CT OT. That's 1K load per plate. Add another 1K per cathode, you can only get half as much power INto the amp, and only half of that can get OUT. So a 40W amp becomes a 10W amp, with 10W useless heat in cathode resistors.

Not a dumb idea. Back before DMMs, 35mV was awful small. We used 10 ohm resistors. The 0.350V extra bias was usually "small" compared to typical 20V-40V biases. But you can read 0.35V on a 1.5V scale. And 10 ohms is small compared to typical plate loads.

But if your DMM isn't sick, 1 Ohm works just fine.

[dbishopbliss]

Thanks for the info and advice.  Makes sense now.  Just goes to show... there's a difference between knowing ohms law and knowing how to use it. 

[12AX7]

I always had a question about this too, but a little different than the OP's. I worried that the chances a couple 1R resistors would be close enough in tolerance to insure accuracy from side to side in my dual bias amp was unlikely. In other words, i figured unless they were exactly the same, you could have one side reading a lot more or less than the other side when they are actually drawing the same. So when i match the bias on both sides i could actually be setting them rather far apart. Not sure if my theory falls short or not in some way, i really hope it does ! but to insure things were as accurate as possible i used 2 big aluminum cased dales that were like 20 watts. Not for the wattage of course, but dales are pretty precision plus i figured the bigger the actual resistive element the more likely the precision would be better.

[stratele52]

12AX7

When you write  ". ...used 2 big ....20 watts  ... the bigger resistive element...."


What do you mean ? Powerfull resistor ( in watts )  or more/ higher resistance ?

IMO if you talk about more powerfull ( watts) this is not mean more précision on reading . 

But if you use let say a 100 ohms resistor instead a 1 ohms you'll have less error in reading .

.05 volts / 1 ohms = 50 ma 
.05 volts /  1.2 ohms =41 ma

.05/ 100 ohms = .05 ma
.05 /102 ohms = .049 ma

The usefull thing by using 1 ohms precison is I let them in circuit for ever , no need to removed them.

[sluckey]

I worried that the chances a couple 1R resistors would be close enough in tolerance to insure accuracy from side to side in my dual bias amp was unlikely.

Not if you use 1% resistors. 1Ω 1W 1% resistors are tiny and cheap too.

Those 20 watt aluminum resistors are way overkill in a 1/2 watt circuit. What's the tolerance on those Dales?

[12AX7]

I can't remember. It's been a long while, i just remember that was all i could find locally that would seem as accurate as i was gonna find around here.

[HotBluePlates]

... But if you use let say a 100 ohms resistor instead a 1 ohms you'll have less error in reading .

.05 volts / 1 ohms = 50 ma 
.05 volts /  1.2 ohms =41 ma

.05/ 100 ohms = .05 ma
.05 /102 ohms = .049 ma ...

But remember, in the way the 1Ω resistor is being used, voltage is not fixed; rather, it varies based on the tube current.

Also, your example was skewed by the difference in resistance you chose. The "1Ω" resistors are 20% apart, while the "100Ω" resistors are 2% apart. As a result, the current differences are 20% and 2% apart, respectively.

So the accuracy of the measurement still follows the basic tolerance of the resistors. And even 10% tolerance parts (or 5%, or 1%) routinely measure closer than their marked tolerances.

Lastly, any given meter might have a basic accuracy poorer than 1% on its d.c. voltage function.

So PRR already covered why older equipment used 10Ω resistors for this function; namely, that the lowest-voltage range on service-grade meters (often 1.5v full-scale) might have difficulty resolving 0.035v, while 0.35 (due to 35 milliamps through a 10Ω resistor) was easy to read. This is rarely a problem with modern meters.

[G._Hoffman]

1% is 1%. 

At 1Ω, that should be ±0.01Ω, which is probably finer than your inexpensive meter can reliably resolve.  Not really much of a concern, particularly given your phase inverter is probably less symmetrical than 1%.

Higher powered resistors MAY give you a better Temperature Coefficient, but the two resistors should change similarly with temperature, as it is unlikely one will be significantly hotter than the other.  A 5% 1Ω carbon film resistor (or at least, the one I just looked up on Mouser) is +350 PPM/C to -450 PPM/C (that's Parts Per Million per degree C).   So, raise the temperature by 20 degrees, and you have ±0.007Ω (give or take).  Which is finer than most $350 Fluke multimeters can resolve reliably.   A 1% 1/2 Watt Metal Film resistor is more like ±50-100 PPM/C.

Oh yeah, and the best guitar sounds you've heard?  The tubes weren't matched, they only had one bias pot, and they were probably using 10% - or even 20% - resistors!  There are some things where you really don't need to worry about it. 

But if you are really concerned about getting those resistors as close to identical as possible, spend $5 on 100 1% 1/2W 1Ω resistors, and sort them for values.  They will probably all be in the 0.991Ω-0.997Ω range, as the ones better than that will have been sorted at the factory for their 0.1% range, and they seldom seem to come in higher than spec (at least for the cheaper brands - I'm told this is less true for the real name brand parts).  This will be cheaper and more effective (at getting the resistors values to match) than buying just about any two higher rated resistors, unless you plan to buy 0.1% resistors at $10-20 each!!!


Gabriel