Different bias readings at different ground points?

[dude]

I added a bias pot to a Fender Champ SE amp, PCB board. Runs 6L6s, the sockets were PCB mounted with long pins off the bd about 1 1/2".
I installed a bias pot. The sockets (6L6's) were off the bd and had long legs. I cut legs pin 8 and installed a 1 ohm 1% resistor between the cut leg. When I check bias right across the resistor I get a lower mA reading, (about 5 mA) then when I use the chassis for the ground point...?

I can't understand the difference, if it's 35mA right across the resistor, it's 40mA using the chassis as ground.

I wanted to put test points on the back of the chassis and use the chassis's ground but 5mA is a big difference, which is the correct reading right across the resistor or using the chassis as ground...?

Don't think I ever seen this before. Continuity from the cut leg ground point and chassis ground is 5 ohms, is this resistance the difference? Should I run wires from the 1 ohm R, both sides to the test points and not use the chassis ground?

al      

[sluckey]

Directly across the resistor is the most accurate reading. When you use some other chassis ground point there will be additional resistance added to the 1Ω resistor. For example, let's say you have 0.2Ω additional resistance. Well, 35mA flowing thru a 1Ω resistor will give you 35mV. That's what you see on your meter. But, that same 35mA flowing thru a 1.2Ω resistor will give you 42mV.

0.2Ω resistance doesn't seem like much, but when you're expecting 1Ω but really have 1.2Ω, that's about a 20% error.

[jjasilli]

Yes, but is seems something else is wrong.  There should not be a 5 Ohm resistance between the bottom of the leg & the chassis -- maybe .25 Ohms tops.  It seems there's a wiring or solder fault.  (Or I'm missing something)

[sluckey]

Yes, but is seems something else is wrong.  There should not be a 5 Ohm resistance between the bottom of the leg & the chassis -- maybe .25 Ohms tops.  It seems there's a wiring or solder fault.  (Or I'm missing something)

Yes, I agree. But 5Ω does not compute the current/milliVolt readings he had. So either you believe the voltage readings he has, or you believe the resistance reading. I chose to believe the voltage reading and hope he just had a poor connection when doing the resistance reading. Heck, even the probe leads are a factor.

[stratele52]

0.2Ω resistance doesn't seem like much, but when you're expecting 1Ω but really have 1.2Ω, that's about a 20% error.

Resistance error + your meter error ; touch your 2 probes together , how do you read ? My Fluke 175 read 0.2 - 0.3 ohms

[dude]

My meter BK2704B, tips together is .1 to .2 ohms but jumps around a lot.

I understand, I didn't really scrape the chassis when I took the readings. I did now and got .5 to 2 ohm jumping around with a new battery. A 1 ohm resistor can't be checked on any meter I used, expensive or not. I usually read a 10 ohm and add the 1 in series to check the accuracy.

This amp must have some cheap traces, just checking resistance on ground traces I get a up to 1.5 ohms in five inches of traces, even scraping the points. 

So the 5 ohm was a little careless reading, more like 2 ohms. But still that's enough to give me 5 mA off on the correct bias. which it still does.

I think I'll find the grounding points and beef them up, although I don't have much hum but do have blow/by.

I'll just run bias wires to the test points on the chassis direct from the resistor, maybe even move the 1 ohm resistors to the test points.

Should I twist the wires if I move the 1 ohms to the test point? 

[HotBluePlates]

... if it's 35mA right across the resistor, it's 40mA using the chassis as ground. ...

Yes, but is seems something else is wrong.  There should not be a 5 Ohm resistance between the bottom of the leg & the chassis -- maybe .25 Ohms tops.  It seems there's a wiring or solder fault.  (Or I'm missing something)

Ohm's Law

Assume the 35mA at the tube is based on measuring 35mV across the 1Ω resistor; also assume the tube current stayed constant even though he measured 40mV using the chassis ground.

So if it's always 35mA, then use ohm's law to calculate resistance:
Measured Voltage/Known Current = (Implied) Measured Resistance
40mV/35mA = 1.143Ω

If we also assume meter lead and contact resistance are the same between the two measurements, then moving to the chassis ground has added 0.143Ω.

It should not matter to the above that we have assumed exactly 1Ω in the original current measurement, as the change of reading between the 2 points is what was found by calculation.

Note that tubes can and do drift, even over short spans of time (especially when new). So our assumption that there was no current drift may not be valid.

[stratele52]

My meter BK2704B, tips together is .1 to .2 ohms but jumps around a lot.

I understand, I didn't really scrape the chassis when I took the readings. I did now and got .5 to 2 ohm jumping around with a new battery. A 1 ohm resistor can't be checked on any meter I used, expensive or not. I usually read a 10 ohm and add the 1 in series to check the accuracy.

This amp must have some cheap traces, just checking resistance on ground traces I get a up to 1.5 ohms in five inches of traces, even scraping the points. 
That is experience, we learn all the time 
So the 5 ohm was a little careless reading, more like 2 ohms. But still that's enough to give me 5 mA off on the correct bias. which it still does.

I think I'll find the grounding points and beef them up, although I don't have much hum but do have blow/by.

I'll just run bias wires to the test points on the chassis direct from the resistor, maybe even move the 1 ohm resistors to the test points.

Should I twist the wires if I move the 1 ohms to the test point?  No

[dude]

Always learning something new.

Thanks for the explanation HBP, the tubes were used old Westinghouse and I played them a while before rechecking the bias at idle. Made sure I had a good contacts. Still till exactly 5mA difference. And with new battery my meter reads .5 to 2 ohms (chassis resistance from pin 8), so 1.14 seems accurate.

Ohms law always tell the truth, 

Thanks 

[HotBluePlates]

... Still till exactly 5mA difference. ...

5mV difference, not 5mA, right?

Because you're measuring millivolts across the 1Ω resistor and mentally converting it to milliamps of current... The distinction is important, because it tells you what's going on (different voltage drop due to added resistance, not changed tube current).

[stratele52]

... Still till exactly 5mA difference. ...

5mV difference, not 5mA, right?

Because you're measuring millivolts across the 1Ω resistor and mentally converting it to milliamps of current... The distinction is important, because it tells you what's going on (different voltage drop due to added resistance, not changed tube current).

Sorry I don't understand very well what you mean .

Across 1 ohms resistor MV and MA are same . The distinction ( if you need one ) is because your voltmeter is on MV scale AND across a resistor ( in parallel )  it is always a voltage reading . Current reading is always in serie , between the resistor and circuit .

[HotBluePlates]

Across 1 ohms resistor MV and MA are same .

But his situation is that he made a 1st reading directly across the 1Ω resistor, and a 2nd reading with one lead on the 1Ω resistor and the other lead on chassis ground. He found a difference in readings of 5m__.

He keeps saying "5mA" but I'm trying to reinforce that if he makes the readings back-to-back, then he is probably seeing a 5mV difference in readings, while tube current did not change.

I also showed above how a 5mV difference in readings could be caused by only a 0.143Ω increase in resistance between the 2 measurement methods.