The subject is a lead acid. The first photo is my trusted BM235 with 3 el-cheapos. Second is an android split screen with the Rudeng charger output (and controls) the bottom is the Aneng9002 voltage meter on the battery. Sofa charging.
Multi-meter try on haul.
- Thread starter venquessa
- Start date
.5% variation between the four meters. For our purposes that's probably close enough. On a daily basis, our LiFePO4 cells shouldn't be seeing voltage above 3.6 volts anyhow. But it's interesting that you have that much variation between four meters.
I bought a voltage checking device to verify that my multimeters were calibrated. All of mine tested well.
voltagestandard.com
I bought a voltage checking device to verify that my multimeters were calibrated. All of mine tested well.
voltagestandard.com
low cost, battery operated, precision DC voltage references from 2.048V to 10V
voltagestandard.com
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 23,143
That's more zeros than I expect (accurately) from a meter, or in a non temperature-controlled environment.
Even your 3458A; It might do 8 digits as a transfer standard for a few minutes, but not absolute as previously calibrated.
(Then again, maybe it typically does, just not guaranteed.)
Try warming test leads with your hand, see if thermoelectric effect of junctions registers a change.
We rented a 3458A for a month. I bought a 3456A off eBay. With 4.096 Vref device, I resistor-divided to 1.024V to use the 6.5 digits of my meter. It agreed with the 8.5 digit meter except LSB toggled one count.
Even your 3458A; It might do 8 digits as a transfer standard for a few minutes, but not absolute as previously calibrated.
(Then again, maybe it typically does, just not guaranteed.)
Try warming test leads with your hand, see if thermoelectric effect of junctions registers a change.
We rented a 3458A for a month. I bought a 3456A off eBay. With 4.096 Vref device, I resistor-divided to 1.024V to use the 6.5 digits of my meter. It agreed with the 8.5 digit meter except LSB toggled one count.
Yes. Precision versus accuracy. It's easy for the former to trick you into a sense of the later.
Voltage... no the whole holy trio V=IR... are rabbit holes.
For most measurements, the best approach is to sample it over time, preferable multiple measurements with different instruments. Then data analysis techniques come into play. A single measurement at a point in time pales in the shadow of a series of measurements made over time. Not least it tends to "bring out the noise".
Most multi-meters do this and show the average. I posted a trace in another thread, but the reality of what is actually going on, needs an oscilloscope. Voltage ripple in AC/DC chargers and inductor switching/ringing/spikes on DC/DC chargers. I found my DC/DC converter when pushed to 20A produces an inductor ring of nearly 400mVpp. However, technically speaking, power supply ripple is usually only stated with a 20Mhz bandwidth limit. The origins of that "standard" elude me. I'm not sure it means that noise over 20Mhz has no effect on equipment.
Voltage... no the whole holy trio V=IR... are rabbit holes.
For most measurements, the best approach is to sample it over time, preferable multiple measurements with different instruments. Then data analysis techniques come into play. A single measurement at a point in time pales in the shadow of a series of measurements made over time. Not least it tends to "bring out the noise".
Most multi-meters do this and show the average. I posted a trace in another thread, but the reality of what is actually going on, needs an oscilloscope. Voltage ripple in AC/DC chargers and inductor switching/ringing/spikes on DC/DC chargers. I found my DC/DC converter when pushed to 20A produces an inductor ring of nearly 400mVpp. However, technically speaking, power supply ripple is usually only stated with a 20Mhz bandwidth limit. The origins of that "standard" elude me. I'm not sure it means that noise over 20Mhz has no effect on equipment.
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 23,143
Vendors of switch-mode power supply ICs give advice on how to reduce bandwidth of oscilloscope (add a capacitor to ground on probe tip) so they won't see the spikes. One was so bad, 900 mVpp on a 1.8V supply, it kept cycling the "power good" signal of a linear regulator on and off. It drove noise into other power supplies, even when input voltage was isolated and only ground plane shared. It made 8Vpp appear on a 10V regulator. A bunch of capacitors, including from output to input, made it tolerable.
Linear supplies also have ripple. The linear regulator may be good, but rectifier/capacitor front end produces 120 Hz harmonics. The best might be a PF corrected 3-phase front end with linear post-regulator. We actually get good results with HV DC into a low-noise SMPS; that reduces or eliminates the 120 Hz we saw.
We use a Keithley SMU. It's spec are extremely low noise "below 10 Hz." It has a 1 kHz burst of noise on the output.
Measuring average DC is easy - you just average. What is difficult is measuring low ripple riding on top of DC. AC coupling will do that, but has low-frequency roll off. More effort to measure ripple down to 1 Hz. Capacitor circuits with low cut-off, or active circuits offsetting DC component.
Linear supplies also have ripple. The linear regulator may be good, but rectifier/capacitor front end produces 120 Hz harmonics. The best might be a PF corrected 3-phase front end with linear post-regulator. We actually get good results with HV DC into a low-noise SMPS; that reduces or eliminates the 120 Hz we saw.
We use a Keithley SMU. It's spec are extremely low noise "below 10 Hz." It has a 1 kHz burst of noise on the output.
Measuring average DC is easy - you just average. What is difficult is measuring low ripple riding on top of DC. AC coupling will do that, but has low-frequency roll off. More effort to measure ripple down to 1 Hz. Capacitor circuits with low cut-off, or active circuits offsetting DC component.
Similar threads
