EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: dSolder on May 09, 2017, 11:15:31 am
-
I was looking for a possible drop-in replacement for the MAX4239 and stumbled across the OPA388 with 10MHz gain. It looks like a new op-amp from TI. The SOT-23-5 pinout is the same as the MAX4239. Would this make a suitable replacement in the ucurrent gold? Also it looks like the SOT-23-5 version is only available in October! |O
http://www.ti.com/lit/ds/symlink/opa388.pdf (http://www.ti.com/lit/ds/symlink/opa388.pdf)
(pasted from the datasheet)
• Ultra-Low Offset Voltage: ±0.25 µV
• Zero-Drift: ±0.005 µV/°C
• Zero-Crossover: 140-dB CMRR True RRIO
• Low Noise: 7.0 nV?Hz at 1 kHz
• No 1/f Noise: 140 nVPP (0.1 Hz to 10 Hz)
• Fast Settling: 2 µs (1 V to 0.01%)
• Gain Bandwidth: 10 MHz
• Single Supply: 2.5 V to 5.5 V
• Dual Supply: ±1.25 V to ±2.75 V
• True Rail-to-Rail Input and Output
• EMI/RFI Filtered Inputs
• Industry-Standard Packages:
– Single in SOIC-8, SOT-23-5, and VSSOP-8
– Dual in SOIC-8 and VSSOP-8
– Quad in SOIC-14 and TSSOP-14
Thanks in advance!
-
Like other low noise AZ OPs, the OPA388 has a relatively high possible bias current (up to 0.7 nA) and a not so perfect 0 offset. So there might be a slight offset of the readings and the supply current is likely higher.
The max4239 is an AZ OP with relatively high noise, but very low bias and very low offset. So the zero is correct even without an adjustment.
A little more similar replacement would be a LTC2050, though slower.
-
Like other low noise AZ OPs, the OPA388 has a relatively high possible bias current (up to 0.7 nA) and a not so perfect 0 offset. So there might be a slight offset of the readings and the supply current is likely higher.
The max4239 is an AZ OP with relatively high noise, but very low bias and very low offset. So the zero is correct even without an adjustment.
A little more similar replacement would be a LTC2050, though slower.
Can you expand on how the max4239 is noisy? Are there any ultra-low VOS opamps with low noise?
-
The max4239 has a specified DC-10 Hz noise of 1.5 µV, thus relative high compared to other AZ OPs. However it also comes with low maximum offset and low typical bias (1 pA).
Usually the lower noise AZ OPs use a higher internal sampling / chopping frequency. This high frequency is somewhat needed to avoid 1/f noise of the CMOS amplifiers used internally. The higher chopping frequency also leads to a higher bias (togehter with larger, lower resistance switches) and also to more residual offset. Lower noise amplifiers also tend to use more current. Even if the GBW specs a relatively low, supply decoupling can be important, as there are some spikes in the supply current.
For the residual offset, the typical values are usually considerably better than maximum values - so chances are that selecting a good one is possible. Even if the maximum specified offset is at 5 µV, chances are good to actually have an offset below 1 µV. Depending on the circuit there could also be the option to do an adjustment, as the offset is usually quite stable.
Ideally for highers performance on an current sensing amplifier one would use different amplifiers for the different current ranges: a low noise version with extra offset adjustment for the higher currents (e.g. > 1 mA). The low current ranges would ideally use a trans-impedance type circuit. This might not need a AZ amplifier at all.