High-current Pulses for Battery Research


For those interested in determining the characteristics of batteries, especially by discharging them to see their changes in impedance, Gamry Instruments offers the Reference™ 3000 potentiostat plus our 30k Booster. This combination of instruments is useful for running stress tests on batteries, by applying rapid, high-current pulses singly or repeatedly.

reference3000 30k booster

Fig. 1. Reference™ 3000 (left) with Reference™ 30k Booster (right), with Test and Cal cell (green).

As a cell, we used our 30k Test and Cal cell. The 30k Test and Cal cell printed circuit board was carefully designed for:

  • True four-terminal sensing on both the Test and Calibration sides of the Cell
  • Minimal inductance in the current-carrying connections
  • Minimal mutual inductance between the current-carrying and sensing circuits.

Gamry Instruments uses the Calibration side of the cell to galvanostatically measure impedance spectra

at frequencies up to 300 kHz with less than 2° of phase error.

This Application Note shows results of our own research into the characteristics of our equipment.


The tests were run on a Reference 3000 and Reference 30k Booster. A Gamry Reference 30k Test and Cal cell was connected to the booster in all tests. The cell connections used the standard cables supplied with the system: a 1 m current-carrying cable and a 1.5 m sense cable.

Most of the testing was done on the Calibration side of the Test and Cal cell. This side is a 200 mΩ resistor with a 2.5 A fast-blow fuse in the counter electrode lead. The fuse protects the 1 W resistor from damage caused by excessive current flow. The resistance of a sample fuse was measured to be 34 mΩ. Fast current pulses larger than the 2.5 A fuse rating will not blow the fuse if the pulses are short enough.

The Test side of the Test and Cal Cell is a 3 mΩ four-terminal resistor. It does not require a fuse because it can safely pass 30 A currents.

Many tests were performed at different CA (control amplifier) Speed settings. The faster the CA Speed is, the better time-resolution is achieved, but at the cost of stability in the amplifier.

Most data were recorded using our signature Framework™ software. Some of the tests required use of a Tektronix 2024C digital oscilloscope instead.


2A Pulses on 200 mΩ Cell, Using the Framework™ Software

Fig. 2 below shows 2 A current steps lasting 200 µs each. Connections were made through the Reference 30k Booster, with the Reference 3000’s 3 A full-scale range.

fig2 cell voltage upper trace

Fig. 2. The cell voltage is the upper trace, and cell current is the lower. Colors correspond to different control-amplifier speeds (CA Speeds):
green are CA Speed Fastest, red are CA Speed Fast, and blue are CA Speed Norm.

Oscilloscope Data on 200 mΩ Taken from the E Monitor BNC

A BNC cable was connected between the Reference 3000 rear panel E Monitor output and the input of the oscilloscope.

Fig. 3 shows the voltage-versus-time curve of a 200 µs, 2 A pulse on the Calibration-side resistor. The graph was recorded in CA Speed Fast.

fig3 data recorded e monitor connector

Fig. 3. Data recorded through the E Monitor connector.

The oscilloscope-screen capture (Fig. 4) shows a very similar waveform.

fig4 data recorded to oscilloscope

Fig. 4. Data recorded to oscilloscope under same conditions as Fig. 5.


Reference 3000 Booster Current Pulses on 200 mΩ

Here (Fig. 5) Gamry Instruments tested a range of current pulses. The smallest current of 2 A did not require the 30k Booster. The larger currents (4 A through 12 A) included the Booster. The curves show no evidence of slew-rate limiting.

fig5 data from range of currents

Fig. 5. Data from a range of currents. The lowest current (2 A) used no 30k Booster; the higher currents (4–12 A) added the Booster.

Fig. 6 shows the effect of CA Speed on 8 A pulses. The blue curve was recorded with CA Speed Fast and red curve was recorded with CA Speed Fastest. The red curve has a squarer response but ringing is seen on both the rising and falling edges.

fig6 data using CA speedFig. 6. Data at 8 A, using the CA Speed Fast (blue) and CA Speed Fastest (red).

The next curve (Fig. 7) shows 4 A and 8 A pulses in CA Speed Fastest. There is still no evidence of slew-rate limitations.

fig7 data using CA fastest

Fig. 7. Data using the CA Speed Fastest , comparing signal responses at 4 A (blue) and 8 A (red).

Results on a 3 mΩ Cell

The connections were switched to the Test side of the Test and Cal Cell. This side is a 3 Ω, 4 W resistor. At 30 A, its power dissipation is only 2.7 W, so it is safe from misconnection and oscillation.

Fig. 8 shows 2 A pulses on this cell. Such current pulses always are applied with the Reference 3000. Voltage is on the top set of traces and current is on the bottom set. The trace colors correspond to CA Speeds: green is Fastest, red is Fast, and blue is Norm.

fig8 data on test sideFig. 8. Data on the Test side of the Test and Cal cell, from 2 A pulses, at CA speeds of Norm (blue), Fast (red) and Fastest (green).

There is a slight indication of ringing in the Fastest CA Speed.

The next graph (Fig. 9) shows 4 A pulses on the Test side resistor at the different CA Speeds. The trace placement and colors are the same as the plot above.

fig9 data test side 4AFig. 9. Data on the Test side of the Test and Cal cell, from 4 A pulses, at CA speeds of Norm (blue), Fast (red) and Fastest (green). Top: Voltage; bottom: Current.

Both Fastest and Fast CA Speeds show ringing in Fig. 9. The oscilloscope capture in Fig. 10 shows the E Monitor waveform for CA Speed Fastest and a 4 A pulse. Ringing is also visible here.

fig10 oscilloscope data

Fig. 10. Oscilloscope data on the Test side of the Test and Cal cell, from 4 A pulses, at the CA speed of Fastest.

50 µs Current Pulses

Fig. 11 shows 50 µs 2 A and 4 A current pulses into the 200 mΩ Calibrate side of the Test and Calibrate cell. The green trace is a 4 A pulse in CA Speed Fastest, the red trace is a 4 A pulse with CA Speed Fast, and the blue trace is a 2 A pulse at CA Speed Fast.

fig11 daton calibrate side

Fig. 11. Data on the Calibrate side of the Test and Cal cell, from 50 µs pulses, at 2 A (blue), 4 A Fast (red) and 4A Fastest (green). Top: Voltage; bottom: current.

The charge lost in the slow rising edge is recovered in the slow falling edge.

30 A pulses

Gamry Instruments’s equipment even allows you to pulse with 30 A current, as Fig. 12 reveals.

fig12 30A pulse

Fig. 12. 30 A pulses using the Reference 3000 and 30k Booster. Red trace is current; blue trace is voltage.


Gamry’s Reference 3000 and Reference 30k Booster can accurately and reliably generate current pulses as short as the high tens of microseconds. Approximate rise times τ are shown in the Table below:

Instrument Cell τ for Fastest τ for Fast τ for Norm
 Reference 3000  200 mΩ  τ < 6 µs ringing  τ ≈ 10 µs  τ ≈ 29 µs
   3 mΩ  τ < 6 µs ringing  τ ≈ 7 µs  τ ≈ 23 µs
 30k Booster  200 mΩ  τ < 5 µs ringing  τ ≈ 7 µs  τ ≈ 17 µs
   3 mΩ  τ < 2 µs ringing  τ ≈ 6 µs  τ ≈ 22 µs


As to which is the optimal CA Speed setting, Gamry Instruments recommends that users run a preliminary test that measures the rise time on a typical cell prior to running a set of experiments.

Galvanostats generally become more stable when cells are capacitive. This effect was not measured in this Application Note.