-
Products & Solutions
-
Support & Training
-
Follow Us
-
- SUBSCRIBE
How Infinity Improves Refresh Rate in Low-Grayscale
Under the traditional PWM scan architecture, the single-pixel visual refresh rate 
in the low-grayscale region inevitably degrades to near (or even below) the input frame rate. This is the root cause of flicker visible to the naked eye and stripe / banding artifacts captured by cameras.
Without breaking through the limits of GCLK and the minimum pulse width—which is, in fact, difficult to overcome from the perspectives of chip process and cost—the hybrid PWM+PAM drive boosts the low-gray visual refresh rate to the same order of magnitude as the nominal (specified) refresh-rate metric.
The pulse structure in the low-gray segment changes from “sparse” pulses to “multiple” pulses. Quoting the formula from the previous section:
1. Under Traditional PWM Mode
When the grayscale level g approaches 1 (near-minimum brightness), the common implementation is:
Only a single narrow pulse exists within one frame:
The current amplitude is fixed at the nominal (rated) current:
The pulse width is already close to the physical lower limit:
In this case, the pixel’s brightness waveform on the time axis consists of narrow pulses at roughly 60 pulses per second (matching the input frame rate). Its equivalent visual refresh rate 
degenerates to nearly the input frame rate 
, which is the root cause of the “flicker” observed in low-grayscale regions.
2. Under Hybrid PWM+PAM Mode
We redesign the pulse structure for the low-grayscale segment as follows:
Each pulse width is fixed within a safe operating region:
A fixed, elevated number of low-grayscale pulses is selected:
For example, 4, 8, or even higher. This enables:
The current amplitude 
is used to “fill in” the brightness:
Thus, the equivalent luminous flux in the low-grayscale segment becomes:
Fig.1 PWM and PWM+PAM Refresh Rate Schematic Diagram
3. Determining the Low-Grayscale Current Amplitude
To ensure that the low-grayscale segment simultaneously satisfies the required refresh-rate performance and tracks the target gamma curve 
, the low-current amplitude 
cannot be chosen arbitrarily. Instead, it must be solved inversely from the equivalent luminous-flux relation:
Hence:
Compared with the low-grayscale behavior of a pure PWM scheme, the PWM+PAM hybrid drive exhibits inherent advantages in the low-grayscale regime. Consider the following assumptions / constraints:
Input frame rate: 
GCLK upper limit: 
Scan count: 
Minimum safe pulse width: 
For a pure PWM implementation targeting 14-bit grayscale, the theoretical low-grayscale LSB pulse width would be approximately:
This already approaches physical limits. When additional real-world factors are taken into account—such as LED turn-on thresholds, nonlinear response in the low-current region of the I-L curve, and other external nonidealities—certain emitter chips may fail to illuminate entirely or only produce faint, unstable light output. Under dark ambient conditions or long-exposure capture, such a refresh characteristic is readily perceived as flicker or scanning stripes / banding artifacts.
By contrast, with PWM+PAM hybrid driving in the low-grayscale segment, assuming a low-gray pulse count of 
and a single-pulse width of 
, even the minimum grayscale level 
can achieve:
Under these conditions, 
low-grayscale refresh far exceeds the critical flicker fusion frequency (CFF) for human vision even in dark environments, so the perception of flicker is virtually eliminated. For cameras operating at 30 / 60 / 120 fps, the beat-frequency interaction between the sampling rate and the LED pulse train is significantly reduced, and low-grayscale stripe / moiré artifacts are substantially mitigated.
Refresh Rate vs. Grayscale: PWM vs. PWM+PAM
Fig.2 Grayscale vs. Refresh Rate Relationship
