Timeseries 16MB limit

Working with Data
, ,
1

I create timeseries with following schema. If I have 90K+ TrendReadings I get the 16MB limit exception.
How to structure the trend POCO to avoid this? What are the best practices? I would need to search the timerange latter.
Thank you.

Schema:

db.trends.findOne()
{
  _id: ObjectId("611bb0c404571e809dda5ab8"),
  EntityId: Binary(Buffer.from("7cae6edc9167304a91f7052137f4bb36", "hex"), 3),
  ExperimentId: ObjectId("611bb0c30000010001001e6e"),
  TrendObjectId: ObjectId("611bb0c30000010001001e6e"),
  Nsamples: 0,
  Interval: 2,
  UnitOfMeasure: '°C',
  TrendReadings: [
    {
      TopicIndex: 0,
      Reading: 15.4664096832275,
      timestamp: ISODate("2013-04-25T19:51:05.000Z")
    },
    { .. etc

C# code

foreach (var item in trendReadings.Values)
{
    _TrendReadingsCollection2.InsertOne(item); //exception if item.Readings.Count > 90K
}