Optimizing a query that selects from a select query

I am trying to count all suspicious deposits from my app.

The query basically selects, counts and groups the results from an inner select query.

Using this query:

SELECT *, count(id) as repeated from (
    SELECT id, userAccount_id, `change_value_amount`,`change_value_currency_code`, JSON_EXTRACT(details, '$.depositId') as depositId From billing.transactions WHERE type_value = 'DEPOSIT_MADE'
) as q1 
where depositId is not null group by depositId having repeated > 1

Now, this query's execution time is 4-6 secs, which is really annoying.

I also tried creating a temporary table and do the outer where clause based from the temp table:

create temporary table if not exists susDeposit as (SELECT id, userAccount_id, `change_value_amount`,`change_value_currency_code`, JSON_EXTRACT(details, '$.depositId') as depositId From billing.transactions WHERE type_value = 'DEPOSIT_MADE');

select id, userAccount_id, change_value_amount, change_value_currency_code, depositId, count(*) as repeated  from susDeposit
where depositId is not null group by depositId having repeated > 1

in workbench, it executes at 1sec max.

but using pdo, it executes at 5secs, idk the reason why the execution time is different in workbench and pdo.

Code is:

$tempTable = "CREATE TEMPORARY TABLE IF NOT EXISTS susDeposit as (SELECT id, userAccount_id, `change_value_amount`,`change_value_currency_code`, JSON_EXTRACT(details, '$.depositId') as depositId From billing.transactions WHERE type_value = 'DEPOSIT_MADE');";
        $finalSelect = "SELECT id, userAccount_id, change_value_amount, change_value_currency_code, depositId, count(*) as repeated  from susDeposit
        where depositId is not null group by depositId having repeated > 1";

        $query = $this->pdo->query($tempTable);
        $query = $this->pdo->query($finalSelect);
        $query->execute();
        $r = $query->fetch();

Is there any way to optimize this query?