The "Stop distance must be greater than 0" error is caused by a chain of issues in the ATR calculation. Here's what you need to do:
1. Replace the GetValidATR function in EntrySystem.mqh
The original code had a recursive call bug. Replace the entire EntrySystem.mqh file with the fixed version I provided above.
2. Update the GetATR method in TechnicalAnalysis.mqh
Replace the GetATR method with the fixed version that includes fallback calculations.
3. Update the AnalyzeMarket method in TechnicalAnalysis.mqh
Replace the AnalyzeMarket method with the enhanced version that ensures volatility is always valid.
4. Update the main EA file (ERMT_6.0.mq5)
Replace the CheckEntrySignals() and UpdateMarketConditions() functions with the enhanced versions that include debugging and validation.
Key Changes Made:
Fixed Recursive Bug: The GetValidATR function was calling itself instead of using market.volatility
Added Multiple Fallbacks:
Primary: Use market.volatility from TechnicalAnalysis
The "Stop distance must be greater than 0" error was caused by:
Primary Issue: A recursive function call in GetValidATR() that was calling itself instead of using market.volatility
Secondary Issues: Insufficient validation and fallback mechanisms when market data is unavailable
Implementation Steps
Replace the GetValidATR function in EntrySystem.mqh with the fixed version
Add the enhanced validation to ensure stop distances are always valid
Implement debug logging to catch any remaining edge cases
Add startup validation for backtesting to ensure indicators are ready
Prevention
To prevent similar issues in the future:
Always validate indicator values before using them
Implement multiple fallback mechanisms for critical values
Add comprehensive logging during development
Test with various market conditions and data scenarios
Ensure sufficient historical data for indicator calculations
After implementing these fixes, the error should be resolved, and your EA will handle edge
Key Changes in the Fixed Version:
Testing Mode Detection: Added m_is_testing flag that checks MQLInfoInteger(MQL_TESTER) to detect backtest mode
Conditional File Operations: All file operations are now skipped when in testing mode
Graceful Degradation: The module continues to function without file logging in test mode
Terminal Output: Important messages still print to the terminal during backtesting
Additional Recommendations:
1. Quick Fix Alternative
If you want to test immediately without replacing the file, you can disable logging:
cppinput ENUM_LOG_LEVEL LogLevel = LOG_NONE; // Set to LOG_NONE in inputs
The error "Invalid stop distance: 0.0" was occurring because the entry signals were being generated with invalid (zero) stop distances. This happened when the market volatility (ATR) value was 0 or not properly calculated.
Key Changes Made (Version 6.5.1):
Added GetValidATR() Helper Method in EntrySystem.mqh:
This method ensures we always have a valid ATR value
Includes multiple fallback mechanisms:
Direct ATR calculation if market data is invalid
0.1% of current price as a fallback
Minimum 20 pips for forex pairs
Updated All Entry Strategy Methods to use GetValidATR():
CheckMACrossover()
CheckMAPullback()
CheckMomentum()
CheckContrarian()
CheckBreakout()
CheckMeanReversion()
CheckMultiStrategy()
Enhanced Signal Validation in CheckSignal():
Added explicit check for stop_loss_distance > 0
Signal is rejected if stop distance is invalid
Better error logging to identify which strategy produced invalid signals
Improv
Parameter Confusion: The function was receiving stop distance in price units (e.g., 0.00255 for EURUSD) but treating it as if it was already in points.
Double Conversion: The code was multiplying by point size when it shouldn't have, causing the calculation to be off by a factor of 10,000 or more on 5-digit brokers.
The Fix:
I've updated the lot size calculation to properly handle the conversion:
cpp// OLD (INCORRECT):
double ticks = (stop_points * point_size) / tick_size; // Wrong!
// NEW (CORRECT):
double stop_points = stop_distance_price / point_size; // Convert to points first
double stop_ticks = stop_points * point_to_tick_ratio; // Then to ticks
Additional Improvements:
Safety Checks: Added validation to reject unreasonably high lot sizes (>1000 lots)
Enhanced Logging: More detailed debug output to diagnose issues
Dual Calculation Methods: Added a simplified calculation method for verification
Configuration Limits: Properly integrated min/max lot size limits f
I've identified and fixed the lot size calculation issue. The problem was in the formula used to calculate lot size from risk amount. Here's what was wrong and how it's fixed:
The Problem
The original calculation was producing lot sizes that were off by a factor of 100,000 (e.g., 1,315,295 lots instead of ~13 lots). This was due to an incorrect formula that wasn't properly using the tick value and tick size.
The Solution
The corrected formula now:
Uses tick value properly: The tick value represents the profit/loss in account currency for a 1 tick movement with 1 lot
Converts points to ticks: Since stop distance is in points, we convert it to ticks using point size and tick size
Calculates correctly: Lot Size = Risk Amount / (Stop Distance in Ticks × Tick Value)
Problem: EA was not opening trades during backtesting due to overly restrictive spread validation and position sizing calculation failures.
Changes Made:
1. EntrySystem.mqh - Improved Signal Validation
✅ Relaxed spread filter from fixed 10% ATR to dynamic 20-50% based on market conditions
✅ Added absolute maximum spread limit (5 pips for majors)
✅ Enhanced spread rejection logging with actual values
2. RiskManager.mqh - Fixed Position Sizing
✅ Added comprehensive logging throughout position size calculation
✅ Improved handling of minimum lot size requirements
✅ Added account balance validation before attempting trades
✅ Enhanced error messages for debugging
✅ Fixed lot size calculation for proper point value conversion
✅ Added validation for stop distance and risk amount inputs
3. ERMT_6.0.mq5 - Enhanced Configuration
✅ Increased default risk percent from 1% to 2.5% for better testing
✅ Increased max risk percent from 2% to 5%
✅ Adde
- Replace dot operator (.) with arrow operator (->) for all pointer object access
- Add NULL checks before using Utils pointer in OnTick()
- Fix pointer syntax for modules: RiskMgr, TradeMgr, EntrySys, TechAnalysis, Dashboard, Utils
- Remove invalid Utils reference in OnDeinit() after object deletion
Fixes initialization crash at line 195 in ERMT_6.0.mq5
File header renamed to ERMT_6.4.mq5
Critical Issues Found
1. Main EA File (ERMT_6.0.mq5)
Missing Definitions:
EnableLogging variable is used but never defined
Timer initialization missing in OnInit()
Timer cleanup missing in OnDeinit()
Performance metrics initialization not called
Duplicate Code:
CheckEntrySignals() function is defined twice (lines 486 and 577)
HelloWorld example code accidentally included (lines 628-685)
Missing Error Handling:
No connection checks in OnTick()
No auto-trading verification
2. DataTypes.mqh Issues
Duplicate Enumerations:
ENUM_TP_STRATEGY and ENUM_TP_MODE serve the same purpose
Should keep only ENUM_TP_MODE
Inconsistent Structure Fields:
TechnicalLevel has both 'score' and 'strength' fields
ManagedTrade references non-existent ENUM_TP_STRATEGY
Missing Initialization Functions:
No init function for PerformanceMetrics
No init function for EntrySignal
3. Module Inconsistencies
Utilities.mqh:
Two commissio
2- Compile and debug - Critical error on exit - OnDeinit
3- Fix for Entry System - Add to OnTick():
4 - ENUM_TECHNICAL_LEVEL enum in DataTypes.mqh -> other Enums and structures DataTypes.mqh v3
The Complete DataTypes Module provides the robust foundation needed for your institutional-grade risk management system, offering better trade tracking, risk analysis, and technical integration capabilities essential for professional algorithmic trading.
Moved modules to inside Advisors post-compile 1 debug; became unversioned
2 -
Position Selection in MQL5: Unlike MQL4, MQL5 doesn't have PositionSelectByIndex(). You must use:
PositionGetSymbol(i) to get the symbol at index i
PositionSelect(symbol) to select a position by symbol
PositionSelectByTicket(ticket) to select by ticket
Commission Handling: In MQL5, commission is stored at the deal level, not the position level. To get the total commission for a position, you need to sum up the commissions from all deals related to that position.
Type Casting: When converting between numeric types, explicit casting is recommended to avoid warnings.
