Cascade Heating Control with LOGO! 0BA5: Parallel Gas Heaters

1. Application Overview

This reference describes a two-stage gas-heater cascade implemented on a Siemens LOGO! 0BA5 base module extended with an AM2 RTD expansion block. Two PT1000 Class A sensors measure flow and return water temperatures of a primary heating loop. A room thermostat provides the demand signal. Stage 1 starts when the room calls for heat. Stage 2 is allowed to start only when the return-water temperature has not reached a configurable set point (default 50 °C) within a configurable window (default 10 minutes). The intent is to maximize single-stage efficiency, reduce burner cycling, and limit the parallel running time of the second appliance.

The same architecture generalises to any LOGO! 5-series base module (0BA5) with the 6ED1055-1MD00-0BA2 RTD expansion. Programming is performed in LOGO!Soft Comfort V5.x; the FBD and ladder fragments in this article are compatible with the 0BA5 instruction set.

2. System Architecture

3. Sensor Scaling on the 0BA5 / AM2 RTD

PT1000 nominal resistance is 1000 Ω at 0 °C with a near-linear slope of approximately 3.85 Ω/°C. The AM2 RTD linearises the resistance and presents the result as a scaled analog value on the LOGO! 0BA5.

For FBD use an Analog Amplifier (B02) block in conjunction with the AI input. Configure sensor type = PT1000 and measurement range = 0…100 °C. The amplifier output becomes a 0–1000 integer equivalent that downstream math blocks (threshold triggers, PI controller) interpret directly.

4. Why Both Flow and Return Are Measured

A common confusion in the source thread is whether the flow sensor is required when the return sensor already gives a meaningful signal. The answer is that the two sensors carry different information:

• Return water (T_ret) is the controlled variable. It indicates how much heat the emitters have already extracted from the loop. If T_ret stalls, the loop is calling for more heat than a single burner can deliver.
• Flow water (T_flow) is a supervisory variable. The flow/return differential (ΔT) is a proxy for instantaneous loop load. A high ΔT with a low T_ret is a robust start condition for a second burner; a low ΔT with a high T_ret means the loop is over-supplied and the demand should be shed instead of added.

A simple two-stage cascade can run on T_ret alone, but the addition of T_flow enables anti-oscillation, lead/lag, and maximum ΔT clamp logic, which prevents short cycling of the second burner when the loop is near steady state.

5. Cascade Control Strategy

The recommended state machine for the heater pair is shown below.

State transitions in plain language:

1. IDLE → STAGE 1: room thermostat closes (I1) and system enable (I2) is present.
2. STAGE 1 → STAGE 1 + 2: timer T1 elapses (10 minutes default) AND T_ret is below set point (50 °C default).
3. STAGE 1 + 2 → COOLDOWN: room thermostat opens, fault, or T_ret ≥ (set point + hysteresis, default 2 °C).
4. COOLDOWN → IDLE: minimum off-time T2 (3 minutes default) expires, both Q1 and Q2 are forced off to allow pump overrun.

6. Building the FBD on LOGO! 0BA5

The program is constructed in the following block order in LOGO!Soft Comfort V5.x. Block numbers shown are illustrative; renumber for your project.

6.1 Inputs and Scaling

B01  Digital Input  I1     -- Room thermostat
B02  Digital Input  I2     -- System enable
B03  Digital Input  I3     -- Fault reset (NC)
B04  Analog Input   AI3    -- PT1000, T_ret (return)
B05  Analog Input   AI4    -- PT1000, T_flow (flow)
B06  Analog Amplifier (PT1000, 0..100 °C) -> AQ3
B07  Analog Amplifier (PT1000, 0..100 °C) -> AQ4

6.2 Threshold Detection

B08  Threshold Trigger   On:  AQ3 >= 500  (T_ret >= 50.0 °C)
                         Off: AQ3 <  480  (hysteresis 2.0 °C)
                         Output -> M1
B09  Threshold Trigger   On:  AQ3 <  500  (T_ret <  50.0 °C)
                         Off: AQ3 >= 500
                         Output -> M2
B10  Threshold Trigger   On:  (AQ4 - AQ3) >= 150  (ΔT >= 15.0 °C)
                         Off: (AQ4 - AQ3) <  130
                         Output -> M3  -- "loop still hungry"

6.3 Time-Delayed Stage-2 Enable

B12  AND              I1, I2            -- master demand
B13  On-Delay         00:10:00          -- 10 minute window
      Triggered by:  B12
      Output:        M4  -- "stage 2 allowed"
B14  AND              M4, M2, M3        -- stage 2 allowed AND
                                    -- T_ret < 50 °C AND
                                    -- loop ΔT > 15 °C
B15  Set/Reset Latch  S = B14, R = B08

6.4 Output Drivers

B16  AND              B12, NOT(B15)     -- Q1 if demand and not in stage 2
B17  Output Q1        = B16
B18  Output Q2        = B15

Block B16 deliberately forces Q1 off while B15 is set if you want exclusive operation, but for parallel/cascade operation with the source's requirement keep B16 = (I1 AND I2) so that Q1 stays on whenever there is demand and the safety contact is closed. Replace B16 with a direct pass-through of B12 if parallel running is desired.

7. PI Controller Variant

The source thread mentions that a PI controller was attempted but abandoned due to unfamiliarity. The PI controller block (B20) on the 0BA5 can be used as a supervisory controller on T_ret with a slower integral action that decides whether stage 2 should be added. A practical implementation:

The PI output is fed into an additional threshold trigger (B21) that turns on stage 2 when the integrated error exceeds 70 % of the controller span. This provides soft staging and avoids hard time-based decisions during steady-state operation.

8. Hardware Wiring

9. Commissioning Procedure

1. Power the LOGO! 0BA5 with the AM2 RTD detached. Verify the base module boots and the LOGO! display shows no expansion error.
2. Connect the AM2 RTD on the expansion bus. Confirm in LOGO!Soft Comfort online view that AQ3 and AQ4 appear in the I/O status list.
3. Apply known temperature sources (ice-water slurry at 0 °C, boiling water at 100 °C) to each PT1000. Read AQ3/AQ4 values. Trim the amplifier offset in the AM2 configuration if the deviation exceeds 0.5 °C.
4. Force I1 on in online test. Confirm Q1 energises within 100 ms and the burner lights.
5. Force I1 off and verify Q1 drops, Q2 drops, and the on-delay B13 resets.
6. Simulate the alarm condition: with I1 on, clamp T_ret to 20 °C using a calibration source. After 10 minutes Q2 must energise. The on-delay B13 must be the only gating element on the cold start.
7. Apply the actual return temperature by running the loop. Verify M1 toggles at the expected set point ± hysteresis.
8. Log the LOGO!Soft Comfort online trace for at least 30 minutes of normal operation; verify no fault, no spurious stage transitions.

10. Verification Checklist

11. Troubleshooting Matrix

12. Safety Considerations

Gas burners require independent flame supervision; the LOGO! must never be the sole safety interlock. The burner controller's own safety chain (air pressure, flame rod, gas valve proving) is authoritative. The LOGO! enable outputs (Q1, Q2) must be wired in series with the burner enable circuit so that a LOGO! fault cannot energise a burner unintentionally. A watchdog or fail-safe relay on Q1 and Q2 is recommended if the burner enable input is not galvanically isolated.

13. Practical Tuning Notes

• Default 10 min delay is a good cold-start value. For well-insulated loops, reduce to 6 min; for sluggish emitters, extend to 15 min.
• Set point 50 °C is appropriate for low-temperature underfloor loops. For radiator loops, raise to 60–65 °C.
• Hysteresis on the disable threshold should be 2–3 °C, not less, to prevent contactor hunting.
• ΔT threshold of 15 °C is conservative. Calibrate it to the actual ΔT observed at design load (typically 10–20 °C for residential heating loops).
• Pump overrun: if the system pump is not controlled by the LOGO!, the 3 min minimum off-time protects the burner; if the pump is also LOGO!-controlled, extend the off-time to 5 min to dissipate loop heat.