Troubleshooting
Start a PID ramp soak profile from ladder logic
Short answer
The profile never starts from ladder because the JMP stage handoff leaves the old stage on for three more scans, which swallows the one-scan PD pulse that must set the ramp/soak start bit. Replace the jump with SET and RST stage logic, then verify bit 11, the +33 octal control word and the +35 error flags before touching the PID instruction or table values.
On this page
- Why the jump kills the one-shot
- Replace the jump with set and reset logic
- Verify the generator enable and the start bit
- Add the offsets in octal
- Read the error flags before editing the table
- Check every table value in its own format
- Test the ladder start, not a forced ramp
- The decision, case by case
- Questions asked
Your ramp/soak profile will not start from ladder because the JMP stage handoff leaves the old stage on for three more scans, and that overlap swallows the single-scan PD pulse that must set the ramp/soak start control bit. Replace the jump with explicit SET and RST stage logic — that is the fix that cleared the DL450 press program behind this article — then work the remaining checks in order: bit 11 generator enable in PID V+0, the ramp/soak control word at PID V+33 octal, the error flags at PID V+35 octal, and the decimal or BCD format of every table value. None of this is a hidden conflict: every bit is documented in the controller's PID chapter and readable live in the programming software before you change one PID parameter.
Why the jump kills the one-shot #
A PD one-shot is edge-triggered: its contact closes for exactly one scan when the driving condition goes false-to-true, and it stays open if the condition is already true. That is what a ramp/soak start needs — one clean pulse into the control word. The manual's example rung for a table based at V1500 is exactly that: a PD contact setting B1533.0, the start bit in the ramp/soak control word.
The JMP instruction in stage logic breaks that pulse in a documented way: when a JMP executes, the former stage stays on for three more scans. Your new stage starts trying to issue its start pulse while the old stage is still live, and the old stage may still drive rung conditions or control bits through that window, so the pulse never lands or is cleared before it counts. From outside it looks like an internal conflict between bit settings; it is scan overlap, and it repeats every time the handoff runs. That is also why a ramp forced from the software works while the ladder start fails — a forced write does not depend on the edge at all.
Replace the jump with set and reset logic #
Make the handoff explicit so exactly one stage is on in every scan:
- Identify the two stage bits in the press-close handoff: the pressure-control stage that is on now and the profile stage that takes over.
- On the transition condition, reset the old-stage bit and set the new-stage bit in one rung group, so the handoff completes inside a single scan.
- Keep the PD one-shot in the new stage, driven by the press-closed edge, still setting B1533.0.
- Audit every rung in the program that sets or resets the ramp/soak control bits, so no other stage can clear the start request during or right after the handoff.
- Leave the PID instruction and the profile table untouched: the handoff is the suspect until one of the checks below proves otherwise.
Rehearse the two-stage handoff in a scratch program before it touches the press — building a throwaway program just to prove transition logic is the whole point of the walkthrough on how to build ladder logic practice projects in STEP 7.
Verify the generator enable and the start bit #
Two bits must be right before any pulse can do anything, and you can read both live in the Memory Editor while the press sits in exactly the state where the profile should run:
- Bit 11 in PID V+0 is the ramp/soak generator enable; for a table based at V1500 that word is V1500 itself, and the DL405 manual lists it on page 8-57. If it is off, the generator is disabled and no start pulse matters — fix the rung or setup that owns the bit first.
- The ramp/soak control bits sit in the word at PID V+33, which works out to V1533 for this table — the arithmetic is in the next section. The PD one-shot writes the start request into that word as B1533.0.
Then watch the chain live: press-closed contact true, PD closes for one scan, B1533.0 sets. If the pulse fires and the bit never sets, some rung is overwriting the word; if the bit sets and the profile stays idle, go to the error flags.
Add the offsets in octal #
The PID data-area documentation mixes decimal and octal offsets, and the ramp/soak offsets are octal. Worked example for this table:
- Control word, PID V+33 octal from a base of V1500: add octal to octal, digit by digit, no carries — 0+3 = 3 in the units column, 0+3 = 3 in the eights column, 5+0 = 5, 1+0 = 1 — which gives 1533, so the word is V1533. Decimal check: octal 33 is (3 x 8) + 3 = 27, and 1500 + 27 = 1527 would point you at V1527, a different word.
- Error flags, PID V+35 octal: the same arithmetic gives V1535, and octal 35 is (3 x 8) + 5 = 29 decimal.
The decision: read the numbering the manual's table uses before you add anything, and convert the offset to match it. The same digits added as decimal land on the wrong word.
Read the error flags before editing the table #
The word at PID V+35 octal — V1535 for this table — holds the ramp/soak programming table error flags, described bit by bit on page 8-58 of the DL405 manual. Read that word before you edit anything: a set flag names the entry or condition the controller rejected, which beats re-checking values by eye. Correct what the flag points at, clear the cause, and retest the start. If no flag is set and the profile still will not run, do not assume the table is valid — compare every entry against the documented formats below.
Check every table value in its own format #
The profile table does not use one number format for all four values:
| Table value | Required format |
|---|---|
| Ramp SP end point | decimal |
| Ramp slope | BCD |
| Soak duration | BCD |
| Soak PV deviation | decimal (optional) |
Worked example: you want a soak duration of 15. In BCD the word must read 0015 in hex — 0001 0101 in the low byte, one nibble per digit, '1' then '5'. Typed instead as the plain binary value of 15, the word reads 000F — 0000 1111 — and the F nibble is not a decimal digit, so the entry is invalid and the error flags catch it. Read the actual words in the programming software and re-enter each value in its documented representation. Do not change units or endpoints to make a bad encoding appear to work: an invalid slope or duration can hold the whole profile idle with bit 11 on and the start bit set, which is exactly why this check comes before you touch the PID instruction.
Test the ladder start, not a forced ramp #
DirectSOFT's PID view writes the same control bits your ladder writes — it does nothing special that ladder cannot do. A forced run is therefore a useful table test and a useless handoff test: if a forced ramp works and the ladder start does not, the fault is in the stage transition or the one-shot, not in the PID instruction.
Run the corrected sequence in order and watch each link:
- The press-closed condition goes true in ladder.
- The PD contact closes for one scan.
- B1533.0 sets and holds.
- The error flags in V1535 stay clear.
- The setpoint leaves its dwell and begins to ramp.
Prove the sequence with the machine locked out to your site's safe-setting procedure or with the ram clear, before a ladder-driven setpoint moves hydraulic pressure under load.
The decision, case by case #
- Forced ramp works, ladder start does not: the stage handoff is the fault — replace the JMP with SET/RST logic and retest before touching anything else.
- Generator enable off: the generator is disabled — fix the rung or setup that owns bit 11 in V1500, then retest.
- The PD pulse fires but B1533.0 never sets: another rung is overwriting the control word — audit every set and reset on the ramp/soak bits.
- B1533.0 sets and the profile stays idle: read the error flags in V1535 and re-enter the flagged value in its documented format.
- Everything above checks out: only now edit the PID instruction or the table values, because the start logic was not the problem.
In the press program behind this article the jump was the whole fault. These checks exist so the next profile that refuses to start costs you minutes, not days.
Questions asked
How do I catch a one-shot pulse that only lasts one scan?
Latch a spare bit from the same PD contact for the test and clear it at the end of the test, so the display shows whether the pulse fired at all. If the latch never sets during the press-close transition, the old stage is still smothering the edge and the handoff needs the set/reset fix.
Why is 000F an invalid BCD value, and how should 15 be entered?
BCD stores each decimal digit in its own four-bit group, so a digit can only be 0000 through 1001, which is 0 through 9. F is 1111, not a digit, so a word reading 000F is invalid and the error flags catch it. A soak duration of 15 belongs in the word as 0015; read back as plain binary that is 21, so trust the BCD display in the programming software.
Where are these bits documented on the current DL205 controllers?
The D2-250-1, D2-260 and D262 user manual covers PID ramp/soak in its chapter 8, programs the ramp/soak control in relay ladder, and lists the ramp/soak flag bits in a table on page 8-60. The DL405 manual behind the addresses in this article documents the generator-enable bit and the control bits on page 8-57 and the error-flags word on page 8-58.
Do I need to retune the PID loop after fixing the ramp/soak start?
No. Ramp/soak only shapes the setpoint the loop tracks; gains, deadband and output limits are untouched by replacing a stage jump with set/reset logic. Retune only if you changed the profile itself, because a steeper ramp slope asks more of the loop than the tuning was proven against, and validate any changed slope on the machine before running product.