CO2 laser instability diagnosis – tests completed and request for your technical opinion

[chrik99]

Hi everyone,

I’m trying to diagnose a very unusual CO2 laser problem and I’m hoping someone here has seen something similar.

The laser has **2 distinct operating states during normal firing**:

* **Stable state:** tube is quiet, PSU is quiet, and output power is good
* **Noisy/unstable state:** tube and PSU both make a high-pitched / buzzing / crackling sound at the same time, and output power drops

Important behavior:

* sometimes the laser starts stable and quiet, then after about **5 seconds** the noisy state appears
* once the noisy state starts, it **does not go away by itself** while firing continues
* I have to stop and restart
* sometimes restarting brings it back to a stable quiet state, sometimes not
* when the system is in the noisy state, the abnormal noise is present across the **whole usable current range**; the sound changes with current/amperage, but it does **not** disappear

A key observation is that the **tube and PSU are always synchronized**:

* when the PSU is noisy, the tube is noisy
* when the PSU is silent, the tube is silent

I uploaded a short video here that clearly shows the **noisy state** and how the **sound changes as amperage changes**:

https://youtu.be/r6nxMVo3qdk?si=PKgK9KwMYaSIh5p5

## What I tested

* Original tube: problem present
* New replacement tube: same problem
* Another new tube: same problem
* Original PSU: same problem
* New PSU: same problem
* Tube connected directly to PSU output (anode + cathode), bypassing the machine’s internal extension wires: same problem
* HV connector replaced: no change
* HV wires removed from the machine and physically separated/isolated: no change
* Cooling tubes separated from the HV wires: no change
* Cathode wire completely replaced with a new one: no change
* FG connected directly to main chassis ground: confirmed
* Ground continuity between PSU case, chassis, signal G, and main machine ground: confirmed
* AC input to PSU: about **117V to ground on each leg** and about **201V across AC1-AC2** (machine is on **120/208V service**)
* Ruida **L** signal: about **4.52V idle**, about **0.257V during firing**
* Ruida **IN** signal: about **0V idle**, about **4.24V during firing**
* **L / IN** behavior in quiet vs noisy firing: **same in both cases**
* **P** signal: no meaningful change
* Conclusion from controller signal testing: the controller signals look normal and appear **identical in both the good and bad states**

## Important note about the PSU TEST button

I also tested the local TEST function directly on the PSU.

What I observed:

* I do **not** see the tube visibly light up
* I do **not** measure any actual laser output power at the end of the tube
* at low and high knob positions, I hear nothing unusual
* at mid-range, I hear a slight crackling/buzzing from the PSU/tube area

So in my setup, the PSU TEST function does **not** seem to create a real laser firing condition. It seems to create some internal activity or partial HV behavior, but not visible tube glow or measurable output power. Because of that, I do **not** treat the PSU TEST result as a true standalone reproduction of the main fault.

## What I think so far

At this point I feel I have mostly ruled out:

* a bad tube alone
* a bad PSU alone
* external HV connector/wire routing alone
* a simple grounding omission
* an obvious controller signal issue

My main remaining theories are:

1. **PSU instability during real firing under load**
2. **Bad PSU/tube interaction under load**
3. **Machine-level electrical issue affecting multiple PSUs**
4. Possibly some form of **internal corona / partial discharge**, although I now doubt the external HV wiring is the main cause because I replaced or isolated so much of it already

## My question

Has anyone seen a CO2 laser behave like this:

* sometimes starts stable
* then flips into a noisy unstable state after a few seconds
* once noisy, stays noisy across the full usable range
* tube and PSU always make noise together
* controller signals look normal in both the good and bad states

Any thoughts are welcome. I’m deep into this one and trying not to keep replacing parts blindly.