Need help with CT6810p-517

[blaine1963]

I have found some doc's showing pin-out and electro/mech specs for 6810p / 6800HP

The pin-out seems to be:
M+, M-
unused unused
AGC_IN
PD-shield
unused PD_Out_A
PD_Out_B

I assume shield carries no signal, but just rf shields the weak current signals.

Diode common (is this common for AGC_IN and output PD's?)

I made a 20mA current source and fed it to AGC_IN (with respect to Diode Common).

I assume PD_A and PD_B are current sinks, and as such should be run through an op-amp I-2-V then V_PDA and V_PDB run through a instrument amp?

Does anybody have a SK of the PD's and IR sources for CT optical feed-back galvos? I have three, and I'm not afraid to take one apart.

[mixedgas]

I have found some doc's showing pin-out and electro/mech specs for 6810p / 6800HP

The pin-out seems to be:
M+, M-
unused unused
AGC_IN
PD-shield
unused PD_Out_A
PD_Out_B

I assume shield carries no signal, but just rf shields the weak current signals.

Diode common (is this common for AGC_IN and output PD's?)

I made a 20mA current source and fed it to AGC_IN (with respect to Diode Common).

I assume PD_A and PD_B are current sinks, and as such should be run through an op-amp I-2-V then V_PDA and V_PDB run through a instrument amp?

Does anybody have a SK of the PD's and IR sources for CT optical feed-back galvos? I have three, and I'm not afraid to take one apart.

I can help you when I get back to my hotel in 6 hours.

Steve

[blaine1963]

So, I opened one up.

It has two ir LED's (Vf about 1.3 v each) in series AG_IN is anode, PD_COMM is cathode.

The PD's are IR, large surface in a bridge configuration (two above vane, two below), the have 0.45 V Vf. All PD anodes are tied to PD_COMM (reverse biased mode).

When the rotor/vane is say full left, the upper left PD and lower right PD get a flood of IR, the opposite PD's get very little. When the vane is neutral, IR fly's all over the place. I was expecting a dual triangle vane, not a simple flat (grayed) vane.

I may not be up at midnight, but Steve, please respond.

[mixedgas]

So, I opened one up.

It has two ir LED's (Vf about 1.3 v each) in series AG_IN is anode, PD_COMM is cathode.

The PD's are IR, large surface in a bridge configuration (two above vane, two below), the have 0.45 V Vf. All PD anodes are tied to PD_COMM (reverse biased mode).

When the rotor/vane is say full left, the upper left PD and lower right PD get a flood of IR, the opposite PD's get very little. When the vane is neutral, IR fly's all over the place. I was expecting a dual triangle vane, not a simple flat (grayed) vane.

I may not be up at midnight, but Steve, please respond.

The AGC network is a linearity correction function of the Cambridge scanner amp. You may use your constant current source for now instead of the AGC, there will be some small amount of distortion in your scanned image without the correction, at wide angle .

I will leave the attached image up for 24 hours under the educational use clause, then take it down.



Steve

[blaine1963]

I had that "demodulator" SK, but thanks anyhow. Now I know what the 1N4148 was actually doing in a "what I thought was a linear" circuit.

I must have had a ground wire miswired or something, as I finally got a PhotoVoltaic (Voltage) output from the PHA and PHB signal. The "position demodulator" uses a voltage gain of -50k/1k (differential mode). The raw voltage from the photocell pairs is common mode of about -300 mV, with 100mV differential. So 100 mV * A(-100) is about +/- 10 V, and that is the ILDA analog input range for a system.

The SK (Ia, Ib) and General Scanning (current position sensors) made assume CT would be current output for the position (even though GS uses var. caps and not var. light).

Has anybody done a 3D (active shutter, circular polarization, Dolby dRdGdB)? That is what I have in mind for these scanners?

[blaine1963]

Steve do you have an table or equation to perform the position re-linerization in firmware?

I want my system to be maximally: digital, FirmWare, fast (feed-forward), accurate, re-configurable.

[blaine1963]

So if I replace the 1N4148 with some resistor, would I get the AGC, without the re-linerization (I'd rather do that in Firm Ware after the ADC)?

[mixedgas]

You don't want to solve the non-linearity in software. Most people can't see it. Its there for printing applications of the galvo. Just bias it at 26 ma. You'll have soon enough problems with solving the equation of motion and FIR filtering in real time. Hint, don't use PWM H-Bridge as the output stage.

Also watch your coil temperature, the coil is a few thousandths of an inch from the rotor. Over temp it and the copper expands into the rotor.



Steve

[blaine1963]

I really wanted full digital. What about Class-D amps will LC filters?

On the coil temp: I purchased three, one shows a nice 2.9 ohms at the coil, the fuse was open. I Opened the coil from the PCB, and still 2.9 ohms, but will not respond to signal-gen output (600 ohms), at +/- 15v. I assume the coil on this one is less copper and more carbon or at some distance, the cross sectional area is near mill. Since these damn things are assemblable, are they not disassemblable? I plan to return the bad unit, since the Ebay add said fully tested and functional, but for future reference, how does someone disassemble one? What is the damn spring near the vane? I assume it just keeps the bearings somewhat seated (and not generating too much mech noise)?

[mixedgas]

The problem with class D is going down to dc. You have to redesign the output filter to Dc, and they tend to have unstable poles/zeros down to dc.

The spring places a much needed preload on the rotor.

Did you replace the fuse on the bad galvo? It looks like a resistor.

Steve