Increased exposure to 1ms.  Camera temperature -0.2°C.

This seems to be a plethora of vapour trails!  The sun has set behind the trees here but still shining on the far hill to the east.

"Huston - we have a problem"...  Something is wrong.  Firstly the camera stopped capturing so I reset everything and now having comms problems "broken pipe" error from Terminal.  I'll go out and switch off the power and look at it tomorrow.

Power off and all safe.  So it's fault finding tomorrow!!  Groan!! 

Collecting data from earlier - from page 17 :-  The circuits for the dew heater and Peltier TEC camera cooler.  The same arrangement of components is used for both, but heater is controlled by GPIO12 and the Peltier TEC by GPIO13.  To run the cooling from a different voltage during daytime I shall have to add extra circuitry.  Probably using P channel power MOSFETs rather than N channel but I'll have to think about it.  I shall also have to run a test to see how much cooling is wanted during daytime.  Hopefully I can use the +5v supply.  One control line should suffice - just switching between daytime and nighttime with long exposures (low and high cooling).

Been out to observatory, disconnected the cables, opened roof and taken ASC down.  Now indoors awaiting testing and fault-finding.

Set up indoors on table and it's all working except for the dew heater.  The cooling is constantly on because the Astroberry driver is not selected ATM.  I'll fault-find the dew heater with my DMM.  I want a lower amount of cooling for daytime and will work on the circuit design.  I'm not happy with the in-line connector I'm using for the cooler and dew heater so I think I'll find something better.

Added the Astroberry Board driver to the startup command and it's working.  I altered the code to suit my controls for camera cooling and dew heater (the advantage of open source).

Been checking the dew heater control circuit and the fault is between the opto-coupler and the MOSFET so I'll shut down and examine it close up.  I want to change the cooling control circuit so I think I shall re-work that area of the RPi HAT - it's very crowded.

Here's a circuit diagram for the camera cooler control.  With GPIO13 at logic 0 (0v) the opto-coupler is off and the P channel MOSFET J438 gate is at 0v with respect to the source and no current is supplied to the TEC from the 13.8v supply but current is supplied through a diode from the +5v line producing a small amount of cooling.  When GPIO13 goes high, current flows through the LED in the opto-coupler, turning the photo-transistor on and taking the MOSFET gate to -13.5v w.r.t. the source turning it on.  The TEC is now supplied from 13.8v and on full.  The diode to the +5v line is reverse biased.

With this circuit the control logic is inverted compared with before and the driver code will need altering to suit.

With the Peltier TEC now being connected to power Gnd instead of the +13.8v rail and the dew heater to the +13.8v rail, I shall need 4 wires from the control box to the ASC unit instead of 3.  I have 4 pin connectors same as I use for the power on my astro imaging control boxes (originally bought for a project I didn't pursue).  These are screw on types like the power connector on the EQ8 mount (except that has 2 pins).  So that's the connectors decided.

Found the reason the dew heater wasn't working - broken connection.  Now to check how much cooling is required for daytime to keep the camera at a reasonable temperature.

Test begun.  Starting temperature 23°C.  Peltier TEC connected to 5v.  Camera temperature coming down... 16°C...  12°C...

Dropped voltage to 4.6v to represent the +5v rail with a Schottky barrier diode in series with the Peltier TEC.  Temperature has crept up to 12.3°C but I doubt it will go up more than a few tenths more.  Imaging at 32µs with gain at 0 as usual for daytime.  Cooler is in still air at around 23°C.  Camera is now reporting 12.5°C.

Camera temperature going down again - still imaging - maybe the ambient has dropped a bit - cloud has arrived.  Camera is now reporting 12.1°C.  No hot pixels or noise to be seen so this will do nicely. 

EDIT - Camera is now reporting 11°C.  Peltier TEC current is just under half an amp.  Cooler feels barely tepid.  Test Successful. 

If I were to use an ordinary rectifier diode it would drop around 0.9v so I've set the volts to 4.1 and the camera temperature has settled down the about 13°C.  Still satisfactory.  Found a couple of 1N4005s - 600v 1A.  Put one in series with the bench PSU and set it to 5v.  Camera still reporting 13°C.  Satisfactory.

Searched high and low for the 4 pin sockets without success then I thought of looking in my astro box and found one on a box I had originally set up as a dew heater distribution box which I no longer want.

Plug connections :-

1. Gnd - Green (Cooler -ve)
2. +13.8v - Yellow
3. Heater - Orange
4. Cooler - Purple

Socket connections (actually the socket has pins so maybe it's a plug ) :-

1. Power Gnd - Green
2. +13.8v - Yellow
3. Heater - Orange
4. Cooler - Blue

Chassis mounting connector (previously called socket) now mounted in side of box, +13.8v and power ground connected.  Next is to change the control circuitry for the cooler.  It looks like the dew heater circuit is easier to change so I might swap the controls over.  I already need to change the driver code for the reversed logic so I can change the controls over at the same time.

Been working on the control circuits on the RPi HAT.  Started by removing redundant components - mostly unused connectors, leaving a nice lot of space for the cooler control circuit.  Nearly finished but stopped for the night.  I'll take and post a photo tomorrow plus a layout diagram probably.

Circuitry completed and now testing.  A couple of photos.

Testing...

Imaging at 32µs with gain at 0 as usual for daytime.  Camera is reporting 12.3°C.  Lower level cooling satisfactory.

Switching to full cooling - nothing happening so I have something wrong.  Checking dew heater - not working - off to investigate - back shortly...

Found a dry joint which I've fixed.  Dew heater now working and so is the control.  One problem fixed.

Cooling control still not working and found reason is that GPIO12 which now controls the cooling is not going high.  Reading 0.15v so not shorted to Gnd.  Going to check the code...

Checked source code and it looks alright.  I'll shut down and unplug the HAT and check the GPIO pin with nothing connected.  A fault on the power side of the opto-coupler cannot affect the data side (RPi) so I can't have blown the GPIO - that's the whole point of using the opto-couplers.

Pin is at logic 0 when it should be at logic 1 (0.005v rather than 3.2v) so something seems to have happened to that GPIO circuit.  There are plenty of spare GPIO lines I'm not using so I guess I'll use another.  Or I could swap RPi boards.  I can try a new RPi board since I have a brand new unused one and see if that GPIO is working without plugging HAT onto it.

New board is the same but I think I'm getting confused with the pin numbers ?  Guess I need a break...

Designed and now printing spacers for mounting the ASC on the ROR.