Gina Posted October 20, 2016 Author Share Posted October 20, 2016 Heavy dew tonight and the dew heating for the dome is not sufficient - not really surprising as it's only a quarter of what it was designed to be! Link to comment Share on other sites More sharing options...
Gina Posted October 20, 2016 Author Share Posted October 20, 2016 I'm wondering why the BC517 isn't saturating. I'm feeding it with +3.3v logic "1" form the GPIO pin through a 6.8K resistor. Being a Darlington transistor the Vbe is going to be twice the usual 0.7v so 1.4v. Current input would therefore be (3.3-1.4)/6.8 = 0.28mA. Collector current required is 200mA and the datasheet says it saturates with 1v Vce with 0.1mA base current for 100mA collector current so more than doubling the Ib should easily double the Ic I thought. Evidently wrong and not enough to saturate, so I'll increase the current. If I reduce the base resistor to about 1K the current will be 1.9mA - that should saturate it well and truly without overloading anything. Link to comment Share on other sites More sharing options...
Gina Posted October 20, 2016 Author Share Posted October 20, 2016 There's something wrong - even after replacing the BC517 with a new one, it's still showing 7v on the collector and that's even without any base drive. Link to comment Share on other sites More sharing options...
Gina Posted October 20, 2016 Author Share Posted October 20, 2016 Been doing further testing and found out a couple of things. Firstly, when first switched on the GPIO outputs 0.6 to 1.0v - the output is neither logic "0" nor logic "1". In this case the BC517 is driven partly on causing a power dissipation far in excess of its rating - I have a small burn on my finger to prove it!! Once the INDI driver is connected in Ekos the voltage on the GPIO pin takes up either 0v or 3.3v (or a tad lower). I thought the GPIO pins were open circuit until the software enabled them as output but apparently not - there is sufficient current to turn the BC517 partly on The upshot of this is that I shall need to redesign the operation of the dew heater control. Tomorrow is another day... Link to comment Share on other sites More sharing options...
Gina Posted October 20, 2016 Author Share Posted October 20, 2016 Had a quick look at the RPi GPIO spec and it seems some pins have an internal pull-up resistor of 1K8 - specifically pins 3 & 5 and I think it's pin 5 that is used for the Astroberry Board Line A and which I'm using for dew heater control. That would explain it. And so to bed... Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 Further research shows that the pins 3 & 5 mentioned above are the chip pins not the GPIO numbers but there may still be a different value pull-up resistor present. The logic of the Astroberry Board INDI driver is actually OFF=+3.3v and ON=0v so I need a different design anyway. Negative logic referenced to the +3.3v rail. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 Removed BC517 and base resistor and powered up. GPIO5 shows 3.13v. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 Here is a new design for the circuit to control the dew heater. NPN transistor and P-Channel power MOSFET. The 10K resistor prevents killing the RPi if the MOSFET should burn out. It also stops parasitic oscillation that can occur with a direct connection. ATM I can't think of a simpler circuit. Unusually the NPN transistor is being used in common-base mode. In the ON state GPIO pin 5 is at 0v and the 2K7 emitter resistor produces about 1mA current which would produce a 1mA collector current in linear mode but in this case the 22K collector resistor means that the transistor saturates and just over 8v is dropped in the 22K resistor. This is well above the "turn-on" -5v gate to source voltage but also well within the -20v max Vgs limit, so the MOSFET turns on, powering the dew heater. In the OFF state GPIO-5 is above 3v and the NPN transistor receives no emitter current as the Vbe is less than 0.6v so the collector current is also zero, the Vgs on the MOSFET is 0v so the MOSFET is OFF and hence dew heater is off. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 A modification of the dew heater circuit could be used to switch the cooling on/off. By rearranging the transistor connections the collector can be made to go down to nearly 0v in the ON state and should be enough to switch the MOSFET which needs -4v Vgs minimum to turn on fully. Only problem is that I'm running out of space on the HAT. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 If I move the step-down buck converter off the HAT and put it to the side I can free up a lot of space and make room for both dew heater and cooling control circuits. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 I don't seem to have any spare NPN transistors left so I've ordered some more. Meanwhile, I've just connected the bottom end of the heater connector to Gnd so the dew heater is on all the time the power is on. That will do for the time being while I construct another HAT. Once that's finished I'll swap HATs and incorporate the present one in a box together with fuses etc. for the main imaging rig. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 ASC outside front window with mast hooked onto gutter and camera 3ft above. 30s exposure. Link to comment Share on other sites More sharing options...
Gina Posted October 21, 2016 Author Share Posted October 21, 2016 60s exposure. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Been playing with KStars/Ekos and have got the ASC images recording as FITS files - worked it out without reading the manual I set up a schedule in Ekos of 100 frames with minimum exposure of 0.001s (appears to be the software limit - camera goes down to 32µs) with a 1s delay. This was then added to the Schedule list by clicking + in the Sequence Queue. Clicking the ► starts the sequence. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Sun is setting, sky is almost completely clear and looking good for imaging tonight I've set up the camera orientated with north upwards. This removes the distant hill with street lights from the FOV so should be better for night viewing. With the current software/driver there's no auto-exposure and a lower limit on exposure of 1ms so not suitable for unattended operation for the whole 24 hours like I can get with Windows software. I shall be enquiring of the authors of the INDI library of drivers if this limit can be altered to suit these relatively new ZWO cameras. Or I might see if I can alter the source code myself. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 This is a screenshot of the video streaming preview image. Exposure now 3s, auto-gain and getting a new image about every 30s. Stars are starting to show. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Looking good tonight. Milky Way showing well. This is a 30 exposure, auto-gain screenshot of Preview in Ekos. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 And here's a FITS image converted to TIFF and slightly gamma stretched in Photoshop. Full resolution. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Here is a 60s FITS image converted to TIFF, stretched a little and saved as PNG. Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Still capturing decent images - here's no.20 Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Still going Link to comment Share on other sites More sharing options...
Gina Posted October 22, 2016 Author Share Posted October 22, 2016 Cloud coming in - looks like my main imaging of Cygnus will be coming to an end soon This all sky camera stays out all night though regardless Link to comment Share on other sites More sharing options...
Gina Posted October 23, 2016 Author Share Posted October 23, 2016 Total cloud cover tonight without much variation in density. Link to comment Share on other sites More sharing options...
Gina Posted October 23, 2016 Author Share Posted October 23, 2016 Rain drops on the dome. So the heating is not enough to dry rain drops off. I might add another heater with controller - I'll think about it and also see if there's room on the HAT. Link to comment Share on other sites More sharing options...
Starlight 1 Posted October 24, 2016 Share Posted October 24, 2016 Hi Gina, still thinking do I get a made one ? or make one that save me 50 % in price. Only thinking as you are on mk 5 and the time you have put in ? PS I see this in FLO under the sky cam . Starlight Xpress Oculus 'OcuPI' Wireless Control Module Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.