Electronics > Beginners
RC glow plug driver and voltmeter: constructive criticism needed
Rerouter:
Sat down and drew up my op amp based approach, Its still there if the op wanted something analog
http://tinyurl.com/ybuqrljt
Full link below if it expires.
--- Code: ---http://www.falstad.com/circuit/circuitjs.html?cct=$+1+0.000005+10.20027730826997+58+5+50%0Aa+-176+144+-64+144+0+5+0+1000000+0.04950889799120179+4.036530011794033%0Aw+-64+144+-64+80+0%0Ar+-176+208+-64+208+0+100000%0Ar+-176+80+-64+80+0+100000%0Ar+-176+80+-288+80+0+1000%0Ar+-288+208+-176+208+0+1000%0Aw+-176+160+-176+208+0%0Aw+-176+80+-176+128+0%0Aw+-176+208+-176+256+0%0Ax+702+293+790+296+4+24+Setpoint%0Ag+-288+80+-320+80+0%0AR+-288+208+-320+208+0+0+40+5+0+0+0.5%0Aw+-64+208+-64+144+0%0A181+32+256+32+336+0+1065.4064433491974+25+5+1+1%0Ag+32+336+32+352+0%0Aw+-176+256+32+256+0%0Af+96+176+32+176+1+1.5+2%0Aw+32+192+32+256+0%0AR+32+160+32+128+0+0+40+5+0+0+0.5%0Ac+288+80+288+144+0+0.00001+0.34995086109383294%0Ag+288+144+288+160+0%0Ar+288+304+192+304+0+10000%0Ar+368+304+288+304+0+10000%0Ar+192+208+288+208+0+10000%0Ar+368+272+288+272+0+1000%0Ac+368+208+288+208+0+0.000001+-3.4175819977216473%0Ag+368+208+368+224+0%0Aw+288+272+288+304+0%0Aw+192+256+192+208+0%0Aw+192+256+192+304+0%0Aw+288+208+288+240+0%0Aa+288+256+192+256+0+5+0+1000000+3.4175819977216473+3.236797676118947%0Aw+288+80+560+80+0%0Aw+560+80+560+256+0%0Aw+96+176+192+176+0%0Ag+368+304+400+304+0%0Aa+560+272+464+272+1+5+0+1000000+0.34991378270596496+0.34995086109383294%0Ac+464+336+560+336+0+1.0000000000000001e-7+3.357925004091883%0Aw+464+336+464+272+0%0Aw+560+336+560+288+0%0Ar+624+288+560+288+0+100%0AR+624+288+672+288+0+0+40+0.35000000000000003+0+0+0.5%0Aw+192+176+192+208+0%0Aw+464+272+368+272+0%0Ar+368+336+288+336+0+10000%0AR+368+336+400+336+0+0+40+5+0+0+0.5%0Af+192+176+192+80+8+1.5+0.02%0Aw+-64+80+176+80+0%0Ax+243+362+431+365+4+24+Minimum%5CsOn%5Cstime%0Aw+288+336+288+304+0%0Ar+208+80+288+80+0+10%0As+32+256+96+256+0+1+true%0AR+96+256+128+256+0+0+40+5+0+0+0.5%0Ao+8+64+0+4099+10+0.0125+0+2+8+3%0Ao+28+64+0+4099+20+0.0015625+1+2+28+3%0Ao+13+64+1+20491+5+12.8+2+1+40%0A
--- End code ---
fmzambon:
Thank you very much Rerouter, that circuit will be very useful :-+
fmzambon:
Hello everyone,
I'm resurrecting this old thread (has it really been one and a half years???) for the benefit of anyone who might find it while looking for a similar circuit.
I ended up building the circuit of the first post. It worked as intended, except for a minor goof-up.
R5 and C1 are intended to form a low-pass filter to smooth out any high frequency noise on the supply voltage. I had thought "470R, 100nF, yeah, an RC filter with a 47ms time constant should be enough to filter out high frequency noise and still have a reasonably fast response for the voltmeter". Except that 470R and 100nF give a time constant of 47us, not 47ms :palm:
Oh well, what's three orders of magnitude?
I didn't spot the mistake right away because the older analogue RC receiver that I used for the tests seemed to have a farly constant current requirement, so the low pass filter wasn't really called into action. But the newer digital receivers seem to have a pulsed power consumption.
The first attachment is the situation with the original component values. The upper trace is the signal before the filter, the lower trace is after. 1X probes, so that's a 70mV voltage dip. The power comes from a 4 cell NiMH battery, so nominal 4.8V
Changed R5 from 470R to 47k and C1 from 100n to 1u (didn't want to go to 470k because I feared that that could be too much impedence for the LM3914). The second image is the result.
When i wiggle the controls to cause some "real" current spikes from the servo motors, the meter still seems quite responsive. So I'd say it's a good compromise.
The last picture is the actual circuit installed in the airplane. Yes, I did fly it, and it did work as intended both on and off the ground :-+
Some improvements should I ever get back to this (or should anyone consider replicating this circuit):
* I included the 5mm round green LED as a power indication, but it is fairly redundant given that one led of the bar is always on. So that could be omitted and replaced with a simple resistor;
* Perhaps one could add a capacitor across the TL1431 voltage reference, as I found that some noise gets through. Obviously it's necessary to read the datasheet to know what values of capacitors will not cause oscillation issues and do some experiments.
And finally, I haven't forgotten the suggestions for more advanced control schemes, such as sensing the glow plug voltage and using that to adjust the amount of power supplied. I fully intend to do that, but it's going to be another project :)
Edit: goofed up the attachments :--
james_s:
Good to see a followup, a lot of threads like this end with no update on the outcome. Looks like a tidy layout, do be careful with those screw terminals though, they have a tendency to work loose. The vibration of a small glow engine also tends to be very good at loosening fasteners.
fmzambon:
--- Quote from: james_s on December 17, 2019, 06:57:44 pm ---Good to see a followup, a lot of threads like this end with no update on the outcome. Looks like a tidy layout, do be careful with those screw terminals though, they have a tendency to work loose. The vibration of a small glow engine also tends to be very good at loosening fasteners.
--- End quote ---
I've already used screw terminals in this very same model, and never had an issue. Perhaps it's because its engine is a twin cylinder boxer engine, which does not vibrate all that much?
Still, I'll keep an eye on them :-+
Navigation
[0] Message Index
[*] Previous page
Go to full version