Electronics > Beginners

RC glow plug driver and voltmeter: constructive criticism needed

<< < (4/5) > >>

fmzambon:

--- Quote from: oldway on April 23, 2018, 02:26:21 pm ---Yes, it is....
What do you think about it ?

--- End quote ---

About the project or about the text-wall first post?  ;D
Joking aside, I need to check that project more carefully. I'm at work now (waiting for an automated test procedure to end), so I don't have that much time at the moment.

fmzambon:
@oldway: So, I read through that thread I linked a few posts back.

It seems like that project is trying to do MUCH more than I'm trying to do here. Hall effect sensors to detect glow plug current, cylinder temperature sensors and, in the end, barometer, airspeed sensor, Bluetooth connection...  :scared:
Basically a miniature ECU, it's only missing fuel mixture control.




--- Quote from: Rerouter on April 23, 2018, 12:24:54 pm ---My thoughs for dumb analog would be

1x Op amp - Howland current source (He provides the measurement current
1x Op amp - Oscialltor (He switches between measure and heat)
2x Op amp - Sample and hold (He holds the measurement result, and gets reset by the oscialltor)
1x op amp - Comparitor (If your below your desired voltage, turn on the heater for the heating cycle)

So how you arrange it is up to you, but for the howland current source, its very easy to make them for above ground, you can probably leave him running if the current is small, the plug heat will be based on its resistance,

So now the meat of it, the oscialltor disables the comparitor (probably via summing with one of its inputs) for the measuring phase, and updates the sample and hold,
Then for the heating phase the sample is fed to the comparitor

This would be a loose PWM, mainly because based on thermal mass, I'm going to guess the response time is multiple seconds, not fractions of a second,
Equally there is no reason the duty cycle has to be 50/50,

I will say this way is fairly idiot proof, if the wiring fails open, it thinks its over temp and never heats, and if its shorted, well its not heating the plug.

--- End quote ---

I was thinking some more about this suggestion. It may not be ideal to incorporate it in the current project, as the driver that I'd like to build is powered by a NiMH cell, so no PWM would be used here. In fact PWM would interrupt power to the plug, thus reducing the RMS voltage applied to the plug itself, which is already marginal.

However in a few months I'll build a glow plug driver for a 5-cylinder radial engine and I was planning to use a lithium battery for that (LiFe battery most likely). Hence I'll need to PWM the plugs and then your idea would mesh in nicely.
If I may, for the first 2 points I think that one may not need "explicit" op amps: I already experimented with the LM334 current source, so maybe I could use that instead of a Howland current source as you proposed (but I didn't know that current source type, so that's nice to know  :-+ ). And the oscillator could very well be a 555 or any other oscillator chip.
In fact I could configure the final op amp as a differential amplifier rather than as a comparator and feed the output of that into the control voltage input of the 555. I need to think more about this, but it could work.


As an aside, I have a clear feeling that I have read kikinou's (the outhor of the other, linked thread) original attempts at powering a glow plug with a PWM waveform using a 555 timer, but I can't find that thread anymore. I clearly remember that someone was burning plugs and that was because he assumed that for a PWM signal Vrms = D*Vin (D is the PWM duty cycle), whereas the correct formula is Vrms = sqrt(D)*Vin. Hence he was using too high a duty cycle and burning out plugs.
I already built a PWM glow plug driver based on a bipolar 555 astable and a bank of mosfets for my engine test stand, and I can confirm that it works just fine. If one uses the correct duty cycle, that is. And if one solders the 555's bypass capacitors correctly... (why is it running at 30+ kHz rather than the designed 350Hz?  :palm: )

oldway:
I can not comment on either your project or that of Kikinou because my experience in model aircraft dating from my youth and we made the circular flight with the engine still at full throttle ...

Apart from the webra boxer 2 cylinders, I do not remember multi-cylinders engines at the time.

My favorite engine was the 2.5cc Cox Olympic, a very powerful engine that ran at 20,000 rpm.

I have to make a suggestion: change the title of your topic so that we know what it is about.

fmzambon:

--- Quote from: oldway on April 23, 2018, 08:00:56 pm ---I can not comment on either your project or that of Kikinou because my experience in model aircraft dating from my youth and we made the circular flight with the engine still at full throttle ...

Apart from the webra boxer 2 cylinders, I do not remember multi-cylinders engines at the time.

My favorite engine was the 2.5cc Cox Olympic, a very powerful engine that ran at 20,000 rpm.

I have to make a suggestion: change the title of your topic so that we know what it is about.

--- End quote ---

Done  :-+

james_s:
A friend of mine has a nitro radial engine for a model airplane, it's an impressive piece of craftsmanship but he's never flown it in a plane, it's too valuable. He had a good point that with a multi-cylinder glow engine it's possible for one or more cylinders to die without it being obvious to the pilot on the ground, until you need to suddenly open the throttle and realize not all the power is available. I suppose that's what this proposed circuit is to prevent.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod