DIY non invasive temperature gauge probe placement ideas...

[jaylee]

You all pretty much know i got sent what i paid for & it works..!

But i have ordered (just before the heads up.) a second unit.. It hasn't arrived yet!
When it does.. The plan is to "pan test" in conjunction with mine to make sure they are calibrated the same, then sent on for further testing to a certain gauge mad Bulgarian forum member... (Mad about gauges, not off the gauge scale insane in the membrane..)

Sorry you guys are getting the hassle from this company!


Kirsty; just for future reference... Could you define what you meant by "haywire"??

From earlier in the thread....

Just wondering if, of the faulty Sure units received, there happens to be one which has not been requested to be returned and whose owner would consider donating to research - either now or once functioning replacement arrives? If not, I'll go buy my own but if surgery is successful, I would gladly return the pimped unit to the donor...

Still hasn't arrived yet...! But when it arrives, if it fails a pan test with my working unit, & it is found to be "on the fritz"..?
I'll turn it over to you...

I don't think "the company" will want to talk to me any more anyway...?!!


Just an idea..! Do you think there could be any chance of tying in the alarm side of things with the existing beeper on the Bongo...??

[tallbongo]

Just an idea..! Do you think there could be any chance of tying in the alarm side of things with the existing beeper on the Bongo...??

Should be relatively simple.

[Driver+Passengers]

If it's just the display you can experiment on mine. I'd like to keep the probes though.

...if it fails a pan test with my working unit, & it is found to be "on the fritz"..?
I'll turn it over to you...

Thanks for the offers both - feel free to PM me if/when appropriate. Just to restate, I'm only after a software-fritzed (ie unsuitable) unit with no probes.

Just an idea..! Do you think there could be any chance of tying in the alarm side of things with the existing beeper on the Bongo...??

Should be relatively simple.

Personally, I wouldn't. If you've got to run a cable to that unit, you might as well not bother and just add a buzzer. But then I'm wanting a 'Benny Hill Buzzer' anyway...! On the other hand, if there's a spare pin and/or spare space, why not? Mine's unplugged atm.

[bikerbob]

would willingly donate my faulty Sure unit for research, regarding a re-programming of the unit, Sure sent by email attatchment a programme for this purpose if its of use to anyone your welcome to have it, it`s not any use to me, however would have to be guided through how to re-send to recipient.

[Driver+Passengers]

would willingly donate my faulty Sure unit for research, regarding a re-programming of the unit, Sure sent by email attatchment a programme for this purpose if its of use to anyone your welcome to have it, it`s not any use to me, however would have to be guided through how to re-send to recipient.

Cheers, Bob. PM sent...

[tallbongo]

When I got a working device I planned to put on an add on board with a buzzer. I would tap into the probes using a home-made Y-lead from the molex connector. One end would go to the sure device and the other the add on board, which would be something like this:

Not fancying working outdoors on my rock n roll bed in the remnants of a hurricane, this afternoon I knocked up a similar circuit from components I had lying around.

Knowing I have a unit that is faulty above 75 deg C I wanted to set the alarm at 70 deg C.

I used an old power supply that was 12.14V (nominally 12V). First I checked voltages across the thermistor at different temperatures. Note this time I got very good correlation between the two probes as I had them suspended in liquid. The extension I added made no difference to the temp readings (but the cheapo ebay leisure battery voltmeter did so I swapped to a proper multimeter).



So 70 deg C (343 deg K) is approx 0.9V.

I then made the previous circuit but without the capacitors and resistors on the op-amp power supplies. R1 was 1.8 Mohm, R2 was 2x68Kohm (all I had), and R3 a 45K pot (nominally 47K). I replaced the diode and buzzer combination on the output with a 5k resistor and a high output red LED. The op-amp was a TL071. Others are probably more suitable for driving a buzzer, but it's a pretty good all-rounder.

I then re-ran the test, with the probes as in pic below. Ch1 was unmodified, Ch2 went through my circuit with the alarm on it.



Here's the circuit



and here's a low bandwidth video of it working. You can see the LED above the display kick in at 70 deg C.

http://www.youtube.com/watch?v=owekmKw3f5Q

In conclusion, I'm now convinced an over temp alarm can be made for this device at very low cost. My circuit will need a voltage regulator due to the low differences in voltage across the thermistor at higher temps, but I think it should be possible to get it to alarm within a couple of degrees of the desired temp. From widdowson's info, about 93 deg C looks good, but I'd appreciate feedback on the best temp.

[nigelgibson]

Just returned this afternoon from an approx 8 mile journey towing large towavan twin wheel box trailer [ empty but still a big 6ft high 10 ft long 5ft wide thing with sticking brakes as well].

temps on my sure gauge were varying between 75 to 85 in the main but strayed to 90 to 91 when i stopped on way home. restarted after 5 mins and temp was at 97.5 when i switched on. dropped near enough immediately to 85 once i got moving and air flow recommenced.

My question is would an alarm set at 93 trigger too early ? my mason is set at 4.5 and didn't go off at 97.5 or in early nineties which is where it was on hills. i thought [ having seen jaylee's figures] mine was normal?

p.s love the work you're doing and put me on list for an alarm if you're marketing them!

[tallbongo]

Just returned this afternoon from an approx 8 mile journey towing large towavan twin wheel box trailer [ empty but still a big 6ft high 10 ft long 5ft wide thing with sticking brakes as well].

temps were varying between 75 to 85 in the main but strayed to 90 to 91 when i stopped on way home. restarted after 5 mins and temp was at 97.5 when i switched on. dropped near enough immediately to 85 once i got moving and air flow recommenced.

My question is would an alarm set at 93 trigger too early ? my mason is set at 4.5 and didn't go off at 97.5 or in early nineties which is where it was on hills. i thought [ having seen jaylee's figures] mine was normal?

I've only seen the graph in the Arduino thread and that's where I plucked the 93 deg from. I'm completely open with regard to what temp is best.

[Driver+Passengers]

You can see the LED above the display kick in at 70 deg C.
...
My circuit will need a voltage regulator due to the low differences in voltage across the thermistor at higher temps, but I think it should be possible to get it to alarm within a couple of degrees of the desired temp.

Great!! Well done - quick work!

Presumably you'd have to calibrate it in water and then install - what if you knock the pot though - you might never know... On another note, perhaps you could use a transistor or another op-amp (or the other channel of the same opamp?) to amplify the voltage and increase the swing, saturating the device at low temps. Nice to see both channels reading the same value - do they still do that when you switch the channel you tap from?

I'm still going to pursue the software approach, but it will mean that the threshold is pre-programmed.

Great to see that though - mightily impressed!

[Driver+Passengers]

My question is would an alarm set at 93 trigger too early ? my mason is set at 4.5 and didn't go off at 97.5 or in early nineties which is where it was on hills. i thought [ having seen jaylee's figures] mine was normal?

I'm not currently an alarm owner, but I believe folk do have their alarms set such they will go off when stopped for a while. Thing is - if you got to this temp while driving, perhaps you'd want to know and back off a bit to be safe - regardless of whether it peaked to this temp when idle/stopped. Like I said, don't own one, so don't really know.

Edit: I understand the thermostat control range is 82/96°C, with the fans kicking in beyond that to assist rad cooling. So in theory, you've still got headroom at these temperatures. But it depends on where you're sensor is, what your sensor is, etc... - "standard" readings for both the TM-2 location and the hose stub (metal) are still being collected as folk go along. YMMV.

[tallbongo]

Presumably you'd have to calibrate it in water and then install - what if you knock the pot though - you might never know...

Yes and no! Does the Mason not have a pot that you can adjust to set your personal alarm level?

It's looking like the thermistors are reasonably good, so there should be little voltage variation between units/probes. The alarm temp could simply be set by resistor ratios with no need for a pot. My original idea was to use the resistors but with a small pot added to allow a small amount of adjustment - say 5 deg or so.

On another note, perhaps you could use a transistor or another op-amp (or the other channel of the same opamp?) to amplify the voltage and increase the swing, saturating the device at low temps.

I'm not sure this would work. It would make the switching point more definite, but I don't think it would increase accuracy (just amplifying any errors?). I did think of putting a capacitor on the output to reduce the on/off oscillation at the switching point.

Nice to see both channels reading the same value - do they still do that when you switch the channel you tap from?

Yes. My previous test where I got a several degrees difference was a load of ****!

I'm still going to pursue the software approach, but it will mean that the threshold is pre-programmed.

Definitely do - there's a lot more possibilities. This was just a simple method that can be plugged in as an add-on without modification of the Sure. Does the PIC have a spare output you can trigger an alarm on? If there is a spare input the alarm needn't be pre-programmed. You could have a setup mode whereby the user can custom set the alarm temp with the addition of a push button.

[missfixit70]

My question is would an alarm set at 93 trigger too early ? my mason is set at 4.5 and didn't go off at 97.5 or in early nineties which is where it was on hills. i thought [ having seen jaylee's figures] mine was normal?

I'm not currently an alarm owner, but I believe folk do have their alarms set such they will go off when stopped for a while. Thing is - if you got to this temp while driving, perhaps you'd want to know and back off a bit to be safe - regardless of whether it peaked to this temp when idle/stopped. Like I said, don't own one, so don't really know.

Edit: I understand the thermostat control range is 82/96°C, with the fans kicking in beyond that to assist rad cooling. So in theory, you've still got headroom at these temperatures. But it depends on where you're sensor is, what your sensor is, etc... - "standard" readings for both the TM-2 location and the hose stub (metal) are still being collected as folk go along. YMMV.

My Mason is set to go off before the fans kick in @ 94 degrees C, I want to know that temps are getting towards fan switching point, so I keep an eye to make sure it does what it should. If it doesn't come down to some degree with first stage (slower speed) rad fans, chances are higher speed fans won't do a lot either as it's likely to be a rad, stat, flow issue that will not be improved by blowing more air over the rad.
If it alarms before first stage fans & they don't control the temp, then you've got a bit more time to do sumat about it, just my opinion
Great work guys

[missfixit70]

When I got a working device I planned to put on an add on board with a buzzer. I would tap into the probes using a home-made Y-lead from the molex connector. One end would go to the sure device and the other the add on board, which would be something like this:

Not fancying working outdoors on my rock n roll bed in the remnants of a hurricane, this afternoon I knocked up a similar circuit from components I had lying around.

Knowing I have a unit that is faulty above 75 deg C I wanted to set the alarm at 70 deg C.

I used an old power supply that was 12.14V (nominally 12V). First I checked voltages across the thermistor at different temperatures. Note this time I got very good correlation between the two probes as I had them suspended in liquid. The extension I added made no difference to the temp readings (but the cheapo ebay leisure battery voltmeter did so I swapped to a proper multimeter).



So 70 deg C (343 deg K) is approx 0.9V.

I then made the previous circuit but without the capacitors and resistors on the op-amp power supplies. R1 was 1.8 Mohm, R2 was 2x68Kohm (all I had), and R3 a 45K pot (nominally 47K). I replaced the diode and buzzer combination on the output with a 5k resistor and a high output red LED. The op-amp was a TL071. Others are probably more suitable for driving a buzzer, but it's a pretty good all-rounder.

I then re-ran the test, with the probes as in pic below. Ch1 was unmodified, Ch2 went through my circuit with the alarm on it.



Here's the circuit



and here's a low bandwidth video of it working. You can see the LED above the display kick in at 70 deg C.

http://www.youtube.com/watch?v=owekmKw3f5Q

In conclusion, I'm now convinced an over temp alarm can be made for this device at very low cost. My circuit will need a voltage regulator due to the low differences in voltage across the thermistor at higher temps, but I think it should be possible to get it to alarm within a couple of degrees of the desired temp. From widdowson's info, about 93 deg C looks good, but I'd appreciate feedback on the best temp.

Not sure if it'll make a difference to the result, but bear in mind, that while driving, with alternator outputting, the voltage will be more like 14 -14.6v.

[tallbongo]

The voltage regulator will take care of that. The fluctuations in battery voltage would have a massive effect so it's a requirement, not a nice to have.

[Rhinoman]

I'm still going to pursue the software approach, but it will mean that the threshold is pre-programmed.

Great to see that though - mightily impressed!

You could store the threshold value in EEPROM and then allow the user to change it via the serial link