Ok latest testing : for target 30k
Left series resistance in Kohms
Right Series resistance in Kohms
Temperature (deg )
28.3
28.8
32.5
28.7
28.7
32.5
28.8
29
32.25
28.9
28.9
32.125
29.2
29.2
31.875
29.3
29.3
31.75
29.8
29.7
31.5
29.6
29.5
31.5
29.5
29.6
31.5
29.8
29.8
31.379
Last edited:
Sadly my progress has been slooow what with the summer months but your efforts are far better than mine, almost as if you are the pro and me the hobbyist ;-).
The board is nice and compact and eludes to more to come, I see an OLED interface and IR control added to the web interface from earlier. Cool.
What RPi is required, I guess a zero w might be viable. Would also be interested in seeing a resistance range plot.
Did you experiment with thermally linking the LDRs in the end?
Is the I2C bus available, ie for a relay interface (input selector)?
Keep up the good work and if you want a tester?
The board is nice and compact and eludes to more to come, I see an OLED interface and IR control added to the web interface from earlier. Cool.
What RPi is required, I guess a zero w might be viable. Would also be interested in seeing a resistance range plot.
Did you experiment with thermally linking the LDRs in the end?
Is the I2C bus available, ie for a relay interface (input selector)?
Keep up the good work and if you want a tester?
Hi Garry
we are now in full testing of the LDR boards. We see some crazy stuff....more on it after testing . (with T vs R)
Anyway we are getting closer and closer. At 30k calibration we get 29.9k (measured on photoresistors with a 0.1% precision voltmeter) ..but excluding T. Today we will get the exact skew of T vs R...but its very big. I am not sure why it was never implemented...its crucial .
Yeap we will need a tester. Will PM you.
we are now in full testing of the LDR boards. We see some crazy stuff....more on it after testing . (with T vs R)
Anyway we are getting closer and closer. At 30k calibration we get 29.9k (measured on photoresistors with a 0.1% precision voltmeter) ..but excluding T. Today we will get the exact skew of T vs R...but its very big. I am not sure why it was never implemented...its crucial .
Yeap we will need a tester. Will PM you.
So far I have got good constant current control of the LDR via a DAC and current source opamp. Where I was going next was getting a ADC to perform the calibration. This works but yep, that was my findings too... the temp skews the calibration. Having said that this is where my project has paused for now, I was looking to get better accuracy before looking a forms of compensation.
On thought was to try and maintain a constant temp perhaps by using a heating resistor with the LDR thermally linked and choosing a target temp higher than ambient. Or go the other way, cooling with a Peltier device. But it is all getting a little silly.
For now I have been running the Arduino LDR, with OLED and source selection on a regular basis and been happy with its day to day operation, dispite clicks and bumps between certain volume steps which away from the bench go unnoticed.
On thought was to try and maintain a constant temp perhaps by using a heating resistor with the LDR thermally linked and choosing a target temp higher than ambient. Or go the other way, cooling with a Peltier device. But it is all getting a little silly.
For now I have been running the Arduino LDR, with OLED and source selection on a regular basis and been happy with its day to day operation, dispite clicks and bumps between certain volume steps which away from the bench go unnoticed.
ok so after more testing...what we see is a clear trend.
Calibration done at 29C ---30K resistance. (read manually 30.1k)
At 30C ...we read manually photo resistor being at 27.K
At 31C ...we read manually photoresistor at 24.7K
At 32C............................................................23K
Some of the results might be skewed because we waited a short time (3 min) for the rise in T and reading (so maybe the heat did not fully propagate)...but what we see so far is a skew of 2000R==2500R per 1C
Calibration done at 29C ---30K resistance. (read manually 30.1k)
At 30C ...we read manually photo resistor being at 27.K
At 31C ...we read manually photoresistor at 24.7K
At 32C............................................................23K
Some of the results might be skewed because we waited a short time (3 min) for the rise in T and reading (so maybe the heat did not fully propagate)...but what we see so far is a skew of 2000R==2500R per 1C
That's at 30k. What might the figures be at, say, 10k and 50k? Is this drift a function of the resistance as well as the temperature? Might it be, for example, a drift of a percentage of the the resistance? So if it's 2k per 1C at 30k (6.7% per 1C) would it be 3.3k at 50k?
Both the series and shunt resistors will be affected by the temperature. If your principal interest is simply the ratio of the shunt and series then *if* the drift is a percentage then they will cancel each other out. (The source will see a change of impedance, of course)
In any event, there is good evidence for ensuring that the two LDRs are at the same temperature.
Might I say, I am following your work with interest (as I have such a project on my 'to-do' list sometime) and look forward to more results from your investigations. Thanks for your work.
Geoff
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.