Guru
Joined: 05/10/2019
Location: United KingdomPosts: 9022
| Posted: 07:45am 28 Sep 2022 |
Ah - it may have internal forward-biased diodes to VCC and you are seeing leakage. You'd be better using a simple BS170, 2N7000 or similar MOSFET on each output, driving the gate directly from the pin, then switching them on & off. You'll get a good open drain output then, at the expense of a bit of board space.
Guru
Joined: 18/11/2011
Location: United KingdomPosts: 4365
| Posted: 07:59am 28 Sep 2022 |
Should R4 be connected to the input (like the other resistors)?
John
Guru
Joined: 11/02/2018
Location: AustraliaPosts: 3348
| Posted: 08:04am 28 Sep 2022 |
Try another pin, perhaps that one has become faulty during testing.
Here is what I get.
> setpin 20,dout <---2.4mV
> pin(20)=1 <---3.3v
> setpin 20,din <---315mV of random noise.
>
Edit.
Are you certain the diode is cathode to +?
Edited 2022-09-28 18:09 by phil99
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 08:21am 28 Sep 2022 |
Sorry John I had left off the junction spot.
OA47
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 08:32am 28 Sep 2022 |
Was using pins 5,6 & 7 for the switching so tested with pin 20 (not used previously)
SetPin 20 dout.........0V
Pin(20)=1 .............3V3
Pin(20)=0 .............0V
SetPin 20 din .........0.5V
All measured to Analog GND
No circuitry attached to Pin 20
Similar results to Phil99
I am using a schottky diode and the anode is definitely to 3V3.
OA47
Guru
Joined: 05/10/2019
Location: United KingdomPosts: 9022
| Posted: 08:39am 28 Sep 2022 |
A DIN pin should be high impedance unless you've used the PULLUP or PULLDOWN option when using SETPIN. Anything else would point to external miswiring (or a damaged pin).
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 09:02am 28 Sep 2022 |
Humble pie for supper
The protection diode is no longer a protection diode but more of a resistor. Will chase round the workshop and see if I can find another.
OA47
Guru
Joined: 05/03/2018
Location: NetherlandsPosts: 6024
| Posted: 09:39am 28 Sep 2022 |
Don't forget: all DOUT pins must be DIN before you can measure how high impedance one single pin is....
Guru
Joined: 05/10/2019
Location: United KingdomPosts: 9022
| Posted: 09:55am 28 Sep 2022 |
Perhaps you need something a bit meatier than a OA47. ;)
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 10:17am 28 Sep 2022 |
Point taken. The original diode was a glass encapsulation.
OA47
Guru
Joined: 05/10/2019
Location: United KingdomPosts: 9022
| Posted: 10:27am 28 Sep 2022 |
Don't let that fool you nowadays. The 1N4148 / 1N914 is a brilliant diode, hence its popularity. :) I do like the old glass diodes though, having got through a few in my time. :)
Guru
Joined: 11/02/2018
Location: AustraliaPosts: 3348
| Posted: 11:39am 28 Sep 2022 |
Volhout said:
" For this diode to work, it should be a shottky diode. And these tend to have high leakage (1uA). "
Could that leakage be the problem?
Temporarily remove it, leave the input open circuit then measure the voltage on the pin in all conditions.
If that is it then some other protection measure will be needed.
Edit
Ignore that I missed the post where you found the problem.
Edited 2022-09-28 21:57 by phil99
Guru
Joined: 11/02/2018
Location: AustraliaPosts: 3348
| Posted: 10:47pm 28 Sep 2022 |
Assuming this input attenuator is for a scope, frequency compensation will be needed to get the best out of it.
Since the Pico input has capacitance you need a capacitive divider that matches the resistive one.
Guru
Joined: 18/11/2011
Location: United KingdomPosts: 4365
| Posted: 07:14am 29 Sep 2022 |
Roughly what value capacitor?
John
Guru
Joined: 07/08/2011
Location: AustraliaPosts: 6570
| Posted: 07:26am 29 Sep 2022 |
To answer that, you have to know the input capacitance of the device the probe is connected to.
Jim
Guru
Joined: 05/03/2018
Location: NetherlandsPosts: 6024
| Posted: 07:32am 29 Sep 2022 |
I am not sure OA47 is going to build an oscilloscope, so the capacitor may not be needed.
About the capacitor: If you want to measure 50/60Hz or DC it is not needed.
The capacitor is needed to balance against the input capacitance of the pico ADC, and the impedance of the circuit board. The balance (ratio) should be identical to the ratio of the resistor divider.
Since the resistor divider is switching, you will have to switch the capacitance also.
What they do in oscilloscopes is that they add capacitors everwhere. Practical values in your case would be (look at the schematics OA47 provides).
If C1 = 15pF, and the input capacitance of the pico and board layout is 7pF (estimate)
Capacitor parallel to R1 is 22pF (1M R1/1M R2 attenuator)
Capacitor parallel to R3 is 200pF (2x100pF) (1M R1/100k R3 attenuator)
Capacitor parallel to R4 is 2200pF (1M R1/10k R4 attenuator)
But again, for DC and 50Hz this is not needed.
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 03:51am 30 Sep 2022 |
Here is a screen shot of the project. Not a desktop DVM but a shelftop DVM, one I can monitor from many meters away.
OA47
Edited 2022-09-30 15:35 by OA47
Guru
Joined: 05/03/2018
Location: NetherlandsPosts: 6024
| Posted: 05:56am 30 Sep 2022 |
Nice project! A monitoring system.
And yes, readable from a distance...
Guru
Joined: 11/04/2012
Location: AustraliaPosts: 1050
| Posted: 06:03am 30 Sep 2022 |
The use of an old LCD TV makes things easy as it can supply the pico from USB, has VGA input and audio input all readily accessible and the system can be powered down via IR remote.
Here is a pic of the prototype pcb so far.
