Anode current and voltage measurement

[Paul Andrews]

I would like to be able to measure the anode voltage and current on one of my clocks and have a permanent display of those values built in to it. I came across this reference design at TI (http://www.ti.com/tool/tida-00528?jktype=design#technicaldocuments) but was wondering if anyone can point me at other possible solutions?

[gregebert]

Anode voltage is easy; just use a resistive divider to get the voltage within the range of your ADC, and be sure to put a filter cap (0.01uF should work) at the ADC input (most ADC inputs appear as a switched-capacitor).

Current is a different animal. The easiest way to dodge this is to use a current-limit circuit at each anode and you really dont need to measure it anymore.

Anode current can be measured at the cathode side by tying all of the cathode drivers together, then connecting to GND thru a sense-resistor. For example, if you use an HV5530 to drive your cathodes, connect all of the VSS pins together, then to GND thru a sense resistor. If you drive 3 tubes from single HV5530, your max current will be about 15mA. Next, select a sense-resistor that will give about 1V drop at 15mA; by ohms law this works to about 67 ohms. Be aware the sense-resistor will affect the signal-integrity into the HV5530, so be sure to use the recommended input levels (12V logic), and this will directly impact your low-level noise margin. But with a 1V drop, the overall impact to noise margin at 12V signal swing is negligible.

The other option is high-side current-monitoring. You could use a cheap analog or digital meter and put in-series with the anode supply. If you want to extract that info for you microcontroller, then your best option is to have an A/D converter on the high-side, and use opto-isolators on the serial interface to the ADC. There will be a fair amount of support circuitry for this, so be warned.

[David Forbes]

You can measure cathode current much easier than anode current, and it's the same thing. Tie all your cathode driver grounds together, and put a small resistance between them and ground. Select the resistance so that it has a bit less than one volt when operating. 

On Sun, May 12, 2019, 5:28 AM Paul Andrews <pa...@nixies.us> wrote:

I would like to be able to measure the anode voltage and current on one of my clocks and have a permanent display of those values built in to it. I came across this reference design at TI (http://www.ti.com/tool/tida-00528?jktype=design#technicaldocuments) but was wondering if anyone can point me at other possible solutions?

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/79373a6a-dc12-4608-90fd-077884c2633b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Bill Notfaded]

This is interesting to me because the current baffled me some. I like how the Nixie tester/healer has the connections to hook up to MM to measure current and voltage. To have this measured by the processor in the clock is neat idea. I also like the idea of controllable current and voltage from software... That's true flexibility.

Bill

[Paul Andrews]

My driver is an HV5523, which is a 5V part. The control signals are already level shifted from 3V3, and so are somewhat round and peak at more like 4.5V. I have an INA250 which I could shove between VSS and ground, the shunt resistor is a few mOhms, so I will try that.

[gregebert]

When measuring milliamp-range current across milliohm resistance, your voltage will be in the microvolt range,  which means that noise can easily creep-in and affect your mesurements.

With TTL-level signals, VIL(max) is 0.8V, so your sensing resistor could probably go up to 0.5V without disastrous effects, but you will need to experiment a bit. I think that works out to about 130 ohms. Also be aware the accuracy of this resistor directly affects the accuracy of your current-reading, so 1% or 0.1% resistors should be used.

Also, you want to take readings with no nixies on to determine the "floor" of your current reading, then subtract that from your actual measurements.

[gregebert]

Doh!  33 ohms, not 130.

[Paul Andrews]

Have you read the INA250 datasheet?

 

From: gregebert
Sent: Monday, May 13, 2019 9:37 AM
To: neonixie-l
Subject: [neonixie-l] Re: Anode current and voltage measurement

 

Doh!  33 ohms, not 130.

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/yhnsLbsIsLY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/db74d8b9-9947-4cea-877d-4b2d989c77af%40googlegroups.com.

[gregebert]

Have you read the INA250 datasheet?

 

I took a quick look thru the datasheet and I'm having doubts that it can measure milliamp currents with accuracy; it's designed for measuring 15 amps.

If there's a different version with an external sense-resistor, then I think you can get it to work.

[Tomasz Kowalczyk]

If you're using direct drive, then the problem can be solved with a differential amplifier with integrated resistors - this is the only way you can achieve high common mode rejection ratio, as any mismatch in the resistor network can really affect your measurements. 
There is a nice article about differential amps and measuring current with higher value shunt resistors (you really should use as high value as possible):
https://www.analog.com/media/en/training-seminars/tutorials/MT-068.pdf

There is an example of a +/- 270V input differential amplifier using AD629.

Why this method is good only for direct drive? Those amps quickly loose their CMRR above few hundred Hertz, which may affect your reading if you're trying to get high resolution and accuracy. 

[gregebert]

Thanks for posting this!  I had no idea there were diff-amps available that could handle such a high input voltage. A single instance is all you really need even if there are multiple tubes, because it's possible to measure each tube individually by blanking all others.

[Tom Harris]

Try the ZXCT1009. It's a current mirror on a single 3 pin package intended for high side current sensing.

https://www.diodes.com/assets/Datasheets/ZXCT1009.pdf

The output voltage can go straight into the ADC input of a microcontroller.

Tom Harris <celep...@gmail.com>

On Tue, 21 May 2019 at 10:08, gregebert <greg...@hotmail.com> wrote:

Thanks for posting this!  I had no idea there were diff-amps available that could handle such a high input voltage. A single instance is all you really need even if there are multiple tubes, because it's possible to measure each tube individually by blanking all others.

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/45f7c438-d911-4e02-9724-da76df1e8ebd%40googlegroups.com.