Meter protection - when diodes won't work - Page 2 - UK Vintage Radio Repair and Restoration Discussion Forum

GMB · 24th Aug 2022, 10:27 pm

I have considered adding a capacitor but that changes the meter's ballistics and can result in unexpected readings from funny waveforms.

Craig Sawyers · 25th Aug 2022, 8:52 am

This is quite a good description about the small current behavior of diodes here https://spinningnumbers.org/a/diode-equation.html

The upshot is that a diode being a clamp is not how it works at low currents.

Suppose you want the diode current to be less than 1/100 of the full scale current, so that the maximum error in reading is 1%, so 50uA/100 = 500nA. For a silicon diode that current will pass for a voltage of ~0.33V. Which is of course just fine. Now suppose we want 3x current overload, and so need a current of 150uA. That is reached with a diode voltage of 0.47V, so a coil power of 0.47 x 150 x 10^-6 = 70uW. The nominal full scale is 50 x 10^-6 x 0.04 = 2uW. So the coil power overload is 35 times.

For a Germanium diode this works out at ~10mV, so might give rise to meter non linearity at 40mV. But if we want three times current overload at 150uA, that is reached at 0.125V, and so a coil power of 18uW, so a factor of 9 times power overload. The problem with a germanium diode, apart from the potential for meter non-linearity, is that the IV characteristics are strongly temperature dependent, more so than silicon.

There is a further complication of a fudge factor called the "ideality factor" that modifies the IV behaviour of a diode

So meter protection with back to back diodes is a little more complicated than might be imagined. It is a fine balancing act between meter non-linearity and protection.

Craig

Craig Sawyers · 28th Aug 2022, 9:08 am

It is worth saying that if you put an opamp between the circuit and the meter, such as you would to match a meter with different characteristics, you can easily implement a current limit or voltage clamp much more effectively than simple back to back diodes.

Craig

GMB · 28th Aug 2022, 10:44 pm

The term "back to back" keeps being mentioned, but just to be clear they are anti-parallel diodes (one clamps the meter in normal polarity and the other copes with reverse voltage). Back to back would not help at all!

Vintage Engr · 30th Aug 2022, 6:33 pm

I've just stumbled across this discussion, and wonder if anyone had ever had a problem with meter protection on the Ferrograph RTS2 audio test set. I believe the original meter 's FSD is 100 microamps.

I had my original RTS2 for around 30 years, and the meter on this unit has always seemed a little under-protected. I now have another unit that I was hoping to swap the meter out of (my original's meter graphics have faded completely & also the movement is very sticky).
I found another RTS2 with what looked like a perfect meter.. The scale was fine, but the meter is o/c.
So before proceeding with a repair, I thought I'd see if anyone had been there first.

David.

24th Aug 2022, 10:27 pm	#21
GMB Dekatron Join Date: Aug 2003 Location: near Reading (and sometimes Torquay) Posts: 3,103	Re: Meter protection - when diodes won't work I have considered adding a capacitor but that changes the meter's ballistics and can result in unexpected readings from funny waveforms.

25th Aug 2022, 8:52 am	#22
Craig Sawyers Dekatron Join Date: Nov 2010 Location: Oxford, UK. Posts: 5,001	Re: Meter protection - when diodes won't work This is quite a good description about the small current behavior of diodes here https://spinningnumbers.org/a/diode-equation.html The upshot is that a diode being a clamp is not how it works at low currents. Suppose you want the diode current to be less than 1/100 of the full scale current, so that the maximum error in reading is 1%, so 50uA/100 = 500nA. For a silicon diode that current will pass for a voltage of ~0.33V. Which is of course just fine. Now suppose we want 3x current overload, and so need a current of 150uA. That is reached with a diode voltage of 0.47V, so a coil power of 0.47 x 150 x 10^-6 = 70uW. The nominal full scale is 50 x 10^-6 x 0.04 = 2uW. So the coil power overload is 35 times. For a Germanium diode this works out at ~10mV, so might give rise to meter non linearity at 40mV. But if we want three times current overload at 150uA, that is reached at 0.125V, and so a coil power of 18uW, so a factor of 9 times power overload. The problem with a germanium diode, apart from the potential for meter non-linearity, is that the IV characteristics are strongly temperature dependent, more so than silicon. There is a further complication of a fudge factor called the "ideality factor" that modifies the IV behaviour of a diode So meter protection with back to back diodes is a little more complicated than might be imagined. It is a fine balancing act between meter non-linearity and protection. Craig __________________ Doomed for a certain term to walk the night

28th Aug 2022, 9:08 am	#23
Craig Sawyers Dekatron Join Date: Nov 2010 Location: Oxford, UK. Posts: 5,001	Re: Meter protection - when diodes won't work It is worth saying that if you put an opamp between the circuit and the meter, such as you would to match a meter with different characteristics, you can easily implement a current limit or voltage clamp much more effectively than simple back to back diodes. Craig __________________ Doomed for a certain term to walk the night

28th Aug 2022, 10:44 pm	#24
GMB Dekatron Join Date: Aug 2003 Location: near Reading (and sometimes Torquay) Posts: 3,103	Re: Meter protection - when diodes won't work The term "back to back" keeps being mentioned, but just to be clear they are anti-parallel diodes (one clamps the meter in normal polarity and the other copes with reverse voltage). Back to back would not help at all!

30th Aug 2022, 6:33 pm	#25
Vintage Engr Heptode Join Date: Jan 2008 Location: Shrewsbury, Shropshire, UK. Posts: 847	Re: Meter protection - when diodes won't work I've just stumbled across this discussion, and wonder if anyone had ever had a problem with meter protection on the Ferrograph RTS2 audio test set. I believe the original meter 's FSD is 100 microamps. I had my original RTS2 for around 30 years, and the meter on this unit has always seemed a little under-protected. I now have another unit that I was hoping to swap the meter out of (my original's meter graphics have faded completely & also the movement is very sticky). I found another RTS2 with what looked like a perfect meter.. The scale was fine, but the meter is o/c. So before proceeding with a repair, I thought I'd see if anyone had been there first. David.