High Voltage measurement

[VictorPS]

Hi,

I have use TV flyback transformer to generate a voltage to DC 1.5KV.
I need to measure the high voltage, but my meter only allow 1000V, so I has use 2x 5.1Mohm resistor to divide the voltage down by half.

By don't consider accuracy, and my meter input impedance, I manage to measure it , show reading of about 600V. (a rough indication.)

But when there is arching effect of the high voltage, my meter show overload, look like it is >1000V (on meter), over the meter measurement range.

Questions:
1. Is a 1.5kV source, when arching/corona effect, the voltage will go very high >1.5KV?

2. Or there is problem in my driver circuit?
I am using PIC MCU (16F877A) PWM to drive MOSFET -> Flyback transformer.
Every time when there is arching, the PIC MCU will reset by itself.
How to prevent it reset? I have even add cap, 0.1uF, RC filter, LC filter to the 5V power supply, no matter how, there is still noise that attack the PIC MCU, especially at MCLR pin.

Thank you,

 

[Styx]

well... actually seeing the driving cct and WTF you are trying todo would be helpful, my crystal ball is in for polishing this week

 

[eblc1388]

Surprise, surprise.

Resistor do have voltage rating. When one approaching the voltage rating the resistance value changes. If one exceeding the voltage rating flashover will result.

When flashover occurs, all hell get loose.

I don't know about the exact voltage rating of 1/4W resistors but I always use 1/4W resistor to work at less than 250V and never have had a flashover.

So use proper expensive HV resistors which can stand your circuit voltage or made one up using many resistors in series.

 

[David Bridgen]

Can't help you with the problem of your circuit.

But, as regards measuring a voltage higher than you meter's maximum .....

A simpler way is to connect in series with the meter a resistance which has the same value as your meter's resistance on its highest range. If it is a digital multimneter it will probably be 10M on all the voltage ranges, but check.
You mulitply the meter reading by 2.

Another way is to use two meters, both on the same voltage range and with the same internal resistance, in series. Add the two readings.

 

[Hero999]

You could measure the meter's impedance using another meter.


Look up the potential divider formula and calulate the divor ratio you require.

You can always connect resistors in series to get a higher voltage rating, for example if you get nine 1kV 10M resistors can connect them in seris with your meter you'll get division by ten so you can measure up to 10kV.

 

[chemelec]

First you need to know the input impedance of your meter.

And when measuring things like a TV Flyback that have a Very Low Output Current, Your meter BETTER BE A Very High Input Impedance, or it will LOAD DOWN the reading and give a False Results, Even with the Series Resistor.
At Least 11 Megohm per volt.

"Check the Manufacturers Data Sheet for your meter".

Next you need a High Value, High Voltage Resistor.
A typical 1/4 watt resistor is rated at 250 Volts Max.
A 1/2 watt is rated at 450 Volts.

You will also need a Very High Value, and High Voltage Resistor. Greater than 100 Megohm, Possibly even Greater than 1 Gigohm.

You Might check out the HV Resistors for sale on my Website.
I May have a Suitable Value.

Gary
*******************************************

Hi,

I have use TV flyback transformer to generate a voltage to DC 1.5KV.
I need to measure the high voltage, but my meter only allow 1000V, so I has use 2x 5.1Mohm resistor to divide the voltage down by half.

By don't consider accuracy, and my meter input impedance, I manage to measure it , show reading of about 600V. (a rough indication.)

But when there is arching effect of the high voltage, my meter show overload, look like it is >1000V (on meter), over the meter measurement range.

Questions:
1. Is a 1.5kV source, when arching/corona effect, the voltage will go very high >1.5KV?

2. Or there is problem in my driver circuit?
I am using PIC MCU (16F877A) PWM to drive MOSFET -> Flyback transformer.
Every time when there is arching, the PIC MCU will reset by itself.
How to prevent it reset? I have even add cap, 0.1uF, RC filter, LC filter to the 5V power supply, no matter how, there is still noise that attack the PIC MCU, especially at MCLR pin.

Thank you,

 

[Nigel Goodwin]

First you need to know the input impedance of your meter.

And when measuring things like a TV Flyback that have a Very Low Output Current, Your meter BETTER BE A Very High Input Impedance, or it will LOAD DOWN the reading and give a False Results, Even with the Series Resistor.
At Least 11 Megohm per volt.

And as you already know, digital meters are almost always 10Megohms in total, so only 10Kohms per volt on the 1000V range - considerably worse then a standard Avo 8.

 

[VictorPS]

Thanks for reply. I will try David suggestion which is the simplest.

P/S: The HV is use to test insulaton of magnet wire.

However, when arching/corona effect, the voltage will go higher ?
To me, I think it shall go lower voltage than original source, but it is not, and because of this, I lost my Fluke73 DMM.

 

[chemelec]

And as you already know, digital meters are almost always 10Megohms in total, so only 10Kohms per volt on the 1000V range - considerably worse then a standard Avo 8.

The Beckman Series 300 & 3000 are 22 Megohm on All Ranges.
So if I set it on a 1 volt range it is 22 meg per volt.

My Fluke can also be set on a One Volt Range with 10 Meg Input resistance

10 K-ohms per volt is rediculous.
Even my Cheap Radioshack Multimeter Does Better than that.

Gary

 

[chemelec]

Not Sure How your Driving that Flyback, But Most TV Flybacks are 15,000 Volts or More.
Not 1,500 volts.
And it is nearly impossible to get ANY amount of Corona Below 5,000 Volts.

Gary

Thanks for reply. I will try David suggestion which is the simplest.

P/S: The HV is use to test insulaton of magnet wire.

However, when arching/corona effect, the voltage will go higher ?
To me, I think it shall go lower voltage than original source, but it is not, and because of this, I lost my Fluke73 DMM.

 

[Nigel Goodwin]

The Beckman Series 300 & 3000 are 22 Megohm on All Ranges.
So if I set it on a 1 volt range it is 22 meg per volt.

My Fluke can also be set on a One Volt Range with 10 Meg Input resistance

10 K-ohms per volt is rediculous.

But presumably that's just what your Fluke is?, and your Beckman is only 22K-ohms per volt - both on 1000V ranges.

It's what happens when you have a fixed input impedance.

Incidently a 40+ year old Avo 8 is 20K-ohms per volt, so twice as good as your Fluke, and almost as good as your Beckman!

 

[Hero999]

I don't care 10M at 1kV only draws 100:mu:A and at 1V it's just 100nA.

If current draw really is a problem then set it to the 20V range and use a 1:100 resistive divider to give a huge 1Ghm: resistance and a 2kV range.

 

[Nigel Goodwin]

I don't care 10M at 1kV only draws 100:mu:A and at 1V it's just 100nA.

You don't need to 'care', but you do need to 'know', and by your response I suspect you didn't?.

The antique Avo 8 will only take 50uA, so is twice as good for high voltages!

If current draw really is a problem then set it to the 20V range and use a 1:100 resistive divider to give a huge 1Ghm: resistance and a 2kV range.

Damn, where did I put my box of 990M-ohm 2000V resistors?

Like I said above, it's something you need to know if you're using a digital meter, and I suspect most people here didn't realise the limitations of digital meters.

 

[chemelec]

Nigel, Normally when I'm doing High voltage Measurements above 1000 volts, I typically add a Series resistance of 1,011,000,000 Ohms in series with the 11,000,000 of the meter. It works great.
(1 gig + 11 meg to get a correct scaling for my meter)

I have used Resistor values up to 10 Tera Ohm, for measuring Electrostatic voltages.
And I have these resistors in 1% tollerances.

Check out that Ohm per Volt.

Sorry but it is Much better than your Antique Avo 8.

Not sure How you think yours is Better?

But Enough said, I'm not into arguing this further.

Merry Christmas......Gary

You don't need to 'care', but you do need to 'know', and by your response I suspect you didn't?.

The antique Avo 8 will only take 50uA, so is twice as good for high voltages!



Damn, where did I put my box of 990M-ohm 2000V resistors?

Like I said above, it's something you need to know if you're using a digital meter, and I suspect most people here didn't realise the limitations of digital meters.

 

[Nigel Goodwin]

Nigel, Normally when I'm doing High voltage Measurements above 1000 volts, I typically add a Series resistance of 1,011,000,000 Ohms in series with the 11,000,000 of the meter. It works great.
(1 gig + 11 meg to get a correct scaling for my meter)

I have used Resistor values up to 10 Tera Ohm, for measuring Electrostatic voltages.
And I have these resistors in 1% tollerances.

Check out that Ohm per Volt.

Sorry but it is Much better than your Antique Avo 8.

Not sure How you think yours is Better?

I don't, and don't even have an Avo 8, but it was THE standard for many years.

I don't see as your's is 'better' if you need an addon to make it better than a 1950's analogue meter (for one specific range)

But it's not about 'better' or 'worse', just about trying to make people aware of the problems of digital meters - people from the pre-digital age (like ourselves) should already be aware of it, but youngsters are almost certainly not.

 

[Hero999]

Well, I've always been aware that digital meters have an impedance of about 10M but I didn't know that some older meters had a higher impedance.

I wouldn't say the older meters were better as they didn't have all the other fancy features like transistor tester, capacitance and frequency that are really handy. They were also very fragile, cumbersome and expensive. Modern digital meters are also cheap, light weight and very robust, it's not surprising that they caught on so quickly.

 

[David Bridgen]

They were also very fragile ..

Some certainly were, but I remember one which wasn't. I can't for the life of me think of the manufacturer. It had taught-band suspension and could withstand an awful lot of physical abuse. It was significantly smaller than the Avo.

It had been demonstrated to the field trials team at M.S.D.S in Frimley (R.I.P.) and was left with them for evaluation. Out of devilment more than real interest it was subject to abuse which was really disgraceful.

If my memory serves me correctly the only damage caused (it had literally been kicked around according to my fellow "engineers") was the switch shaft being broken. After an application of Araldite it worked perfectly.

Based on this I subsequently bought one and it went with me all over the world. I have no idea how, when or why I parted with it - one of the daftest decisions I've ever made.

 

[Leftyretro]

There is no question that the modern DVM are wonderful, cost effective and accurate given it's made by a quality manufacture. I love the various Fluke meters I own and wouldn't trade them ever for the old Simpson multimeter and various manufactures VTVM meters I used many decades ago.

However with all the automatic features such a auto-ranging, true RMS, etc. many newcomers that lack good grounding in electronics fundementals and tend to trust the numbers being displayed even when they might be wrong such as due to input impedeance when measuring high impedeance circuits. Or why one might want to measure a DC voltage using the AC mode in the meter to find out if there is a ripple voltage content or not.

It comes down to having a good understanding of electronics fundementals. Any instrument can be used incorrectly.

I don't think many newcomers understand the danger they can do to the meter, the circuit, or themselves with say attempting to measure a voltage while having the meter leads in the current position, a lot can happen before that 10 meter fuse blows open.

Lefty

 

[Nigel Goodwin]

Well, I've always been aware that digital meters have an impedance of about 10M but I didn't know that some older meters had a higher impedance.

Exactly, you're too young!

Digital meters generally have a fixed input impedance of 10M-ohms, which sounds a lot - and it is on the low voltage ranges. Older analogue meters consisted of a micro-ammeter (usually 50uA FSD) fed through different valure series resistors for the different ranges, So an analogue meter has a different impedance on every range, so it's expressed as ohms per volt - with a 50uA based meter giving 20K-ohms per volt.

So on a 10V range it will be only 200K-ohms (10V x 20K-ohms), but on 1000V it will be 20M-ohms.

This difference is crucial for some applications, for example measuring and setting the first anode voltage on a CRT - an analogue meter is high enough to measure it with no noticeable effect, but a digital meter has a considerable effect on the reading. An Avo 8 (or 9) also has a 2500V range, which increases the impedance to 50M-ohms as well.

While I still have an old Avo at work, I use a digital meter almost exclusively (the Avo is under a bench gathering dust), so adjust the first anode voltages by an alternative method. The only time I use the old Avo is for measuring the magnetron voltage in a Microwave oven, where the 2500V range is 'just' enough.

I wouldn't say the older meters were better as they didn't have all the other fancy features like transistor tester, capacitance and frequency that are really handy. They were also very fragile, cumbersome and expensive. Modern digital meters are also cheap, light weight and very robust, it's not surprising that they caught on so quickly.

I never said 'better', I said 'better for some purposes'

You need to understand the limitations of your instruments, and the reasons for said limitations - digital readouts tend to make people forget it might not be reading what it's displaying!.

 

[chemelec]

I don't see as your's is 'better' if you need an addon to make it better than a 1950's analogue meter (for one specific range)

But it's not about 'better' or 'worse', just about trying to make people aware of the problems of digital meters - people from the pre-digital age (like ourselves) should already be aware of it, but youngsters are almost certainly not.

Even for That Avo 8 Meter, you would need an "External Resistive Probe" for Such LOW CURRENT, High Voltage Measurements.
50 ua is still a lot of current in these situations.

But YES I Totally Agree that there are MANY ADVANTAGES to the "Older Analogue meters". I Have many of them which I Still use for various purposes. Especially for Audio signal Amps. No Update Times like on Digital ones, so an Analogue meter will Average the readings.

My Regret is in the Forseeable Future, I will have to Dispose of all this Electronic stuff, as well as my metal and wood working equipment.
That will be Interesting as this is a TINY TOWN with little if any interest in this here. Especially the electronic parts and equipment. And Shipping to the USA is Not Practical.

Take care......Gary