need help with an anlogue 360-440Hz 200v frequency meter

Hi i need help with a frequency meter im trying to build, i have a 100uA anologue movement with a resistance of around 2K ohms and a frequency transducer circuit to make the instrument read 340-440hz at 200v on the scale ,but i am having problems with it it is very unstable, everytime i calibrate it then connect it to equipment to check it 30 mins later it is out of cal and i would like some help in making it more stable mb by using different components or adding a couple more. I can submit pictures of circuit layout and component list if someone out there is prepared to help.Also when i set the voltage at the start of calibrating do i need to adjust it back to 200v after the instrument is powered up or do i leave it alone as there will be a substantial voltage drop once meter is connected and powered up? If i check voltage after bringing it back up to 200v while meter is disconnected it reads 440v surely this cant be right!! I want instrument to work at 200v so do i leave voltage alone after i have connected it at 200v , the multimeter says its only 95-98v can this be right? please get back to me.

Can you post a schematic? (That would be far more helpful than a layout picture/comonent list.)

need help with an anlogue 360-400Hz 200v frequency meter

hi here is the circuit, just a note i already had to replace R2 with a 5k6 just to get it to calibrate, RV1 is trimmed to get the 360hz reading and RV2 is trimmed to get the FS of 440hz just keep trimming each one till top and bottom readings are bang on then rest of scale should follow, but for some reason it doesnt stay calibrated , leave it for 30 mins then go back to it to recheck it and its way off, i even tried powering it up using a fluke 5101b calibrator at 200v 400hz which i know is dead accurate and it reads way off there too. I have this posted around a few places and i tell you who ever comes up with a solution that works and can be repeated will get a little reward

You say "frequency meter". Do you want the meter to read 360hz =0 and 400hz to read full scale?

You said " unstable" , " 30 minutes" Are any of the parts hot?
Is C1 250V?
What is the voltages on C2, C4? Are they below 16V and 6.3V?

frequency meter

ronsimpson said:

You say "frequency meter". Do you want the meter to read 360hz =0 and 400hz to read full scale?

You said " unstable" , " 30 minutes" Are any of the parts hot?
Is C1 250V?
What is the voltages on C2, C4? Are they below 16V and 6.3V?

Click to expand...

when i say unstable i mean it doesnt retain its readings after 30 mins of being disconnected then reconnecting it the test kit. the only component that gets hot is the 27k 1W resistor R1 which do get warm when in use .C1 was 275v i have also tried a 1.25kv capacitor in there with the same result. I cant tell you the voltages on C2 and C4 till i get back to work and reconnect it to their equipment, i guess i just put a multimeter across the 2 capacitors while its powered up + to + and - to -. Will post back info when i find it thnx so far. I will also post a picture of the dial, instrument 0 is not 360hz their is a small gap from zero power off line to first reading of 360hz.

it's been a while since i worked with 400hz equipment. such equipment is usually military or avionics. so 400hz is the center of the scale, 360hz on the left side, and 440 on the right. zener diodes have a temperature coefficient which depends on the particular zener voltage. zeners rated below 6V have a negative tempco and operate as true zeners. zeners rated above 6V operate in avalanche mode, and have a positive tempco. the interesting thing is that 6V zeners have a tempco of about 0. so replace each 13V zener with two 6V zeners in series, and it should eliminate temperature drift. (actually it's 5.6V zeners that have a zero coefficient, but a 6V zener has a close to zero one)

what is happening is the 13V zeners are heating up and their breakdown voltage drops (maybe as much as a volt). when you disconnect and then later reconnect the device, the zener voltage is back up to 13V.

Simulation in LTSpice shows that although the temperature dependence of the zeners has a slight effect on stability, a far greater effect is caused by variation in the leakage resistance or ESR of C2 or C4; particularly C2. I would suspect one of those is the cause of your problem. Try replacing both, with known good quality caps.

I agree that C2, C4 could be the problem. There are two voltage dividers C1C2 and C2C4. C2 and C4 were never built for stability. They are probably made to filter a power supply where changes in capacitance and ESR will not make much much of an effect. Not knowing the voltages on C2, C4 I have asked for the voltage on these parts.

hi i can have a go at replacing each 13v diode with 2 5.6v ones i have access to these but i cant use 6v ones without having to order them i will also have a go at replacing C2 and C4 but i dont know good quality ones as this sort of technology changes frequently. The ones in the circuit are tantalum bead ones can i replace them with electrolytic caps as that has a + and a - on them too or will they not work the same. Sorry no pics yet of dial or voltages across C2 and C4 as i have been busy, will have a look tomorrow and post again. Sorry im a bit vague but capacitor technology is a bit of a mystery to me. and my electronic skills are not the best. I wouldnt know a good quality cap if it bit me on the ass.. I did start to suspect the zener diodes as i have had trouble with them in the past with other curcuits, i tested (10) 20v zeners and only 3 of them actually broke down at 19.7v which was near to what i needed others were way off. I think they are made in such high volumes these days that they have had to drop the accuracy as they cant test so many on the production line. Years ago they they were far more accurate. With C2 and C4 will it matter if the voltage of them is slightly higher than stated on the drawing and can i use electrolytic instead of tantalum bead.

With C2 and C4 will it matter if the voltage of them is slightly higher than stated on the drawing and can i use electrolytic instead of tantalum bead.

Click to expand...

A higher voltage rating may well be better. I would have suggested using tantalum beads, but you've already got those. Maybe you have defective ones. Try substituting other caps (tants or aluminium) of the specified values and see if you get better stability.
As for the zeners, the value isn't critical and the pots should cope with calibration.

https://www.electro-tech-online.com/custompdfs/2012/01/203.pdf

+0.7%/degree c. with the self heating a zener undergoes, and you could have a 10% change in zener voltage. that's not insignifigant

It's a poor design. You could use a voltage regulator and fixed period monostable, that wil give a fairly precise (filtered) analog voltage of 400Hz= 50% voltage (total output voltage range of 45% to 55%).

Then output through the milliampmeter to the centre of a voltage divider, and trim the divider to give mid scale = 400Hz.

As both the monostable and the voltage divider are run from a regulated voltage the precision and repeatability will be pretty good.

I agree with Mr RB. Should be much more stable (assuming the cap used for the monostable timer is a good'n )

another way to do it would be to use a 4046 PLL, and buffer the control voltage. the buffered control voltage would then drive the meter.

Ok guys slow down please my electronics skills are very basic to say the least, Mr RB has an idea but i cant do circuit design to that degree could you possibly show me the circuit you are talking about. I have some more info for you all i tried the 2x 5.6v zeners in series to replace each of the 13v ones that just made it more inaccurate after a 30min period of disconnection. I measured the voltage across C2 and that was 4.14v and voltage across C4 was 0.104v, im going to take some photos on monday of whats happening some b4 pics of instrument calibrated and after 30 mins of bieng disconnected. I even waited 30 secs after reconnecting to see if the readings would rise allowing time for caps to charge but it didnt help. Like mrRB says it is an old circuit about 20years old if someone could show me a more modern one that would not cost too much in components as budget is tight. where the pointer rests under no load is no reading then there is a small gap to where 360hz is then 440hz is FS. Nobody has said yet as wether i have to increase the voltage to 200v after connecting the meter up. I know i have to set it to 200v b4 i connect it but do i have to readjust to 200v once i have connected it up, because if i measure voltage after meter is disconnected i get a volt reading of 440v this cant be right surely????? thnx again for your help so far.

exactly what is this hooked up to? are you adjusting for 200V and then hooking up the frequency meter, then adjusting for 200V again? then after measuring the frequency, you disconnect the meter and readjust for 200V? that sounds like what you are doing. if so, the meter is loading the circuit you are measuring. the loading is causing a voltage error, and readjusting the voltage after you remove the frequency meter is probably why it's reading wrong when you reconnect it. so it may not be so much a thermal problem, but a loading problem (causing a procedural problem)

eddi40 said:

Ok guys slow down please my electronics skills are very basic to say the least, Mr RB has an idea but i cant do circuit design to that degree could you possibly show me the circuit you are talking about.
...

Click to expand...

Hi Eddi40, I'm a bit reluctant to give a circuit since this is a 200v application and you said you have very basic electronics skills. Apart from the safety aspect if your circuit is not built right the very first time the 200v will very likely damage it and give you a fault finding nightmare.

How about a ready-made (and safer) solution? You can buy off-the shelf pocket frequency meters, or even many multimeters these days have a frequency measuring mode. All you would need is a couple of resistors to drop the 200v to a lower voltage, and use the ready-made frequency meter to give you a nice accurate reading of the Hz.

the equipment im using is about 50-60 years old with an external toroidal transformer wired up to boost the machines maximum voltage of 50v to 200v im starting to think because the machine has not been used for at least 10years that it may be faulty but i cant be sure because i dont know if there is a calibrating house out there that would want to touch it because of its age and lack of info, i cant even find a name on it. You will see on monday night when i post pics. I was told once that this piece of kit came fro ma 1950s bomber as part of its radio equipment but i dont know, mb 1 of u guys will recognise it. All i know it has been used in the past for a firm called Hawker Siddley and it did the job for them.

hi im gonna post pictures of the equipment monday night, firstly i plug equipment in turn it on wait a few minutes for it to warm up set voltage to 200v set frequency to 440hz connect instrument to equipment then adjust voltage back up to 200v because it has dropped to around 110v then once that is done i then proceed to calibrate instrument, it sets ok so then i disconnect instrument not touching any settings on equipment leave it on then go back to instrument after 30 mins hook it back up to same already set equipment and the readings are different, the only way to get it reading right again is to slightly correct voltage and readjust RV1 and RV2 till it reads right again, it has to be rechecked because it is an hermetically sealed instrument and if it is wrong it is a lot of hassle to get it open and correct it. As to mr RB's reponse this is an hermetically sealed instrument for rough outdoor/military use and i cant use what he has suggested.I have found a frequency circuit i know i can build using 555 timer chip can anyone modify it to work at the frequencys and voltage im trying to get i know i can build this i have built circuits with 555 timer chips before and i have built tachos for military use and fuel guages for military use , when i say my electronics skills are basic i mean i dont know complex stuff with micro-controllers PIC's and such like . Also this circuit im going to post looks like it has 4 connections i need it to be self powered like the one i was trying to use ie- 2 exterior terminals. I also notice with this circuit it only has one pot i guess this is trimmed while input is at mid scale then rest of scale is ok. If i have decent drawings and instructions i can build most things i have built digital instruments in the past with proper drawings and boards and had no problems , its when components change or have markings wrong that throws a spanner in the works, like for instance the 2 tantalum beads on the circuit i am trying to use they came in from RS with a stripe and a + sign on the same bloody terminal it is one or the other it cant be both.

Hi Eddi40, ok if you are determined to make a freq meter it is pretty straightforward.

The 555 is wired as a timed period "monostable" which looks like the schematic you posted (but that schematic is too tiny to be sure).

The 555 monostable must be run from a regulated supply voltage (which it looks like in your schematic). Also as people have said you need a timing cap that is not temperature sensitive.

Then your circuit shows a meter that reads frequency from zero-max. To fix that you disconnect the bottom wire of the meter, and connect that meter wire to the middle of a voltage divider (trimpot etc). That voltage divider sets your minimum freq that is shown at meter min scale. Then the frequency shown at meter max scale is roughly set by your choice of the two filter resistors (before the meter).

If you have a larger better copy of that schematic please post it so people can help with sugestions.