EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: mullecy on August 31, 2017, 01:54:08 pm
-
I have a few DMMs and my preferred ones have a nice bright OLED display. :-DMM
The problem is that they use almost 40mA.
I'm sick of buying expensive 9V batteries and NiMH ones just last a few hours. :--
Knowing that I almost never use these except on the bench I'm planning to replace the 9V battery with an external PSU. ;)
To do that I'll use a small transformer, a bridge rectifier, big caps and a 9V regulator. I'll probably have to add some filtering (common mode choke, X and Y caps before the transformer, LC T filter and a ferrite after the regulator). I'll also add fuses in case something goes wrong.
I'll have to do a small hole in the back of the DMM for the wires.
Any suggestions? :-//
PS: the idea is to keep it really floating and avoid the noise from a switching psu.
-
Wouldn't it be easier just to find a 9V linear wall wart?
I see them all the time at thrift stores for almost nothing cost wise.
-
The difficult part is to keep it really floating. Usually a good conventional transformer with separate windings should be the better separation than a switched mode supply. The extra weight might also have an advantage, so the case is sitting on the desk and not pulled around by the cables.
Due to insulation issues an external wall wart is not a good option. It would need an extra level of insulation (e.g. a DCDC converter). The ready made DC/DC converters are often rather noisy and thus not that suitable.
The advantage of a classical supply is that one usually does not need the Y -cap. One might get away even without the class X cap at the primary, as the usually should not be much RF interference from such a low power meter.
Another option might be using rechargeable batteries. Some meters might work with 2 LI cells.
-
The difficult part is to keep it really floating. Usually a good conventional transformer with separate windings should be the better separation than a switched mode supply. The extra weight might also have an advantage, so the case is sitting on the desk and not pulled around by the cables.
Due to insulation issues an external wall wart is not a good option. It would need an extra level of insulation (e.g. a DCDC converter). The ready made DC/DC converters are often rather noisy and thus not that suitable.
The advantage of a classical supply is that one usually does not need the Y -cap. One might get away even without the class X cap at the primary, as the usually should not be much RF interference from such a low power meter.
Another option might be using rechargeable batteries. Some meters might work with 2 LI cells.
Linear wall wart = isolated, no Y cap.
Look for a big, heavy one. Your clue as to a big-ass iron core transformer. I see them all the time for $2-3 in the second hand stores and have a few at various voltages for just this purpose. I would stay away from any SMPS wall wart for obvious reasons.
-
Rather interesting, I wouldn't go for the cheapest linear wal-wart or dare to suggest to go with the cheapest wung-ho-low no specification units. It is a DMM after all, the most abused electrical meter type. Also to be noted is that when you do this modification, your meters CAT ratings etc. are all gone, the end results is 100% up to you to be safe in every situation where you or someone else will be using it.
-
How about using two rechargeable batteries which are swapped in and out by a latching relay. When one battery is charging the other is connected to the DMM side. When the voltage of the DMM side drops below a certain threshold, the relay will swap the current battery and the already charged battery. No problems with galvanic isolation. If the battery drainage is not too high, the relay logic control can be done manually once every while.
-
Even on this case the relays must be withstand the maximum voltage of the highest measuring range ie. 750 volts (= mystery waveform AC) continuous, plus safety margin, plus transients (now thinking of it even this is not enough in case of phase shift between DUT and meter power source). Or in case of ie. 250V AC or DC range, minimum starting point should be mains phase-to-phase voltage. Also the switching arroy must be safe brake-before-make type and all the possible fail states must be considered ie. blackout of the diatribution network and power up and down routines/states. ..etc
If safety aspect is pushed aside anything works, but..How about using two rechargeable batteries which are swapped in and out by a latching relay. When one battery is charging the other is connected to the DMM side. When the voltage of the DMM side drops below a certain threshold, the relay will swap the current battery and the already charged battery. No problems with galvanic isolation. If the battery drainage is not too high, the relay logic control can be done manually once every while.
With a quick analyse.
-
How about a cheap and easy to find USB power supply, and add to that a common DC DC converter, like what used to be common in network cards, that run off 5V and provide an isolated ( with 2kV test rating) 9V supply that will power the meter. Simple, robust and should be easy to buy online as well, or find an old NE2000 clone network card ( with coax socket on it) and harvest the converter from it.
-
AFAIK network card don't have a full DC/DC converter, but just an isolation transformer for the high frequency signal. In theory one could use this type of transformer to build a DC/DC converter. Still I am not sure the isolation rating is sufficient - it could limit the safe voltage limit.
Using just a wall wart would connect the wall-wart output side and the usually thin cable connected to the DMM part. So no more high voltage measurements. So it would need a wall-ward and an additional low EMI DC/DC converter. It might work with an AC wall- wart and than an additional 50/60 Hz transformer (or 2 in series to get something like a 12 V to 12 V transformer with good isolation) inside the bench-top case.
-
Rather interesting, I wouldn't go for the cheapest linear wal-wart or dare to suggest to go with the cheapest wung-ho-low no specification units. It is a DMM after all, the most abused electrical meter type. Also to be noted is that when you do this modification, your meters CAT ratings etc. are all gone, the end results is 100% up to you to be safe in every situation where you or someone else will be using it.
While I understand and support your basic approach, I have noted that the best WWs are often sold cheaply, because they are bigger and heavier (transformer-based).
-
Rather interesting, I wouldn't go for the cheapest linear wal-wart or dare to suggest to go with the cheapest wung-ho-low no specification units. It is a DMM after all, the most abused electrical meter type. Also to be noted is that when you do this modification, your meters CAT ratings etc. are all gone, the end results is 100% up to you to be safe in every situation where you or someone else will be using it.
While I understand and support your basic approach, I have noted that the best WWs are often sold cheaply, because they are bigger and heavier (transformer-based).
Well with this kind of case where one do need a trustworth separation/insulation, the typical transformer in those units might not have proper mechanical separation and electrical streght between core and windings. In example the separation of windings from each other is one sheet of rice paper for cutting the costs. Looking at safety electrical perspective some level of galvanic isolation they almost always provide (even a sheet of rice paper) if the windings do not share a common point by construction. Also the specifications of the unit might be false or non-existent. I'm not saying you can not get a good unit that would meet the needs even on this kind of case with pennies, but do you want to gamble if not necessary for some obscure reason?
-
AFAIK, there is nothing you can do to make a hand held meter equivalent to a properly designed bench meter. The negative on the battery(ies) of a multimeter is almost not equal to the Gnd/Neg jack on the meter. Any potential applied to that jack will be present on the negative terminal on the battery. This causes an unsafe situation. If that is not important to you, then just use an analogue power supply with a transformer to power the meter.
If safety is important, then you need to make a box that contains the meter and the power supplies so that the cord to the wall only carries the house power to the box. The transformer used in the power supply needs to be an isolation type that has at least the same isolation of the highest voltages that might be used in testing the CAT rating of the meter. Only with the multimeter and the power supply inside the same insulated box will you have something safe.
-
How about using two rechargeable batteries which are swapped in and out by a latching relay. When one battery is charging the other is connected to the DMM side. When the voltage of the DMM side drops below a certain threshold, the relay will swap the current battery and the already charged battery. No problems with galvanic isolation. If the battery drainage is not too high, the relay logic control can be done manually once every while.
Good idea but the problem is safety.
The internals of the DMM may be live if I probe mains voltage.
So the battery itself could be 9V on top of 220VAC...
So batteries and switch should not be accessible if I use the meter.
BTW my DMM has a neat feature: you can't open the battery compartment if probes are inserted.
So the whole PSU could be at a high voltage when I use the DMM so I must make sure everything is isolated and can't be touched.
That's why I'll probably use silicone wires to make sure that if my soldering iron touch them I don't damage them.
The problem with switching PSUs is that there's some leakage from the transformer due to parasitic capacitive coupling between the primary and secondary. That's why it's standard to add a Y cap between primary and secondary side to prevent EMI. The drawback is that the secondary is connected to line or neutral through that cap (depending on the way the plug is connected).
If you measure the AC voltage between the output of that PSU and ground you'll find maybe 100V... and on some devices you can even feel a tickle...
Ex: this cheap sig gen uses one and has some problems: https://youtu.be/9o5MzTOzZo4
and the PSU is changef for a linear one: https://youtu.be/ML-lmuHoh-0
-
AFAIK, there is nothing you can do to make a hand held meter equivalent to a properly designed bench meter. The negative on the battery(ies) of a multimeter is almost not equal to the Gnd/Neg jack on the meter. Any potential applied to that jack will be present on the negative terminal on the battery. This causes an unsafe situation. If that is not important to you, then just use an analogue power supply with a transformer to power the meter.
If safety is important, then you need to make a box that contains the meter and the power supplies so that the cord to the wall only carries the house power to the box. The transformer used in the power supply needs to be an isolation type that has at least the same isolation of the highest voltages that might be used in testing the CAT rating of the meter. Only with the multimeter and the power supply inside the same insulated box will you have something safe.
Of course it's not 100% clean and it will not be like a bench DMM.
The problem is that for some reason bench DMMs are much more expensive even if you take poor accuracy ones. :'(
So my goal is to be as close as possible to the safety provided by a bench DMM.
My initial idea was to have the DMM on one side and the PSU elsewhere and good wires with good isolation properties.
An "improved" version would be to build a box with the DMM and its PSU embedded in it. with just the screen controls and posts accessible from outside. But that's more work and I'll call that v2.0 ;)
-
I'd try to improve the battery situation instead of trying to make a mains transformer.
ie. Improve the capacity and make it really fast/easy to swap them.
You can get USB battery packs with huge capacities these days.
https://encrypted.google.com/search?q=usb+battery+pack
Get two packs with 12000mAh capacity each and switch between them. Swapping packs will take a couple of seconds (it's just a USB cable) and you'll only need to do it a couple of times a month even if you use the meter all day every day. Just remember to disconnect the probes from mains AC while you're swapping the battery, OK?
Boosting the voltage from 5V to 9V is left as an exercise for the reader. Does the meter have an internal booster? What's the cutoff voltage?
-
I'd try to improve the battery situation instead of trying to make a mains transformer.
ie. Improve the capacity and make it really fast/easy to swap them.
You can get USB battery packs with huge capacities these days.
https://encrypted.google.com/search?q=usb+battery+pack (https://encrypted.google.com/search?q=usb+battery+pack)
Get two packs with 12000mAh capacity each and switch between them. Swapping packs will take a couple of seconds (it's just a USB cable) and you'll only need to do it a couple of times a month even if you use the meter all day every day.
Boosting the voltage from 5V to 9V is left as an exercise for the reader. Does the meter have an internal booster? What's the cutoff voltage?
I understand your point but if I modify the DMM it's not "half-way". Also I don't really like the idea of having a battery pack hanging on the back of the DMM with potentially lethal voltage on the USB plug. Power bank are not designed to be floating at high voltage and any metal part accessible to the user would be a hazard.
Another problem is the switching noise from the dc-dc converter. Of course it could be filtered but I don't want to have strange measurements because I have a noisy 9v.
The DMM is probably not designed to deal with that (it's designed to use a battery) so the behavior is unknown.
If I go that path I would probably use a 3S or 4S lipo pack with a linear regulator to get a clean 9V.
The problem with batteries is also that they die when you use them. (I know that sentence seems obvious). So you always end-up using the DMM then discover that you are out of power in the middle of a measure.
Also the DMM has some logging features (more or less useless on NiMH batteries) and I want to be able to leave it in use for hours without having to worry about the state of batteries.
If I use a good cable between the DMM and the PSU this part is not more dangerous (it is actually less or equal) than any power cable connected to mains in the lab...
So if all powered parts are out of reach in a nice PSU box and a good cable going to the DMM and the back cover of the DMM has just a small hole to let the cable enters I feel not that bad.
BTW you can find good little transformers that can deal with 4200Vrms between primary and secondary for $5:
https://www.electronicsdatasheets.com/manufacturers/zettler-magnetics/parts/bv302s12010 (https://www.electronicsdatasheets.com/manufacturers/zettler-magnetics/parts/bv302s12010)
-
You seem to be intent on doing what you want and ignoring advice from experienced people. Good luck.
-
I would also suggest just a normal (old fashioned) transformer.
That case it IS isolated.
The only problem is getting the ripple and noise as low as possible.
So you'll need some pretty good regulators.
-
You seem to be intent on doing what you want and ignoring advice from experienced people. Good luck.
Thank you for your suggestions.
Well I have a pretty good idea of the result I want.
I understand the advantages of your proposal but also its drawbacks.
I have a few reasons to prefer a mains powered system.
So I guess I was expecting advises on the implementation of my proposal and not really an alternate solution altogether.
If I have enough time I'll try to implement a solution with just a bigger battery.
-
I'll post a picture tomorrow of externel DIY power supply modified multimeter ..ew multitester, which will make even cpt.Kirk to facepalm. It is vintage and done by previus owner. As it still is working I suppose the original propeller head is now safely retired. It does have 1000V range. ::)
-
battery-powered solution:
-
I don't really like the idea of having a battery pack hanging on the back of the DMM with potentially lethal voltage on the USB plug.
Put the battery+USB plug somewhere where you can't touch anything, eg. in a closed box.
Also the DMM has some logging features (more or less useless on NiMH batteries) and I want to be able to leave it in use for hours without having to worry about the state of batteries.
If you have two batteries then you can simply swap them before anything critical.
-
I don't really like the idea of having a battery pack hanging on the back of the DMM with potentially lethal voltage on the USB plug.
Put the battery+USB plug somewhere where you can't touch anything, eg. in a closed box.
Also the DMM has some logging features (more or less useless on NiMH batteries) and I want to be able to leave it in use for hours without having to worry about the state of batteries.
If you have two batteries then you can simply swap them before anything critical.
It seems easy to monitor the battery it it's not that obvious: the DMM only sees regulated 9V and has no idea that the battery is going down.
A solution could be to lower the regulated voltage when the battery is under a certain level but that's a bit in contradiction with a properly regulated voltage for the DMM.
Another one could be to have a voltmeter inside the box with a transparent window to monitor the battery voltage ( ex: https://www.banggood.com/0_28-Inch-2_5V-30V-Mini-Digital-Voltmeter-p-974258.html (https://www.banggood.com/0_28-Inch-2_5V-30V-Mini-Digital-Voltmeter-p-974258.html) ) (with another switch to activate it only on demand).
-
Sealed 12 volt SLA battery and a proper charger for it
Bodge up a connector (alligator clip leads) to your meter battery compartment
Drop the voltage with resistor to 6 or 9 volts or whatever your meter requires
If your meter needs 6 volts, get a 6 volt SLA battery and in you're in business immediately
You get instant 'floating isolation', clean power, lasts ages even with OLED power suckers!
Recharge the SLA battery once in a blue moon and ensure it never drops below 12.5 volts over time
And DO NOT forget a diode and fuse arrangement for this 'conversion' :scared:
-
Even if you avoid SMPS and choose to use a "proper" transformer, just using any old one isn't good enough.
A multimeter must be truly floating. For this reason, the AC line transformer used in professionally designed bench meters is a special type of transformer. Not only does it have some guaranteed voltage isolation between primary and secondary, but also ultra-low capacitance from secondary to primary and to ground. This is usually achieved with a extra robust insulation materials and a special metal shield placed in between the primary and secondary windings. A basic wall-wart transformer you pull out of your box of junk will never be sufficient. IF you can accept that your CATIII or CATIV meter will now be no better than CATI (extra low voltage) safe and maybe produce noisy readings, and may induce AC noise through the test leads, then maybe it's OK to try. Otherwise, find another solution, like a proper used bench meter with LED or VFD display.
-
Well don't cut corners and accept that the device is working under its nominal specs.
Below is one old meter with DIY external supply. I suspect it have been a workhorse for some radio and TV repair shop somewhere. Definedly not something I would put in to a circuit with more that ELV rating.
The transformer is some mystery transformer, propably from 1970s, but it looks like a well build and might even be a small isolation type. The Cap is old Siemens&Helske electrolytic from Finnish factory and the regulator is goldplated (SIC!) LM7918. The mainsplug is hacksaw modified to fit in new grounded sockets and the "high-speck" powercord have two layers of PVC insulation ... one layer for each wire :-DD. The low voltage cable is thin coax.
-
The common mode isolation of whatever power supply you come up with will be a major problem in at least the sense of safety.
A small AC power transformer with a high isolation rating followed by a linear regulator is one way to do it. One way to find these transformers are to look for ones intended for medical applications:
http://www.mouser.com/Power/Transformers/Power-Transformers/_/N-8u9n5Z1yzvvqx?P=1z0wcwh&Keyword=medical&FS=True (http://www.mouser.com/Power/Transformers/Power-Transformers/_/N-8u9n5Z1yzvvqx?P=1z0wcwh&Keyword=medical&FS=True)
Another option is to use a high frequency inverter with a transformer that has high isolation. Most of my bench DMMs do it this way. You could wind the transformer yourself however Coilcraft now carries them:
http://www.coilcraft.com/prod_isolation.cfm (http://www.coilcraft.com/prod_isolation.cfm)
-
Well don't cut corners and accept that the device is working under its nominal specs.
Below is one old meter with DIY external supply. I suspect it have been a workhorse for some radio and TV repair shop somewhere. Definedly not something I would put in to a circuit with more that ELV rating.
The transformer is some mystery transformer, propably from 1970s, but it looks like a well build and might even be a small isolation type. The Cap is old Siemens&Helske electrolytic from Finnish factory and the regulator is goldplated (SIC!) LM7918. The mainsplug is hacksaw modified to fit in new grounded sockets and the "high-speck" powercord have two layers of PVC insulation ... one layer for each wire :-DD. The low voltage cable is thin coax.
Well, that transformer is a Phillips part, probably used in a radio or other device as power transformer, and generally those wound core transformers are pretty good, and low capacitance as well with separate primary and secondary windings. Rest like mains cable is appropriate for being from the 1970's to mid 1980's, generally double isolated wire was not too common prior to the EU harmonising wiring to be double insulated, and that insulation likely uses PCB filled PVC wire as well, as it still is flexible.
Transformer with high isolation is the issue, as you need a transformer that has separate primary and secondary bobbins, not ones with them wound over each other, and they have to be rated to 2kV withstand.
-
How did you find out that it is Philips transformer, old cumulative knowledge maybe?
-
I'm sick of buying expensive 9V batteries and NiMH ones just last a few hours. :--
Any suggestions? :-//
PS: the idea is to keep it really floating and avoid the noise from a switching psu.
What NiMH and what charger? Do you have decent of each, a lot of store bought batteries and chargers are crap.
Though and easier option IMHO would be to just attach a 6 AA holder on the back.
-
I have seen a Y-type cap and series-resistor between earth-GND and multimeter GND on bench DMM's (mains-powered, floating). I think for common-mode noise, even though a 50/60Hz power transformer is used. 1000pF 6kV and 47R 1/2W.
Agilent does not use this, but they have an electrostatic shield between primary/secondary.
A switching PSU would be awful.
-
I had an idea to have the maximum possible isolation and low capacitance between the 9v supply and mains and ground:
Transfer energy between the primary side and secondary side with something else than a transformer. :scared:
Initially I though of a cheap Chinese power led a small photovoltaic panel in a box but the problem is efficiency. Probably in the order of a few percents.
So even if I need only 40mA*9V=0.36W I would use 10-30W. So a huge waste and lots of heat in a box. :palm:
Then I though of a mechanical way: a small DC motor connected through a non conductive shaft to a generator (basically a 3 phase motor) and a bridge rectifier, caps...
The problem is probably noise (sound*) then but I don't see how I could be more isolated and if the rotation speed is high enough it means that the ripple after rectifier is not at 50 or 60Hz but at a much higher frequency (1-10 kHz) so it's easy to filter it. Of course the efficiency won't be very high but I hope to get more than 10%. Finding a 3W motor and generator is easy and cheap ;D
*: Note that the photovoltaic solution is probably noisy too because of all the heat I'll have to use a fan
to eugenine: you are right the easiest solution would be to use a 7xAA pack of NiMH in a non-conductive box and I would have 50 hours of battery life (2000mAh / 40mA) or 3x18650 Liion for 70 hours.
Yet, if I leave the meter on for days to log stuff, it's a bit short.
Also the 9v NiMH battery I currently use are 200mAh with low self discharge rate. I know I can find 300mAh ones. But this is a ridiculously low lifetime of 4 to 6 hours. There are 400mAh liion ones but I don't like the presence of a boost converter and the noise that comes with it for a small gain at the end of the day: only 8 hours if the capacity is really what is advertised (but probably not).
-
The LED - PV solution might not be that bad. IR LEDs at 850-950 nm are relatively efficient, and also Si PV cells have about the best efficiency in this range. So maybe an 5-10% overall efficiency could be possible.
For good isolation there are acoustic transformers. Essentially using a piezo to emit ultrasound, sending it trough an isolating material (like quarz glass) and than use another piezo to convert back to electric power. It's kind of "noisy", but more like in the MHz range.
There are also special transformers with good isolation. I have used some at about 10 kV - though considerably larger than a DMM.
For low power this would be like rater large opening ferrite cores - similar to those found in TVs from the 1970s.
For the motor version there would be noise, but also vibrations, that are somewhat difficult to filter and can effect a meter.
-
It seems to be almost a trend on this forum to make things over complicated? :-//
Just look how regular benchtop multimeters are being build.
Use a similar transformer (which is pretty much ANY transformer nowadays) and done.
Next project please.
-
I'm wondering if you could mod the meter with a Qi charging coil perhaps?
And either charge an internal battery, or even power it directly?
-
It seems to be almost a trend on this forum to make things over complicated? :-//
Just look how regular benchtop multimeters are being build.
Use a similar transformer (which is pretty much ANY transformer nowadays) and done.
Next project please.
Thank you for your answer. You are right it's over-engineering ;)
But concerning transformer, I agree that it's easy to find 4000V insulation but the question raised by some other contributors here was the capacitance between the 2 windings.
You essentially end up with a capacitor connecting the DMM power rails to mains and earth.
So "any transformer" is probably not that good.
Of course bench multimeter manufacturers(of decent quality) can chose ones with the lowest possible capacitance. The problem is that this information is often missing in the datasheet of transformers you can buy as a hobbyist.
I measured a small transformer I have 230V-2x15V 8VA.
The capacitance between primary and secondary between 26 and 44 pF (DMM and small component tester results), also between the 2 secondaries, it's in the 800 to 900 pF range.
-
As far as I know, most/all (print) transformers have a 4kV isolation.
Maybe the very old ones didn't?
To be very honest, I do even think that the whole capacitance thing is extremely far fetched.
I find it very hard to believe that DMM transformers are 'that special'.
Especially for the price these things cost.
But some people seem to be overly concerned about pretty much everything.
(no, i am not gonna go that road again)
You can even go for just a small SMPS module, which are totally isolated from mains.
Only need to take care of the noise of course.
-
I'm wondering if you could mod the meter with a Qi charging coil perhaps?
And either charge an internal battery, or even power it directly?
This is a great idea. :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+
Manufacturers could actually have that as a standard feature in a future generation of products!
Il would only cost a few $ to do that.
charger $2.50: https://www.banggood.com/Qi-Wireless-Charger-PCBA-Circuit-Board-Coil-Charging-p-1000423.html (https://www.banggood.com/Qi-Wireless-Charger-PCBA-Circuit-Board-Coil-Charging-p-1000423.html)
receiver $2.50: https://www.banggood.com/Universal-Wireless-Qi-Charger-Power-Receiver-Charging-Adapter-USB-For-Samsung-Xiaomi-p-1123760.html (https://www.banggood.com/Universal-Wireless-Qi-Charger-Power-Receiver-Charging-Adapter-USB-For-Samsung-Xiaomi-p-1123760.html)
dc-dc $2.50: https://www.banggood.com/DC-3_3V-3_7V-5V-6V-to-12V-Boost-Voltage-Regulator-Module-Converter-Step-up-Power-Supply-p-1136710.html (https://www.banggood.com/DC-3_3V-3_7V-5V-6V-to-12V-Boost-Voltage-Regulator-Module-Converter-Step-up-Power-Supply-p-1136710.html)
(to convert 5V to 12V then some filtering and a 7809)
total: around $10 ;D
The receiver could probably stay inside the battery compartment so no need for holes in it, so basically no modification of the DMM.
-
I could never get more than 150ma or so out of a Qi no matter how well coupled the system was.
-
I only need 40mA ;D
-
Then you should be good to go. :-+
-
I had thought about the Qi idea for a couple of years, but never got around to testing the idea. I suspect that it will be a bit difficult to get a clean reading multimeter with such a strong magnetic oscillating field at high frequencies so close to the sensitive circuitry inside the meter plus any dc-dc converter rf hash.
-
It seems to be almost a trend on this forum to make things over complicated? :-//
Just look how regular benchtop multimeters are being build.
Use a similar transformer (which is pretty much ANY transformer nowadays) and done.
Next project please.
Thank you for your answer. You are right it's over-engineering ;)
But concerning transformer, I agree that it's easy to find 4000V insulation but the question raised by some other contributors here was the capacitance between the 2 windings.
You essentially end up with a capacitor connecting the DMM power rails to mains and earth.
So "any transformer" is probably not that good.
since the required power is really small, just get a couple of small and cheap PCB mount transformer
choose the first to be 1:1 with a centre tapped sec. and another one to provide the voltage you need
then connect the sec. of the first transformer to the prim. of the second one with the center to ground and get higher insulation, lower noise and less capacitive coupling
-
I did a similar thing with an old Marconi moisture meter, got tired of the weekly purchase of batteries for it, and as it used a D cell for the directly heated filaments, and needed a 90V HT supply, provided by 10 PM3 blocks in series, i made a battery eliminator using a lot of small 1W transformers I had to hand ( bought on auction for under $1 for a gross of them, still NIB) took 2 and paralleled the secondaries, to give 12VAC at 200mA, enough to rectify and use a LM317 to drive the heaters from 1V2 as heater supply, and used another 2 in reverse. 12V windings in series, and 220VAC windings in parallel to give roughly 100VAC to the bridge rectifier, 22uF 400V capacitor and then a stack of 400mW zener diodes and a resistor to give 90VDC as anode supply. Transformers are all same lot, so reasonably well matched, and with 12R secondary DC resistance no real issue with circulating currents either. Then left it on most of the time to keep stable, as it was basically a valve conductivity meter, used to set a particular moisture content in feedstock. Was a lot more stable than using batteries as well, as it did not drift with cells ageing.
As to how i know it is a Phillips transformer, I have similar ones from Phillips and Bang and Olofsen ( rebadged Phillips) audio and TV equipment, which used them as either mains transformers, or for the standby transformer on units with a remote control, as they are low loss, very robust and most importantly make no noise at all from core or windings. They also provide good noise immunity from mains transients, as you did not want RFI on the audio from the TV in standby, as the speakers and amplifier in many B&O TV sets were always powered as part of the Beolink system, so they could play audio coming in even with the TV set in standby and with the window blinds closed.
-
It seems to be almost a trend on this forum to make things over complicated? :-//
Just look how regular benchtop multimeters are being build.
Use a similar transformer (which is pretty much ANY transformer nowadays) and done.
They use a high isolation transformer for this so it is not just any transformer at least if they want the common input to be rated the same as the positive input.
since the required power is really small, just get a couple of small and cheap PCB mount transformer
choose the first to be 1:1 with a center tapped sec. and another one to provide the voltage you need
then connect the sec. of the first transformer to the prim. of the second one with the center to ground and get higher insulation, lower noise and less capacitive coupling
One of my bench multimeters works this way with a separate small 60 Hz transformer on the analog board but the transformer is a custom one. My others use custom high frequency inverter transformers like the Coilcraft ones I linked which I actually think is a better way to go.
With the center floating winding wired to ground, the capacitive coupling and isolation is that of the second transformer.
-
Usually the very small ( < 3 VA) transformer have relatively high no load loss. So using several of them instead of a single suitable one can increase the loss a lot - the special standby mode transformers are a little better, but usually not that much.
For a moisture meter a normal transformer is OK. However with a voltmeter / DMM a better isolation might be needed, as the transformer's isolation can see a higher voltage (mains + measured voltage worst case) and one might have contact to the terminals.
QI charging might work, but it could well cause to much interference to do a measurement while measuring.
The charger part from an old electric toothbrush might be an option too. It works similar, but a more closed magnetic field.
-
Just use a piezoeletric / piezo-ceramic transformer: For example this design will provide 10V to 15V DC/DC-converter with 20 kV isolation:
http://www.edn.com/design/power-management/4413068/Some-thoughts-on-DC-DC-converters--part-four (http://www.edn.com/design/power-management/4413068/Some-thoughts-on-DC-DC-converters--part-four)
-
to eugenine: you are right the easiest solution would be to use a 7xAA pack of NiMH in a non-conductive box and I would have 50 hours of battery life (2000mAh / 40mA) or 3x18650 Liion for 70 hours.
Yet, if I leave the meter on for days to log stuff, it's a bit short.
I didn't see in the first post needing days, in that case https://www.batteriesplus.com/battery/sla-sealed-lead-acid/werker/12/99=166 (https://www.batteriesplus.com/battery/sla-sealed-lead-acid/werker/12/99=166) :)