Fluke 45 transfomer foobar

[kosmonooit]

Greetings all,

My Fluke 45 desktop DMM stopped working - the input power transformer is blowing fuses, seems its toast. Sustained high supply voltages (+400V)  here might well have popped it. Nominal is 230v AC. I read that it has an embedded MOV but its all potted up as a unit, so don't think I can fix that. 

Now to find a replacement. The Fluke 45 parts list has it as a Power Transformer, 35W, 90-264V AC, 14.6-43vAC, Part No 609088.   But the output has three wires so is that two taps for 2x 14.6-43 VAC?  Since I can't apply an input I can't measure, and I can't see to find a wiring diagram or any other info, although I do have the service manual. Looking the the PCB both sides seem symmetrical, the center wire seems the neutral / common. 35VA or bigger rating should do?
 
Anyone have an idea of this setup and/or suggestions for a replacement power transformer?

This setup has the mains power always applied to the input of the transformer so its never really Off as such.  I will fit a toggle switch at the back from the AC Socket, also going to have to do some MOV input protection.

Thanks.

[xrunner]

Here's the schematic of the transformer -

[SeanB]

Yes, time to go buy a few surge plugs from Ellies, or spring for the Clearline DIN rail mounted MOV units that you fit to your main board, which will at least provide protection against Eskom's voltage surges. Generally you want them anyway, I have a good number already, and have done DIY using 275VAC MOV's in a plug, that you plug in. Have had a few sacrifice themselves to protect equipment already.

But before you condem the transformer check that MOV across it, it might have died to protect the transformer, as the MOV normally fails short circuit, unless there is enough current to blow it open. Easy to replace the MOV with a new one, and thank Fluke for including it as well.

[kosmonooit]

Thanks for the circuit diagram xrunner, it seems that center tap / wire just goes off the battery supply option so the listed secondary voltage should be across the outer two wires (?) and I need not bother with the  center, I have  been struggling to find a transformer here with those specs ie 2x24v.


Ja Sean those DIN Clearline surge protection units are excellent, I use only them at client installations, but you know the old story about the plumbers own plumbing...  My problem is Joburg City Power, unspeakable mess at the moment, I am just about off grid anyway,

It seems the mov is deeply embedded in this OEM xformer and its properly potted closed. There is a metal  cover that I could try to cut off to see if I can get to the mov, but looks unlikely since I can see the primary wires going in to the wrapping of the coil paper.

230-24v AC 35VA  xformer should be easy to find.  Chint make a NDK-50VA-24V, 50VA but that should not be an issue. I normally get chint from  Key Electrical in Midrand so will be contacting the on Monday.

Anyway hope the problem stops here, I am missing those VFD numerals.

[alm]

The original transformer will likely have an electrostatic shield, if not two, between primary and secondary to reduce capacitive coupling and hence common mode noise. The isolation voltage between primary and secondary might also be a problem for a standard transformer if the analog circuits are powered straight from the secondary windings without extra transformer (I haven't used the 45).

[kosmonooit]

Thanks - the OEM transformer is encased in a steel plate covering each side of it that is stamped to the shape of the coils and firmly potted in, would that be a Faraday Cage idea for EMI?  Not sure about the plate between the coils, perhaps time for a dissection ...

I could  make a small sheet metal box for a off the shelf replacement, I am struggling to a original replacement that is feasible to source. I did find a used one available from ebay from China for like US$70, that's not really an option.

Is there a supplier in the USA for new or even NOS Fluke parts? I haven't seemed to be able to fins anything.

Thanks for the help.

[alm]

Thanks - the OEM transformer is encased in a steel plate covering each side of it that is stamped to the shape of the coils and firmly potted in, would that be a Faraday Cage idea for EMI?  Not sure about the plate between the coils, perhaps time for a dissection ...

No, the shield would be internal, wound between the primary and secondary winding. You can see the shield in the schematic (figure 9-1 from the service manual) as attached to earth. The only external sign would be an extra terminal that's attached to earth.

[SeanB]

No you need the centre tap, it is going to be used to generate split rails, plus with the charger to do the battery. The transformer is only a 12VAC unit, so you probably can use a smaller transformer, like

https://www.mantech.co.za/ProductInfo.aspx?Item=50M0034

to replace it. 12VAC between the centre taps and the centre common, and 500ma rating, which is likely more than enough for the meter. It says 15VA, though likely that is more for the battery charging. Do you have the optional true RMS option installed.

Otherwise

https://www.mantech.co.za/ProductInfo.aspx?Item=65M1113-H

Though you will have to check it will actually fit in the volume there, and Mantech is reasonably close to you anyway.

[kosmonooit]

Thanks I'll order from Mantech and give it a go. I am regular there, and all their predecessors. I will have to live with some common mode noise until I find a decent OEM replacement.

As I understand it True RMS AC is a standard feature in the  Fluke 45.

[Paceguy]

Best to find someone with a 45 that's been parted out and see if they will sell you the transformer.

[robert.rozee]

hi kosmonooit,
    as you've noted, the operating input voltage range of the Fluke 45's transformer is quite wide, 90v to 264v. while i'm no expert on mains transformer design, it seems that this particular one has been over-engineered in some special way to withstand such a wide input range. any other off-the-shelf transformer is likely to require fitting a 115/230v switch.

i own a fluke 45 myself, and with 115vac input the rectified output voltage measures around 15v. with 230vac input it was a little over 30v. also, i have successfully powered the meter from 2x 18650 Li-Ion cells, producing around 7.5v, without any issues. note that there is a further isolated [addendum: am not so sure that it is isolated!] switching PSU on the meter's main PCB that takes this wide range of DC voltages and converts it into the various exact voltage rails used internally by the meter.


given the above, my inclination would be to look at getting a small switchmode module, of reasonably quality, similar to this one:
https://nz.rs-online.com/web/p/switching-power-supplies/6210607
or
https://nz.rs-online.com/web/p/switching-power-supplies/1065824
and fit it in place of the blown mains transformer. you can wire the -ve side to where the transformer's centre-tap was connected, and the +ve side to either (or both) of the end connections.

i believe that there is sufficient 'other electrical separation' in the meter for this 'downgrade' to have no material impact on the performance of what is only a 4.5 digit meter.


cheers,
rob   :-)

[kosmonooit]

Thanks Robert, I've used lots of Meanwell power supplies in all sorts of projects, good stuff! 

Question is how would I wire the single rail DC output into the Fluke 45?

[robert.rozee]

Question is how would I wire the single rail DC output into the Fluke 45?

have a look at the schematic in Reply #1. P7/J7 has 3 pins, the middle one goes to ground, while the top one goes to the anode of CR2 and the bottom one goes to the anode of CR3. the cathodes of CR2 and CR3 are tied together and feed C27 (the filter capacitor) and this then connects to pin 5 of S1 - the front panel on/off switch.

note: there is an additional wire coming off the cathode of CR3, but you can ignore this. it goes via a small diode and 5k6 resistor to a 10uF smoothing capacitor and then on to the micro that controls the whole meter, enabling the meter to tell if it is powered up but there is no AC coming out of the mains transformer. this is so the meter can decide if it needs to do anything with the 'low battery' annunciator on the VFD display.

if you connect together the top and bottom pins (ie, outside pair) of P7/J7, you can feed in here anything from around +7v up to a maximum of +35v, with the negative return going to the middle pin.  CR2 and CR3 will be permanently conducting, but this doesn't matter as the energy loss is minimal at the 200mA or so (@15v) the meter consumes when turned on.


so, assuming you are happy that the mains transformer is shot:

1. cut off at the transformer body the three wires going to J7. on my meter, these are two RED wires and a YELLOW wire. connect together the two  RED wires, and run them to the +ve output terminal on your Meanwell PSU. connect the single YELLOW wire to the -ve output terminal on your Meanwell PSU.

2. unplug the two black wires coming from the old transformer body that are plugged into the mains socket on the meter's back. connect PHASE and NEUTRAL inputs of your Meanwell PSU in their place, ensuring PHASE goes via the fuse.

3. assuming your meter does not have the battery option fitted, there should be more than enough room to accommodate the new PSU module. if there is a battery option fitted, and the battery is not dead, then you have quite a bit less space to play with.


the Meanwell PSU will be powered up permanently, even when the meter is turned off and no current is being drawn from the output. i am assuming it is designed to handled this - hence my suggestion of getting "a small switchmode module of reasonably quality".


cheers,
rob   :-)

[Kleinstein]

A DMM is something where I would prefer a classical old style transformer. This is because of less EMI and better isolation for RF interference.  Even a good medical grade SMPS is still noisy compared to simple ol d iron.

[robert.rozee]

A DMM is something where I would prefer a classical old style transformer. This is because of less EMI and better isolation for RF interference.  Even a good medical grade SMPS is still noisy compared to simple ol d iron.

as designed, the Fluke 45's mains transformer (producing between 15v and 30v after rectification) is followed by an LM3578 configured as a 5 volt buck converter (40kHz), which then feeds into a 4-transistor multivibrator (30kHz) driving a step-up transformer with multiple outputs for in-guard and out-guard circuitry.

as there is already plenty of 30kHz and 40kHz noise floating around in there, hopefully, given the relatively modest resolution of the meter, an extra switching converter will not adversely affect performance in the owner's specific usage case. while an exact replacement for Fluke's bespoke mains transformer would be nice, the alternative of an off-the-shelf SMPS is better than nothing.


cheers,
rob   :-)

[Kleinstein]

Withpout special shields a main SMPS produces quite some common mode signal. For this reason they usually come with a common mode choke and class Y capacitors from the primary to secondary side. The common mode problem problem is different from the interference the buck converter produces. Also the step up transformer with lower voltage produces less CM signal.

I would consider a normal transformer much better than a standard SMPS. If at all one would want a medical grade SMPS with extra shields.

[robert.rozee]

I would consider a normal transformer much better than a standard SMPS. If at all one would want a medical grade SMPS with extra shields.

hi Kleinstein,
    i've been doing a little more research, and it looks like, on-the-whole, one can take a 230v transformer and run it on 110v with the downside being just reduced output voltage and no additional current output available at that lower voltage.

lets say that at 264v AC the transformer you select outputs 30v when rectified (this allows a small margin below the 35v rating of C27). at 90v AC, the rectified output would drop to a tad over 10v. these limits of available voltage (10-30v DC) are within the operating range of the Fluke 45's buck converter.

however, the meter draws (according to my measurements) a bit less than 200mA @ 15v DC, so 3 watts. at 10v DC (our minimum voltage) this would go up to around 300mA, so therefore the transformer would need to be sized to be capable of supplying 300mA at 30v (when fed from 264v AC). this gives a about a 10VA required rating, lets say 12VA to be on the safe side. looking at it, this does seems doable, and without the battery option fitted there is plenty of space available.

the question remains, would a 'regular' 12VA transformer be suitably screened, etc to fit your PSU preference? or would the original poster still need to obtain something with additional shielding, etc?


personally, i'd still go for a "small switchmode module of reasonably quality" and evaluate its performance. i suspect it would prove to be quite adequate.


cheers,
rob   :-)

[kosmonooit]

Thanks for the help all, what I did was get a 2x12V  24 VA transformer as Sean suggested. I soldered a MOV across the 230V input terminals and I also replaced the IEC socket on the back panel with a type that has a line filter (which has to be a good thing right?) and thus had to fit a panel mount fuse. I also fitted a big ON/OFF toggle switch on the rear panel  to switch off power to the transformer so its not powered up all the time, which is an issue since power can go south here very quickly.  I've had 410V for days/weeks

The unit seems to work okay now and its back in action.

I must say the build quality of this Fluke 45 is exceptional, my hacks have made functional but it would have been nice to keep it pristine. That was not possible since I could not source an OEM transformer and a used one was not not really an option since they are powered up all the time, beside the cost of sourcing that. The old transformer is a thing of beauty, and will keep it for that reason.

 

[Kleinstein]

the question remains, would a 'regular' 12VA transformer be suitably screened, etc to fit your PSU preference? or would the original poster still need to obtain something with additional shielding, etc?

personally, i'd still go for a "small switchmode module of reasonably quality" and evaluate its performance. i suspect it would prove to be quite adequate.

The regular transformers have no extra screen. A screen is more like an exception. One can still get regulator transformers with seprate sections for the primary and secodary. That form is relatively common and give low coupling capacitance and good isolation at least.  A standard mains transformer is still better isolation than most SMPS, maybe with the exception of some medical grade SMPS that have shields. So the natural choice for likely good enough is a old style mains transformer with separated windings.  The idea of using a 230 V rated transformer with 110 V can help to reduce the no load loss and can make sense for very low power.  It makes less sense with a 10 or 12 VA 230 V transformer instead of a regular 115 V 6 VA transformer.

[robert.rozee]

[...] I also fitted a big ON/OFF toggle switch on the rear panel  to switch off power to the transformer so its not powered up all the time, which is an issue since power can go south here very quickly.  I've had 410V for days (230V Nominal)

The unit seems to work okay now and its back in action[...]

410v suggests that your grid supply is arriving in your area as 3-phase, 410v between each phase. houses down the street will be supplied with a single phase (houses alternating through the three, so you may be connected to L2, while your neighbour to the left is connect to L1 and your neighbour to the right is connect to L3) and all houses will share a common neutral - at the street this neutral wire will be a lighter grade than the phases, as across multiple dwellings it carries minimal current.

the fault that sees your house being fed 410v is when the neutral wire is, somewhere, being shorted to one of the phases. this will mean 1/3 of the houses in your street will receive no power, while the other 2/3 will receive 410v! if this is something the lines company will/can not fix, i'd be powering everything sensitive from batteries...

which leads into the fact that the Fluke 45 will be quite happy working off batteries. a 12v gel cell feeding in place of the transformer will work perfectly well. then when you are not using the meter disconnect the gel cell and recharge it.


cheers,
rob   :-)

[SeanB]

400VAC is because some lowlife stole the neutral and grounding busbars at the local substation, a very common thing in South Africa.