Fluke 8600A Battery / Power Supply Conversion

[xrunner]

New project arrived yesterday - a Fluke 8600A w/battery option.

This unit is extremely clean, and it really has only one small problem - it doesn't work. But I believe the reason it doesn't work is because this unit has the battery supply option, and the batteries are totally shot. Some of them look to be original NiCads. Now you may ask, well, just flip a switch inside to make it run off the mains, take out the batteries, and call it a day.

'Fraid it doesn't work like that. 

It has to have the batteries to work because it essentially runs off of 5V, and has a  DC-DC converter that provides +/- 15V all the time, powered by the batteries, all the while the batteries are charging. No battery - no worky.

Well, why don't I replace the batteries then? Because I'm not interested in running it off NiCads and I don't want to buy any D-size NiCads, especially for a unit that might sit for a long while without being used. Any future use would be for a benchtop DMM anyway. No, I'm going to convert it to let it run off the mains all the time. I also want the conversion to be minimally invasive to allow some future owner to restore it to it's original condition easily, if they so desire. I've researched some other people's ideas, some here on this forum, but I have another idea also.

Feel free to take a look at the power supply schematic while I do some investigation and testing myself. 

[Seekonk]

Should work if you pull the batteries and place a 4700uF cap in its place.  Interesting design, what size uf is C201?  That might have to be increased with an additional parallel cap.  Watch how hot the zener gets, especially OFF.  Might want to disconnect the OFF winding of the transformer.  Do some tests and report back.

[xrunner]

Should work if you pull the batteries and place a 4700uF cap in its place.

It might at that, but I'm still keen on another idea. 

Interesting design, what size uf is C201?  That might have to be increased with an additional parallel cap. 

It's 5 uF. According to the manual it's a cap that has to be changed out depending on the operating voltage and line freq. , of course it's the correct one for 120V - 60 Hz, but I may not have to worry about it if my idea pans out.

Watch how hot the zener gets, especially OFF.  Might want to disconnect the OFF winding of the transformer.  Do some tests and report back.

I'm testing my idea this afternoon, will report back ... 

[Seekonk]

If you notice there is no dropping resistor for the zener that prevents over voltage to the batteries.  The 5uf capacitor performs that function of the voltage drop, that is the simplistic explanation.  You have to be careful with capacitors in series with inductors.  I have these nice German fans that are 220V.  To use them on 120V I place a 4uF in series and that gives me 190V on the motor because the RL is nearing resonance which will cause the highest voltage on the inductor and capacitor in series.  Note also when the unit is off it switches to a higher winding to charge because there is no current draw from the meter electronics.  I would disconnect that winding if is going to be plugged in all the time.

[SeanB]

Replace CR22 with a 5V1 25W zener on a smallish heatsink, so that it provides the required 5V rail. Parallel a 1000uF 25V capacitor across it as well, and remove CR21 and CR23 as well, and disconnect the red terminal on the transformer. Place a 120V MOV across red and black on the transformer primary, this will help with high voltage spikes.

[xrunner]

Replace CR22 with a 5V1 25W zener on a smallish heatsink, so that it provides the required 5V rail. Parallel a 1000uF 25V capacitor across it as well, and remove CR21 and CR23 as well,...

Thanks for the idea. Here's what I've done: I, of course, had it disconnected from the mains. I then removed all the batteries, which were turned in for recycling at Home Depot. I then simply connected a bench PS set to 5V to the appropriate terminals on the board and the meter worked perfectly. I noted that the current draw was 430 mA. So this tells me which direction I want to go. It's favorable to my initial idea. 

[SeanB]

That is only if you are not willing to remoce CR22, replace it with a 1000uF 25V capacitor and then snip one lead of L2 and place a 7805 on a heatsink in series with the lead, and with the case connected to the common of the power supply on the transformer yellow lead. Then you can leave the primary side as is, and it will have a regulated 5V rail and a lower power draw from the mains to boot.

[edavid]

Should work if you pull the batteries and place a 4700uF cap in its place.

This is incorrect, you also need a regulator.  A Chinese "LM2596" stepdown module should work well.

[SeanB]

Should work if you pull the batteries and place a 4700uF cap in its place.

This is incorrect, you also need a regulator.  A Chinese "LM2596" stepdown module should work well.

There is a zener regulator in there, CR22, but I am wary of running it at 400mA of current, as it probably is a 1W zener. It will probably go short circuit after a while, but if it goes open circuit bye bye meter.

[edavid]

Should work if you pull the batteries and place a 4700uF cap in its place.

This is incorrect, you also need a regulator.  A Chinese "LM2596" stepdown module should work well.

There is a zener regulator in there, CR22, but I am wary of running it at 400mA of current, as it probably is a 1W zener. It will probably go short circuit after a while, but if it goes open circuit bye bye meter.

The manual says that CR22 is a UZ8720 20V 1W zener.  The supply voltage goes up to 7-8V without batteries, which is not fatal, but not so great either.  As I said, add a regulator.

[xrunner]

The solution I'm going to use is coming out of this box, but there's more to it than just soldering up a couple of wires. 

[edavid]

The solution I'm going to use is coming out of this box, but there's more to it than just soldering up a couple of wires. 

The trouble with that is that some of the 8600A DC-DC converter modules are not really happy at 5.0V, they need 5.5V or so.

[xrunner]

The trouble with that is that some of the 8600A DC-DC converter modules are not really happy at 5.0V, they need 5.5V or so.

Oh I already checked it out with 5V from my bench supply - it's happy. 

But I have found a candidate. I've patched it in and the meter works perfectly. I'm going to integrate the little board into the meter, this is just a test. The little switcher can handle 700 mA A. It ran all afternoon with no problems at all. The meter will still use the same line cord as usual.

[xrunner]

I removed the little supply from it's "wart" and will be mounting it inside the meter along with the mounting parts for the batteries, so as to keep all the original parts inside. It turns out I have to do little to make this work. First off I'm pulling the input wires for the transformer and bundling them inside some heat shrink - they will be unused. The mods I'm making are shown in the schematic.

[SeanB]

Add the optional MOV to that little board, it will help it last longer.

[Seekonk]

Good refurb. I think you made the best decision.  I thought about the many ways to convert this Li batteries if it were mine since I have stacks of used batteries.  This is a good bench meter, but too clunky these days to take out in the field.

[xrunner]

Add the optional MOV to that little board, it will help it last longer.

OK good idea.

Good refurb. I think you made the best decision.  I thought about the many ways to convert this Li batteries if it were mine since I have stacks of used batteries.  This is a good bench meter, but too clunky these days to take out in the field.

Thanks, yea these days who would drag that into the field with batteries - the thing was pretty heavy with all those D cell NiCads. But it's a really good bench meter so why not convert it? I was very surprised when I got this one because it's so clean. The inside looks like it just came off the factory floor. On the top back it has "USN" stenciled on it, but I would never have believed it was in the possession of the Navy. It's just not been abused at all. Well, except for the V-ohn and common jacks. The inside is so worn the banana plugs do not fit in at all tightly anymore. I'm going to change them out with new ones. Maybe it was in some lab for years and just sat on a bench, someone plugging and unplugging various things? 

[Seekonk]

Since you have a stack of wall warts, it is easy to change the voltage up to several volts.  That opto isolator likely has a zener in series in which case a diode or two in series will bump it up a little.  OR they have a TL431 that gives you lots of options for voltage.

Some here talk about crappy wall warts cause they draw half the rated current and the voltage drops.  Look for a resistor on the output that is a current sense feeding a transistor to the opto.  Remember, most of these are chargers and they limit the output in the case of dead or damaged batteries.

I do a lot with solar and many of these modules will work down to 50V at reduced current.  Nice for an isolated high side driver or operating a micro.

[xrunner]

So here is the final installation. I kept all the holder parts for the NiCads, and used one of them to hold the PS board off the circuit board

I also needed to replace the banana jacks on the fromt due to extreme wear, so I disassembled the instrument and pulled the front panel so I can replace the jacks and re-work the wiring to them.

[PlaneSailing]

Case temperature after Fluke 8600A line only conversion ?

   Was given a non working Fluke 8600A with the battery option, and I converted it to line only. Took out the NiCads and plastic holder pieces, and saw much corrosion of the PCB traces.  2200uF 25V and an LM7805 in their "line option" PCB locations of C103 & U102, custom heat sink of bent aluminum, a few cut traces & jumpers, and it now works well.  The display was also "loopy" till re-soldered the connections from the flexible display PCB to the mother board.

   Had to remove the yellow wire from the center tap of the transformer primary side, and also CR22 (mislabeled as CR21 on schematic) & CR 23 to fully remove the line voltage from the transformer when turned off.  A 0.1uF on the output of the LM7805 to try to remove the 50mV pk-pk 10kHz regulator switching noise resulted in oscillation of the new 5V regulator (against the L2 coil maybe ?). Otherwise all well, so just did not use a 0.1uF.

   It's a cute little 4.5 digit meter - not as good as my HP 3466A, or as much character as my Solartron 7045... but really nice.

   Question : Can someone with a "natural" line only 8600A, or the battery option version, tell me if the outside top of the plastic case over the transformer gets warm ?.  Mine does, a noticeable 120 degF (measured with IR thermometer) after 30 minutes at the hottest spot on the center back top of the case. I "think" it's the transformer.  If it's not supposed to be that warm, plan to drill a few 1/4" holes in the back of the case, thanks.

[PlaneSailing]

Yes, aware the 2 versions (battery or line only) of the 8600A use different transformers.  If someone with either version and an IR thermometer can measure the hottest spot on the top of the case 30 minutes after power on, at normal room temp, would be grateful.  Then I can tell if my observed 120 degF after converting a battery version to line operation is "normal", or something I have to work on, thanks.

[Jeff eelcr]

I have the line version not an -01 option I can check tuesday next day at work
If no one else does it first.
Jeff

[BobRyan]

I have the line-powered version as well. I've left it on ACV/AUTO for the last hour and the hottest spot I can find is 100.5*F (rear and just left of IEC-C14 input while viewing the display).
That said, I still wouldn't worry much about 120*F.

[edavid]

I checked a couple of line units, and they both reached 105F over the transformer.  Ambient temp was 60F.

[PlaneSailing]

My thanks to BobRyan and edavid for measuring 100 to 105 degF case temperature over the line transformer on their 8600As.  After my battery option to line power conversion I measure 120 degF outside case temp over the transformer.  Took off the case, and the transformer itself is the same 120 deg. after 30 minutes.

I am worried about the plastic case coloring/melting, and wonder if there is anything I can do to draw less line current and thus run the transformer "cooler" ?.  Measure 8.2V DC after the rectifying diodes and CR27 7.5V zener controlling the Q43 triac, presumably they are included in the original battery powered circuit to hold the voltage at 8.2 so as to not damage the nicads ?.  Thus the triac/zener is "shorting" the transformer secondary voltage above 8.2V both cycles of the rectified AC.  As I have removed the nicads, maybe removing the triac and zener also would help reduce line current ?.

An alternate idea is to improve the cooling by drilling holes in the back of the case either side of the power socket, but don't think it would help, as measure the same 120 degF with no case......  But then could install a small 40mm 5V fan to force air in one side, around the transformer and out the other side.  Getting far away from the original design, but the inside is already "hacked up".  Appreciate any advice, thanks.

[Side note:  Meter was reading DCV, DC current and resistances 1% low (did not try ACV or ACI) that could not be calibrated out.  So I soldered a 470 Ohm resistor in the A/D reference chain (in series with R60) to lower the 1V/0.1V reference - a calibration, and the readings are now where they should be.  A great little meter.]