Fluke 731B NIMH battery pack

[VintageNut]

This is long overdue. I soldered together a 12S1P NIMH battery pack and installed it in the first of two units. Awaiting gold plated crimp terminals for the second pack.

The 731B has been operating for a week on the new battery pack. It is great to be able to unplug the 731B have it continue to run.

The batteries are 2aH which is a big upgrade from the OEM 450 maH battery pack. The pack removed was NiCd and was date 1981. Probably dead for 20+ years.

In February I will be taking an 8+ hour drive to a calibration lab to have both of the 731Bs and a 5450A calibrated for the first time ever since I bought them. The plan is to have these calibrated once or twice a year to make these part of a real calibration capability at my home office. This will allow me to check voltages and currents of my other instruments at home.

The DMM7510 tells me that the 731B has a standard deviation of 2.6 uV out of 10V over several days of acquisition.  0.26 ppm is much better than any measurement instrument that I have. 

[Conrad Hoffman]

I made up a new NiCad pack for mine some years ago, and really need to open it up and check the condition. At the time I wanted to use NiMH, but read that they won't tolerate a float or trickle charge for long periods. Is that still an issue, because I don't think the unit has any intelligent charge control.

[VintageNut]

I made up a new NiCad pack for mine some years ago, and really need to open it up and check the condition. At the time I wanted to use NiMH, but read that they won't tolerate a float or trickle charge for long periods. Is that still an issue, because I don't think the unit has any intelligent charge control.

Good question. I do not have an authoritative opinion. I have read dissenting opinions concerning NiMH inside the 731B. The reason that I chose NiMH over NiCd is that NiCd have memory and I have had nothing but bad experiences with NiCd when used as a power tool energy source.

Only time will tell. Power goes out occasionally in my area and I want the 731B units to stay powered 24/7/365 for the foreseeable future until such a time that I no longer want to have calibrated equipment in my hone/office/lab.

For the short term, I wanted packs that would allow me the 8+ hour drive to the cal lab that I have access to.

What I would REALLY enjoy is the DMM7510 to be running on batteries.

[dr.diesel]

What I would REALLY enjoy is the DMM7510 to be running on batteries.

That would be the same a running with line sync off right?  I have a dedicated UPS bus here at my house, but have left my sensitive T&M gear off it and on standard utility power due to those concerns.

I've also wondered if by having two phases at some random offset in the same general area is affecting my very low measurements.

[VintageNut]

What I would REALLY enjoy is the DMM7510 to be running on batteries.

That would be the same a running with line sync off right?  I have a dedicated UPS bus here at my house, but have left my sensitive T&M gear off it and on standard utility power due to those concerns.

I've also wondered if by having two phases at some random offset in the same general area is affecting my very low measurements.

Yes. Good points. It may be more useful to acquire an AC power supply that is pure sine wave and completely isolated from the spikes and droops of the power grid.

[Andreas]

I made up a new NiCad pack for mine some years ago, and really need to open it up and check the condition. At the time I wanted to use NiMH, but read that they won't tolerate a float or trickle charge for long periods. Is that still an issue, because I don't think the unit has any intelligent charge control.

Hello,

NiMH have around factor 4 higher capacity than (old) NiCd batteries with the same size.
So with the C/10 trickle charge rate a NiCD will get around 50 mA.
The NiMH of the same size with 200 mA will naturally go much hotter (get cooked).
Some manufacturers of NiMH cells allow a trickle charge with C/100 ... C/50 or even C/30.
This makes sense to me since the same package as the old NiCD will not be able to
dissipate more than the 50mA without getting too hot.

At room/lab temperature I do the following for my references:
I do a CC/CV charging with C/10 up to C/4 limited to around 1.43 .. 1.45 V per cell.
So for a pack of 12 cells I use a power supply of 12*1.45V for the cells + 0.7V diode drop
for a series diode = 17.1V in total.
The final value of the voltage has to be adjusted to around C/50 (30-50mA)
at the hottest room temperature when the battery is fully charged.

My first pack lived 3.5 years (until in a very hot summer the trickle current went to around 80mA).
The 2nd pack is now running 6 years.

With best regards

Andreas

[VintageNut]

I made up a new NiCad pack for mine some years ago, and really need to open it up and check the condition. At the time I wanted to use NiMH, but read that they won't tolerate a float or trickle charge for long periods. Is that still an issue, because I don't think the unit has any intelligent charge control.

Hello,

NiMH have around factor 4 higher capacity than (old) NiCd batteries with the same size.
So with the C/10 trickle charge rate a NiCD will get around 50 mA.
The NiMH of the same size with 200 mA will naturally go much hotter (get cooked).
Some manufacturers of NiMH cells allow a trickle charge with C/100 ... C/50 or even C/30.
This makes sense to me since the same package as the old NiCD will not be able to
dissipate more than the 50mA without getting too hot.

At room/lab temperature I do the following for my references:
I do a CC/CV charging with C/10 up to C/4 limited to around 1.43 .. 1.45 V per cell.
So for a pack of 12 cells I use a power supply of 12*1.45V for the cells + 0.7V diode drop
for a series diode = 17.1V in total.
The final value of the voltage has to be adjusted to around C/50 (30-50mA)
at the hottest room temperature when the battery is fully charged.

My first pack lived 3.5 years (until in a very hot summer the trickle current went to around 80mA).
The 2nd pack is now running 6 years.

With best regards

Andreas

This is good information. I suppose I should monitor the trickle current as well as possibly monitor the cell temperature.

[Pipelie]

I replaced with 4S lithium battery pack, it works well.
 

[VintageNut]

I replaced with 4S lithium battery pack, it works well.

Did you use a battery management system board? Did you modify charge current?

[SeanB]

I see you used the RS tagged cells there. I have used them, and they are reasonably good, though I strongly suggest taking the pack out, removing the tape and using some kapton tape to hold the pack together, some more tape at the cell ends under the tabs to increase insulation thickness and then another tape layer on the end for insulation, then put a large heatshrink sleeve ( transparent one if possible, you can then see the green crud and replace the cells before they rot apart) and shrink the sections into blocks, then put back with an inline connector ( pity you cannot get those nice gold plated hermaphrodite connectors easy any more, they are so sexy) so the inevitable corrosion is limited to the battery only.

[HighVoltage]

This was my replacement of the batteries, worked really well.

https://www.eevblog.com/forum/repair/fluke-731b-dc-reference-standard-repair/msg1073583/#msg1073583

[dr.diesel]

Both of my 731Bs have been powered with NiMH batteries for weeks now.  If the trickle charge heating is an issue I cannot detect it, internal amb temp is hovering at 76F and the batteries are not distinguishable from that.

Perhaps not an issue?

[orin]

Note that the original charging circuit on the 731B is a diode and 510 ohm resistor from the rectified, but unfiltered secondary voltage from the power transformer.  It therefore isn't a constant trickle charge, but a series of pulses.

Also, depending on the exact voltage provided by the pre-regulator, with no battery or a fully charged battery, these pulses may bypass the pre-regulator and cause noise on the output.

I designed a high-side switch to switch the battery in when the pre-regulator voltage drops significantly.  It prevents the noise from the charging circuit leaking into the reference power supply.

[Pipelie]

I replaced with 4S lithium battery pack, it works well.

Did you use a battery management system board? Did you modify charge current?

Hi,
I don't use battery management system board, only a simply/General Protection board.
the batteries i use are 4 lithium(4500 mAh) battery in series.
there is no need to modify in the charge current. the charge current is around 10mA~20mA, it will take about 2 weeks for full charge.
The final voltage of each battery should be less than 4V for safety reason, since the battery working in floating charge mode. So, the Vz of zener diode  should be tested and selected.

here is the schematic:
D0,C0 and the fuse are part of the battery pack.

EDIT: the fuse is 1/16A, not 0.5A, connect to the additional connector for fast charge through a Li-ion Battery Charger.

[Pipelie]

any advice for improvement is welcome. 

P.S  you can hear  the "acoustic noise"  come from inside the unit? if you don't install the battery. and the frequency could be some point around 100 Hz to 10 kHz?it‘s relevant to the value of C5 .
because without the battery, the power supply regulator oscillate. you can pull out the battery and power on your unit, connect your oscilloscope probe to the test point C, you will see what I'm talking about.


http://www.murata.com/en-global/products/capacitor/mlcc/solution/naki?intcid5=com_xxx_xxx_cmn_hd_xxx

[Pipelie]

another modification
 

[Micke]

Do you think there would be any risks/disadvantages of increasing the size of capacitor C5 from the original 150µF?
I have my first 731B unit on its way  , and planned to do re-capping anyway of C5, so why not increase size to like 470µF or even more to (hopefully) increase stability on the output?
As I can see it, the battery charging circuit will not be affected.
Any thoughts?

[Pipelie]

Hi, Mike

I'm already replaced C5 to a 1000uF Axial Electrolytic Capacitor which I happen to have it in my table.
there is a resistor R33 in series, I don't think there is  risks/disadvantages of increasing the size of capacitor C5.
but this modification will not  increase stability on the output.

[Micke]

Hi Pipelie!

Hmm, yes you are right, looking at the schematic the diode CR8 anyway will happily feed possible ripple regardless of capacitor C5...
So it must mean that if you do not have a battery connected, you should remove CR8, preventing mains ripple to be seen on the output.
I wonder why they have not put an extra capacitor after R1?