Chris,

A serious question !  Out of curiosity, why are you changing all these diodes as
it is unlikely that they would all fail at the same time.  Do you know for
certain that they and the other parts are bad ?

The only output on the 731B that has current capability is the 10 volt output.
All the other voltages are delveloped via resistive dividers and will not handle
any loading.  I.E., they were made to drive a very high impedance, like a Fluke
845 null meter.  Generally, 1.xxx voltage outputs are useless these days unless
you have some "Standard Cells" laying around.

The TD8253 diodes can be replaced with 1n4148 or 1n4448 (preferred) diodes.

For the 1N967B, it is available and here is a web address for that:

http://store.americanmicrosemiconductor.com/1n967b.html?gclid=CMTmqY2w4bMCFYF7Qgod_RIAQg

As for the FET current limiter, that is going to be a little more tricky.  No
idea about a replacement at this point.  You could just use a resistor of the
right value to limit the current through the zener.  That particular stage is
just a "pre" regulator to provide roughly 17 volts to the real action further
down stream.  Yes, having a more stable 17 volts is better.  The purpose of CR10
is to force a fixed amount of current through CR11 irrespective of input voltage
variations; within reason of course.  A number of parameters would have to known
to find a suitable replacement.  But, again, this is not a real critical circuit.

In the case of the Bourns' pot, any modern replacement part that may be fitted is
going to run more than $30.  What makes you think the pot is bad ? ? ?

Good luck with the project,

Bill....WB6BNQ

Christopher Brown wrote:

On 11/21/12 5:04 PM, WB6BNQ wrote:

The delta pot is flat out broken, the tail end is broken off.  Turns
counter looks OK though.

The power switch is broken off.

The power and adjust board has been /worked on/.  Resistors and diodes
replaced, etc.  Based on the workmanship, I trust this about as far as I
can throw my truck.

The hard fail is the UV pot, could not find a xref and didn't know about
the original unit to guess about a modern equiv.

I am aware of the output capability, this is my second 731B.

I have null meters and a pair of 752 dividers that need a standard cell
or suitable substitute to calibrate the input.

Hi Chris,

Do you mean the Fluke 750A ?  It has connections for standard cells.  The 752 is a ratio
device designed to be used with a 10 volt reference standard.  No need for the 1.xx outputs
from the 731B or the use of standard cells..

The 750A standard cell arrangement can be bypassed and set up for use with just a 10 volt
reference using the original banana jacks for the reference and null meter as normal.  All
that is needed is to (non destructively) remove one connection and add one wire.

Bill....WB6BNQ

Christopher Brown wrote:

On 11/21/12 6:44 PM, WB6BNQ wrote:

Your are correct, not the 752 (would like a 752 and a 720) but do have 2
750A.

That sounds like a very worthwhile mod.  How is it done and how does it
change operation?

As far as the 731B(s), I have 2.

One is in good working order and overall great shape.  Came out of a
standards lab in the czech republic that recently switched to a 732A.

Have a replacement NiCd pack currently cycling.  Other than cleaning,
and replacing the dead pack the only changes planned are replacing the
IEC socket with a small filtered socket (have a box of them) and
replacing R30 (if needed) to adjust the float current for the new pack
(old is 450mah, new is 2100mah.

For the second (received this morning), intend to do the same, and
repair.  I may just hard-wire around the failed switch, but beyond that,
not fixing it would just bug me.

The ref board is in good shape and does not appear to have been messed
with.  The main issues being the pot and the P/S section.

Near as I can tell the Fairchild 1N5248B is a good match for CR1, and
for 2 cents each I can just grab a batch and select.

I have a bunch of 4148 and 1n914 fairchild parts around here somewhere,
curious as to suggestion of 4448 over 4148?  Looks like 2pf v.s. 4pf and
the 4148 may start conducting a few hundred mv earlier from the datasheet.

Thanks
Chris - WL7CLA

Hi Chris,

If you study the schematic you will see that S3, the front panel lock-open-momentary switch, is
in series between the standard cell (J10) and the null meter (J8).  So what is happening is the
voltage from the 750A goes to one side of the null meter and the reference (Standard Cell) goes
to the other side of the null meter.

Remove the wire from S1201 going to J7 and cover it with heat shrink.  Then connect a wire
between J7 and the 10 volt TAP (top of R12).  From this point on, the "Standard Cell" banana
jacks is where you connect the 731B 10 volt source, the null meter goes to its jacks and the S3
switch is still normal.  The kelvin decades for the 1.xx voltage are no longer in the circuit and
have no effect.

Should you ever want to sell the 750A (you shouldn't) it is easy to reverse back to the factory
arrangement.  If you don't have a manual for the 750A, let me know and I will forward it to you.

Bill....WB6BNQ

Christopher Brown wrote:

I have a 750 too but it needs some rare 6 V batterys to make -6-0-+6V. The adaption to 10V sounds interesting but i do have standard cells and like to use it for that. I do not use the 750 at the moment because without the batterys it won't work. You have an alernative solution for that too ?
I measure the cells with a >10 GOhm 7,5 digit meter, write down the reading, then zero the meter in the 1V range with that, set my 720 to that value and connect the meter, then connect my 332 and set the 10V or 100V so the meter reads zero again. And finaly undo the zero and measure the output of the 720 again to heck if the reading is the same as I wrote down.

Fred PA4TIM

Op 22 nov. 2012 om 07:43 heeft WB6BNQ wb6bnq@cox.net het volgende geschreven:

Current regulator diodes are typically JFETs with gate-source tied so
that their Idss is more or less constant over a wide voltage range.
Anything in the 1 mA to 50 mA range can be replicated by selecting
common JFETs for the right Idss, up to 40 V or so, or by selecting
one with slightly higher current, then adding series R in the source
to degenerate it to the right amount. They are not very precise, but
much better than resistive biasing over wide voltage ranges. If
higher current is needed, they can be paralleled. If less than 5 V
overhead, or more than 40 V compliance are needed, it's better to go
with an active circuit with appropriate devices.

Ed

At 02:48 PM 11/21/2012, you wrote:

Hi Fred,

Well, it seems to me that the only time the relay is an issue is if the coarse adjust pot is not fully clockwise.  If you look at the schematic, the relay is bypassed with R101 (coarse) is in the full CW position (a detented switch position) actuating S101.  Near as I can tell, the coarse CW position completely bypasses the protection circuit.  So, you should not
need batteries.

The coarse/fine controls are used when you have a fixed voltage source.  As you indicate you have 332 then you can adjust it and do not need the coarse/fine function.  Of course you would be responsible for making sure of INPUT / OUTPUT switches are set to the right scales for the voltages being used.

But let me make the standard disclaimer - - -  It has been a long time since I used my 750A - - - so I could be wrong !

Please let me know if I am all wet.

Bill....WB6BNQ

Fred Schneider wrote:

This is correct, the switch portion of of the course pot bypasses the
protection circuit.

I have one 750 with good batteries, and one without.  The one without is
currently supplying ~ 100mv to a 3456 running on the bench.  Don't have
a 332 or 335 at the moment, so am feeding 10V out of 731 into the 750
input.  One needs to set the input voltage to 1kv to prevent the divider
from loading the 731 down, but was the quickest way I had to provide a
stable ~ 100mv

I was thinking of Alkaline AAs, maybe a pair of 4 cell holders.

On 11/21/12 11:03 PM, WB6BNQ wrote: