Biddle 72-432 Wheatstone Bridge

[Wallace Gasiewicz]

I bought a nice looking and apparently a well working Biddle Wheatstone Bridge for a few bucks at a Hamfest.
I bought it for the primary function of using it as a resistance standard. It has seperate connections for use as a resistance box and for Wheatstone Bridge.
I know it is not a "standard" but just wondered if any one has experience with this unit as far as accuracy
It compares well to my calibrated HP 3456

[Vgkid]

I have no experience with this unit , but the specs list it as being 0.05% accurate as a decade resistor box. So the resistors should at least be that accurate, if not better. So the switch contact resistance(switch zero) will be an issue.
Have fun.

[Wallace Gasiewicz]

Biddle does say that the switch resistance is meaningful at low resistances and actually give a value. I do not remember it but as I recall they were pretty close.
I think that the unit exceeds the .05 accuracy. I know most folks here use more accurate stuff, but this is pretty good for me.
I am quite happy with my $25 purchase.

[mendip_discovery]

Remember to exercise it regularly and the zero will have a nominal amount but with regular exercise and measuring it will be fairly predictable. The bridge I have in the lab has been fairly good over the years.

[trobbins]

I've found it a nice pass-time to compare all the step elements in such vintage bridges/decade boxes, and identify if any individual steps have drifted or been stressed compared to others in each decade.  It's likely that the relative tolerance of steps can be tweaked to well below 0.01% for the upper ranges with your HP 3456, and you may get consistent switch resistance characteristic to give some confidence.

[Wallace Gasiewicz]

trobbins:

I did what you have suggested and the Biddle seems to be quite accurate in all ranges.
Maybe to .01% Better than .05 % Pretty good for my stuff.
Also all of the switches seem to be very clean. And it seemed to get more consistent with more use of the switches, as you pointed out.
Nice big wire wound resistors that have some temp consistency, probably because of their size mostly.

It appears I have something that is quite usable for my purposes, and it is in a nice box with nice connectors!!

[trobbins]

I am in same position.  A 1930's Gambrell 4-decade box where all settings are <0.05% (relative) above 4 ohm (9,999 ohm total), and a similar performing mid 1950's PYE Wheatstone with 4-decade arm.  Your coils are likely manganin so tempco likely to be pretty good, and may be non-inductively wound (the Gambrell coils were, and I tweaked the bottom 3-decades to minimal phase shift across the audio band).  I have found it onerous to confirm low tempco's as I would need to thermally stabilise my DMM (Keithley 197) and then independently do large temperature steps of the DUT - but WIP.

[trobbins]

I just modified my 4-decade Gambrell box to also function as a Hamon 10:1 and 100:1 divider.  The box is quite old and I've had to repair it, so no issue with not retaining its original patina !

The 100 and 1k decade steps had sequential groups of 3 that were less than 0.01% relative tolerance except for one group of 1k steps that were 0.05% low, so I thought that was a good start given the 1 and 10 decades would make 99 of the 100 ohm bottom leg of the Hamon dividers.  So I added three 4mm banana terminals to the front face, with one terminal connected to the link from 100 to 1k decades, and one terminal connected to the link between 10 and 100 decades, and I wired in a tweaked 1 ohm from the new Hamon 0V terminal to the existing LO terminal.  The 100 decade steps were shorted as per Conrad Hoffman's article, and a 6 digit meter used to tweak the 1ohm bottom link to give equal voltage division.  The 1k decade was then grouped and shorted and only needed 0.6 ohm added between the end of the 1k decade and the existing HI terminal to give equal voltage division with the lower three decades + 1 ohm. 

I can re-check the equal voltage division configurations if needed, but haven't tried to make that a simple process as there are other low confidence aspects of any measurement that are likely going to dominate (such as tempco of the manganin coils, and available 4mm terminals and banana plug measurement cables, and uncal'd meters).

If I get keen I will also try and similarly modify my PYE 7440 wheatstone bridge (https://www.dalmura.com.au/static/PYE%20Portable%20Wheatstone%20Bridge%207440.pdf) as that also includes  a 4 decade multiply/divide set of dividers that I may be able to cross-check and tweak using the Hamon accuracy performance.

[Stray Electron]

I bought a nice looking and apparently a well working Biddle Wheatstone Bridge for a few bucks at a Hamfest.
I bought it for the primary function of using it as a resistance standard. It has seperate connections for use as a resistance box and for Wheatstone Bridge.
I know it is not a "standard" but just wondered if any one has experience with this unit as far as accuracy
It compares well to my calibrated HP 3456

  Pictures, man, pictures!

[trobbins]

My PYE 7440 bridge is quite basic in format, which was a benefit in disguise as I recently tweaked its 4-decade arm resistors, as well as its multiplier/divider resistors, for better performance.  I wanted to go through the tweaking process using the Hamon technique as a poor-man's way of effectively gaining at least an order of magnitude improvement in accuracy.   I'm not sure if my process would be applicable to the Biddle bridge even though the 4 decades and multiplier/divider ratio capabilities are effectively the same, as the service schematic shows a more complex switching arrangement of resistor elements.

I started my tweaking by setting up a Hamon 9:1 resistance ratio using three parallel groups of 3 steps in series from the Hundreds decade (using Conrad Hoffman's Hamon divider guideline) to compare to the lower decades, and then to do similar Hamon comparison with the highest Thousands decade.  This was relatively easy as the individual steps in the decades were still pretty closely matched (0.02% original spec), and the decade switches are robust with consistent contact resistance and allow easy access to step terminals (to connect in Hamon shorting links and to tweak any particular step).  1:1 voltage comparison's used a Picotest M3510A which has sufficient resolution (0.1uV) and quite a low bias current to be confident with resistance comparison matching. 

I then 1:1 comparison matched the '10th' step in the 2nd top decade so that I could configure 10:1 and 100:1 resistance dividers to then use as one side of a basic bridge ( simplified schematic below) where the multiplier/divider resistors are on the other side (these are individual 1, 10, 100 and 1000 ohm resistors, rather than a series connection as shown in schematic).  I could then tweak the multiplier/divider resistances based on a schedule of null voltage comparisons to 10:1 and 100:1 decade dividers, and then use 1:1 voltage comparisons to tweak between the multiplier and divider resistors.  I still have the 1 ohm multiplier/divider resistors to assess, so nearly finished. 

This should allow me to more confidently assess the status of the Sullivan Volt Ratio Box ( https://www.eevblog.com/forum/metrology/hw-sullivan-volt-ratio-box-from-1964/ ) using a bridge null configuration with the PYE bridge decades set up for 10:1 and 100:1 voltage ratios. 

PYE 7440 details document, including operating manual and the comparison processes used.  https://dalmura.com.au/static/PYE%20Portable%20Wheatstone%20Bridge%207440.pdf