Alternative detectors for vintage dc and ac bridges

[trobbins]

I'm sure there must be a few vintage bridges with detectors or galvanometers that haven't survived over the decades, or there is a need to enhance their performance.  So this thread has an aim of referencing any efforts to fix or replace or improve the situation.

My first foray into this topic was preparing a preamp for a galvanometer in a PYE 7440 bridge from the mid 1950's (https://www.eevblog.com/forum/metrology/null-detector-based-on-conrad-hoffman-design-check/msg5480467/#msg5480467. 

I recently made another such standalone preamp to use with an Aneng AN8009 meter that improved its raw sensitivity from 5uVdc to at least 1uVdc.  That was used with an ESI combo of 230B guarded wheatstone bridge and an 803 generator/detector which had a poorly performing ss detector from the early 1970's.  The detector preamp is powered from a USB battery bank to allow non-mains powering.

It may just be ESI, but a mid 1960's 300 PVB (potentiometric voltmeter-bridge) I checked out earlier this year also had a poor detector, and ESI don't make it easy to repair their detector circuitry (not a simple assortment of bjt's and an unidentified chopper relay, and also sealed in a metal can).  So whilst I've found the internal bridge and other parts etc of these vintage instruments are still typically in excellent condition/accuracy, the detector portion may make them less appealing to use.

I have an early 1960's  ESI 250 DA RLC bridge on the bench at the moment, and its magic eye detector is not working.  As an easy (for me) workaround, I have been using an audio soundcard and spectrum analyser software (REW) to act as an external detector.  That has worked very well, as a modern audio soundcard has quite a low noise floor, and the benefit of fft spectrum plotting to null the specific generator frequency from the ESI 250DA makes it easy to resolve a null condition to the level of resolution that this bridge can achieve (0.1 to 0.3% accuracy).

Looking at the later ESI 250 DE bridge, I note that it uses a generator isolation transformer to allow an external generator to easy excite the bridge without the grounding issues that the DA model has.  The benefit of such an isolating transformer would allow a modern soundcard to also drive the bridge at any specific test frequency, and to link the generator and detector for better fft capabilities.

I haven't noticed anyone using a soundcard/software for such use, and there may be other interesting ways that some have kept using vintage lab grade equipment, so would be pleased to hear of any.

Ciao, Tim

[Conrad Hoffman]

I keep using the vintage oscillators and detectors with my bridges due to the HV capability that some had. FWIW, I look at my L&N top-hat galvanometers on the shelf and think they're perform way better than originally, using a laser pointer and reflection off a mirror on a far wall. Microvolts with no electronic at all!

[kleiner Rainer]

Hi Tim,

seen this?

https://www.sillanumsoft.org/

The software grew out of a PhD thesis at an Italian university. My plan is to use it with my ESI 250DE as external detector. I still use a Samsung N150 netbook (14 years old!), and due to its excellent soundcard with real stereo input it should work well as detector.
Maybe you could use the 250DA bridge circuitry in combination with this software and enjoy the precision of the ESI bridge with the measurement capability of the software. I will certainly look into this.

Greetings,

Rainer DG1SMD

[trobbins]

Hi Rainer, it has been nice to see the evolution of software for soundcard devices over the years.  I have been using REW for about 8 years now and luckily it has grown in capability and performance for the bench tool that I use it for.  I find its spectrum analyser capabilities easy to use, and it provides enhanced ways to lower the noise floor and process the fundamental.

I think you will find the soundcard generator and input (as detector) an excellent tool with the DE.  I use a similarly old soundcard (EMU0404) which has an impressively low noise floor.  Given you may want the generator to provide a level greater than circa 1-2Vrms straight from the soundcard, it could be worthwhile making a simple 2x9V battery powered linear amp to get to circa 5-10Vrms to provide wider signal to noise for the likely excellent 250DE nulling performance.  I used the 250DA's internal generator although rarely needed to turn that up beyond half way (~5Vrms on DUT) to easily confirm I was nulling to the minutest adjustment of dekastat and D-Q.

It sort of parallel's Conrad's suggestion to use a simple visible laser pointer with mirrored galvanometers, and the laser can double its use for your cat as well.

I was certainly an instant fan of the modern auto-zero opamps as a way to easily provide battery powered pre-amplification to then input to any reasonable DVM, without the need for a more exotic vintage null or nV-uV meter.