Electronics > Projects, Designs, and Technical Stuff
Calculating values in voltage reference higher current series pass transistor
(1/2) > >>
dlucy67:
I'm replacing a precision voltage reference circuit in an old school differential voltmeter, I've chosen a very precise voltage reference device, but I need to increase the available output current.

The Fluke 893A that I'm restoring uses a precision 11 and 1.1 VDC voltage reference at its heart. From there it can measure DC or AC from 1mV to 1100V. Pretty cool, but the voltage reference circuit is hard to repair. So, I'm building a replacement reference with modern devices.

Analog's LT1236A will give me very, very precise (for my needs) reference and they are stackable in shunt configuration, so I can put a 5.000 VDC device on top of a 10.000 VDC to yeild 15.000 VDC... and send that through a precision (0.05% each resistor) voltage divider network... and pull out the 11.00 VDC and 1.100 VDC that I need. (see the attached 2xLT1236A PNG)

However, the LT1236 wasn't designed to drive the rest of a unit like the 893A as it only handles 10 mA. So, I need, I believe, a current boosting design like those in the common linear voltage regulator device data sheets to provide whatever current the rest of the diff voltmeter needs above 10 mA. Right?

I should be able to handle at least 1 A, right, if I add a series pass transistor to my above 2x LT1236 design, just like the LM317 or MC78xx regulators do, right? (see the two attached GIFs for the LM317 and MC78xx)

If so, which of the two attached designs is better? The LM317 example with a single power transistor as a switch and three paralleled LM195s seems like a waste. The example with a single MJ2955 seems like a more-modern, fewer-component design. Is one significantly better than the other? Will either work for my current boosting with LT1236 shunts needs?

If so, how do I calculate the sense resistor (between the power source and the LT1236) and the base resistor for the series pass transistor? I'm planning to use a readily-available higher-current transistor like the MJE15032 or MJE15033 with a Vbe of 1.0 VDC.

LT1236 datasheet https://www.analog.com/media/en/technical-documentation/data-sheets/lt1236.pdf
RandallMcRee:
Neither.
Follow the recommendations in the datasheet page 8, 9. No sense reinventing the wheel.

In a similar design I am using an OPA189 and the LT1010. Works well, so far, and has some measure of output protection.

To get the precision possible from an LT1236 you need to use some sort of force/sense discipline, because at >10mA you will have microvolt accuracy only at one point. In a differential voltmeter this presumably wants to be the top of the first voltage divider.

Good luck,
Randall

Edited to add schematic of my buffer. The OPA189 preserves low noise and precision of input. LT1010 has 150mA. I goofed on the kicad library so the LT1010 pin numbers are incorrect.
dlucy67:
Doh! Thanks for reeling me back in!

You’re using the OPA device as a buffer after the LT output?

Help me understand the second part of your reply:

Force sensor?

Top of the first voltage divider? Ok, I can easily duplicate the design so I have a second circuit that generates just the 1.1 reference (I.e. one circuit with LT1236A-10 + LT1236A-5 into divider 1 with out out of 11... and then a second circuit with just one LT1236-5 and a second divider to generate the 1.1). Would this overcome the point you are making?
RandallMcRee:
I'm building a precision differential voltmeter from scratch. If you precisely generate 10 volts +- 1ppm you will find that a simple strip of wire or a ground loop will degrade that accuracy.  But you typically only need that precision at a specific point and this is easily accomplished using the opamp correctly. The idea is illustrated in the electrically identical schematic below.

Also see Jim Williams AN-86. He has some comments on this whole problem. If you start googling you will find lots and lots of interesting material.

Hope this is helpful,
Randall

Edit: ok, williams DN-11 is the one I was *really* thinking of. Warning: reading this may cause volt-nuttery. Read at your own risk. https://xdevs.com/doc/Fluke/720A/doc/dsol11.pdf
Tj138waterboy:
Also have a gutted partially disassembeled 931b meter if you have schematics could you please share as well. Every wire in my unit has been desoldered from pcb and I havent been able to find any schematics.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod