Skip to content

DIY Hioki DM7275/76 Temp Probe (Z2001)

Some time ago, I acquired a Hioki DM7275 Precision DC Voltmeter.
I primarily use it for measuring and logging my DIY build LM399, 10V precision Voltage Reference.

Everything is working beautifully, but there was one missing piece. The DM7275 has an option to measure and log temperature along with voltage. To enable temperature measurement, a temperature probe is required (go figure).
Hioki sells this probe under the part number Z2001.

Depending on your location, the Z2001 price can vary from 88€ to 150€ (approximately 95$ to 160$). This is not a bad price for a company, especially when excluding VAT. However, as a private buyer from Europe, the price is around 150€, which is quite a lot for me.

After some research, I have concluded that the Z2001 probe uses a 2.186 kOhm NTC resistor type probe.

The correct part should be: NTCDS3HG222HC4NB, priced at 3.09€ per piece at the moment of writing this article.
The only place I could order this part was Mouser. So I ordered two NTCs (just in case I mess it up) and a 3.5mm mono headphone jack.
In total, including shipping, it cost me 31€ (which includes 20€ for shipping costs plus VAT).
The cable I used was just some random USB cable I had laying around.

ATTENTION!

As the commenter on the post below, Lymex Zhang, mentioned, the Shibaura PB-36 would indeed be a much better choice for this build. Although I won’t alter the information in the already written article, I recommend reading the Appendix at the end of the article for the correct information.

Parts:

Probe assembly:

It turned out that after I assembled everything, the probe was off by almost exactly 0.5°C (0.9F). This discrepancy occurred because I hadn’t paid attention to the cable resistance. The initial crappy USB cable had a resistance of 3.5 Ohms. Realizing this, I found another, better cable lying around: a Coaxial RG174Q with a total resistance of 0.4 Ohms over a length of 2 meters (6.56 feet). After redoing all the work with the improved cable, everything was great.
NOT!

The measuring error was even larger now. I retraced my steps and attached the resistor decade to the Hioki. I simulated a 2.186 kOhm resistance, and the instrument measured exactly 25.0°C.
Since the instrument was properly calibrated, the error was indeed caused by my probe. After some measurement and quick math, I realized that the probe was “missing” 39 Ohms. To rectify this, I added a 39Ohm, 0805 SMD resistor, and the problem was solved.

Finished Probe:

Library:

2.186kOhm NTC: NTCDS3HG222HC4NB, datasheet
3.5mm Mono Phone Plug: KM2PB-AU

It is a bit challenging to compare these two probes because the K-type thermocouple reacts to every disturbance in the air, whereas my probe is slower, which makes it well-suited for this application.

APPENDIX!!!

The chosen TDK sensor was not a perfect choice for this purpose. However it was the only one I could acquire with 2.186kOhm resistance on 25°C.

Out of the two purchased NTCs, one was deviating by 0.5°C, and the other by 0.8°C. If I bought more of them, maybe I would get lucky enough to hit a closer match. However, it’s also possible that the situation could be worse.

Considering the intended purpose (logging the voltage reference), I don’t need to measure extreme temperatures.
A range of 10°C to 35°C is more than sufficient. By adding the 39-ohm resistor in series, I sacrificed some linearity but gained precision within my useful range. The manufacturer itself states a tolerance of ±0.5°C, and I achieved ±0.2°C.

Better NTC option: Shibaura PB-36 in not attainable by us mere mortals so if you need a linear temperature measurement through entire range of instrument, I suggest buying the original Hioki Z-2001 temperature probe.

3 Comments

  1. lymex zhang lymex zhang

    The TDK NTC used is not the best.
    A much better choice is Shibaura PB-36:
    https://www.shibauraelectronics.com/pdf/shibaura_products_catalog.pdf
    For one it is radial, and it is more precise to replace Z2001 than NTCD 3HG222.
    Take 90.0 degree C for example, RM3545 requires 280.3 to 281.9Ω, PB-36 is just fit with 0.7 to 1Ω wire resistance, while NTCD 3HG222 is way off(284.9Ω).
    Zlymex

    • Lajt Lajt

      Dear Lymex Zhang,
      Thank you for this comment and information.
      After doing more extensive measurement, I can confirm that on higher temperatures, my probe was quite off the mark.
      I have updated the article with more precise information.

  2. SS SS

    @lymex Shibaura PB-36
    is unobtanium

Leave a Reply

Your email address will not be published. Required fields are marked *