Author Topic: Burster KELVIMAT 4323 teardown  (Read 1238 times)

0 Members and 1 Guest are viewing this topic.

Offline doktor pyta

  • Frequent Contributor
  • **
  • Posts: 411
  • Country: pl
    • RF Scientific
Burster KELVIMAT 4323 teardown
« on: September 15, 2020, 10:51:40 am »
What does Isotech / ASL F150, Cropico 3000 , Burster 4323 have in common?
They share common design.
I suspect ASL to be former designer because they had experience in making highest class thermometry bridges.
I wrote 'had' because ASL was acquired by Tinsley, Tinsley by Wika.
Also the Isotech/ ASL user's manual is most comprehensive among others.

Item bought as parts unit.
After examination I figured out that there are two SLA batteries -one for analog, one for digital and one lithium for RTC/ NVRAM- and what is important, the instrument won't turn on if the akkus are discharged. Even if it is mains powered.
I applied external power supply (to mimic charged akkus) and saw this:

Below photos of the guts.

PCBs are nice, tidy, no bodge wiring. But these two akkus... well let it be.

ADC section based on HI7190IP.

Custom wire wound reference resistor 400 ohm.

Cropico name on the PCB.

« Last Edit: January 05, 2021, 08:29:32 am by doktor pyta »
The following users thanked this post: TiN, Echo88, Vgkid, lukier, bsw_m, capt bullshot, LaserEng, wolfy007, Kosmic, TUMEMBER

Online Vgkid

  • Super Contributor
  • ***
  • Posts: 2619
  • Country: us
Re: Burster KELVIMAT 4323 teardown
« Reply #1 on: September 16, 2020, 04:40:03 am »
Thanks for the teardown. Interesting that they use a custom made resistor. My azonix uses a vishay foil(non hermetic)
If you own any North Hills Electronics gear, message me. L&N Fan

Online dietert1

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: de
    • CADT Homepage
Re: Burster KELVIMAT 4323 teardown
« Reply #2 on: September 16, 2020, 02:28:02 pm »
How did you find out? We have some Arroyo TecSource controllers that also have a 24 bit ADC (ADS1256) but i could not tell yet which resistor i should upgrade. All SMD 0805 - about 100 of them.

Here is a test log i got recently. I used 2x 10 KOhm thermistors, one for the Arroyo TecPak, one for a HP3456A resistance measurement. The TecPak is setup for 22 °C with Autotune and it is cooling since ambient temperature is above that, maybe 27 °C. Vertical scale is 1 Ohm which corresponds to 2 mK.

The HP 3456A measurement exhibits an equivalent temperature noise of 0.033 mK on a 5 minute time scale. Its resolution is 10 mOhm, equivalent to 0.02 mK, So there is very little extra noise. But when looking at a log of several hours the ambient temperature process adds some mK. Resistance is going up during night when ambient is cooling down. That drift probably stems from the TecSource reference resistor and/or from imperfect thermal coupling between the two thermistors. Maybe i have to find and upgrade the reference resistor.

Regards, Dieter

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo