EEVblog Electronics Community Forum

Products => Test Equipment => Topic started by: ketil b on October 02, 2015, 11:51:12 pm

Title: Hioki 3540 milliohmmeter teardown
Post by: ketil b on October 02, 2015, 11:51:12 pm

Hi

Got this meter of the old eBay for not much, the seller said it not working and would not power up.

I got it open and there was a lose pin-header, bit of solder and all is good

The specks are nothing to amazing but still a useful meter

The interesting thing about it is it has two ADCs a MAX135 multi slope, that is measuring the sense inputs and an AD7715 Sigma-Delta, that look to be measuring the drop on the ref resistors. that is for the temperature sensor input.

Main chips

REF193 (http://www.analog.com/en/products/linear-products/voltage-references/ref193.html) 3.0V Voltage Reference
MAX135 (https://www.maximintegrated.com/en/products/analog/data-converters/analog-to-digital-converters/MAX135.html) 15-Bit Multi-slope ADC
AD7715 (http://www.analog.com/en/products/analog-to-digital-converters/ad-converters/ad7715.html) 16-Bit Sigma-Delta ADC
PGA204 (http://www.ti.com/product/pga204) Programmable-gain instrumentation amplifier from TI but this one is marked as Burr-Brown

Thanks

ketil

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: ketil b on October 02, 2015, 11:56:02 pm

And the rest of the pics

Thanks

ketil

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: retiredcaps on October 03, 2015, 12:14:09 am

Thanks for sharing those pics. 

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: GlowingGhoul on October 03, 2015, 01:07:05 am

Interesting. Thanks for the teardown.

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: JoeB83 on October 03, 2015, 01:52:46 am

Neat, thanks for the tear down.

What are the gray boxes with banana jacks in your first picture? Reference resistors?

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: ketil b on October 03, 2015, 05:18:09 am

Thanks all

I think I may reverse engineer the analogue front end, it look like a relatively simple circuit and it would be interesting to see it.

Quote from: JoeB83 on October 03, 2015, 01:52:46 am

What are the gray boxes with banana jacks in your first picture? Reference resistors?

They are just that, along the same lines as Dave's 1K and 10k resistors in a box, but I got a bit carried away and they go from 0.1 ohm to 1M ohm in decade steps.

thanks

ketil

edit:spelling

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: ketil b on October 07, 2015, 10:22:45 am

Hi

So I reverse engineered the analogy side of this meter and I have to say that it is a very well engineered bit of kit.

The layout of the bord is almost a work of art.

I have added a pic that shows the star grounding for the voltage reference.

thanks

ketil

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: ketil b on October 08, 2015, 03:25:50 am

Added all the files to Github (https://github.com/ketil/Hioki3540)

thanks

ketil

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: pigrew on March 05, 2017, 08:14:21 pm

It looks like the GitHub address has changed (the old link is bad).

Currently, https://github.com/lukier/Hioki3540 (https://github.com/lukier/Hioki3540) is a working link.

I purchased a pair of these on eBay. Both look like they had severe overvoltages applied to the DC input. Further diagnosis is needed. Also, their PCB layouts are different from eachother (one is about five years newer).

As for the damage, both have burnt out input inductors (there is no fuse, or maybe the inductor is the fuse?), series resistor R133 had smoke around it, and the two series Schottky diodes are dead (they became shorts). Also, the ground via closest to L109 is gone. I can't think of a good reason that this particular via would be different than the rest.  I'll probably just put random-value replacement components for L109, R133, D105, and D106, and see what happens. I can't imagine that their values matter so much.

EDIT: Seems that the buck converter isn't regulating. Output voltage rises above 5V as the input voltage is raised. Will try replacing converter. Charge pump seems to be working properly.

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: 1audio on March 06, 2017, 06:43:57 am

The description suggests someone tried to measure the impedance of the AC line with pretty spectacular results. Probably looking for new employment now. Good luck with them.

Title: Re: Hioki 3540 milliohmmeter teardown
Post by: pigrew on March 06, 2017, 06:16:47 pm

In order to keep this thread as a teardown thread (and not a repair thread), here's a photo of the newer revision mainboard. I noticed the following changes:

• Display now uses only a flat flex cable, the two pairs of ground wires were removed. The ferrite bead around the flex cable is also not there.
• Extra resistors close to IC103 (93LC66) EEPROM.
• New testpoints (TP101, TP102), near the beeper, perhaps for debugging display voltages (complete guess)?
• C127 (close to IC102) is now physically much larger.

There's a lot of flux residue on the bottom of the board, but I guess that's OK since it's measuring low resistances.

My other unit (with the same sort of damage) is a working board (measured resistances are accurate) if the input DC voltage is adjusted to
 about 6.5V (so that the 5V rail is 5V). I have not yet tried this board. I will try replacing the buck converter IC and the fried diodes and inductor (with a 3.3 uH ferrite, since that's what I have).

EDIT: The extra cables to the display are grounds, not power lines.