Hioki HiTESTER 3238 bench DMM, no measurement reading, been dreading this moment

[RichardM]

Hi All,

My HiTESTER 3238 has stopped taking measurements. It boots and runs the self checks ok it seems, just like normal. It responds to keyboard presses and commands via RS232 witht the exception of measurement commands (it locks/hangs when requesting a measurement). The digit display has no leds lit or any readings when turned on. Normally, after the self-test boot, it shows random readings with relays clicking with the test leads open. I hear none of that relay clicking now. Also, in diode mode it does not output a voltage to the test leads.

It seems like Hioki do not publish their service manuals so that is a problem for a hobbysist. Any suggestions on how to start would be welcomed. I have probed quite a few logic ics's and 5V is there as is the 7V refererence.

Richard

[coromonadalix]

some small infos on a thread i started

absolutely no service manual can be provided by hioki,  they dont want ...   it is not a normal meter,  with say lm399 or else

its more like an advanced ADC with an mcu  etc ...

you have to snif out  how its made, check the supply lines   ... you have at least 4 supply lines,  on symetrical and the others are single supply lines

show us the inside like i did

[RichardM]

I will post some photos. Looks very similar to the inside of yours. I guess I will begin by tracing the inputs and try to see.if the relays are working or powered first. Soooo many small 3pin SOT packages to try and ID..... The reference is an LT1021.

It has been a great meter.

[coromonadalix]

why are you thinking this way .... thinking inputs parts

the 1st thing is to check supply lines, caps etc ....  theses are old units ... some are old as 20 years in some series ...

cold be an xformer coil who busted, some regulator who died  etc ...

in my meter 
you have +8v and -8v symetrical supply (ic71 and ic72) and +5vdc (ic74) coming from the xformer secondaries / blue wires
around 7vdc for the 2 small xformer wires who goes on a rectifier bridge, and regulated (ic75)   at 5vdc  for the cpu section

blue wire pin 1 and 2 = 4.8 ohms
blue wires pin 3 4 5 =  3-5 = 14 ohms,    3-4 or 4-5 is 7 ohms
two small yellow "ish" wire = 0.9 ohms


some front pcb section ground is not common with othe meter ground sections, that's normal,  you have to know where your probing is grounded

[RichardM]

Thanks

Per my first post I have checked the 5v supply line and 7v ref, bla bla 😉

I will check the transformerand caps like you suggest. The unit is basically functional, it just does not take ant readings or display them on the front panel.

[RichardM]

I have attached some pictures of my unit.

The relays at the input are TX2-L2-4.5V-1 that have a direction indicator ? Is that a small LED that should illuminate ? It does not currently and I have no idea whether id did previously as I have never had the unit open.

I have checked the electrolytic caps as best I can in circuit and they appear to be ok. 5V rail ok as checked on most logic IC's, OP27 has 15V. There are many SOT-23 ? packages around the input relays labelled A1, LY, JY etc.

As a hobbyist, really hard to troubleshoot with no manual or schematic.....

Regards

[coromonadalix]

ok   more parts installed than mine for sure  loll

Main mcu is the one on the rear with a white sticker on it,  i have identified many parts on my 3237 thread

its a basic meter in a sense, many switchings  ....

ill try to get you some switching logic table later


when the dmm start, does it display its hw version for a brief moment ?  and pls  update your avatar   location,  your meter is set on 220vac settings  right ?  to be sure

can you gently remove the shield pcb   and show us   under it  ...

[RichardM]

Great thanks

The meter starts fine and reports HW 2.08. Controllable over serial.

Avatar changed, I am in Australia.

The last time the unit was working I was feeding in high AC voltages which is why I was looking at the inputs.

Do your relays illuminate?

Regards

[RichardM]

Image attached under the shield

[coromonadalix]

relays illuminate   no  nothing 

im not sure to understand ?


im gonna try to make the 5x optocouplers and the 2 relays chart

ok  i think this puppy use the ic14 ic15 ic16   as the main control from the mcu ? 
they are 8 bit  74hc595  shifft registers, it will be a mess to decipher,  tons of probing, you have to do all modes, plus all ranges ... means 24 outputs / control lines with theses

ic12 ic40 ??

the meter case is the mcu (logic) ground, not related to the front input circuits
the ground in the input circuits is the one who go on the shield pcb, snapped on top of the main pcb


the 5x big optocouplers seems to go on the rear ext i/o connector

[RichardM]

My relays have a tiny led indicator and I wasn’t sure whether this was supposed to illuminate. You can see the tiny led in the photo above.

[coromonadalix]

sorry  cant see what you mean

and sorry   wont help anymore than this,  i'll close up the 3237 i have to sell it in the future, don't want to keep it, since mine it the basic model

[factory]

The last time the unit was working I was feeding in high AC voltages which is why I was looking at the inputs.

I've heard of DMM inputs being damaged, while measuring higher voltage transformer outputs, when the transformer is switched off giving a spike of higher voltage.

David

[RichardM]

Thanks David,

I suspect that is exactly what has caused my failure. I was measuring output from an old transformer taken from a console tube radio. Looks like I am on my own . Still curious about the led on these relays, I may source a couple to test.

[coromonadalix]

you will have to check all the inputs protection devices, movs gaz tube discharge if any, voltages divider

it could have created a spark, and could have gone far in the circuit ....

[Kleinstein]

The relays at the input are TX2-L2-4.5V-1 that have a direction indicator ? ....

There are many SOT-23 ? packages around the input relays labelled A1, LY, JY etc.

A LED indicator would be an exception / rare. Chances are this is just plastic mold, maybe a kind of plastic "rivit".
The relay seems to be a latching one and as such may need several transistors for driving the coils. The SOT23 parts can be many different parts: BJTs, JFETs, MOSFETs, diodes and zener and even some resistor network come in SOT23.  Chances are the ones near the realys are BJTs to drive the relay coils and maybe diodes for clamping / free wheeling.


Turning of a higher voltage transformer can cause a high voltage spike and this could well be the cause of damage. Knowing what likely caused the damage helps to look at the right area (e.g. input protection and initial switching). It is still odd to get it hanging and no dispay. This would be more than just an input damage - more like the controller on the input side not running.

[coromonadalix]

Display Drivers : 2x  TD62083F  8CH DARLINGTON SINK DRIVER

means  it could be directly driven by the mcu  or pass thru some shift registers ....

the problem is to get an general idea of the front end section, supply lines etc ...     this meter as tons of sot23

you have +8 -8vdc lines and 5vdc supply 

[Kleinstein]

The pictures are not very good. So one can not really see much of the chips. Especially the big one in the center (some SO16? wide or so) could be relevant to see if this is a µC or ADC chip.
A point to check would be if there is a µC on the input side at all, or all the control is via the OK and shift registers.
There may also be parts on the other side, as there are quite some active parts, but few resistors / capacitors.


A lot of the SOT23 ones are JY , which should be a SK208 JFET, likely used for analog switching.  These would usually need some LM339 / LM393 or NPN transitors for control.

[coromonadalix]

https://www.eevblog.com/forum/testgear/hioki-3237-5-5-digits-meter-teardown/

i have id'ed some parts

[Kleinstein]

The pictures in the other thread are much better. There is a ADS1250 ADC chip and as far as I can see no µC at the input side part. This could also explain the rather large number of opto-couplers.
A point worth looking at would be the interface of the ADC1250 - this would tell if the ADC is active at all.

Quite some of the JY Jfets are used as diodes.

[coromonadalix]

this is  a simple adc on steroids with an dmm input frontend,  nothing special in a sense, good reference an goo adc  and voila, some dmm can be made

for sure there is a tons of parts, could be mosfets fets  ... as voltage switchers,  divider  etc ...

yes you could sniff out the adc activities, since all theses parts datasheet are available,   

sadly only the mcu could be problematic, i did search a programmer for theses, and good luck to find one, it will cost $$$ and this was decades ago

[RichardM]

Yes, I am certainly no photographer. To be honest, some of the marking on the sot23 packages are almost impossible to read regardless. You need the exact angle of lighting to read the markings, or a much better camera than my phone has.

The sot23 nearest to the inputs and connected to the 2 NAIS black relays are "LY" ,"EY", marked as Q1xx (FET's or BJT?), "A1" and "30Z" marked as Dx (diodes?). There are also sot23's on the underside.

All I have done at this stage is remove the big green varistor and it measures open or very high ? as does AG1 which is a "spark gap" or "gas tube" that also measures open/very high. I put 4.5V into the 2 NAIS relays and they do still appear to latch.

Just to recap, the meter is all but functional (ie boots, responnds to rs232 commands) except it does not display any readings on the large LED segment display and I do not hear the relays switching back and forth like normal ?. The front panel buttons work and I can switch between DCV , ACV, OHMS etc and the relevant units are illuminated etc.

Regards,

[coromonadalix]

could it be the flex cable is not well inserted,    had fun with it when i dismantled my meter, bad contacts

prety dumb display  with 2 driver ic's     pretty basic  see my 3237 thread   there is not much

[RichardM]

I do not think it is the flex cable but will check again. I removed the flex and cleaned it and the contacts with contact cleaner. I guess I could also look to see if the flex signals change when I apply a voltage to the front terminals ?