Time Electronics 5075 7.5 / 8.5 digit multimeter repair

[NyquistNoise]

Hi all,

Thought I'd document a repair I did recently to a bit of an "unknown" brand in the world of metrology. It seems there's little information out there on these (dare I say) quite fantastic multimeters. With their ability to go all the way up to 10kV and 30A with very high resolution they're imo very useful meters. I believe the datasheet was something like 18ppm/yr.

First of all, I'm sure with a quick Google you'll notice this is a 7.5 digit meter, which is true (kinda...). Later I'll describe how you can "unlock" this to an 8.5 digit meter which doesn't seem to be listed in any manuals anywhere. This is a true last digit and not just some approximation using some foolery from what I can tell.

I bought this particular unit off eBay for a couple £100 broken without much description of the exact fault, but for the price I thought I'd give it a go. It of course lacked a recent cal (2012) and the Cal key was also missing (a 2.5mm shorted jack) but this could overlooked. I gave it a quick test and all appeared to be working well comparing it to my MF calibrator so wasn't sure what the problem was (and was worrying I'd be unable to justify tearing it apart!)

Until I checked the capacitance range. Completely dead.

I've attached some images, the back "blocks" are of similar construction to other Time Electronics gear - I'm tempted to x-ray the A/D module as it's potted, but the AC module was not!

While I was in there, I dumped the content of any eeproms and ICs I could find. There is a battery, but from the manual, this appears to be RTC only and not backing up any vital cal data and the like. The rest of the digital board is so so, bunch of logic, power supply bits and GPIB circuitry. There was a curious looking burnt resistor, but this was for the frequency and unrelated. I changed it for a 0.5W because it gave me the chills looking all brown even if it was still in tolerance.

On the analog side we have two blocks on separate PCBs, an amplifier on its own board, some optos, attenuator sections and capacitance/frequency/diode sections.
As I'm sure many of you will notice, there are VPG resistors everywhere. Underneath the board is a large shunt resistor (not pictured) for the high current shunt.
Skipping over the boring diagnostic processes, the capacitance section is at the top left of the board. It appears that REG3 was dead and so was TR30. I suspect one took the other out, but as to who was the culprit, we will never know...
Changing these got the capacitance range up and running again, so now it'll be left to settle for a while, then will receive a full cal!

On to the fun unlock, it appears that with the Cal key in, I was able to hold MENU while the meter booted up. This enabled a bunch of options in the menu. The one to look for here is "MAX RESOLUTION". This enabled "8 DIGIT RESOLUTION" under "N DIGITS" button. Make sure you change this and save your changes before exiting!
I believe this is a "real" and not averaged digit, as you can change the LSD (8th digit) on all the ranges completely, and the sample time does not change. This is no solartron sample time either, this meter does many samples a second!
If anyone is interested in manuals, or has anything else to add, feel free to PM me! I may add to this post if I think of anything else and will probably x-ray the potted section.

[alm]

Thanks for posting the tear down and the repair report! The potted modules and lack of easily available documentation has made me hesitant to buy their equipment.

Since you mention the meter apparently showing up to 8.5 digits, do you know anything about the linearity or transfer accuracy, either from specifications or testing it yourself if you have the time and equipment?

Could you upload the firmware images and manuals you have to one or more of these places if you haven't already?
http://www.ko4bb.com/getsimple/index.php?id=upload-instructions
https://xdevs.com/contact/
http://bama.edebris.com/

[Kleinstein]

What reference is the meter using ?  I don't see a LM399 (nor really good for 7-8 digits) or LTZ1000.

It is interesting to have ranges in 1 - 3 - 10 sequence, which is not very common anymore. That was more like a thing of the analog meters.

[NyquistNoise]

I can't say Ive done any testing personally and unfortunately don't have the equipment to do so to a 7.5/8.5 digit meter. I will probably be sending it off to a cal lab in the coming month or so and may do some testing at that point as I have no trustworthy calibration equipment that have a recent cal. I'm tempted to measure it's tempco as the meter is supposedly temperature controlled...
I'll upload those at some point today, for the time being I've attached the manual here.

[NyquistNoise]

Hard to say, the manual suggests this is in the potted box which of course appears so, there's no ref on the analog board!
I believe in other time electronics gear they use some strange diode arrangement, but in this case there's no clue in the manual. I'll X-ray it and see what's inside - I'm so curious as to the ADC used in the block.

[Kleinstein]

The manual says the signal is scaled to a 2.5 V range and a sigma delta converter is used. This would suggest using an integrated ADC chip (e.g. like LTC2440). In this case it would also make sense to have the reference close to the ADC inside the box.
Some of the SD ADC chip are quite good when it comes to noise, but the INL can be a weakness. It may need some numerical corrections to get a good lineariry. So the internal cal constants may be for more than just offset and scale factor.

Testing linearity is a bit tricky. One can still do a few tests with relatively little euquipment.  One is the classical turn over test. So see if the positive and negative reading give the same result. A slightly more difficult test uses 2 equal sources (e.g. 2x 5 V) in series and than checks if the sum really reads as the sum.

[Flukeman]

Hi, I've dropped you a message about some issues I'm currently having with my 5075.

The unit would sometimes reset if hit with 1kV straight off and usually had to be built up to.

I've been adjusting it this week at a UKAS lab over here after I replaced the AC module a few years ago (and never got around to it). Adjustment went fine, I need to do a few of the lower resistance ranges when I go back in a week or so.

The issue now, whilst checking the 5025 last night on it is that the DC ranges have a fault. 10VDC is fine, anything beyond that is over range all the way to 1kV range on the meter, so something has gone 😔

I've been out of the game a good 4 years and haven't seen the inside of the meter for probably 5 so hoping you can help 🙂

(Edit, confirmed this is just DCV affected)
(The 8th digit enabled)
(C5 and C6 on the AC module picture you posted, well mine were totally burnt out. Thankfully a rather nice chap was able to provide a replacement. I also have a new A2D module sitting here, I hope the one inside hasn't gone, given it worked until last night).

[coromonadalix]

thks for the teardown,

I would be careful around thoses flex pcb's,  allways be sure to pull the strain relief,  nothing elses

Had bad surprises in the past with theses, sometimes  the pins moves  etc ...


to me,  look's like  some  Prema meter   loll  5017 or 6001    nice looking ...