Fluke 185 Review

[DK ekectro]

 with modifying this dump file, the problem is that I can't change the label h01 to h02 ... I've tried several programming applications but I miss the last three column to display the address, so at the top of the program I only have; 0123456789ABCDEF columns, but I'm still missing these next three columns; 012, where in the end is the data needed for the modification ... Can anyone help with this problem ...? Thank you!

[shakalnokturn]

 :)All Hex editors won't have the same number of columns, it's just a question of display layout.
Look for address location Hex:600 no matter how many columns there are in your editor, if you're really as bad as I am for looking through hex files count back from an easily identifiable position such as start of serial number.
While you're at it could you please upload a copy of your original EEPROM contents if available? It could help understanding organisation for people in need of manual readjustment.

Edit: I figured you were using my memory dump anyway so if you really can't do the job with your hex editor, I can send you the modified file.

[shakalnokturn]

I have just received a Tektronix TX3 with a lowish serial number (<2000), F.W. version 0.0716. Funny thing about it is the back label is factory marked "DEMO NOT FOR SALE" above the serial.

Any other TX1 or TX3 owners noticed this?

[shakalnokturn]

My two meters side by side comparing the modified Fluke 183's uncalibrated temperature readings to the long ago factory calibrated TX3's.

Note that the meters do have some differences, I prefer the plastics used and keyboard colours on the Fluke. The LCD is better on the Tek though.

I also noticed in the Tek that other than the spring contact there is a pogo-pin contact to the back shield near the relay, the pin is free to drop out and be lost (at the worse left floating in the DMM). I lightly squashed the end of the barrel on mine to hold it captive. As I remember it was already missing on the 183 leaving me puzzled to the purpose of the part at the time I opened it.

[DK ekectro]

Hello everyone! First, I would like to very much thank the shakalnocturn member for the detailed description and presentation of everything he has done so far regarding this work on the TX1/TX3 measuring instrument. I believe that, like me, it helped everyone a lot. I bought a used TX1 but in very bad shape and with a UNCAL problem. Thanks to a member  shakalnocturn I made a complete modification to make the TX3 officially now... Everything works great except that there are deviations in some measurements, for example 80mV-100mV difference in AC and DC in all rankings, other measuring areas are not quite accurate either. Maybe someone has a very original dump file of TX3 so let me try it with him because the one that is currently in it is from Fluke 185. who knows...? I can live with that but after so much effort to paint the case again, polish the plastic of the screen, print the markings( it's not all over yet), I'd be sorry if some other instrument that was paid $20 in a supermarket showed more accurately than it...!

[shakalnokturn]

Wow! Nice job on restoring the plastics!

When I get round to it I'll dump the EEPROM from my TX3 and post it here.

It isn't really a TX3 dump you need, no matter the dump you may be better or worse off than now. Your voltage measurements could be closer if you're lucky but you could also end up worse off on other functions.

The correct way to do things would be going through the software assisted adjustment procedure.
I'm pretty confident there are a couple of people around that know the procedure and have the software, I tried asking on Tektronix@groups.io without much luck.

The more painful option is identifying memory locations in EEPROM for each adjustment and tweaking them by trial and error until gathering enough understanding of gain and offset.

[DK ekectro]

Thank you. I had no problem removing the old paint with an IPA and a soft cloth. I have a friend who prints various pictures on T-shirts, lighters and other items, so he made the print that can be seen in the picture, because of the equipment he has it was difficult to make a print at the area between keys and the rotary switch, so we will do it with stickers...I don't know why but I just love this instrument, it's simple and yet it has all the functions I need in working with electronics. And I believe that there are people who know the calibration process or even who could share information about which items in the eeprom file should and how they should be changed for certain measuring areas... It would be a shame if such old classics became useless...

[shakalnokturn]

Finally found a good enough excuse to drag the TX3 out and dump the 25LC160 contents.
Still looking for the calibration commands over infrared. Anyone?

[jchw4]

I took some high resolution photos of my Tektronix TX3 PCB.
These are aligned, you can overlay them and do some tracing.

Components side.

[jchw4]

I took some high resolution photos of my Tektronix TX3 PCB.
These are aligned, you can overlay them and do some tracing.

Sockets side.

[jchw4]

A few close up photos of nice resistors.

[jchw4]

A few close up photos of nice resistors.

[jchw4]

A few close up photos of nice resistors.

[ErnestB]

Hallo, I have read this post and am curious if there is any progress on the calibration procedure of the Fluke 185 / Tek. TX3? I got one 185 in physical good condition, but with UNCAL massage after the power-on. It was used in my company (by me for several years), but several years ago it came back from the calibration service with a vague massage that it couldn't be calibrated or something similar my company wrote it off, and it could end up in a "dumpster".  The meter is still working quite good, but I am not sure which ranges of functions are actually off, and by how much. And of course, there is the UNCAL massage at the power up every time. Anyway, it would be interested to find out how does that calibration procedure actually works. As already stated above, the Service Manual is very clear about it: if you want to calibrate it send it to us, there is no other way. The thing is of course, these meters economic value at this moment would be much lower then the cost of the official calibration. So, if I can help in any way with mine 185 let me know. I am not such a hacker (yet) though as you guys but with some guidance maybe I could contribute something. We would need calibration procedure description, including function-key-combination and the proper set of commands needed, with some cal voltage references. Greetings from Holland.

[bdunham7]

the Service Manual is very clear about it: if you want to calibrate it send it to us, there is no other way.

The calibration manual specifies the exact procedure, unfortunately it appears to be done through the IR port and with MetCal software only.  It certainly doesn't need to go back to Fluke, there are many calibration facilities  that have these tools.

https://www.industrialcalibration.co.uk/downloads/Fluke%20183-185%20(PN%201610436%20Rev%201%206-02)%20Calibration%20Manual.pdf

[ErnestB]

Thank you for responding and posting the document. I have read this Cal manual and the part regarding the Calibration Adjustment is very short and does not describe anything specific:
"The Fluke 183 and 185 require adjustment only when they fail the performance tests or have been repaired. Software adjustment is performed through the meter’s IR port using a semi-automated Fluke Met/Cal procedure and Fluke 5520A. The meters have no physical adjustments and cannot be manually adjusted".
I have seen what Fluke 5520A costs    30-40k And maybe/suppose other "calibration facilities" still could do the calibration on this discontinued model, the expected price would be above the economical value, so that would not be an option I suppose...
Maybe/probably the software "talks" to 185/TX3 as well with the 5520A during this semi-automated Fluke Met/Cal procedure, and change the settings on 5520A accordingly, ingaging various voltages and currents on it's output during the procedure.

[ErnestB]

There is MET/CAL software Reference manual document 1252 pages long: http://download.caltech.se/download/fluke/metcal/metro_meeting/ReferenceManualV71.pdf

[bdunham7]

I think the manual is specific enough, it just doesn't seem substantial because there isn't much to it.  You connect the meter and the 5520A to your PC and run the MetCal procedure.  It prompts you to select a certain range, then it tells the 5520A to output a specific stimulus and then after the reading stabilizes it it tells the 185 to store a new calibration constant for that particular range and reading.  Rinse and repeat until done.

I wouldn't assume the calibration is more than the value of the meter without a quote.  I think maybe here in the US the price might be $80 or so?  If the meter isn't worth that, then nobody really cares.  What I would worry about is that calibration might not work--it might have some fault that prevents the process from completing or storing the constants.  Then you'd be out the calibration cost and still have no meter.

If you could reverse engineer the MetCal commands, perhaps you could come up with some alternative.  However, I think you'd need a complete working setup and meter to do that and their really aren't too many of those about.

[ErnestB]

Good point: "Then you'd be out the calibration cost and still have no meter". So yeah.. I don't know. There has to be a (valid) reason, I suppose, that the calibration facility couldn't calibrate this meter anymore in the first place. Maybe the eeprom (or some sections) is damaged in some way.. But the meter seems to works fine, so far I have seen, but of course I have not checked every range and every function to abs values and tolerances.

[ErnestB]

...and of course there could be various reasons (beside the eeprom) why this 185 have not "made through" the official CAL. I will try to check myself so far I can on different ranges and functions to see how far off the stated accuracy actually it is at this moment, and then make a conclusion.

[ErnestB]

I know, long shot, but would somebody maybe know what value and type the 2 capacitors C99 and C100 are?
I suspect these throwing off my ACV reading. Thanx in advance.
Edit: I measure (in circuit) about 45pF each.

[shakalnokturn]

I don't know if he values but the values you have measured seem consistent with placement in circuit and ceramic colour.

They're there for frequency compensation on the higher frequencies. Unless they've both gone resistive there's little chance they'll be causing measurement error at 100Hz.

I'd be more suspicious of the opto-isolated analog switches if anything.

[ErnestB]

Hi, thanks for the answer and valuable insight. What I have found when doing some tests today is that the measured value strongly decay with frequency. I did the test on about 100, 1kHz, 10kHz en 20kHz. Fron 1kHz to 10kHz bigger jump down. Within this freq. range the decay was about 7.5% on both 0.5V and 5V range (0.4V and 4Vrms sinus used). First I was thinking that the main coupling capacitor ESR was high, but when I shorted that C I have not seen much difference. Than I was thinking about those capacitors parallel to every "range resistor" (1k, 10k, 100k and 1M). But because both 0.5V and 5V range behaving almost the same that was less obvious. So I have de-soldered those 2 AC compensating capacitors and have experimented with some regular through hole ceramic capacitors instead. Some values made it more freq. depended. At the end used 2x 10p and I have to say I do not have a very good capacitor meter for lower values, so the original one's could also be about 10p.
So maybe you know more about the capacitors used in this meter, as you said regarding the color? Are those MLCC? As I see the compensation is very critical. I could imagine that if the capacitor would have some ESR, also freq. dependable then it can throw off the whole AC measurement. Maybe for that 10k could use adjustable resistor to try to adjust the values...

[helius]

The ceramic compensating capacitors are in a measurement circuit, so they are going to be C0G (NP0) caps. The color of this material is typically gray. When people talk about "MLCC" they are usually referring to barium titanate based dielectrics which are more beige or brown in color. Those Class II dielectrics are not very stable and not suitable imho. (See the recent thread on "MLCC X5R Capacitor Degradation")

[ErnestB]

Thanks. These capacitors are not beige indeed, more gray. Indeed NPO capacitors would be necessary here. But I have to figure out the values, as I think those are responsible for the AC measurement being out of specs. Frequency response of this 4 component network have to be as flat as possible between 40Hz and 20KHz, and they probably calibrate the last part (offset) in the firmware. I can not do that, but I could place a small Pot. instead of that 10k resistor to tweak the last part.. But I suppose it will not be easy, as the input and output impedance plays the role...
Edit: hmm, and what am I compensating for anyway :-) so this is a very specific case high pass filter, that works in conjunction with the rest of the measuring circuit. And you have different ranges too....