SOLARTRON 7150 Replace mains filter now ? and calibration

[dimlow]

Just got a Solartron 7150 from ebay, i read that these can have problems with mains filters spilling their guts. I have had it powered up now for two days straight and there have been no signs of any problem with it so far. Should i be replacing the mains filter as a preventative measure now ? or should i take the risk and leave it as is ?

In my view it needs new LEDS, new banana jacks to take shrouded probe jacks, plus this filter.

I have done basic checks on resistance, current and voltage and they are all pretty close to my other meters so i think its OK. I'm kind of a perfectionist and i want at least one meter i can rely on. My other meters all give me various different readings and i was looking for a meter i could rely on to give me accuracy. I got the 7150 because its quoted DC accuracy was 0.002% far better than any of my other meters (0.8%) and because i cant afford anything better. I know it's old but have read that many of these meters are still within their 90 day calibration after all these years. I hope i made the right choice.

I was thinking, i should get it calibrated first to check it out? Then if all is well,  spend time and money updating it, Or should i spend the time and effort getting it updated then, calibrated ? Should i even bother to get it calibrated ?

EDIT: It was last calibrated 2 Jan 2003, Its an Ex MOD unit ( not sure if they look after these things or if they abuse them )

[SeanB]

Bench meters tend to get the kid glove treatment. Best to replace the filter caps now, and do the LEd's if needed ( you probably are wanting more brightness from them ) and the sockets, then send it in for certification. That will tell you if there are any issues, and will show that most probably there has been little drift since last cal.

Handheld meters tend to have a hard life, often dropped and often going swimming in assorted fluids.

[Bored@Work]

Replace the filter. I neglected doing this with the last 7150plus I got. I knew the mains filter could blow up, but this time I was to lazy. A few weeks later I had to pay the price, the magic smoke escaped, making a real mess.

[jucole]

I have a 7150 which I bought on Ebay last year; the seller mentioned that dc volts was slightly out but the others were pretty good.  I had heard about the mains filter but my thinking was there was no point spending time replacing things until it had been through calibration.   I took it to be calibrated late December; and it turned out they had to do a small repair as the error was too much for the offset to make good;  maybe someone had been hammering it with large dc voltages over the years;  anyway,  it all turned out ok;  I posted a pic of the cal results in one of these forum posts somewhere.   Mine now is my "gold transfer standard" ;-) to use as reference for my other meters.

If you're asking the question though "Should i even bother to get it calibrated ?"    my view is it's the electronic project that dictates the level of accuracy you need, but it sounds like a great meter though! ;-)

[ecat]

I posted a pic of the cal results in one of these forum posts somewhere.

So you did many thanks, I had totally forgotten.
https://www.eevblog.com/forum/chat/calibration/msg178924/#msg178924
Looking good.

[Wuerstchenhund]

Just got a Solartron 7150 from ebay, i read that these can have problems with mains filters spilling their guts.

This is correct, however the problem is not limited to Solartron DMMs but any device which uses mains filters made by the Swiss company Schaffner and which is 20 or more years old. After all these years, one of the capacitors dries out and blows, which results in a very nasty mess.

I have had it powered up now for two days straight and there have been no signs of any problem with it so far. Should i be replacing the mains filter as a preventative measure now ? or should i take the risk and leave it as is ?

Replace it, better sooner than later. These old filters are ticking time bombs. Also be careful when buying a replacement. Quite often you can find cheap 'old stock' but these filters are usually as old as the one in your DMM and not much better. Make sure you actually get a brand new one (can be checked on the label, if the filter is Made in Thailand its new, older filters are Made in Switzerland).

In my view it needs new LEDS, new banana jacks to take shrouded probe jacks, plus this filter.

Regarding the LEDs, I guess you want a brighter display? Well, there is not too much you can do. I have replaced the LEDs in my 7150 Plus with bright white LEDs, it's an improvement but you won't get the same brightness as other LEDs have.

[dimlow]

I opened the meter up today, its a beauty inside, so clean,looking like new after all these years. So took your advice's and I decided to do the mods before getting it calibrated. I wont be using it again until i have changed the filters. I ordered a new IEC filter from RS http://export.rsdelivers.com/product/schaffner/fn372-2-21/fn372-2fuse-inlet-filter-2a-5x20mm/4655193.aspx I hope this is not new old stock , i have no way of checking until it arrives.


The Images below inspired me to want to change the back light LEDS in the LCD. I have some super bright blue LEDS i want to try in it. You will also notice in the image with the blue LEDs that the guy also changed the banana plug jacks. I ordered some replacements for them today as all my test leads are shrouded and will not insert into the meter unless i cut off the shrouding. Parts will arrive early next week i hope, I will post some pictures of the inside for electronics pron fans out there when i open it up again.

[dimlow]

Just noticed that RS claim RoHS Status Compliant, So will this mean its new stock ? As RoHS Status Compliance dates from 2002 on-wards.

[dimlow]

Changed the mains filter today, i got a replacement from RS. If anyone is interested the filter is a FN327-2-21 RS 4655193.

It was a direct drop in replacement. The only thing i had to do was remove a resistor from the old filter and solder it onto the new. (not in any photo's) I also had to swap out the fuses.

Whilst i had the unit open, i changed the LEDs to bright blue. The old LEDs were quite difficult to get out initially, but once i had learnt the method for one, the rest got easier.

I posted full pictures on my Picasa Web here.

Below are a small amount of the pictures i took for the hardware porn junkies.

Getting Started


Top Cover Off


Bottom Cover Off


First Glimpse of some parts


A Closer Look


The Bottom PCB


Open wide


The Problem Mains Filter


Mains Filters Side By Side, Note the new filter is made in Thailand (I got the right one)


Installed and Ready to Test


Working Yay!


LEDs Removed


New LEDs Installed


Back at home waiting for the next upgrade, the new banana jacks. Still waiting on the delivery from china.

[jucole]

Great pictures!! 

[ecat]

Very nice pictures indeed. While I'm not a fan of blue displays yours tends more to the funky side of purple so I'll let you off

I've had my 7150+ for about six months now and I've not so much as waved a screwdriver in its direction let alone taken it apart. I decided to wait until I had some sort of reference to compare it against to see how well it has held up since its last calibration in December 1997 and to ensure any opening and twiddling were not detrimental.

Just last week I got hold of a unused Vishay RCK02 0.01% resistor ( http://www.vishaypg.com/foil-resistors/list/product-63214/ ), not ideal as 5k0000 is near the bottom end of the 7150s 20k range but the price was right. How does the little 7150 with a 15 year old calibration hold up? How does 5.00000k sound ? lol.

Okay, there is some noise in the least significant digit of up to +5/-3 counts but, I think that's pretty good going. For those curious, the Vishay checks out on my Datron 1061 at 4.9999.

[dimlow]

Yep, my old Nikon Coolpix 995, take quite nice pictures.

The back-light is in fact blue, but because i took the photo with the flash on, the color is off purple in the images.

I have a couple of 1 ohm resistors and checked them and got 0.00101k on the display that was nice to see. My calibration is only ten years out of date. I was thinking of getting it done until i found out its going to cost more than the meter. So have come up with an idea, not sure if it will work. https://www.eevblog.com/forum/testgear/is-this-simple-ohms-law-test-for-calibration-of-multimeter-useful/

[dimlow]

Today the meter is going off for calibration here They will do ukas for £40 and non ukas testing for £30. So going for the non ukas unless they convince me otherwise. They will Also adjust if its out of calibration.

[jucole]

Today the meter is going off for calibration here They will do ukas for £40 and non ukas testing for £30. So going for the non ukas unless they convince me otherwise. They will Also adjust if its out of calibration.

Great stuff! and sounds like a good price as well!

[dimlow]

Not So Great stuff, i just got back from the lab, there was some confusion as to what meter they thought they were getting, the sales man i spoke to on the phone ( i thought he was the service manager) quoted me the price for a standard handheld multimeter. In the end full calibration is going to cost me £65. That includes adjustment if needed, also if the meter is faulty in anyway, they will not charge if they cant get it into calibration. So in the end i elected to go for the full calibration. Didn't really want to spend that amount but i guess it will give me confidence in the meter.

On the plus side, My son and i also got a nice guided tour around the Lab. Well in fact we saw three labs in total, a repair lab, a standard calibration lab, and their ukas Lab, they sure had spent some cash on the ukas lab. lots of shiny new equipment with flashing lights and screens, it looked the business. I'm confident that they will do a good job.

Could have the meter back by Friday if all goes well.

[ecat]

That's still not a bad price is it? And a free tour

Even if you were to attempt a diy job the standard 7150 can only be calibrated via the IEEE interface, IIRC, so that would be another outlay. Good luck to father and beastie.

 

[dimlow]

Got a call this morning. "Hello, we tried to calibrate the meter this morning. Unfortunately it would not switch on, the meter is completely dead", " OK, i will come and collect it then"

Shock horror! WTF!

Grabbed the keys for the car, went to say good bye to the kids, picked up the phone and as walking to the car, got another call from them. As a final check the service manager tried the meter again and found it to be fully working so they are going to run some tests to see what the problem might be and if all is OK they will calibrate. I explained to them i had changed the mains filter and connector. I believe that they simply didn't push the mains cord in correctly it being a new connector and a little stiff. Service manager agreed with me and said they will call back if there are any more problems. Its been two hour now and no call, so i expect it will be fine.

This is getting to be quite a journey. I really hope this all turns out ok

[dimlow]

The meter is back from calibration but, i need some help evaluating the results. Basically i believe they have calibrated it on a machine that does not have the accuracy needed to test it.

Attached are the Calibration results.
Here is a link to the Calibrators Data sheet.
Here is a link to the manual of the 7150 with the specs.

It passed all the tests they did on it.

When i first looked at the Calibration certificate the first thing i see it the wide Tolerances in DC voltage. Now from the specs of the meter it has in 0.2 DC volts range, a specification of 0.002% of the reading +5 counts in 6.5 digit mode.
If i calculate that to a voltage tolerance i get +/- 4uV + 5 Counts, So that's a max Tolerance of +1uV to +9uV.  The tolerance on the sheet is 24.7uV and thier uncertainties are +/- 30ppm, so is this right ? I dont think so. How could they possible reach the accuracy of the meter if the meter is more accurate than the calibrator they used.

I hope I'm wrong and someone can help my figure this out as i have not done this before


Edit: removed Cal certificate, because it had personal information on it

[jucole]

ok, this is the way I think it works but I could be wrong.  Taking the dc 0.2v range as an example; on your results, the "tolerance" shows how close the reading has to be to "PASS";  the "gotcha" for old meters is the age; if you look at the 7150 manual, section 10.2, and "limits of error"; you have to factor in not the 0.002+5 counts but  0.01+5 counts for every 2 years since it was made + the temp co + the wind direction etc.

Now in your results the 190.000mV applied voltage seems odd but when you think about it, but if they applied the full-scale of the meter which is 199.999mV,  and say your error was +1uV or more it would probably return OL (overload) so you would never conveniently know your error;  so the test voltage of 190.000mV is a good fixed value to set the GPIB software to use as a final result for a nice printout.

[dimlow]

I agree mostly with what you are saying, but the point i was trying to make is that the uncertainties they have are +/- 30 ppm where as the solartron is accurate to +/- 4 ppm plus ten counts.

Lets put that into real numbers.

Calibrator voltage output  is set to 0.190 000V but it could give values of 0.189 970 to 0.190 030
Solartron when supplied with 0.190 000V could read values of 0.189 996 to 0.190 014

The Solartons accuracy is inside the accuracy of the uncertainties they used.

The 200 mV, 2V and 20V ranges are all effected because their error in these ranges is larger than the Solartrons accuracy, shouldn't Calibrations be done with a machine that is 10 time more accurate than the machine they are calibrating ?

But still im not 100% sure im doing it right, that's why i need someone in the know to come and look at the data sheet and the calibrations to see if i need to go back to the Calibrators and talk about this to them. I did talk on the phone to them and the guy immediately offered my money back. Now should i take it ?

[dimlow]

Just made a little spread sheet comparing the Calibrated Tolerances from Alpha Electronic to the Tolerances in the manual. looks interesting



As you can see the Meter would have failed calibration, but if you take Alpha's 30 ppm of their test equipment uncertainty into account the meter could be in calibration. if you look at the 2V range ( 1.99999 ) the reading is 1.900071 if you subtract the 30 ppm it would be a pass not a fail.

Alpha Electronics Tolerances are way out at the lower ranges , but as they get higher they get closer to the Stated tolerances in the manual.

[robrenz]

AFAIK you are not calculating your meter spec properly. The plus X counts is added to the % of reading in both directions.  Your lowest range shown above would be +/- 9µV  not -4µV + 9µV.

This does not change the validity of you questioning the quality of this calibration, AFAIK the TUR should be at least 4:1

[dimlow]

Ah, thank you! I will change the spread sheet and see if it makes me Happier , But i think what its come down to is they have calibrated a 20ppm meter with a 30ppm Calibrator, Am i right in thinking that ?

Edit: Just had a look at the spead sheet again and my error still would have made no difference to to results, its has still failed according to the specs in  the  manual, and i think its because of their test equipment.

[Chasm]

Without looking at the actual specs and using a bad analogy as starting point...
Trying to check the dimensional stability of a vernier caliper with a tape measure will only find gross errors. Very gross errors.

You obviously can't actually calibrate an instrument to it's full specifications with an calibrator of lower accuracy.
You can do a sanity check to find gross errors.

Or you could throw the initial specs of the device under test out of the window and calibrate to a different, lower standard.

-> This question is the point of an often repeated sentence in the Aligent cal lab videos: "Your cal lab is NIST certified. Nice. But what exactly is you cal lab actually NIST certifed for?"

[Chasm says: 10V DC, +/-1% ?  ]

[dimlow]

Without looking at the actual specs and using a bad analogy as starting point...
Trying to check the dimensional stability of a vernier caliper with a tape measure will only find gross errors. Very gross errors.

I love this analogy, i think i may use it when i email them tomorrow

Just for completeness here is the new sheet