Repairing A Workhorse Bench Meter

In today’s market, and expensive high-precision bench meter will have a host of features: graphs, alarms, averaging, and more. It will probably even use an operating system. However, old meters can still get the job done at a price that you can actually afford. A case in point is the Fluke 8842A, solid meters with 5.5 digits of resolution and the ability to do two or four wire resistance measurements.  They are built like tanks and are surprisingly affordable, especially if you consider what they went for when new. [Illya Tsemenko] recently updated a log about repairing such a meter, and there is a lot of good information about them if you own one or are thinking about one.

The biggest problem with repairing these meters is that there are several custom parts including the display that are essentially unavailable. For that reason, [Illya] took a meter with a broken display and used it to source parts for another meter.

Someone had already worked on this meter and there were some suspicious mods. For example, a matched set of JFETs gave way to four individual FETs. Perhaps they were hand matched, but perhaps they were not, which could cause lots of measurement errors.

The meter uses a patented recirculating remainder A/D technique. The patents give a good overview of how it works and are linked in the post. The issue with the meter turned out to be cracked glass in a thin film resistor network that had caused the resistors to change values. This is one of those components that would be very hard to source, but luckily it was good in the donor meter.

These are great old meters and well worth their cost and bench space. If you want something more steampunk, maybe use a clock as a bench meter. Of course, for many uses, a cheap meter will do just fine.

23 thoughts on “Repairing A Workhorse Bench Meter

  1. A fine meter that a one time I used at work. I have never found the combination of functionality, repairability, and price to put more than 4-1/2 digits on my personal bench. Price being key. My bench workhorse is still an 8000A, and an HP3475 (3-1/4 digit autoranging from the paleolithic era) is my secondary. 4-1/2 digits come in the modern import form factor, having been free from a contract job.

    This is a very nice writeup, and makes my slightly (very slightly) more willing to consider gear of the era in need of repair. Thanks for the link.

    1. I actually have one of these on my bench and I use it often. But I will admit, I don’t really use the precision as often it just is a solid and handy meter. it is nice though to be able to see very small changes in things even though the absolute value is usually not very important, sometimes the trend does matter.

  2. Several months back, I purchased an HP 3455A 5.5-6.5 digit voltmeter which cost around the price of a new car when it was produced, but now can frequently be had for ~$100 shipped. The internals were super clean, and while I don’t know the accuracy of its AC measurements (which I currently don’t care about), or resistance, I did compare DC to a 10V voltage reference from VoltageStandard and it was bang on. While bulky, these precision bench meters from the late 1970s to 80s are impossible to beat price-wise when you want accuracy and high resolution.

    It’s actually pretty incredible how much you can get for $100, at least in the US. It’s the amount I frequently buy once expensive equipment for.

    1. I have a couple of 3455’s and they are workhorses. HP made an extremely large number of them so they’re easy to find repair parts for. I’m also fond of the 3468a. (I used to assemble and test these for HP as my high school job.) They’re still in the $200 range, but they also offer inline current measurement up to 2A, which I really appreciate.
      I’m seeing 34401a’s for under $400. Those are very recent and really lovely machines, although they’re much harder to repair than the 3455’s and 3468’s.

      1. When I was looking for a bench meter, I wasn’t dead set on a particular brand or model, but noticed the abundance of HP gear which tends to move the average selling price downwards. While I could occasionally find some other makes and models similarly priced, it’s nice to have a bigger pool to choose from should finding replacement parts be necessary as you mentioned. With the 3455, I liked the visibility (and serviceability) of an LED display over the LCD on the newer models, and the physical adjustments for calibration over the battery-backed electronic calibration found in the later models. For my price point, I felt like I’d be taking a greater gamble on something like an older 3457/3478 which could have had a dead battery and thus, its cal settings wiped out.

        BTW, sounds like you had a pretty cool job in high school.

    2. I would love an meter from the HP3456 series but they are just too big, and at home I have yet to need their accuracy. The vast majority of the stuff I do at home is well serviced from one of the many free DVM’s that harbor freight used to hand out. I do have a 5.5 digit fluke that a friend gave me and said it needed new caps. One of these days. Perhaps. I would hate to put the $$ into it to find out it is really the big custom chip that is toasted. At least it is not physically large.

  3. Is there such a thing as a late developing dyslexic? Just because even hours after seeing the title for the first time, my brain keeps tripping up on it as I scroll up and down the front page and I keep trying to read it as repairing a workbench horse meter or something.

  4. I bought one of these on EBay last year after seeing them touted on EEVBlog. Good tips on replacing the receptacle and inspecting the capacitors.
    Mine hasn’t been calibrated since 1985. Is it more than the $100-$200 that these multimeters cost? I am in San Jose, CA and see a couple local shops that I will call. Just curious what other home workbench people do.

  5. Can someone enlighten me on what purpose these highly accurate meters might be used for outside a laboratory or a place that builds/tests OTHER similarly accurate equipment?
    Nice as they are they would be far and away beyond the ‘need’ of 99.9% of engineers.

    Not saying I wouldn’t want one though!

    1. I have a 4.5 digit handheld DMM, and in my experience the absolute value is rarely useful.

      I mostly appreciate being able to see small differences, like seeing the forward voltage of an LED drop as the temperature increases.

      Also, being able to measure down to microvolts is frequently useful to measure voltage drop across a wire or PCB plane, or measuring microamps in battery powered devices.

    2. That’s pretty much their purpose. The fact that you can this kind of equipment for the price of a lower-end Fluke handheld adds to their appeal. Granted, that’s not an apples-to-apples comparison, but if all of your measurements take place on a bench, then why not get something like this?

    3. If someone elsewhere that wants to duplicate your results, then you want to make sure that your meter is tied to a standard. Duplicating results by definition also apply to doing open source work.

      Doing proper measurements without accurate results is a waste of time. measurement =/= quick check.

  6. Another interesting fact that not many people know, ADC of this meter is used almost with zero changes (even using same custom chip and resistor array!) in mighty Fluke 5700 series calibrators as precision null-meter. And stability of reference, performance in front-end components used by 8840A and 8842A is well worthy of 6.5-digit meter. Nice little fast meter that could!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.