How to convert a Fluke 77 IV average responding multimeter into a True RMS model 177 for 10 cents.
Should also work for converting a Fluke 83V to 87V
True RMS measurement vs Average Responding
And calibration of a Fluke meter.
00:00 - Hacking a Fluke 77 IV Multimeter
00:35 - Is the Fluke 70 series discontinued?
01:15 - Fluke History
03:10 - I noticed something...
04:49 - The AD737 True RMS converter chip
06:04 - Theory of Operation
07:21 - What is Crest factor
08:58 - How do they do it?
09:25 - Fluke 87V Schematic and the custom ASIC
10:53 - Let's modify it!
12:58 - A sine wave is NOT a crest factor of 1
14:01 - Soldering
15:15 - Will it work?
16:46 - Calibration procedure
20:09 - Winner Winner Chicken Dinner
Forum: https://www.eevblog.com/forum/blog/eevblog-1448-convert-a-fluke-77-iv-to-true-rms-for-10-cents/
Subscribe on Odysee: https://odysee.com/ @eevblog:7
EEVblog Web Site: http://www.eevblog.com
The 2nd EEVblog Channel: http://www.youtube.com/EEVblog2
EEVdiscover: https://www.youtube.com/eevdiscover
Support the EEVblog through Patreon! http://www.patreon.com/eevblog
AliExpress Affiliate: http://s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
Donate With Bitcoin & Other Crypto Currencies!
https://www.eevblog.com/crypto-currency/
T-Shirts: http://teespring.com/stores/eevblog
#ElectronicsCreators #Fluke #Hack

Hi. In a previous video, I did a teardown of the Fluke 77.4 and compared it with the Brim and Bm786 Ah meter and I noticed something really interesting in this. So in a minute, I'm going to show you how to turn an average responding Fluke 77 4 into a true Rms fluke 177 and do it for about 10 cents. Stick around.

But um, curiously, here on the website, I didn't actually know. But I think Fluke have technically discontinued the 70 series the venerable 70 Series. I think it's actually been discontinued because if you go to their digital multimeter page here, there is no Fluke 70 Series anymore. They've got the 170 series.

They've got 175, 177, and 179. but that's it. There is no 7. like it's still there on the site if you actually search for it.

But I think as I talked about in the video previous to that about why flukes are so expensive in that, Um, yeah, this is appealing to legacy customers only because the Fluke 70 Series were average responding meters. They were not true Rms meters. So a slight bit of history here. Um, the Fluke 79 Series 3.

That actually was true Rms. but they don't sell the 79 anymore that was replaced by the 179 or the 170 series. In fact, the entire 70 series has been replaced by the 170 series True Rms. But there are customers who still want average responding meters instead of true Rms.

As I said, they have these built. the readings for these average responding meters built into test procedures in the military and government and organization they don't want to change all their test procedures. They want an average responding meter, so Fluke will still sell you one. But basically nobody buys the 70 series anymore.

You buy the 170 series so you buy the 177 and it's exactly the same meter as the 77 4. Except it's true Rms. Now when they actually moved over to the 170 series, this was about 2001 and they dropped the uh, you know, the slash like the Mark Iv and everything from it. And so the 177 you buy today is still the same 177 from 2001..

So it's a 20 year old model. Now there is no 177 Mark Ii or whatever. it's still the 177. So 20 year old model now.

But they kept selling these Uh 70 series. They the Mark 3 was the first series fluke that actually looked like this with the new over molded design and that's what the 170 series stem from. And then when they moved to the Mark Iv, the only one you could get was the 77 Mark Iv. They just didn't bother with the previous versions because they knew they were phasing out.

they wanted to go to push everyone to the 170 series tour Rms, but they kept one model in the seven in the line as an average responding. but I think it's gone because like it's just not on their main page. Anyway, I actually forget how long these like the 170 series has been around for 20 years now. That is absolutely incredible.

Anyway, during the tear down video of the 77 4, I noticed something interesting in the chips here. And uh, this is my, um, high-res photo available on my Flickr account. I put all my high-res photos on my Flickr account. Go check that out anyway.
Um, if you want to know how I get these excellent Pcb uh, photos, uh, head on over. It's actually on my second channel. I did a part three to my Pcb Photography Lightbox and this is how I get the excellent photos. so I tell you how to make one in there.

anyway. Um, yeah, I noticed something really interesting. We've got our Msp430 processor over here. We've got a 20 bit Delta Sigma converter in here.

The reference is in there somewhere. We've got the Fluke multimeter chips, which by the way, if you search the number here, you can actually still find and you can. actually looks like you can buy this chip from third-party um, vendors Abs. Absolutely amazing.

Anyway, the interesting thing is up here. This is an Ad737 That's a true Rms converter chip used in like a ton of multimeters on the market. What's it doing in a Fluke 77? That is an average responding meter. Why does it have the chip there? Now, as I noted in my teardown video, the Pcb actually is the Fluke 170 Series.

There it is. Fluke 170 X up there. Um, it's exactly the same Pcb. And sure enough, I just looked for some tear down photos, found them on the Eev blog forum.

Everything's on the Ev blog forum. Um, and sure enough, the 170 and the 79 have this exact same components except one. And sure enough, the Ad737. It's a true Rms to Dc converter used in multimeters.

So what gives? Why would they go to the expense to put this chip in an average responding meter? Well, consistency and build is one thing. Of course, fluke make a quite a decent margin on these things. So maybe the cost of a 737 is not a big deal. They just have the same build.

But remember when I said there is actually one component build difference between the 77 and the 177. this is where you go to the features over here. Look at This computes true Rms, average rectified value, and absolute value. It's actually got three different modes.

this chip. It can actually do average responding. multimedia functionality itself. Of course, most average responding meters on the market.

the real cheap ones. They do this inside the multimeter chipset itself, but the Fluke chipset obviously doesn't do that. In fact, we can have a look at the block diagram in a minute. But what they're doing in the 77 and other model flukes.

as we'll look at, they're actually using this Rms converter chip in the average responding mode, so I won't go into huge detail how this works. I'll link in the data sheet down below if you want to have a read for yourself. But basically so what we've got here. This is our input signal here and this is our input fet, a buffer and that goes into effectively what is a full wave bridge rectifier and that gives you an absolute value.
It takes any negative stuff and puts it and flips the negative half up to the positive half and then that goes into. Then there's an Rms core down here and that's what does the true Rms conversion part. Using an averaging capacitor down here and then up here, you can choose either Ac or Dc feedback path here. So if you bypass this Rms circuit down here, then you'll get a Dc average response just like any Dc average responding meter.

And of course you should know your peak to Rms conversion factors. And of course Rms is 0.707 and your average is 0.636 That's the Dc average of the peak value. So what they do is they calculate a scale factor in here of in this case 1.11 which is 0.707 divided by 636.. of course all it does is basically it squares the signal.

It takes the average and then contains the square root, Rms root, mean squared, and any Rms converter will have a maximum crest factor. It can actually uh, tolerate. and you know, like a pulse. Like a very narrow pulse down here.

Like an extreme crest factor. And your true Rms multimeter probably can't handle, uh, that sort of thing. But anyway. um, you can actually get errors.

But what happens in your side? Your multimeter. If you've got an average responding meter, they it's actually calibrated to match the average to give you an average responding for a perfect sine wave. So here it is: undistorted sine wave. So perfect sine wave.

If you've got a true Rms meter, it gives you 0.707 of the peak value, and an average responding meter will also give you 0.707 There's no error whatsoever, and they give you the error over here. Zero. But you put in any other waveform, a square wave, a triangle wave, noise recta, a pulse wave. You know, Scr switching waveform which is like a switching um thing.

and then, no, it's not a sine wave. So you're going to get an error on your average responding meter. and that can be. Some of these errors can be pretty high, right? You know it's It's not a huge amount if you're talking about, say, a triangle wave or something like that, like 3.8 percent.

It's not much, but the further away it gets from that ideal sine wave, the more error you're going to get. This is why people use true Rms meters, and they're pretty much the standard these days except on really low cost meters. Because most multimeters that need true Rms function, they have to spend more in their bill of materials to get this Ad737 chip, but fluke because their meters are so high price and they design it in it. They're just using the chip anyway and they're using it in average mode.

so how do you do that? Well, it's real simple. you simply remove the averaging capacitor and if you do that, it basically passes straight through and you get the average mode. So surprise, surprise. What do we find in our average responding Fluke 77.

Yep, a capacitor right there missing the 33 micro farad capacitor. And that is the only difference between a Fluke 77 and a Fluke 170 series. But it's not the only meter that does this. Check out the Fluke 87.
We've actually got the service manual here and I'll link it in down below and we've got the full schematics. Check this out! Beautiful back. When you can get schematics, I cannot find a Fluke 77 schematic or a 170 series schematic, but it's going to be very similar to this. And there's that Fluke custom Asic I showed you before and um yeah, I think you can actually buy this um on the market.

So yeah yeah. I don't know if you can buy it from Fluke, but other suppliers have it but looks of it. Now The good thing about this schematic is that they tell you look for the Model 87 only they do this. This is it.

Looks like that's doing. Locate near the V terminal. they are using that as a temperature sensor. Yeah, yeah.

thermal uh, compensation there. And there's an 800 Hertz Filter here, which is specific to the Model 87 only, and then Model 83 only here. But bingo. Same thing happens here.

There is your averaging capacitor on your Ad737 Model 87 Only 33 microfarads. so that's what's missing from our Fluke 77 to turn it into a Fluke 177. And I haven't checked, but I'm willing to bet this is also the same on the Fluke 27 Series 2 as well, because that's an average responding version of the Fluke 28. I bet you they're doing exactly the same thing.

A missing capacitor. That's it. They're designed for different markets. All right.

let's do a test before we do modification here. I've got three different meters: the True Rms Fluke 87.5 the 77.4 we're going to modify, and the Uh 17b here, and both of these are average responding meters. They are calibrated for an average response of a sine wave, so these should read the same. They might have different, like upper frequency limits and stuff, but at a reasonable frequency.

I'm going to use Uh 88 Hertz here. why not? So that is well within um, the specifications of these meters. So I'm feeding a one volt Rms sine wave here into all three of them in parallel. As you can see, they all read identically.

17b is a bit lower because this is not a very super duper accurate meter and these are within one least significant digit count. So that's for a sine wave and that's what. Even though these are average responding meters, and we are not going to get an average response of the sine wave of zero because that's a mathematical average, The how the averaging response in these works is we saw before the full wave bridge rectifier and then it's calibrated to give you the the average result based on a sine wave. But what happens when you don't use a sine wave? Well, we have the handy table here.

From the datasheet, you can actually, um, calculate these yourself. So for an undistorted or perfect sine wave here, um, we should of course get exactly the same value this zero percent error between an Rms meter and an average responding meter because these are calibrated for a sine wave so you get zero percent error. But if we change this to a symmetrical square wave ie. 5050 duty cycle square wave, then we should get an error of plus 11 percent and sine to square.
it should stay at one volt Rms. This remains the same. Bingo. There's our 11 error very close to it.

So a triangle wave, We're looking at a minus a 3.8 percent error, so it should be negative. Uh, yep, that's about a a negative uh, 3.8 percent error. Is it not? Now a lot of people actually make the mistake of thinking that a perfect sine wave actually has a crest factor of one. Now of course, the crest factor is uh, the peak value divided by the Rms value the biggest.

The peak to Rms value um, is, of course 1.4104 You should know that for your peak to Rms conversions, only a square wave actually has a perfect crest factor. A good data sheet for a meter will actually have and specify the maximum crest factor for its a true Rms measurement chip. Let's play around with some other waveforms here. Let's do a Prbs which is a pseudo random Yeah, we're not going to bother to calculate it, but both are reading the same and it's reading highest.

So we'll see what that measures after the modification. And I've got a pulse here, which is, uh, set to one millisecond, uh, pulse time. And as you can see, even the fluke 87.5 can't handle that horrible uh, crest factor there. So it's 0.565 But the average responding meters are even worse.

But it'll be interesting to see after modification if this goes up to match the true Rms meter. So let's solder in the capacitor onto the handy pad which we have down here. It's even marked fantastically. This positive on this side.

Go to my Avx Tantalum sample kit. Very handy. Now the Fluke 87.5 schematic and parts list says it's a 33 mic. Same as the analog devices data sheet.

It says it's a 16 volt, 300 milliohms the best I've got down here. That's not a biggie. Um, is this 33 mic? that's a B case and that's only 10 volts. But considering that this is a 9 volt battery, it'll be good enough for Australia I think so.

I don't know what the Esr this is. I don't think it's going to matter, so let's use one of those. All right, let's get some freshy on there, shall we? I think that's uh, that could be an A. But anyway, got the polarity correct there? No wackers.

She'll be right, because tantalums, unlike electrolytic capacitors, they have the mark up there for the anode um, not the cathode. so the positive, whereas electrolytic caps, um, have it have the black mark on the uh. On the negative side, that's rather annoying. But there we go.

No workers. All right. We have our brand new Fluke 177 meter here. However, now that I think about it, um, before I cross my fingers and power this up, I don't think we're going to get away with this without recalibrating this meter because even though it's calibrated with a sine wave, the internal scale factor of the conversion is going to be different.
So I reckon we're going to end up with um, the scale factor of uh for the Rmf 0.7 which is uh 0.707 divided by the yeah divided by the Dc average, which is uh 0.636 So I reckon we're going to end up with an error of 1.1 Have I got that in the right direction? I think So. I think we're going to be. It should read up by Um 11. Yes, same as it does with a square wave, isn't it? So let's power this sucker on.

of course. it's going to work. It doesn't say 177? That would be nice. Um, so yeah.

I don't know. Maybe we can mod the firmware something like that anyway. and if you mod the firmware, by the way, we could turn this into a 179 by adding temperature measurement because I think it's all in there. But yeah, it's just the firmware you pay extra for anyway.

Um, yeah, let's plug it in. Here we go. I got my leads. Bingo.

1.11 So yeah, um, before I just measure the other stuff, I'll go to the calibration, uh, manual and uh, we'll have to enter the cal mode on the back and uh, we'll have to recalibrate. Now to do this, you got to switch it to Millivolts over here and then you got a probe in the backside. Oh, yep, There we go. Got it.

Tada, we're in now. Unfortunately, the steps we want are the Ac volt steps here six and seven, and also, of course, um, amps as well. So you've got, uh, Ac down here for the 400 milliamps and the 6 amp range as well. Now I'm hoping that we can actually bypass these other steps because I don't want to have to actually calibrate all the other ranges.

I just want to be able to calibrate that one. Um, yeah, I'm going to see if I can like bypass this and only calibrate the ranges we want. I bet ya, Murphy says we can't do that. Yeah, okay right.

So this reads the live reading on the input. It's uncalibrated. Okay, so press and hold this button to display the required input min max. So there you go.

So it's telling us we have to feed in 600 millivolts. Dc great, but it says press the yellow button, store calibration advanced to the next step. This button is also used to exit calibration mode. Can we just like think I'm going to be forced to actually calibrate this entire meter? Damn it.

Now I've got nothing plugged into it and if I press that to go to the next one, it just double beeps at me. So it's smart enough to know no, you idiot. Um, you haven't got anything plugged in and if I maybe try and hold that down, Does that do anything? Nope. Does range do anything? No, Ah nah nah, I'm I'm forced into it.

I don't know. Back Light Yeah. the backlight still independently works. Um, no, no, I'm forced to calibrate this whole damn thing.
Damn it. So anyway, lucky I have my uh calibrators over here Ac and Dc and well, I've got various Olmski things and like you know, this is not a high spec meter and of course I can compare it against my Uh seven and a half digit uh, job is over here. The bar graph actually still works. It does the business.

So there you go. So I'm feeding in 600 millivolts. Um, that's what we have to feed in and that's what we're live reading. uh, at the moment.

but of course it's uncalibrated. Um, it says it goes into uncalibrated mode when it shows you that live reading, but we know we're feeding in, uh, the precise value and that, well, there it is over on there. That's good enough for Australia, definitely. So I whacked that in and Boom! we go to step two.

Anyway, I'm up to the 60 volt step and uh, well, that only goes up to 10 volts. So yeah, I'm going to have to use some of my other standards over here high voltage, ah, supply. So I'm up to step six. I switched over to my Ac volts uh, standard here and I can generate both required 600 millivolts at 60 hertz and also 600 volts or 60 Hertz and not everyone.

unfortunately is going to have, um, this bit of kit. and likewise, here we have 660 volts live. but I'm feeding in 600. so I'm gonna calibrate that sucker and we're up to Owmski.

Well, dumbass Dave tried to cheat, didn't He and I exited that cow procedure thinking it would have stored all those previous steps in the E-squared prom. And Bob's your uncle, right? I wouldn't have to. I could just like get a quick result just to show that it worked and I could do the current range later. Um, it? Yeah, Nah.

winner winner chicken dinner. I got it. Um, it was not easy getting. Uh, the Ac current source.

Uh, particularly the six amp Ac at 60 Hertz current source. But I was able to, uh, cobble it together here in the lab. and I got it. Um, and here it is.

It's how true Rms. There you go. One volt Rms. Um, it's recalibrated.

You see how it was off before it's now recalibrated. So now we can fiddle around to see if it matches this. So let's go wave form. Let's go.

Our square wave. Oh, look at that. like a bought one. Like a bought one True Rms.

Not that average responding rubbish ramp. Look at that. It was matching this before. Now it's matching this ah, Bobby Dazzler triangle wave we're getting before.

How about that pseudo-random binary sequence we got before? That's not I'm I'm still going to call that. I mean, this might have a different response to this, perhaps, but it's certainly not. Um, higher like we'll get him before. So that's all right.

There we go. Got an exponential rise function there? Um, they're just the first one that came off the rank and um. yep. Sure enough, there's that one millisecond pulse we had before that's not too shabby.

So there you go. I. I did it. I converted a Fluke 77 series 4 into a Fluke 177.
I don't think there's any other differences really. I think it's just that the 170, all of the 170 series every model is a true Rms, whereas the original 7 series was average responding, but it has the true Rms capable chip in it. Don't recommend doing this at home unless you have the ability to calibrate all those ranges. So if you know of a way to skip those steps in there because that's annoying because sometimes you might just have like one range that's out and you just want to fix that one range.

In this case, that would have been really nice. That being said, it's not easy to get. uh, six amps, 50 Hertz? You know you have to like bodge something with a transformer and a you know, big variable resistor load or something like that. Um, which is basically how I got it.

I used my um, variable frequency um Ac and it it was four amps maximum um on the nameplate but I managed to get six amps out of it and I just put it into a two ohm load and I just adjusted the output voltage until I um got the you know near enough and it measured on the seven half digit meter and Bob's your uncle I was only get it but yeah, you got to complete every single step and then it says end, uh on the end of it and then you press the yellow button and then just boop. It goes back to normal and it doesn't so you can't just go halfway through or do an individual step if you know how. I'll leave it in the comments and you should be able to do that. I think with the Uh 80 series and probably the Uh 20 series as well the 2728.

So yeah, if you do know, let us know in the comments down below. but I hope you found that interesting and useful fitted. Please give it a big a thumbs up as always comment down below. Catch you next time you.


Avatar photo

By YTB

5 thoughts on “Eevblog 1448 – convert a fluke 77 iv to true rms for 10 cents!”
  1. Avataaar/Circle Created with python_avatars WacKEDmaN says:

    i found the same thing the other day when looking for the 70 series… it all seems to be 170 series now…

  2. Avataaar/Circle Created with python_avatars the4armedmonk says:

    33

  3. Avataaar/Circle Created with python_avatars Victor Luis says:

    Greetings from brazil

  4. Avataaar/Circle Created with python_avatars Kendall Barboza says:

    Decime

  5. Avataaar/Circle Created with python_avatars armandine2 says:

    Good product placement for the EDU33212A – great to see a successful upgrade too.

Leave a Reply

Your email address will not be published. Required fields are marked *