Equipement list for a professional level lab

[LabSpokane]

But none of us know what you want to do with the gear and what your requirements are for accuracy or range.

And neither does he, I'm sure.  Most of us don't get an email from the future exactly describing the problems we will face with an indented BOM of what we'll need to solve it. You must be blessed with more prescience than the rest of us.

[EEVblog]

I find the diversity of opinions on this matter quite interesting....doesn't seem like there is really a cut and dry answer and perhaps the original question is "too cut and dry"    Interesting either way...

For me the answer depends on if it's a fixed lab budget or not.
If you know you have a $$$$$ to spend then you try and cram as much potentially useful stuff in there as possible. i.e. I'd rather have a spectrum analyser and a scope, than just a kick arse scope and no spectrum analyser.
If it's "tell us what you want and how much it will cost" then it's a bit trickier. My approach there is to list all the different gear I want, and then upscale a few of those items, e.g. a better scope at double the price. If they come back and say "total is too expensive" (they never say "too much gear", or "this line item costs too much") then you scale back to the original scope you were happy with to begin with and save some big bucks and you look like a hero.
But if you don't have that "wank buffer" in there then you risk losing out on gear or what you want/need when they hard ball you. You need that room to move down a bit and still get what you want.

[TunerSandwich]

I find the diversity of opinions on this matter quite interesting....doesn't seem like there is really a cut and dry answer and perhaps the original question is "too cut and dry"    Interesting either way...

For me the answer depends on if it's a fixed lab budget or not.
If you know you have a $$$$$ to spend then you try and cram as much potentially useful stuff in there as possible. i.e. I'd rather have a spectrum analyser and a scope, than just a kick arse scope and no spectrum analyser.
If it's "tell us what you want and how much it will cost" then it's a bit trickier. My approach there is to list all the different gear I want, and then upscale a few of those items, e.g. a better scope at double the price. If they come back and say "total is too expensive" (they never say "too much gear", or "this line item costs too much") then you scale back to the original scope you were happy with to begin with and save some big bucks and you look like a hero.
But if you don't have that "wank buffer" in there then you risk losing out on gear or what you want/need when they hard ball you. You need that room to move down a bit and still get what you want.

Of course, that is how ALL hand me down budgets work....but the real question is what does one actually "need" to accomplish the goals set by the bean counters.....that is the part that most people seem to miss.....what I really need is a xyz box so my abc widget doesn't tear someones face off.....but someone else is paying for it, so fuck it....imma buy hookers and blow 

[Howardlong]

And it ain't just having some snazzy T&M gear. You need to know how to use it too. That is what is really concerning me about this thread, it sounds dangerously to me like it might just be a trophy lab.

Haven't read the whole thread, but from what I gather the OP has been asked by his company set up a lab using decent gear, and that's what he's looking for.
Nothing wrong with that, and not a "trophy lab" at all. This happens all the time in industry. Usually you are given a budget to spend on a new lab and they leave it up to you to decide what you want.
At very large companies we'd often get that with every new major contract. They win a new contract and $$$$$ is set aside for a new lab gear to do it.

Thanks Dave, its exactly that ! 

If that is the case, then the contract will detail the project, and you'll have a budget. Right now there is little project detail and no budget specified.

All we know is it's a remote control in the 433MHz ISM band. Is it just the transmit side, or is it both ends? Is it bidirectional? If it's in the 433MHz ISM band then it will not be wide band or high bit rate. What is the interface to the radio? Are you designing more than just the RF bit, and if so, what is the nature of that? Is this an R&D or production project?

I'm now wondering if the contract is yet to be awarded and as part of the RFP the prospective customer needs to know what's in your lab (or isn't, as the case may be!)

[tszaboo]

Maybe it is better to get an annual equipment budget. There is sense in buying equipment when you need it because at that point you'll know what you need.

Not to mention, if you need something like twice a year for a week, like RF precompliance testing, it makes more sense to rent it for the two week than actually buy it. That is why the budget would be better.

[G0HZU]

But none of us know what you want to do with the gear and what your requirements are for accuracy or range.

And neither does he, I'm sure.  Most of us don't get an email from the future exactly describing the problems we will face with an indented BOM of what we'll need to solve it. You must be blessed with more prescience than the rest of us.

Maybe you can see more into the future than me? I was merely suggesting caution with the budget until a better sense of direction was known.

But you are the one suggesting (in post #19) he spunks up nearly £2000 for a pair of exotic DMMs and Benchvue. If someone bought those meters for our R&D labs they would sit there on a storeroom shelf unloved (too big/bulky and non portable with a cluttered UI) and would eventually end up on a test engineer's bench as an expensive bit of eye candy.

[Lagaffe]

If that is the case, then the contract will detail the project, and you'll have a budget. Right now there is little project detail and no budget specified.

All we know is it's a remote control in the 433MHz ISM band. Is it just the transmit side, or is it both ends? Is it bidirectional? If it's in the 433MHz ISM band then it will not be wide band or high bit rate. What is the interface to the radio? Are you designing more than just the RF bit, and if so, what is the nature of that? Is this an R&D or production project?

I'm now wondering if the contract is yet to be awarded and as part of the RFP the prospective customer needs to know what's in your lab (or isn't, as the case may be!)

I never said it was for some contract or outsourcing or anything like that ... I also never mentioned that this was/is a RF project.

If you read back some pages, you'll find that I've said:
"The equipment will be used mainly for robotics and automation, but this includes almost everything you have in electronics. Starting from batteries, you have analog and power electronics, going to some gyros and accelerometers and you have digital and signal processing, making something remote controlled and you have RF ... sensors and actuators and there will be some network (CAN, most likely) ... and so it goes. Even for me it is difficult to make a very specific topic 

A scope with a large bw is good, but I guess I'll go for a scope with good sampling rate and memory depth ... at least for now. But with a bw >= 200Mhz"

Even before this, I also said:
Research, design, prototyping, everthing from scrach till a final product. RF will be used, but I guess for some ASK 433.92MHz remote control ... nothing on the GHz side of spectrum.

Of course, I haven't been so specific about this project, and I can't ... yah company policies suck, I know !

About the RF part, I'm not even considering designing it, I don't think there will be a need for a high bit rate or even transmitting power. And for that I can get some certified RF modules (go to any electronic suppliers company and you get it by the dozen!) ... and if they are already certified, one less thing to worry. I even have dropped the option of getting a spectrum analyzer, and instead renting one, or just have it integrated on the scope.

Are you woring about the final value for all the gear that I've mentioned? ... well, the mechanical part of the thing is at least one order of magnitude higher. If someone, at the company, gave you a rough idea of what they are planning to start developing, and ask you "make a list of good quality equipment that you need" ... what would you do?, wait until you have a list of all the specifics an then (only then) present a list of equipment ? ... ask for some rudimentar gear because you don't really know what you need, and you don't want the company to spend money on "fullish things" ? Ask for a future budjet ... and hopping that when you'll need something this budjet still exist ?

I don't know about you. As for me, I'll try to get the most diversified equipment that I can possible get, at the end, they might say "too expensive" and then, I'll try to find something "less expensive". If I don't ask, I'll not get! IMO, its better to bargain and get something, than let the oportunity pass and regreat it later. And no this is not a trophy lab !

[LabSpokane]

But you are the one suggesting (in post #19) he spunks up nearly £2000 for a pair of exotic DMMs and Benchvue. If someone bought those meters for our R&D labs they would sit there on a storeroom shelf unloved (too big/bulky and non portable with a cluttered UI) and would eventually end up on a test engineer's bench as an expensive bit of eye candy.

A recommendation based on using those meters in the past and today. I'm looking at a pair of said "trophy meters" right now. I love them and use them frequently.  They're great general purpose tools.  Paid for out of my own pocket I like them so much. I have a pair of handhelds as well. Each serves a purpose and neither can completely replace the other.

The OP asked for advice. Instead, you have decided to pass judgement.  Your present job is not representative of everyone's.  Your work environment and managerial directives are not universally representative either.  And the static signals you use your handheld DMM to measure with such great success might not be the signals the OP's robot is barfing out.

And once again, the cost of this stuff is trivial in the big picture.

[Howardlong]

If that is the case, then the contract will detail the project, and you'll have a budget. Right now there is little project detail and no budget specified.

All we know is it's a remote control in the 433MHz ISM band. Is it just the transmit side, or is it both ends? Is it bidirectional? If it's in the 433MHz ISM band then it will not be wide band or high bit rate. What is the interface to the radio? Are you designing more than just the RF bit, and if so, what is the nature of that? Is this an R&D or production project?

I'm now wondering if the contract is yet to be awarded and as part of the RFP the prospective customer needs to know what's in your lab (or isn't, as the case may be!)

I never said it was for some contract or outsourcing or anything like that ... I also never mentioned that this was/is a RF project.

If you read back some pages, you'll find that I've said:
"The equipment will be used mainly for robotics and automation, but this includes almost everything you have in electronics. Starting from batteries, you have analog and power electronics, going to some gyros and accelerometers and you have digital and signal processing, making something remote controlled and you have RF ... sensors and actuators and there will be some network (CAN, most likely) ... and so it goes. Even for me it is difficult to make a very specific topic 

...

Even before this, I also said:
Research, design, prototyping, everthing from scrach till a final product. RF will be used, but I guess for some ASK 433.92MHz remote control ... nothing on the GHz side of spectrum.

Of course, I haven't been so specific about this project, and I can't ... yah company policies suck, I know !

The first description you gave of the project was regarding a 433MHz remote control, so I'd assumed this was the fundamental part of the project. I do fully accept that I did miss the other description about robotics and automation part.

And no this is not a trophy lab !

Don't fight it! Tere are plenty of us with trophy labs on this forum! (Maybe we've just coined a new phrase too).

[nctnico]

And you don't have to use every piece of equipment every day for it to pay for itself.

[Dave Turner]

When my bosses used to ask me how long something would take I'd always use the 'Scotty' principle - double what I thought likely and add a bit - it's amazing how practical that works out in practice.

The same was true of the equipment budget; double what you really think and add a bit. The bean counters will bleat and either you'll end up with just enough to do the work or with a bit of luck your lab will truly gain ability. Depending on circumstance I'd always lobby for a yearly equipment budget for replacement/upgrade plus a slush fund.

It worked for me in a company network environment where historically a lack of sufficient yearly investment caused a nasty financial shock when having to completely reinvest due to hopelessly outdated equipment.

[G0HZU]

And the static signals you use your handheld DMM to measure with such great success might not be the signals the OP's robot is barfing out.

That's why I'd expect an engineer to design adequate BITE into a system such that they didn't need lots of lab grade DMMs for logging etc.

I really don't think we would want to rely on BenchVue for stuff like this either. That stuff (i.e. Benchvue and the exotic DMM) is better suited to the test department or maybe customer support if someone wanted to do some basic logging of a faulty unit without needing much in the way of programming skills.

The OP asked for advice. Instead, you have decided to pass judgement.

I gave pretty good advice in my first post. Most of my other posts have been trying to preach common sense about how little extra value there is to be gained from buying expensive bench DMMs and calibration standards

[LabSpokane]

And what do you do when you're stuck with someone else's design and have to characterize and integrate it Mr. Smartypants? 

============================================================================

And for the love of all that is Holy, do not confuse the me or anyone else here with TunerSandwich. 

[G0HZU]

And what do you do when you're stuck with someone else's design and have to characterize and integrate it Mr. Smartypants?

Most stuff comes with some kind of an ICD and if it doesn't then I really don't see how having a 6.5 digit bench DMM is going to add much value if you had to reverse engineer something.

[LabSpokane]

And what do you do when you're stuck with someone else's design and have to characterize and integrate it Mr. Smartypants?

Most stuff comes with some kind of an ICD and if it doesn't then I really don't see how having a 6.5 digit bench DMM is going to add much value if you had to reverse engineer something.

You are so fortunate to have everything work according to the data sheet.  Many of us yearn to reach your Utopian plane. 

[Howardlong]

Have to say I'm with G0HZU on this, I've never been able to figure out practically speaking why you need a precision DMM to 6.5 digits unless it's for some certification box ticking.

I can however understand why you need accurate frequency sources to rubidium standards, there is a very practical use for those in RF particularly in the GHz.

If it's not for procedural form filling, what do you do with this DMM accuracy? It's not a criticism, I'm really just genuinely interested to know.

[LabSpokane]

You two are absolutely fixated on the number of digits of the meter. The accuracy is a side benefit for the few times that its needed (characterizing sensor drift is one time it's handy). It wasn't a primary purchasing criteria.  It's the functionality, display capabilities, and refresh rate that are what makes the reviled, opulent, TrophyMeter9000 a good tool.  It's just hilarious to me how much you two are railing on me for this meter. In between posts today, I've been working on an RTC and using the charting to quickly see what's going on with the battery bus. I can get a nice real time chart of its behavior in seconds after hooking up the probes. Meanwhile, my scope is free to pull data off the I2C bus.

If I had used my handheld DMM for this purpose I couldn't have seen what was happening because it's update rate is too slow. And I would have needed to record and plot the data separately. The all in one device is just faster, better, and more convenient.

Next, I'm sure you're going to tell me how wrong and unnecessary it is to have serial decoding on a scope. 

[Howardlong]

Ok, so,it's not the accuracy, it's the logging capability.

Thank you for filling me in.

I get serial decoding, been using it all day myself in fact on my Infiniium trophy scope.

[LabSpokane]

The 34461a is essentially a very slow, high-accuracy DSO with a really poor UI (for a DSO). It's very useful for me and allows me to maximize my time solving problems, not charting data.

[G0HZU]

You two are absolutely fixated on the number of digits of the meter. The accuracy is a side benefit for the few times that its needed (characterizing sensor drift is one time it's handy). It wasn't a primary purchasing criteria.  It's the functionality, display capabilities, and refresh rate that are what makes the reviled, opulent, TrophyMeter9000 a good tool.  It's just hilarious to me how much you two are railing on me for this meter. In between posts today, I've been working on an RTC and using the charting to quickly see what's going on with the battery bus. I can get a nice real time chart of its behavior in seconds after hooking up the probes. Meanwhile, my scope is free to pull data off the I2C bus.

If I had used my handheld DMM for this purpose I couldn't have seen what was happening because it's update rate is too slow. And I would have needed to record and plot the data separately. The all in one device is just faster, better, and more convenient.

 

It's not exactly research and development is it? Bench DMMs are great for T&M applications but experience of working in R&D labs (and watching the work patterns of hundreds of engineers over 25 years) has shown me that the demand for bench DMMs is very low.

However, I can understand why a decent DMM is needed for IC design and for critical design work for the development of formal T&M gear and for its versatility in a test or ATE environment but I suspect that your two meters are probably better than any DMMs we have ever used (as in needed) in our engineering labs in over 25 years.

As for high speed logging, we tend to leapfrog the DMM and use multiple fast ADCs that allow the captured data to be post processed using all kinds of tools including DSP, SW tools or spreadsheets etc. There could be dozens of things logged in parallel like this. To do this with DMMs would mean slower data capture rates and lots of £££ DMMs taking up lots of space...


 

[Lagaffe]

All this rattling about the 34461A reminds me of a conversation I had many (many) years ago, about a PC (Amstrad 1512, if you are in to those things) that had fit an optional hard drive, of a whooping 20MB (yes MB !!) ... and someone saying: "why you want such a big hard drive ? .. you will never be able to fill it!"

For those who are not happy with choosing a bench MM like the 34461A ... well, folks, you are all free to use whatever you can or want - that is why there are so many products on the market. If you are worried about the price, then sorry, but one only worries about the price when he can't afford it. And I'm even more sorry to tell you, that if you think 860€ is too much for a compan to spend on a quality multimeter like this one, well, what can I say, its better to change business, and grow potatoes ... or something.

[Lagaffe]

When my bosses used to ask me how long something would take I'd always use the 'Scotty' principle - double what I thought likely and add a bit - it's amazing how practical that works out in practice.

The same was true of the equipment budget; double what you really think and add a bit. The bean counters will bleat and either you'll end up with just enough to do the work or with a bit of luck your lab will truly gain ability. Depending on circumstance I'd always lobby for a yearly equipment budget for replacement/upgrade plus a slush fund.

It worked for me in a company network environment where historically a lack of sufficient yearly investment caused a nasty financial shock when having to completely reinvest due to hopelessly outdated equipment.

That "Scotty" principle is genious  ... I guess I'll start to use it too   

[Howardlong]

All this rattling about the 34461A reminds me of a conversation I had many (many) years ago, about a PC (Amstrad 1512, if you are in to those things) that had fit an optional hard drive, of a whooping 20MB (yes MB !!) ... and someone saying: "why you want such a big hard drive ? .. you will never be able to fill it!"

I do indeed remember those Amstrads, and PCs before them. I had one on my desk for a year or so at one company I worked at, as well as some Z80 cross assembly and Microsoft Pascal, I did some ORCAD PCB layout on one, we used to think it was amazing, even with its 640x200 CGA graphics.

For those who are not happy with choosing a bench MM like the 34461A ... well, folks, you are all free to use whatever you can or want - that is why there are so many products on the market. If you are worried about the price, then sorry, but one only worries about the price when he can't afford it. And I'm even more sorry to tell you, that if you think 860€ is too much for a compan to spend on a quality multimeter like this one, well, what can I say, its better to change business, and grow potatoes ... or something.

By that logic, I might as well go and buy myself a hairdryer. I am bald, but maybe one day it'll grow back and I might be able to use it one day. Or perhaps not.

As I run my own business, I buy all my own test equipment, so unlike your situation it's my money I'm spending and not someone else's. Perhaps I have a different way of looking at it. And that works both ways: if I can see value and RoI in having a piece of equipment for a project, than I will buy it there and then, I won't wait for some beancounter to sign it off. But I also won't buy it if I don't see a need for it, or there is better value elsewhere or doing it differently. I could be using it to go on holiday or something. 

That's why I questioning why you think you need it, it is not meant to put you on the defensive, if anything I might learn something about it, and perhaps change my mind and get one myself, you never know, I might have a use for one and didn't know it.