Test Equipment Anonymous (TEA) group therapy thread

[bd139]

Technically this is work and budgeted as it’s a commercial project. The only thing is I’m funding the initial costs and time myself. Have done it for a couple of software projects. This one is hybrid software and hardware so requires more toys.

[Specmaster]

Technically this is work and budgeted as it’s a commercial project. The only thing is I’m funding the initial costs and time myself. Have done it for a couple of software projects. This one is hybrid software and hardware so requires more toys.

All the more reason to ensure that it's reliable and capable of doing what you need or otherwise you'll be as frustrated as you are with your Rigol now. Either way it's your own money not your bosses, and I would want to get the best bang per buck I could as it will more than likely have to last you longer than a company owned one. Again the car scenario your Fiat is far older then company cars tend to be, they are at worst 4 years old.

Sent from my SM-J510FN using Tapatalk

[Berni]

Ah yeah i had a look on ebay and indeed the DSO6000 scopes are going a bit high lately. Tho a used X3000 is just as good if you patiently watch for a good deal to come around.

Probes aren't that big of a issue tho since the probes you get with these scopes are the lowest end probes they make. They can be had for a reasonable price and they are not bad probes at all. The more problematic are the digital MSO probes, those can be quite rare and expensive. On top of that there are multiple types, so you need to get the right one for your scope family.

As for buying a new model i think its reasonably safe with Keysight since they aren't known that much for shipping buggy products(Well sure they did have a occasional slip up but its rare). But on some of the chinese test gear manufacturers there have been some pretty major problems, sometimes they would get fixed a long time later with a software patch, some not and some might be hardware issues.

[bd139]

Exactly the point. The Rigol has a few annoying bugs. The Siglent scopes have hardware and software bugs. General Chinese support is abysmal. Once it’s been thrown over the fence you tend to have to live with it or get one or two word responses. If something from Keysight isn’t behaving, your chances of getting it fixed are more realistic.

Case in point, my DG1022Z arb has an interesting problem where at very slow (200mHz) square waves it gives sinusoidal wiggles on the leading edge for no apparent reason. Makes it bloody useless as a pulse generator. Waiting for a 33120A to come along instead.

[Berni]

Well on the topic of bugs the Hantek signal generator i had was quite an example. Hardware wise the signal gens are very nice, especially for the price.

I bought a brand new one and once i got it i found that it was dead on arival. No matter what you do there is no output signal and fails self test on FPGA. Via the seller i eventually got to someone at Hantek and they said that this is a hardware fault and needs to be returned. Eventually i convinced them to give me a firmware update file and after going to the procedure my signal gen came to life and passed self test. (Looks like they screwed up the FPGA image flashing at the factory and nobody noticed). After playing with it for some time i managed to find a combination of FM modulation settings that causes it to lock up and reboot. Because i set it to remember the last configuration meant that these settings got loaded in and crashed it before you can do anything so it was stuck in a eternal reboot loop. Eventually fixed it by doing a factory reset using a key combo. Then i found that most modulation settings cause weird bugs or cause it to output garbage. Sometimes you could get it to output the wrong frequency on long square wave settings. Occasionally i managed to press the output enable button, it caused it to light up but there was no output, so i pressed it again to disable it and press it again to enable it again, that time it lit up like before but also the relay clicked to actually enable the output. When you did get it to work how you wanted it still had a bad case of the typical DDS jitter at high frequencies.

After a few months i had enough and sold it to a coworker. He seamed happy with it tho.

[bd139]

You have more patience than me. I’d have widlarised it before it got passed on

[Specmaster]

I really don't believe that the SMD technology is as good as the TH tech is. Its primary reason is to reduce the amount of labour required in the production process and secondly to make someone far more money as result of less labour costs and also lower components cost. Yes there is also the spin off in as much as it reduces the physical size of the final product and that suits a lot of people, not least the consumer in a lot of cases.

It tends to get way to complicated for its own good, eg FPGA is a classic example and then there is level of software integration that a lot of things rely on to function correctly, a bit too much ripple on the supply line and it freezes or crashes out all together or other such matters that simply was not there a few years ago.

Things are just these days sometimes too complicated for their own good and as a result often fail in service. I love to go military airshows and these days becoming all too common at such events for a booked and advertised plane not to make it to the show at all because it went "tech" at the last minute.

Bd139 said awhile ago that his meter was a 4.5 digit one because that was adequate for his requirements and I tend to agree with that. I have a 6.5 digit meter and what do I tend to use the most?, either a 4 digit handheld or a 4.5 digit bench meter, why, because the 6.5 digit one is just way to complicated for general day to day use and in practise the extra resolution it affords in reality is not justified for the vast majority of users, it is a vanity thing, bragging rights etc. I got it because it looks very impressive all those digits and precision it affords but the fact is I don't need it, nothing I do justifies it in any shape or form but I still have and I expect that I'm far being the only one who can say that. 

[Mr. Scram]

Ah yeah i had a look on ebay and indeed the DSO6000 scopes are going a bit high lately. Tho a used X3000 is just as good if you patiently watch for a good deal to come around.

Probes aren't that big of a issue tho since the probes you get with these scopes are the lowest end probes they make. They can be had for a reasonable price and they are not bad probes at all. The more problematic are the digital MSO probes, those can be quite rare and expensive. On top of that there are multiple types, so you need to get the right one for your scope family.

As for buying a new model i think its reasonably safe with Keysight since they aren't known that much for shipping buggy products(Well sure they did have a occasional slip up but its rare). But on some of the chinese test gear manufacturers there have been some pretty major problems, sometimes they would get fixed a long time later with a software patch, some not and some might be hardware issues.

I learned about the history of the U1272A right after purchasing an older one.  Even though it seems to be an amazing meter, the repeated problems and recall don't do much for the confidence. At least they took the high road, maybe after a few "nudges" from the customers at large.

[bd139]

I really don't believe that the SMD technology is as good as the TH tech is. Its primary reason is to reduce the amount of labour required in the production process and secondly to make someone far more money as result of less labour costs and also lower components cost. Yes there is also the spin off in as much as it reduces the physical size of the final product and that suits a lot of people, not least the consumer in a lot of cases.

It tends to get way to complicated for its own good, eg FPGA is a classic example and then there is level of software integration that a lot of things rely on to function correctly, a bit too much ripple on the supply line and it freezes or crashes out all together or other such matters that simply was not there a few years ago.

Things are just these days sometimes too complicated for their own good and as a result often fail in service. I love to go military airshows and these days becoming all too common at such events for a booked and advertised plane not to make it to the show at all because it went "tech" at the last minute.

Bd139 said awhile ago that his meter was a 4.5 digit one because that was adequate for his requirements and I tend to agree with that. I have a 6.5 digit meter and what do I tend to use the most?, either a 4 digit handheld or a 4.5 digit bench meter, why, because the 6.5 digit one is just way to complicated for general day to day use and in practise the extra resolution it affords in reality is not justified for the vast majority of users, it is a vanity thing, bragging rights etc. I got it because it looks very impressive all those digits and precision it affords but the fact is I don't need it, nothing I do justifies it in any shape or form but I still have and I expect that I'm far being the only one who can say that. 

Well not strictly true with respect to SMD. It's a matter of frequency and parasitics more than anything else. You have to make stuff smaller and with less parasitic inductance (leads) and capacitance (encapsulation/construction) to push frequency up. A TH resistors is only a resistor for a relatively low frequency span. If you think of a typical RF termination for example, say 50 ohms, that TH resistor might have 1pF of capacitance. At 2.4GHz, that's ~70? impedance from that capacitor in parallel with your 50 ohm resistor. It's not 50 ohms any more and your transmission line is no longer terminated and/or you have an unexpected filter with the transmission line impedance. This is pretty heavily apparent in pretty much every circuit I've built upwards of about 50MHz. I built a couple of 2m (144MHz) receivers (DC+superhet single conversion) and even at that relatively low speed they are touchy as fuck about impedances. My little naff 40m DC receiver is even sensitive to shielding placement at 7MHz.

Your 1740A's hybrids are the other option which is basically an encapsulated SMD board. In fact some people have replaced hybrids in Philips scopes with a few SMD parts

Smaller, faster, less power, cheaper, more reliable. This is sacrificing only maintenance. I've got to jump on it because I can't argue with it. I prefer TH dead bug myself but it doesn't scale to the requirements of products these days.

FPGAs are quite reliable, well Xilinx ones anyway. We put them in flying exploding things. I rather like them if I'm honest - they also weigh a lot less than the 50 odd tubes of 74LS/HC logic I have in the cupboard. Disclaimer: when I was at university VHDL and FPGAs were the thing of the hour.

As for 4.5 digits, it's good enough for most cases. I'll pick up a 34401A if I see one going cheap but it's not because it's 6.5 digits but because it's Keysight/Agilent/HP 

Ah yeah i had a look on ebay and indeed the DSO6000 scopes are going a bit high lately. Tho a used X3000 is just as good if you patiently watch for a good deal to come around.

Probes aren't that big of a issue tho since the probes you get with these scopes are the lowest end probes they make. They can be had for a reasonable price and they are not bad probes at all. The more problematic are the digital MSO probes, those can be quite rare and expensive. On top of that there are multiple types, so you need to get the right one for your scope family.

As for buying a new model i think its reasonably safe with Keysight since they aren't known that much for shipping buggy products(Well sure they did have a occasional slip up but its rare). But on some of the chinese test gear manufacturers there have been some pretty major problems, sometimes they would get fixed a long time later with a software patch, some not and some might be hardware issues.

I learned about the history of the U1272A right after purchasing an older one.  Even though it seems to be an amazing meter, the repeated problems and recall don't do much for the confidence. At least they took the high road, maybe after a few "nudges" from the customers at large.

Any references for the U1272A problems? I got a U1241C so I'm worried now 

[Mr. Scram]

Old and new.

[nfmax]

Five years warranty from Keysight on the X2000 scopes

[bd139]

That’s pretty good.

£2700 over 5 years is £45 a month. Probably at least a £800 return at the end as well bringing aggregate cost to £32/month which is nothing.

[Specmaster]

Well not strictly true with respect to SMD. It's a matter of frequency and parasitics more than anything else. You have to make stuff smaller and with less parasitic inductance (leads) and capacitance (encapsulation/construction) to push frequency up. A TH resistors is only a resistor for a relatively low frequency span. If you think of a typical RF termination for example, say 50 ohms, that TH resistor might have 1pF of capacitance. At 2.4GHz, that's ~70? impedance from that capacitor in parallel with your 50 ohm resistor. It's not 50 ohms any more and your transmission line is no longer terminated and/or you have an unexpected filter with the transmission line impedance. This is pretty heavily apparent in pretty much every circuit I've built upwards of about 50MHz. I built a couple of 2m (144MHz) receivers (DC+superhet single conversion) and even at that relatively low speed they are touchy as fuck about impedances. My little naff 40m DC receiver is even sensitive to shielding placement at 7MHz.

Your 1740A's hybrids are the other option which is basically an encapsulated SMD board. In fact some people have replaced hybrids in Philips scopes with a few SMD parts

Smaller, faster, less power, cheaper, more reliable. This is sacrificing only maintenance. I've got to jump on it because I can't argue with it. I prefer TH dead bug myself but it doesn't scale to the requirements of products these days.

FPGAs are quite reliable, well Xilinx ones anyway. We put them in flying exploding things. I rather like them if I'm honest - they also weigh a lot less than the 50 odd tubes of 74LS/HC logic I have in the cupboard. Disclaimer: when I was at university VHDL and FPGAs were the thing of the hour.

As for 4.5 digits, it's good enough for most cases. I'll pick up a 34401A if I see one going cheap but it's not because it's 6.5 digits but because it's Keysight/Agilent/HP 

Maybe I got that wrong, as I don't get involved in HF stuff really I've never looked into it to appreciate the problems but when it comes down to audio equipment like amplifiers you don't expect to find that kind of frequencies anyway but you still find SMD there when its not really required.

[bd139]

That’s just because it’s smaller and cheaper there. Also some of the newer parts are SMD only.

[Specmaster]

That’s just because it’s smaller and cheaper there. Also some of the newer parts are SMD only.

Granted, but it's not cheaper to the end user I doubt [emoji41]

Sent from my SM-J510FN using Tapatalk

[Specmaster]

That’s just because it’s smaller and cheaper there. Also some of the newer parts are SMD only.

Granted, but it's not cheaper to the end user I doubt [emoji41]

Sent from my SM-J510FN using Tapatalk

I ordered a decimal thumbwheel switch on Ebay and I took delivery of a bcd version, bummer. The seller didn't even list a bcd one!

Sent from my SM-J510FN using Tapatalk

[Berni]

Any references for the U1272A problems? I got a U1241C so I'm worried now 

Yep thats the EMC problem they have. I don't think the  U1241C is affected by it but i know that the entire U127x series is.

I have the OLED version U1273A that is affected and Keysight responded quickly by offering people replacement units once they revise the PCB to fix it or a U1281A as a replacement. I picked the later and indeed received one in the mail a few weeks later. Amazing support from Keysight

[bd139]

I ordered a decimal thumbwheel switch on Ebay and I took delivery of a bcd version, bummer. The seller didn't even list a bcd one!

Claim-o-clock!

I ordered some shitty boost converters a while ago and got buck converters. Took me a while to work that one out

Yep thats the EMC problem they have. I don't think the  U1241C is affected by it but i know that the entire U127x series is.

I have the OLED version U1273A that is affected and Keysight responded quickly by offering people replacement units once they revise the PCB to fix it or a U1281A as a replacement. I picked the later and indeed received one in the mail a few weeks later. Amazing support from Keysight

And that's why you buy Keysight 

[tggzzz]

I really don't believe that the SMD technology is as good as the TH tech is. Its primary reason is to reduce the amount of labour required in the production process and secondly to make someone far more money as result of less labour costs and also lower components cost. Yes there is also the spin off in as much as it reduces the physical size of the final product and that suits a lot of people, not least the consumer in a lot of cases.

Thus spake someone that only works at low frequencies!

Compare, for example, the parasitic inductance in the leads of a PTH resistor or a DIP IC with that of SMD resistors and ICs.

Modern jellybean logic can easily switch 50mA per output in <1ns, a di/dt of 50MA/s. A DIP IC would have at least 10mm leads => 10nH => 500mV induced voltage per output. Of course it isn't quite as bad as that, but SMD is a definite advantage!

[Specmaster]

Claim has already been submitted, can't make a decade box with that combination.

[Specmaster]

I really don't believe that the SMD technology is as good as the TH tech is. Its primary reason is to reduce the amount of labour required in the production process and secondly to make someone far more money as result of less labour costs and also lower components cost. Yes there is also the spin off in as much as it reduces the physical size of the final product and that suits a lot of people, not least the consumer in a lot of cases.

Thus spake someone that only works at low frequencies!

Compare, for example, the parasitic inductance in the leads of a PTH resistor or a DIP IC with that of SMD resistors and ICs.

Modern jellybean logic can easily switch 50mA per output in <1ns, a di/dt of 50MA/s. A DIP IC would have at least 10mm leads => 10nH => 500mV induced voltage per output. Of course it isn't quite as bad as that, but SMD is a definite advantage!

Very true, I do work at low frequencies, I overheat at higher ones 

[tggzzz]

As for 4.5 digits, it's good enough for most cases.

They are like slide rules: schoolkids and undergrads have 12" slide rules, grad students have 24" sliderules, engineers have up to 34' sliderules (note the change of units!), but professors and chief engineers have 6" sliderules. http://www.johnwolff.id.au/calculators/sliderules/sliderules.htm#Cylindrical

Myself, I use a 3.5 digit handheld DMM for faultfinding, but I also have a 8 digit DMM just because.

[iainwhite]

Hopefully it’ll be launched before I buy

Refurb is a good option. Definitely a possibility.

Like a car, I personally wouldn't buy a new model that has just been released, you will suffer from all the niggling little bugs etc that new models have. Wait till   they have been tested in anger and problems resolved, or buy a previous model refurbished by the makers with a warranty for the same reasons, bugs have been sorted and you will have a good scope.

Sent from my SM-J510FN using Tapatalk

Do you think that Keysight might do an introductory package discount like R&S did for the RTB2004 ?  That might make the early-adopter risk a bit more acceptable.

[bd139]

As for 4.5 digits, it's good enough for most cases.

They are like slide rules: schoolkids and undergrads have 12" slide rules, grad students have 24" sliderules, engineers have up to 34' sliderules (note the change of units!), but professors and chief engineers have 6" sliderules. http://www.johnwolff.id.au/calculators/sliderules/sliderules.htm#Cylindrical

Myself, I use a 3.5 digit handheld DMM for faultfinding, but I also have a 8 digit DMM just because.

That's about right.

I've got a Fluke 8024B which I used for a lot of diagnostics for years. The continuity was orders of magnitude better than anything else out there. The U1241C has taken place however as it has as good as if not better continuity response as well as it auto ranges, doesn't predate CAT rating, has a screen which isn't slowly dying and doesn't have a spider living in the DC hole

Still really like my GW Instek. Super-happy with that purchase.

So many cool things around now. Spoiled for choice!

[Specmaster]

As for 4.5 digits, it's good enough for most cases.

They are like slide rules: schoolkids and undergrads have 12" slide rules, grad students have 24" sliderules, engineers have up to 34' sliderules (note the change of units!), but professors and chief engineers have 6" sliderules. http://www.johnwolff.id.au/calculators/sliderules/sliderules.htm#Cylindrical

Myself, I use a 3.5 digit handheld DMM for faultfinding, but I also have a 8 digit DMM just because.

That's about right.

I've got a Fluke 8024B which I used for a lot of diagnostics for years. The continuity was orders of magnitude better than anything else out there. The U1241C has taken place however as it has as good as if not better continuity response as well as it auto ranges, doesn't predate CAT rating, has a screen which isn't slowly dying and doesn't have a spider living in the DC hole

Still really like my GW Instek. Super-happy with that purchase.

So many cool things around now. Spoiled for choice!

So then, would that make you a professor or a chief engineer