It would be nice to measure a 100 nF, a 4 / 4.7 uF and 10 uF MLCC at 0 VDC and 12 and 40 VDC to see how much the capacitance value decreases. When they get the graphing fixed would be a good time to show this.
You are right, with such a screen, ALL graph should be displayable on it instead of using a spreadsheet. An improvement that will come for sure. Hopefully...
I checked out what R&S writes about the LCX100 and LCX200 and it looks like they are doing a full frontal assault on Keysight. The LCX100 is the direct competitor model to the E4980AL and LCX200 is trying to compete with the E4980A. If the LCX200 can actually hold up what it promises, then you get insane value. Quarter of the price of the E4980A and it even goes up to 10 MHz. If you're in the semicondutor fab business, then it'd be the LCR meter to get. IF it hold ups what it promises. I have a hunch they made these to specifically wrestle away Infineon in Munich as customer from Keysight.
Does the binning feature work with DC bias enabled? I couldn't tell from the UI. If it does, that'd be an extremely useful function for testing whether a particular capacitor part is suitable for a particular PDN scenario.
@@Thesignalpath In particular, if you could set a very specific frequency of interest and a DC bias, and then bin based on both capacitance and impedance, that'd be excellent for evaluating MLCCs for compact DC-DC designs. Even the best vendor databases like Murata's SimSurfing usually can't tell you what the combined tolerances are at a given frequency and DC offset, let alone provide a bell curve.
It would be interesting to see if the transformer option is usable to find problems like a shorted winding. It is besides that probably useable to see if a SMPS transformer is one with an airgap.
I am using BK Precision to measure the capacitance of my fabricated piezoelectric film. I am curious to choose the value of capacitance in terms of parallel or series capacitance values and parallel resistor value or series resistance value from the measurement. Also, I have noticed from the manual that a lower capacitor value needs a higher frequency for better accuracy. Can you please share some of your opinions on this matter? Thank you.
Shahriar, thank you for this content! Do the LCX200 and the E4980A share the exact same architecture, meaning are they both 100% based on auto balancing bridge architectures?
How do they achieve test frequency up to 10MHz? It's probably impossible in the simplest form of auto balancing bridge (op-amp as I-V converter). Is there some more complicated auto balancing bridge or some frequency down conversion? How does the thing work??
It is possible to use auto-balancing bride up to this frequency, it is the case for the Hioki IM3536 for example. On that instrument (Hioki) a specific TIA is used for each range. Under sampling ts the most common technique used today to deal with high frequency signals. This is done directly by the use of low jitter sampling ADCs.
@@lelabodemichel5162 What kind of op-amp could be used to ensure perfect balance at 10MHz without phase errors? It seems impossible. According to keysight Impedance Measurement Handbook the simple op-amp bridge is practical up to about 100kHz. Above that some weird tricks are used, like null detector with rf mixers for the bridge and frequency conversion further in the signal path. Example is described in service manual of HP 4192A LF. I wonder if the same technique or something else is in this meter.
@@asdfzxcv6268It is the case because I did the complete reverse engineering of the Hioki IM3536. Each TIA is complex, for some ranges it uses two or three op-amps with complex local feedbacks. The LCX uses a TIA for low frequencies only. I am not allowed to give details regarding the functioning of this device.
I'm curious if this LCR meter can measure the extremely low impedance provided by C0G capacitors such as a 10...56 nF 600V in 1210 case accurately. This is of interest when those capacitors are being used in LLC, LCC and similar switching power supply resonant networks with currents up to 5 Arms or more trough a single capacitor.
I suspect that the measurement fixture may be the limiting factor there, which may require a custom fixture to get the Kelvin connection right at the end caps of the DUT. This could be a simple PCB with some miniature RF connectors on it that the DUT gets soldered to.
@@Thesignalpath Capacitors down to a few pico farads might work. But as I wrote I'm interested in low ESR measurements in the range of 1 mOhm and even lower for C0G in the nF range to replace film capacitors in high power resonant circuits in LLC and LCC resonant power supplies.
The front panel suggests that the shields of the BNC are all coupled together and tied to ground, is that true? As far as I know the Agilent ones are not connected to ground such that a reversed current can be sent over the current guard terminals, to minimize the magnetic field out of the wires in order to minimize coupling from the current terminals to the potential terminals. Would you prefer this LCX200 over the E4980A (or: LCX100 over E4980AL)?
I can't speak for the Agilent, but I can confirm for you that the outer shieldings of the BNC terminals are all coupled and tied to the grounding jack at the front.
Hi Shahriar, I’m looking to buy a bench top LCR meter capable of measuring small (100fF) capacitors accurately. I was thinking of the Keysight E4980A but maybe there is a system from R&S or Hioki at a similar price that has as good or even better accuracy and/or better/more functionality. Could I ask what system you would recommend? Thanks
Nice video and device! Great that you pointed out where the software needs to be improved. I'd like to see more about what you can diagnose in-circuit with such an LCR meter. For example, could you detect a transformer with a shorted winding in circuit, etc.?
@Shahriar, as per @lelabodemichel5162 reply below: "The LCX uses a TIA for low frequencies only." Could you explore it further how the LCX works? That would be really interesting and there seems to be no such information available at this time.
To answer my unanswered question, for the sake of anyone else who orders a LCX but has to wait for their fixtures for a month: Don’t worry if you perform an “Open Correction - Full” and “Correction Failed!” comes up at 66% on the second sweep through with the terminals open. This is expected behaviour, and it will complete only with a fixture attached.
True, but my concern was more based around the strange behaviour of it still completing 1 round of the correction and then failing at a specific point during the 2nd round.@@lelabodemichel5162
I have The multifunctional Bode 100 Vector Network Analyzer from omicron lab , which It make most of these measurement and more with much better accuracy Could you please review it. I am sure you going to be like it😊😊
7:58 Another reason for putting the USB port on a separate board: you don't have to replace the whole computer if the USB port gets yoinked and damaged.
Looks like the same graphing GUI as their NGU401 series SMUs. In that case, I wouldn't expect much improvement given how long I've been waiting for them to drastically improve it. Hopefully I'm wrong.
@@ThesignalpathRoll off starting at 200-300 Hz about -4dB and shelve it there. The sibilants are there and OK, but there is too much low-mid which muffles your voice. This may not sound exactly like you, but the clarity improvement would be great.
i think the simple reason for this is, that those devices are very probably designed for some specific customers, which obviously need some features more than the others, so those features which you find as buggy are simply not thought of completely, because of lesser importance for the delivery of the product, so it starts to pay of asap
If the capacitor is nonlinear, C(V) depends on DCV, then there will be self mixing of the AC to yield a DC level. I predict that this might account for the discrepancy between the DCV measurements... depending on the capacitor's dielectric.
I love this video, my comment is not a complain, just some things that are my opinion. Maybe I'm wrong. (I collect old LCR and impedance bridges and my hobby is measuring components what I use for my work, repairing boards without schematics on component level) Why do you not use D in the primary field ? That is the standard used by most manufacturers for capacitors. Many decades that was the only option on LCR meters and bridges. 1 kHz was the standard frequency but since electrolitics became so big this moved to 100Hz (or 120 in the USA) The DC bias is usd for electrolitycs, so they never see a reversed voltage but for most for modern caramic SMD caps that can have high Volt-co's Just to be clear for readers, the instrument measures the loss angle and presents that as ESR or EPR (or DF). Cp is normally only used as the ratio between reactance and resistance becomes very big. It the example the 10k Rp measured is the parallel combination of the R of the resistor and the Rp of the cap itself. For a good cap D is very low so Rp is very high. That is the reason the meter showed Rp as 10k (Or it measurement functions are very advanced and it does multiple measurements) often Cp is only calculated from the Cs measurements, not measured. I think the term comes from the way bridges worked. If the DF was very high you switched to the Cp function where they changed the connection of components in the oposite side of the bridge so it was possible to zero the bridge again. That was a variable cap and a variable potentiometer. One function put them in series, the other parallel (the same for inductors) Rp is not the same as the leakage resistance used for electrolityc caps. There the leakage is DC leakage. Just for those who think this instrument can measure that too. Resonance it the point the reactance is zero, it has nothing to do with the ESR. It is a widespread misunderstanding the ESR peaks low at the SRF. The cause is probably a misunderstanding of the way LCR meters measure ESR (calculate ESR from the phase angle) . For those interested Dissipation Factor (D 1/Q, tan d, or calculate ESR from this) is the ratio between reactance and resistance. A modern LCR meter like this is an impedance meter. Just what I think: If you use the low output impedance option it is more like a VNA, if you use the standard high impedance it is an impedance meter. (or a VNA + IV option) The low impedance output should be the best choice for impedance around the output resistance (Like a VNA is best for measurements around 50 ohm)
To complete this answer this lcr meter is an impedance meter in all modes. The small source resistance permits to measure smaller impédances by using higher test current. All parameters are computed from the impedance like all lcr meters. Auto balancing technique is used below a certain frequency only. Above this frequency another technique is used.
It would be nice to measure a 100 nF, a 4 / 4.7 uF and 10 uF MLCC at 0 VDC and 12 and 40 VDC to see how much the capacitance value decreases. When they get the graphing fixed would be a good time to show this.
So clean and pretty!
Excellent review as always. Thanks.
I'm not the fan of Rohde & Schwarz but such an amazing review make me a fan of them
You are right, with such a screen, ALL graph should be displayable on it instead of using a spreadsheet. An improvement that will come for sure. Hopefully...
I checked out what R&S writes about the LCX100 and LCX200 and it looks like they are doing a full frontal assault on Keysight. The LCX100 is the direct competitor model to the E4980AL and LCX200 is trying to compete with the E4980A. If the LCX200 can actually hold up what it promises, then you get insane value. Quarter of the price of the E4980A and it even goes up to 10 MHz.
If you're in the semicondutor fab business, then it'd be the LCR meter to get. IF it hold ups what it promises. I have a hunch they made these to specifically wrestle away Infineon in Munich as customer from Keysight.
Nice video and a great tool for any lab! I wish I had such of think.
Thanks for the review
thanks.
very very good rode for all test equpment device
R&S have released a PC based application sweep tool for the LCX100, LCX200 which you may find interesting. 1GP132 released 21 Aug 2024.
Thanks!
Very inspiring
Does the binning feature work with DC bias enabled? I couldn't tell from the UI. If it does, that'd be an extremely useful function for testing whether a particular capacitor part is suitable for a particular PDN scenario.
Good point, I believe you can.
@@Thesignalpath In particular, if you could set a very specific frequency of interest and a DC bias, and then bin based on both capacitance and impedance, that'd be excellent for evaluating MLCCs for compact DC-DC designs. Even the best vendor databases like Murata's SimSurfing usually can't tell you what the combined tolerances are at a given frequency and DC offset, let alone provide a bell curve.
I have been caught out be MLCCs that degrade with age and DC bias.
isn't the text at about 28:00 the wall clock time the sweep was done?
It would be interesting to see if the transformer option is usable to find problems like a shorted winding. It is besides that probably useable to see if a SMPS transformer is one with an airgap.
I would hope so.
I am using BK Precision to measure the capacitance of my fabricated piezoelectric film. I am curious to choose the value of capacitance in terms of parallel or series capacitance values and parallel resistor value or series resistance value from the measurement. Also, I have noticed from the manual that a lower capacitor value needs a higher frequency for better accuracy. Can you please share some of your opinions on this matter? Thank you.
Is the distance between the front panel BNC connectors the same as on Hp/Agilent/Keysight meters, e.g. 7/8" or 22.2 mm?
Yes, these are standard spacing.
Shahriar, thank you for this content! Do the LCX200 and the E4980A share the exact same architecture, meaning are they both 100% based on auto balancing bridge architectures?
Where did he explain the function?
I watched #73 where he didn't explain a Wheatstone bridge.
That value you couldn't quite decipher was just the wall clock time.
Does anybody know what font is used on the instruments? It's very elegant and visible. Was it a custom job or something like Fira Sans?
Frutiger perhaps?
Possibly customized variants of Corpid (RS Corpid, RS Corpid Condensed), and a "RS-Condensed" by Ascender
I believe it's Linotype Univers 420 Condensed - that's one of our standard corporate fonts and it used in many of our products, manuals, setc.
Excellent instrument! Now where to get an extra 20 grand to buy one!
How do they achieve test frequency up to 10MHz? It's probably impossible in the simplest form of auto balancing bridge (op-amp as I-V converter). Is there some more complicated auto balancing bridge or some frequency down conversion? How does the thing work??
It is possible to use auto-balancing bride up to this frequency, it is the case for the Hioki IM3536 for example. On that instrument (Hioki) a specific TIA is used for each range.
Under sampling ts the most common technique used today to deal with high frequency signals. This is done directly by the use of low jitter sampling ADCs.
@@lelabodemichel5162 What kind of op-amp could be used to ensure perfect balance at 10MHz without phase errors? It seems impossible. According to keysight Impedance Measurement Handbook the simple op-amp bridge is practical up to about 100kHz. Above that some weird tricks are used, like null detector with rf mixers for the bridge and frequency conversion further in the signal path. Example is described in service manual of HP 4192A LF. I wonder if the same technique or something else is in this meter.
@@asdfzxcv6268It is the case because I did the complete reverse engineering of the Hioki IM3536. Each TIA is complex, for some ranges it uses two or three op-amps with complex local feedbacks.
The LCX uses a TIA for low frequencies only. I am not allowed to give details regarding the functioning of this device.
@@lelabodemichel5162 Hi, can you please comment on the technical level of Hioki LCR Meter in the industry?
I'm curious if this LCR meter can measure the extremely low impedance provided by C0G capacitors such as a 10...56 nF 600V in 1210 case accurately. This is of interest when those capacitors are being used in LLC, LCC and similar switching power supply resonant networks with currents up to 5 Arms or more trough a single capacitor.
I have measured down to picofarads without any issues.
I suspect that the measurement fixture may be the limiting factor there, which may require a custom fixture to get the Kelvin connection right at the end caps of the DUT. This could be a simple PCB with some miniature RF connectors on it that the DUT gets soldered to.
@@Thesignalpath Capacitors down to a few pico farads might work. But as I wrote I'm interested in low ESR measurements in the range of 1 mOhm and even lower for C0G in the nF range to replace film capacitors in high power resonant circuits in LLC and LCC resonant power supplies.
The front panel suggests that the shields of the BNC are all coupled together and tied to ground, is that true? As far as I know the Agilent ones are not connected to ground such that a reversed current can be sent over the current guard terminals, to minimize the magnetic field out of the wires in order to minimize coupling from the current terminals to the potential terminals.
Would you prefer this LCX200 over the E4980A (or: LCX100 over E4980AL)?
I can't speak for the Agilent, but I can confirm for you that the outer shieldings of the BNC terminals are all coupled and tied to the grounding jack at the front.
Hi Shahriar, I’m looking to buy a bench top LCR meter capable of measuring small (100fF) capacitors accurately. I was thinking of the Keysight E4980A but maybe there is a system from R&S or Hioki at a similar price that has as good or even better accuracy and/or better/more functionality. Could I ask what system you would recommend? Thanks
Nice video and device! Great that you pointed out where the software needs to be improved.
I'd like to see more about what you can diagnose in-circuit with such an LCR meter. For example, could you detect a transformer with a shorted winding in circuit, etc.?
Very nice. R & S always is. But for that price I think I'll need to stick to a NanoVNA, a HV supply, and my 'scope.
I see people say this all the time. "Ma" or my-noun. Like ma'democracy. Why?
Can you link the video you talked about auto balancing bridges?
I believe that would be TSP 193 ?
@Shahriar, as per @lelabodemichel5162 reply below: "The LCX uses a TIA for low frequencies only." Could you explore it further how the LCX works? That would be really interesting and there seems to be no such information available at this time.
Nice info, thanks :)
To answer my unanswered question, for the sake of anyone else who orders a LCX but has to wait for their fixtures for a month: Don’t worry if you perform an “Open Correction - Full” and “Correction Failed!” comes up at 66% on the second sweep through with the terminals open. This is expected behaviour, and it will complete only with a fixture attached.
If the test fixture is not connected Hpot is not connected to Hcur, therefore the measured voltage is totally wrong and the calibration fails.
True, but my concern was more based around the strange behaviour of it still completing 1 round of the correction and then failing at a specific point during the 2nd round.@@lelabodemichel5162
Did this one emerge from the Zurich instruments acquisition?
No - I have something else from Zurich I will show later on.
I have The multifunctional Bode 100 Vector Network Analyzer from omicron lab , which It make most of these measurement and more with much better accuracy
Could you please review it.
I am sure you going to be like it😊😊
7:46 the USB port is probably on a daughter board to reduce the repair cost in the event the USB port is damaged.
That also makes sense.
Compared to other R&S devices you've looked at, the firmware on this one seems unusually rough around the edges.
7:58 Another reason for putting the USB port on a separate board: you don't have to replace the whole computer if the USB port gets yoinked and damaged.
That is a very good point.
Nice to see an instrument at a reasonable price but the graphing needs some serious work right away...cheers and Happy Holidays !!
Looks like the same graphing GUI as their NGU401 series SMUs. In that case, I wouldn't expect much improvement given how long I've been waiting for them to drastically improve it. Hopefully I'm wrong.
15k€ reasonable price? I'd stick with 30-50yo junk from ebay for a while longer ;-D
Oh ? I saw it on there web site for 4 grand I will check !! LOL !@@hinz1
If only Shariar's mic quality were half as good as his test gear. My ears aren't great, but the audio top end seems to be more than a tad attenuated.
I am always looking for ways to make the recording better. I can adjust the equalizer. Are you saying the sound feel dull?
@@ThesignalpathRoll off starting at 200-300 Hz about -4dB and shelve it there. The sibilants are there and OK, but there is too much low-mid which muffles your voice. This may not sound exactly like you, but the clarity improvement would be great.
I've thought about mentioning it for months. Highs are missing, muffled sounding. But thanks for the excellent content!
Thanks - I’ll try to fix it going forward.
Why do R&S, Tek, and Keysight release >$20,000 "laboratory" test equipment with incomplete, missing feature and buggy internal software ???
Good one. Was also wondering about that. Feels like the software is not on par with the hardware design.
i think the simple reason for this is, that those devices are very probably designed for some specific customers, which obviously need some features more than the others, so those features which you find as buggy are simply not thought of completely, because of lesser importance for the delivery of the product, so it starts to pay of asap
@@timeltdme4355 Nice attempt at rationalization. They don't care and have no integrity in their work. Period.
@@timeltdme4355or, you know, they're just trying to put in the least amount of effort and make the most amount of money. you sound like an apple fan
👍👍👌👌
If the capacitor is nonlinear, C(V) depends on DCV, then there will be self mixing of the AC to yield a DC level. I predict that this might account for the discrepancy between the DCV measurements... depending on the capacitor's dielectric.
R&S make gorgeous instruments, sadly not affordable for the average person for home use.
I want one
it would be nice to see a bit more about the fixtures .. how they work and how they are constructed like te first you do ...
thx for the video.
We actually have an introductory video on the fixtures: ruclips.net/video/YdSY5ImW9A0/видео.html
I love this video, my comment is not a complain, just some things that are my opinion. Maybe I'm wrong. (I collect old LCR and impedance bridges and my hobby is measuring components what I use for my work, repairing boards without schematics on component level)
Why do you not use D in the primary field ? That is the standard used by most manufacturers for capacitors. Many decades that was the only option on LCR meters and bridges. 1 kHz was the standard frequency but since electrolitics became so big this moved to 100Hz (or 120 in the USA)
The DC bias is usd for electrolitycs, so they never see a reversed voltage but for most for modern caramic SMD caps that can have high Volt-co's
Just to be clear for readers, the instrument measures the loss angle and presents that as ESR or EPR (or DF). Cp is normally only used as the ratio between reactance and resistance becomes very big. It the example the 10k Rp measured is the parallel combination of the R of the resistor and the Rp of the cap itself. For a good cap D is very low so Rp is very high. That is the reason the meter showed Rp as 10k (Or it measurement functions are very advanced and it does multiple measurements) often Cp is only calculated from the Cs measurements, not measured. I think the term comes from the way bridges worked. If the DF was very high you switched to the Cp function where they changed the connection of components in the oposite side of the bridge so it was possible to zero the bridge again. That was a variable cap and a variable potentiometer. One function put them in series, the other parallel (the same for inductors)
Rp is not the same as the leakage resistance used for electrolityc caps. There the leakage is DC leakage. Just for those who think this instrument can measure that too.
Resonance it the point the reactance is zero, it has nothing to do with the ESR. It is a widespread misunderstanding the ESR peaks low at the SRF. The cause is probably a misunderstanding of the way LCR meters measure ESR (calculate ESR from the phase angle) . For those interested Dissipation Factor (D 1/Q, tan d, or calculate ESR from this) is the ratio between reactance and resistance. A modern LCR meter like this is an impedance meter.
Just what I think: If you use the low output impedance option it is more like a VNA, if you use the standard high impedance it is an impedance meter. (or a VNA + IV option) The low impedance output should be the best choice for impedance around the output resistance (Like a VNA is best for measurements around 50 ohm)
To complete this answer this lcr meter is an impedance meter in all modes. The small source resistance permits to measure smaller impédances by using higher test current. All parameters are computed from the impedance like all lcr meters. Auto balancing technique is used below a certain frequency only. Above this frequency another technique is used.
Interesting device. But as long as R&S is not hackable, it will only be an option for companies. 😕
another high end box I'll never use or likely see....