DIY Analog ESR Meter design

[CMTan]

I have designed a analog ESR meter which is simple and buildable.  The original design was from Luden which I like because it is simple and technically sound.  However, there are several issue with the original design which I highlighted.  Being a design engineer myself, I have to correct the flaws and redesigned it, but the functional block remains intact.  Please refer to this blog if you want to build it.  I have also posted some question for anyone who want to think deeper into the issues and design.

https://electronicsdesignfun.blogspot.com/p/esr-meter.html

This is a work in progress blog, so I will update when put everything together.  For now, you see the evaluation and the final schematics.

Let me know if you have questions.

[mawyatt]

Interesting project!!

One question, since you are creating the signal of interest and "transmitting" such at one probe tip, why not synchronously demodulate on the receiving section probe? The benefits of synchronous technique are ideal for this application and very simple to implement.

Best,

[m3vuv]

I built that design a few years ago,sort of worked with mods,i found the low end of the scale to squashed up!,i then built the j_diddy 5 transistor design from here,that works perfect,actualy fed it with 12v and used a big 1ma fsd meter movement.https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/

[CMTan]

Hi Mawyatt,
Thank you for your question, but dont understand the question.
Could you elaborate more clearly what you meant by synchronous technique?  Perhaps a circuit will be helpful.

[CMTan]

Hi m3vuv,
The original design is suppose to have a mid scale at 10 ohms, but the actual build by Luden shows 5 ohms, so apparently the schematics on the website is not the final one used in his build up.  There are issue which I mentioned in my blog.  The main reason why you faced a compression on the low resistance scale was because the gain of 39 used was too large.  You need to reduce the gain to 20 (max) for his design.  With a high gain of 39, the op output will exceed max Vo swing resulting in amplitude compression.  I faced the same issue when I first build it, which lead me to investigate the design in more detail, then I found more issues with the design.  Even with that, you still need to change the op amp to TL082.  TL062 will work poorly because of the poor gain-bandwidth product.

[pete g]

would like to build your esr meter. could you possibly design a pc board for it?

[mawyatt]

Hi Mawyatt,
Thank you for your question, but dont understand the question.
Could you elaborate more clearly what you meant by synchronous technique?  Perhaps a circuit will be helpful.

There's some discussion on synchronous demodulation on this thread regarding an ESR meter, of course a Google search will provide much more detail and theory.

https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/

Best,

[RoGeorge]

Nice blog! 

About the ESR meter, voltage spikes on the primary of the transformer might be a problem.

Transformers are bidirectional, and this one has a ratio of 20.  This implies that, if a capacitor accidentally remains charged let's say at 24V, then the transformer will generate a spike of 20 x 12V = 240V in its primary turn, at the output of the first opamp. (12V x 20 and not 24V because of the two series 10 ohms resistors, R5 and R6, in parallel with the capacitor to be tested)

It happens that long time ago I've collected a bunch of ESR schematics found online, then extracted the measuring principle from each, being interested mostly in the "input stage", or how the capacitor under test is connected to the measuring circuit.  There were probably 20-30 ESR schematics in total, and the attachment is a draft of what I've scribbled back then.  In fact, there were only 5 types or so in all those collected schematics:



Synchronous detection will fit well in an ESR meter, indeed, as mawyatt was saying.  An example of such a schematics happens to be in the upper right corner of the attached draft scribbling, though that particular example is not directly applicable to your ESR.

To make a synchronous demodulator in principle, a multiplier will be needed, but any switch is also a multiplier with 0 or 1, so the schematic in the upper corner is using controlled switches (CMOS transmission gates from the 4066 series IIRC) to implement synchronous detection of the signal on the capacitor.  In other schematics/configurations of synchronous demodulators, switches can be implemented with diodes (e.g. a diodes ring demodulator), or with transistors as well.

[Jay_Diddy_B]

Hi CMTan and the group,

I suggest that you take a look at this thread:


https://www.eevblog.com/forum/projects/esr-meter-adapter-design-and-construction/

In the thread I share a design that uses a synchronous detector. It also has very good protection against applied voltages or charged capacitors.

Best regards,
Jay_Diddy_B

[CMTan]

Thanks Jay_Diddy_B,
I did not know this is called a synchronous design.  This design is very ingenious.  Basically it is inverting the amplifier on the differential amplifier to have either an inverting or non inverting gain, effectively serving as a AC to DC conversion so that there is no need for the AC to DC conversion stage.   Very smart indeed.

My design was meant to be simple, but more importantly to highlight the issue of Luden's design which could be the reason, some of you are not able to get it to work properly.  It is definite not as efficient as the synchronous design, because only one half of the signal is used for metering and requires a AC to DC conversion. 

[RoGeorge]

"Synchronous design" is usually a dedicated idiom used in digital design, and it is about something different than the trick used in the suggested ESR schematics.

The closest term for the trick in those switches-based ESR schematics is "synchronous rectification".  Depending on how one want wants to look at it, that trick can also be called a "synchronous detector", or "synchronous demodulation".

[CMTan]

Hi RoGeorge,
I totally agree with you.

[CMTan]

Thanks Jay_Diddy_B,
I have also reviewed at your 5 transistor design, and I love the simplicity of the design and the use of jelly bean components.

I plan to redo my design and use a similar functional block without the transformer, but instead of transistor, I will use only cheap op-amp (TL082) to create the differential amp.  I also want to improve the current capability of the metering which is relatively low in the original design with op amp buffering.

Thanks for sharing the 5 transistor design.   

[CMTan]

I am more interested in the design aspect than the construction aspect.  That's the reason I build it on breadboard since it is only one-off for me.  You may want to attempt to design the PCB, I have updated the design which requires no transformer now and also it has a wider lower range similar to the 5 transistor design by Jay_Diddy_B.

[kripton2035]

if possible you should add some protection to your design.
some people measure the esr of 400v charged capacitors in power supplies ...

[CMTan]

Thanks kripton2035,
I have added 2 TVS diodes on the test points connected back to back.  In the circuit, it is showing 1N4007, but in the actual I used TVS diodes which are meant for high instantaneous discharge.  Is that enough?  What is your suggestion if it is not.

[Jay_Diddy_B]

CMTan,
The back-to-back diodes provide shunt protection by rapidly discharging the capacitor being tested. By AC coupling the source and the sensing circuits, series protection can be used to protect against applied voltage.
In this thread I shared another ESR meter using 6 transistors. It has an output proportional to ESR.

Link: https://www.eevblog.com/forum/projects/10-transistor-esr-meter-adapter/msg3033064/#msg3033064

This design uses sinewave excitation.

Regards,
Jay_Diddy_B

[CMTan]

Thanks Jay_Diddy_B,
Thanks for explaining the protection scheme, they are already in the original design which I did not clearly explain.

I could not find the 6 transistor design with output proportional to ESR.

I saw this which has the same concept of having output proportional to ESR but using synchronous rectification.
https://www.eevblog.com/forum/projects/10-transistor-esr-meter-adapter/msg3033064/#msg3033064

Let me know where is the 6 transistor design with output proportional to ESR.

[CMTan]

Jay_Diddy_B,
I found the 6 transistor design.  Thanks.

Could you explain how you achieve this linear transfer function from ESR to Voltage?


Is the schematic marked in red box to tweak the linearity of the transfer function?

[Jay_Diddy_B]

Hi CMTan and the group,

I will use LTspice to illustrate how the left hand side, the ac current source of the 6 transistor ESR meter works.



The dc current in Q2 collector is determined by R2,R3 and R6. With V1 = 9V, the voltage across R2 is 9V x (R2/(R2 + R3)) = 1.8V. The voltage across R6 = 1.8V - Vbe = 1.2V. Q2 emitter current = 1.2/200 = 6mA. Q2 collector current = Q2 emitter current. Assuming infinite gain.

Q1, Q3, L1 and C1 form a 100kHz oscillator that modulates the current source.

The section Q1 - Q3 can be simplified to:



Q4 Circuit



The circuit around Q4 is to act as self-biasing current sink to provide a path for the dc current in Q2 collector.

The current from the source is separated, the DC component flows through the Q4 circuit and the AC component flows through a coupling network to the capacitor being tested.
A small portion of the AC signal flows in Q4. The equivalent resistance of the circuit is about 33k Ohms.

The left side of the circuit can be simplified to:



It is an ideal current source with an output resistance of 6.67k Ohms.

This is very large compared to the values of the ESR being measured.

Regards,
Jay_Diddy_B

[CMTan]

Thanks Jay_Diddy_B,
It's interesting how you create the constant current AC source which is what I expect to be done to have a linear transfer function from ESR to voltage.  Great to learn something from you.

[cnqhdszq]

This design（Beijibing ESR meter design） is better for members ,but there are 2  transformers in it .
This design have 7 transistors,  Any general purpose transistors can be OK! The diodes  are  Ge diode or Schottky diodes  .
The power  is 1 AA 1.5V battery（it  can work under 1.2V so you may use NI-Cadium rechargeble  battery ）, 80mA 。You can change the middle scale by replace the R6( or R7 ) .  the value of R6( or R7) is  the middle scale value .
The maximum voltage on the capacitor under test is 70mV. This allows for in-circuit testing
The capacitor  can be  tested without discharged.



http://www.crystalradio.cn/forum.php?mod=viewthread&tid=668385&highlight=%B3%AC%BC%B6%2B%2BESR