Featured
Latest
Share:
Notifications
Clear all

Forum 141

[Sticky] Homebrew ESR meter.

16 Posts
6 Users
0 Likes
7,143 Views
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

From a pile of components to a working ESR meter in a week (and almost zero cost).

I found a circuit on the Internet and built a prototype around two years ago which worked first time so I decided to build it in permanent form. When clearing out a workshop last year, amongst the bits and pieces was the remains of a home brew capacitance meter. The PCB was smashed and the meter was damaged. However the case was good and I had a 50uA meter that would fit perfectly. I decided that I would recycle the case and build the ESR meter into it.

I found a suitable sized piece of Veroboard and started building last weekend. Tonight the circuit was completed and tested and the results were as good as the prototype. I still need to finish the final assembly, tidy the wiring and make up a scale for the meter. The original circuit calls for a 500uA meter. I had a good new 50uA meter so added a shunt using a precision resistor (think it was 160 ohms).

This is a simple device that will show faulty electrolytics in a power supply without having to remove them from the circuit. Ideal for switched mode power supplies in modern equipment. I used the prototype many times to locate dodgy caps in DVD players and a couple of LCD tv's. More recently I used it on my own Freeview box that developed a problem due to faulty power supply caps.

Some pictures showing the progress so far. Hopefully (time permitting) the final assembly should be finished soon.

It's fair to say that an ESR meter is of no real use for vintage repairs and restoration. However they are almost essential for the repair of modern switch mode supplies as used in digi-boxes and DVD players which we now have to use with our vintage sets.

 
Posted : 27/07/2014 11:33 pm
Refugee
(@refugee)
Posts: 4055
Noble Member Registered
 

Have you asked Chris if he could host the circuit diagram in the library?
I have looked into making such a device myself following the process I went through to make a transistor tester almost 40 years ago but this time with access to many modern components.
The first thing it must be able to do is have a linear read out eliminating the need for a special meter scale to be made up. This will also open up the option of a cheap panel DMM for the display.
This would require a couple of op-amps and a FET extra.
To get the real thing it will also have to have adjustable frequency and ripple currant settings so that the data sheet values can be entered for the component under test.
As it is already testing reactive components It might as well go the whole way and do transformer turns ratio as most of the parts are in there already.
I do need to do something about my current 3 instrument "lash-up".

 
Posted : 28/07/2014 12:41 am
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

Why complicate things Ref? This is merely an ESR meter. It isn't a professional device but perfectly adequate for the job....simple and cheap. Even if all the components were bought new, I doubt it would cost more than £25. One advantage of this circuit is that it does not require an oscillator transformer to be wound. The oscillator is built around an op amp running at around 112Khz...a simple R/C network running into a balanced bridge. The only slightly critical part is the requirement to use 1% tolerance resistors for the bridge.....

 
Posted : 28/07/2014 12:53 am
Refugee
(@refugee)
Posts: 4055
Noble Member Registered
 

Why complicate things Ref? This is merely an ESR meter. It isn't a professional device .....

Well the transistor tester I built in 1975 sprouted an op-amp and LEDs and still works and is able to produce readings that can be plotted on paper to give a plot like a single trace on a curve tracer.
Now look back at what one of those beasts would have set you back in 1975 :cch
The poor thing has had to extend its collector currant ranges from 1ma to 100ma right up to 2 amps with the aid of a bench power supply at almost 40 years old and it did it without missing a beat.
I just feel like living 1975 again and there is no better way of doing it.
If you were in paid employment while you were making the bespoke scale what would you be paid after tax?
If that is more than the cost of the extra components to be able to use a cheap panel DMM then the extra components have paid for themselves before first light.
Op-amps are sold by the 500mm length tube these days and they are cheap too.
It is at the thinking stage to see what other extra functions can be added to what is in effect a multi-purpose chassis with other future home brew projects in mind.

 
Posted : 28/07/2014 2:08 am
Anonymous
(@anonymous)
Posts: 16832
Group Deactivated Account
 

Hi Rich
I was bought a Peak ESR meter when I left work, and use it quite a lot. One thing I have found is that the smaller the capacitor value, and also to an extent the lower the voltage rating, the larger the ESR. I recently bought some low value yet relatively expensive Vishay caps and the spec. was quoting an ESR for;
6.8uf = 21.1 ohms and a
47uf = 4.8 ohms
In theory both these values have failed any ESR test before they even leave the factory.

In reality when I checked them with the Peak their ESR was a lot lower but no better than the original 40 year old capacitors. Just under 1 ohm for the 6.8uf and about .5 ohm for the 47uf.

How do you judge what is good and bad? What are the limits set on your home brew meter?

Is there a chart anywhere that will enable you to check what is acceptable for any given value of capacitor.

Your one looks impressive, but it would be interesting to know more about how it works and how you decide what is a bad cap.
Thanks

 
Posted : 28/07/2014 9:13 am
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

Hi Rich

How do you judge what is good and bad? What are the limits set on your home brew meter?

Good question. There is no calibration as such other than full-scale max adjust when the leads are shorted. The prototype didn't even have a meter fitted...I used my analogue multimeter set to 500uA and after checking some new caps and some known dud ones, I was able to tell just by looking at the meter, what was good and bad. My plan is to go through a large number of elec caps from old stock and new stock and compare the ESR against the 'loss' control on the Marconi bridge. I've also got a couple of dud SM power supplies.

I think the meter indication is somewhat non-linear and the 'good' area is probably about the last half-inch or so of the scale. 6.8 ohms connected across the terminals indicates about 3/4 scale and anything above 8 ohms indicates virtually nothing.

You're right with low values. They do have a higher ESR. Checked some new 1uF up to 10uF last night and all the lower value types gave just under full-scale.

Is there a chart anywhere that will enable you to check what is acceptable for any given value of capacitor.

Don't know. I've not looked for one but there may be something which could be used.

 
Posted : 28/07/2014 10:06 am
Anonymous
(@anonymous)
Posts: 16832
Group Deactivated Account
 

Ah, it helps when you read the instructions. I cant find the paper version I got but found this PDF version for the Peak ESR70
On page 9 it gives a similar table to the one posted by Alistair.

Mike

 
Posted : 28/07/2014 6:20 pm
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

A small update. The recycled case had a third hole at the top that had originally been used for a range switch and I had been wondering what to do to make use of it. My first thought was a simple green LED 'on' indicator but then thought that would be a bit boring. After some thought, I decided a 'battery low' indicator might be more interesting. Many years ago I built such a circuit which worked well and could be adjusted for various voltage limits. This evening the extra circuitry was built and tested and the extra LED was fitted to the spare hole next to the 'shorted cap' indicator. Two extra transistors, three resistors, a pre-set pot, a 47uF cap and a diode are now fitted to the Veroboard which fills the space at the end.

I need to finalise the 'cut-off' voltage of the battery but at the moment the circuit is adjusted to light the LED at around 7.5 volts which is about the on-load voltage of a 'flat' 9 volt battery. I'm not sure how low the voltage needs to go before the ESR meter starts to give poor readings so I need to run it on an adjustable power supply to find out and then adjust the 'battery low' indicator to light just before that point.

I've taken the opportunity to tidy up the wiring and add heatshrink to some of the terminals. The wiring will be tied and secured and most likely the finished Veroboard circuit will be sprayed with a protective lacquer. Last job will be to make up a new scale for the meter.

 
Posted : 31/07/2014 12:17 am
Terrykc
(@terrykc)
Posts: 4005
Member Rest in Peace
 

... I'm not sure how low the voltage needs to go before the ESR meter starts to give poor readings so I need to run it on an adjustable power supply to find out and then adjust the 'battery low' indicator to light just before that point ...

This assumes that your supply line is very well decoupled as the PSU test would ignore the increasing series resistance of the decaying battery which might be more important than the actual voltage ...

When all else fails, read the instructions

 
Posted : 31/07/2014 11:39 am
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

That's a good point Terry!

Maybe I should just use a low value pot in series with the battery and then adjust the pot while monitoring the voltage.....? :aab

 
Posted : 31/07/2014 11:52 am
Terrykc
(@terrykc)
Posts: 4005
Member Rest in Peace
 

Rich - that's what I'd do ...!

When all else fails, read the instructions

 
Posted : 31/07/2014 1:05 pm
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

Well it's pretty much finished electrically. Just tidying up to do. I've been going through NOS and new stock and found a couple of duff capacitors in NOS. Not surprising really as they've been lying about unused since the 70's. I might try reforming them and see what happens. I also set the 'battery low' indicator to come on at around 7.8 volts but may revise it with use.

Anyway a few pics to show present state. It's been an interesting diversion from all things vintage although this could hardly be called State of the Art technology! At least I found a use for the 50uA meter I've had lying about for years.

 
Posted : 01/08/2014 12:20 am
jimmc101
(@jimmc101)
Posts: 98
Estimable Member Registered
 

Is your meter based on this http://www.qsl.net/iz7ath/web/02_brew/15_lab/06_esr/index.htm design?
If so the original design had a problem which was discussed on the other forum http://www.vintage-radio.net/forum/showthread.php?t=56933.

As Terry hinted a nice large capacitor across the battery (after the switch) would 'stiffen' the supply at 100kHz and remove the need to worry about the battery's internal resistance.

Jim

 
Posted : 01/08/2014 12:24 am
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

Hi Jim. Yes that's the one. I knew I'd seen some mods described somewhere but couldn't remember where! I might try some of those changes to improve the meter reading.

Thanks for the links.

 
Posted : 01/08/2014 9:11 am
Dr Wobble
(@dr-wobble)
Posts: 702
Honorable Member Registered
 

Nice job Rich. I thought at first it was the same as David E4GBT AKA Yorkie version, see over on UKVRR, but its a little different.

I,ve found the ESR meter I have invaluable as it tests cap in circuit. so for audio devices where there a lot of small electrolytic caps on the PCB, its great to see at a glance what caps are suspect.

However, ESR meters don't give a 100% pass/ not pass verdict. In some cases you have to to try other tests. For instance, I had a suspect cap the other day on a small mixing desk that showed in spec ESR for a 1000uf 25v cap, but testing on the ohm's range of a DMM showed it to be faulty. It read 1.5meg straight away rather than a few k with the value going up over time.

Andy.

Curiously curious

 
Posted : 01/08/2014 4:08 pm
sideband
(@sideband)
Posts: 4216
Famed Member Moderator
Topic starter
 

I've had a further play with this over the weekend. One thing that became apparent was that the readings were very cramped at one end of the scale although in practice it didn't really matter too much. However after reading through some of the mods described here I decided to try and improve the readings. One other oddity was that shorted leads gave a higher resistance reading than reading than good cap.....

Firstly, the lower values of resistors forming the bridge circuit were changed from 22 ohm to 15 ohm. This allowed the meter to read around 7.5 ohm nearer three-quarter-scale rather than 11 ohms so it opened out the lower readings. This also cured the anomaly with shorted leads.

Tonight I actually used the ESR meter to locate a faulty cap in a Freeview box.This was my own box which had some caps replaced a few months ago. However I only used the visual means, i.e my eyes, to locate and replace five swollen caps. I decided to check the remaining two caps which 'looked ok'. After taking the box to bits, I actually did a quick check on the caps I'd replaced and they all passed. However one of the two remaining caps measured as high ESR so was replaced. The remaining original seemed perfect so I left it (besides I'm low on 100uF 25V 105 degree types).

The Freeview box is now working correctly it seems and the intermittent lock-out appears to have been fixed (so far).

 
Posted : 04/08/2014 12:57 am
Share: