In circuit capacitor testers advice please.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Regarding in circuit capacitor testers, I can appreciate from the advice above that testing capacitors in circuit is not always a good idea but Amazon keep emailing me with capacitor testing products. One such is the Atlas ESR70 which claims to automatically discharge capacitors in circuit when testing and gets positive reviews. Atlas ESR70: Replacement Range Hood Knobs: Amazon.com: Industrial & Scientific
So my question is can an in circuit capacitor tester still be a useful device that MAY find that one dud capacitor and POSSIBLY save time and hassle removing capacitors?
Regarding my boombox, perhaps I should start another thread on that under analogue audio and keep this thread for advice for testing capacitors in circuit?
 
Last edited:
Ferguson 3T18 boombox stopped working?

Hi, I have a 1980's boombox that suddenly stopped working. My partner was listening to it at the time, apparently it made a very loud buzzing noise for about a second then nothing. Now when I switch it on the speakers make the usual popping sound as it powers up and the lights still work and the tape still goes round RR, FF etc... So I suspect the problem is on the audio board? I bought this boombox as spares or repair on Ebay a few years ago and it was displaying all the same symptoms as it is now that it is not working again ( I had it professionally repaired at the time but the guy who fixed it unfortunately no longer does repairs :( ) . I don't know what he did but one thing I do know he did was replace the ceramic capacitor (picture with blue arrow with an electrolytic capacitor (picture with black arrow)), is it likely that this capacitor has failed again due to a weakness in the circuit or is it likely something else (I was told on another thread where I mentioned about this boombox (under equipment & tools section In circuit capacitor testers advice please. ) that the amplifier chip may have fried. Any ideas?
 

Attachments

  • ferguson 3t18 016.jpg
    ferguson 3t18 016.jpg
    625 KB · Views: 124
  • inside ferguson arrow 3t18 010.jpg
    inside ferguson arrow 3t18 010.jpg
    693.5 KB · Views: 130
  • inside ferguson black arrow 3t18 018.jpg
    inside ferguson black arrow 3t18 018.jpg
    335.4 KB · Views: 120
  • inside ferguson 3t18 007.jpg
    inside ferguson 3t18 007.jpg
    682.2 KB · Views: 113
  • inside ferguson 3t18 011.jpg
    inside ferguson 3t18 011.jpg
    674.5 KB · Views: 127
  • inside ferguson 3t18 004.jpg
    inside ferguson 3t18 004.jpg
    614.4 KB · Views: 58
Last edited:
Administrator
Joined 2007
Paid Member
My own personal opinion is that a capacitance tester would be just about bottom of the list of test equipment to own.

If the result appears good do you take that on trust ? and if its bad, is it bad because of interaction with other components.

What is the test voltage and current ? It is easy to degrade (degrade... not destroy or make non functional) sensitive semiconductor junctions by applying unknown reverse polarity test voltages.

I would say a good cap tester is more of a luxury and something to be used for work other than faultfinding.
 
Administrator
Joined 2004
Paid Member
Hi soul music,
It does sound like you have a power supply problem. However, if t was being played loud at the time, a blown output chip is equally likely.

Just in case you haven't heard it enough, you can not reliably test capacitors in circuit! I own and use an expensive LCR meter by Hewlett Packard. It does not test capacitors in circuit either despite being a few thousand dollars with levels adjustable down to 50 mV.

And one for all you happy in circuit testing people ... ESR is not what you need to test! The important characteristic is called dissipation, or the opposite of "Q". These values change with frequency in case you needed to know. I have 4 test frequencies that pretty much covers the audio range.

Huntron trackers are designed to compare two circuits against each other. You make a measurement of a known good circuit, then test the defective one. You are looking for the same "picture".

The other acceptable type of really good capacitor test is done on a capacitance bridge. No, they don't measure ESR either. They do measure "Q". The capacitance bridge does not test capacitors in circuit.

You need a GOOD meter, and an oscilloscope to troubleshoot. Not a whiz-bang "capacitor checker". Save your pennies for something that matters. Even a soldering station would be a good purchase. A digital display model can be had for ~ $100 or thereabouts. A "hobby iron" (used to call those wood burning irons - seriously!) is the greatest threat to equipment.

I hope that helps you in some way.

-Chris
 
Hi soul music,
It does sound like you have a power supply problem. However, if t was being played loud at the time, a blown output chip is equally likely.

Just in case you haven't heard it enough, you can not reliably test capacitors in circuit! I own and use an expensive LCR meter by Hewlett Packard. It does not test capacitors in circuit either despite being a few thousand dollars with levels adjustable down to 50 mV.

And one for all you happy in circuit testing people ... ESR is not what you need to test! The important characteristic is called dissipation, or the opposite of "Q". These values change with frequency in case you needed to know. I have 4 test frequencies that pretty much covers the audio range.

Huntron trackers are designed to compare two circuits against each other. You make a measurement of a known good circuit, then test the defective one. You are looking for the same "picture".

The other acceptable type of really good capacitor test is done on a capacitance bridge. No, they don't measure ESR either. They do measure "Q". The capacitance bridge does not test capacitors in circuit.

You need a GOOD meter, and an oscilloscope to troubleshoot. Not a whiz-bang "capacitor checker". Save your pennies for something that matters. Even a soldering station would be a good purchase. A digital display model can be had for ~ $100 or thereabouts. A "hobby iron" (used to call those wood burning irons - seriously!) is the greatest threat to equipment.

I hope that helps you in some way.

-Chris
The Der EE DE-5000 measures Q, D and ESR. For low inductance and ultra low ESR parts, such as those used in high freq SMPS, having the D function at 100kHz is really useful for testing capacitor health. That being said, i entirely agree that here are other areas which are a better value for spending $100.
 
Administrator
Joined 2004
Paid Member
Hi Pedro,
The HP 4263A tests impedance as well, and phase angle too. With all that available, still the most useful information is capacitance and D. Phase angle is related to D and Q, so the measurements are similar information presented differently.

-Chris
 
Oh dear, this is turning out more complicated than I thought, I was hoping that it would be a matter of replacing a cap or two.
Regarding cheap soldering irons, is the problem unregulated heat that may damage a component? My cheap iron is only 30w and I,m careful not to apply too much heat to a join.
 
Administrator
Joined 2004
Paid Member
Hi soul music,
Uncontrolled heat is one issue, you can't control it well enough for starters. AC leakage current is the other issue. You can damage (destroy) other chips on the PCB using one of those due to leakage currents.

Look, your soldering iron is what can change this from a hobby that isn't that satisfying into one that is. One rule I have is: Don't fight with your tools or equipment! If you can't easily create good joints, then you can't enjoy the hobby.

Those irons really were for woodburning. It was a fad hobby years ago where you would finish a piece of wood, then burn a pattern into it - like an animal (horses were popular), sailing ships. Just whatever. When that happy died off they quickly repurposed those irons for electronics simply because some folks found they could avoid buying an expensive soldering station with one of those. Back then, a Weller station cost about $120 Canadian. I know because I bought a bunch for techs in my shop. The original for me was a Weller WCP. These days I would highly recommend an eastern brand such as the Quick I'm using, Hakko, Solomon and a host of others. These will set you back around $100 today. This is easily your most important purchase before you attempt to repair your blaster.

One quick test for an output chip is the voltage on the output pins. If the chip uses one supply voltage (likely), then expect to see 1/2 Vcc on each output. If the chip uses a bipolar supply (+ & -), expect the outputs to sit within a few 10's of mV from ground potential. These are higher quality ICs used in very expensive boom boxes or lifestyle systems. You need a meter with sharp probe tips to prevent sliding off the test point and shorting the connections - probably blowing the chip in the process. Keysight is finally providing really nice probes like this. Note that a set of good meter probes will cost anywhere from $35 to $80. Not available off Ebay normally.

-Chris
 
Hi soul music


One quick test for an output chip is the voltage on the output pins. If the chip uses one supply voltage (likely), then expect to see 1/2 Vcc on each output. If the chip uses a bipolar supply (+ & -), expect the outputs to sit within a few 10's of mV from ground potential. These are higher quality ICs used in very expensive boom boxes or lifestyle systems. You need a meter with sharp probe tips to prevent sliding off the test point and shorting the connections - probably blowing the chip in the process. Keysight is finally providing really nice probes like this. Note that a set of good meter probes will cost anywhere from $35 to $80. Not available off Ebay normally.

-Chris

Thanks, I will invest in a proper soldering station and some better probes for my multi meter. I cannot see the amp chip in my photo but I,m guessing the heat sink is obscuring it (I will have to open it up again and look inside) from a bit of research online I believe it to be a Rohm BA536 amplifier chip.
 
Administrator
Joined 2004
Paid Member
Hi soul music,
This would be one of those chips where you will see 1/2 Vcc on the output pins. If you find the chip to be bad, then clean the old thermal compound off and use fresh. If it has a silpad, any damage at all means you need to replace it with either Mica (if you can find one that fits) or another silpad (same thing as Mica). If Mica is used, clean it very carefully taking care not to bend it. Clean it on a flat surface on top of some cardboard, like a cereal box cut open. Thermal compound is messy, so clean any up immediately.

Last point. Do not over-tighten the screws. Snug them down and give them another 1/4 turn. If you have a torque wrench, they probably want to "see" 6 ~ 8 inch-pounds, or something like that. Don't crank them down tightly, that's a great way to snap the die inside or strip the threaded holes. Don't use too much grease either. Only a little should squeeze out when tightened. By little, I mean a thin bead around the edge of the chip. A little excess in other words, but not too dry. If you use too much thermal compound (heat sink grease), it may crack the die as well. Besides, the goal is to rid that space of air pockets. Not to use the grease as an intermediary for the heat.

Those two things you decided on will go a long way to make this and future endeavours successful. I wish I could remember the number for the Keysight probes, but they come packed with the current meters, in my case a U1273A. I don't know if the less expensive meters come with different leads of not. They are insulated right up to the sharp point leaving about 1/8" exposed (very approximate). There is a probe kit (#34138A) that comes with the normal probe tips, but they have some slip over needle probes insulated all the way up tp about 1/8", they are around 2 1/2" long so they extend the total probe length by that amount. They also have small grab adapters too. One point, all current probes that are any good are shrouded at the plug end. Most cheap meters take Banana plugs without the shroud. The shroud protects you in case a lead comes out of the meter while you are connected to something live. Fluke and Keysight / Agilent / HP meters have been like this for a long time. If you see someone selling a working HP 974A meter, GRAB IT!!!! These are excellent meters with a basic DC accuracy of 0.05% and AC response to about 100 KHz (rms chip HF response) will most of the other meters are not good much more than a few hundred Hz.

One thing people don't generally say is that the only meters that "hold" their calibration are those from Fluke and Keysight / Agilent / HP. So when the time comes when you can buy a better meter (> $200), buy one of these two brands. The more you spend (within reason of course) the better purchase you will be making. I have meters older than 20 years and I still use them with confidence.

I'm also a calibration technician, so my experiences cover all meter models from most manufacturers. Companies don't change their corporate personality, so what was true of older products still holds true today. I can't stress how important it is for a meter to hold it's calibration (stay in tolerance) over long periods of time (years of easy service) that you are likely to put a meter through. They are also a bit tougher to kill. Cheap ones are easy to damage - terminally.

Best, Chris
 
Fluke and Keysight / Agilent / HP meters have been like this for a long time. If you see someone selling a working HP 974A meter, GRAB IT!!!! These are excellent meters with a basic DC accuracy of 0.05% and AC response to about 100 KHz (rms chip HF response) will most of the other meters are not good much more than a few hundred Hz.

One thing people don't generally say is that the only meters that "hold" their calibration are those from Fluke and Keysight / Agilent / HP. So when the time comes when you can buy a better meter (> $200), buy one of these two brands. The more you spend (within reason of course) the better purchase you will be making. I have meters older than 20 years and I still use them with confidence.

I'm also a calibration technician, so my experiences cover all meter models from most manufacturers. Companies don't change their corporate personality, so what was true of older products still holds true today. I can't stress how important it is for a meter to hold it's calibration (stay in tolerance) over long periods of time (years of easy service) that you are likely to put a meter through. They are also a bit tougher to kill. Cheap ones are easy to damage - terminally.

Best, Chris


Overall a good overview. Though I would add a few other companies to that list: Brymen (OEM for a few Greenlee and Amprobe models in the USA), Yokigawa, and Hioki. All have demonstrated their abilities to make high quality meters over the past decade or more. That being said, out of this limited few, this is the main reason to go with a high quality manufacturer. You want a device that stays in cal for many years. You won't get that from the basic Chinese OEMs.

You make a good point about AC voltage response for TrueRMS, most people don't pay attention to that figure. Many meters top out at just above 400Hz and don't spec any higher. For example, even the venerable Fluke 87V is only spec'd to 5Khz. The Brymen BM869S that I use is spec'd out to 100khz for the AC ranges. It makes life very easy when I need to look at higher freq signals in certain circuits (though the UL-Listed Cat IV 1kV is nice from a safety aspect).
 
Administrator
Joined 2004
Paid Member
Hi Pedro,
Meters that remain in cal over time normally use an input divider made on one ceramic wafer. This construction makes them all track with temperature, and also reduces capacitance. Most meters use a basic resistor string. Many use those inexpensive 1% metal film resistors that can't track with temperature and have loose drift specifications. You get the idea.

I'm not familiar with Brymen, and Hioki will not remain in tolerance. Yokigawa will, but I'm more familiar with their more expensive bench products. The Greenlee and Amprobe products I've seen really aren't much better than the mass meter market. They will not hold their cal like a Fluke or Keysight ...

I didn't mention Keithley, and they also make excellent products. Bench meters and such. I use the industry standard HP 34401A on my bench, as well as 3457A and 3456A. At least service information is easy to find for HP products.

-Chris
 
Hi Pedro,
Meters that remain in cal over time normally use an input divider made on one ceramic wafer. This construction makes them all track with temperature, and also reduces capacitance. Most meters use a basic resistor string. Many use those inexpensive 1% metal film resistors that can't track with temperature and have loose drift specifications. You get the idea.

I'm not familiar with Brymen, and Hioki will not remain in tolerance. Yokigawa will, but I'm more familiar with their more expensive bench products. The Greenlee and Amprobe products I've seen really aren't much better than the mass meter market. They will not hold their cal like a Fluke or Keysight ...

I didn't mention Keithley, and they also make excellent products. Bench meters and such. I use the industry standard HP 34401A on my bench, as well as 3457A and 3456A. At least service information is easy to find for HP products.

-Chris

Amprobe is owned by Fluke. I guess they didn't bother upgrading the QC in the acquisition. Depending on the model, you either get a Fluke design manufactured by Uni-T (just like the Fluke 11x series), a Brymen design, a Wavetek design, a Meterman design, or possibly others. Most people think quite highly of the 5x0 series (other than the 560 and 570 because Fluke screwed up royally on the firmware) as they are the only Fluke designed Amprobe (that I know of). From a safety perspective, the AM-510 is just about the best value on the market. As for the Brymen designs (AM-2x0 series and AM-1x0 series) they are very well thought of and bilt. No clue about the various other designs though.

That is REALLY surprising about Hioki. I know a good number of people (mostly grey beard RF guys) that swear by them. My biggest problem with the Japanese OEMs is that they often seem to be 5-10yrs behind the times on features, UI, etc.

Brymens are amazing pieces of kit, I would put them slightly above HP/Agilent/Keysight in build quality and roughly on par with Fluke: above in some areas like input protection and below in others (they use their own proprietary ASICs, which are not as fully integrated as Fluke's). My BM869S has been rock solid (or at minimum trending almost exactly with my 34401A). As mentioned, it has absolutely EXCELLENT input protection, I would say by far the best in the business. It, like all Brymen meters, uses the laser trimmed ceramic hybrid for the input divider. It also has a metal can TrueRMS chip, one of the few metal cans I have seen in a handheld.
Imgur: The most awesome images on the Internet
Imgur: The most awesome images on the Internet
Imgur: The most awesome images on the Internet

Yeah, I was only considering handhelds. For bench, I used to have 3 Keithleys but I got rid of two of them (two 199 scanning multimeters with the integrated scan cards) when I got my 34401A. I still have my Keithley 236 Source Meter, looking to make a test jig but I am having a hard time finding affordable triax cables/connectors. Without those, it gets REALLY spendy REALLY fast. As for my 34401A, I love that device, it has been rock solid. I had it cal'ed and adjusted last year when a cal lab was running a special on a NIST Cal&Adj with data and return shipping on the 34401A for $75. Couldn't beat that peace of mind.

Anyways, I think I have derailed OP's discussion enough. :D
 
Administrator
Joined 2004
Paid Member
Hi Pedro,
If anything, I hope our discussion has given members some food for thought. When you consider that a meter is something you will own for many years, sometimes up to the point of our death, looking at them seriously makes a lot of sense. This long ownership cycles means that either you will fight with and be lied to for that period of time, or it will compliment your work and easily catch problems missed by other meters.

For handheld meters, the HP 974A is solid, as solid as any higher end Fluke I've ever seen or worked on before. I use both here, I have an 83 and a sick 87 as well, never mind repairing and calibrating literally tons of them. The current batch of Keysight / Agilent meters are again as solid as any new Fluke meter. I'd be happy with either, but I'll admit that I'm happier with the Keysight and Agilent I got recently (U1273A). The only thing I would change is the OLED display. It's a bit harder on battery life, but it is extremely easy to see and make out what is on the display. They even have a bluetooth module that snaps onto the back of the meter to convert the LED power into a bluetooth. It's nice that the additional bit of kit will snap onto other meters in the same line, so you can move those capabilities to any instrument in that line. I also got the USB to optical link cable for these meters. The current clamp-on meter (U1213A) I have uses the same accessories, so a lot of trading of bits as they are required. Setting up the bluetooth was stone simple, as I would expect from Fluke as well.

The only other brand I saw that was inexpensive but held it's calibration was from Escort (I think that was the name). Agilent bought that brand - which allowed them to come out with extremely good lower end meters with very similar quality as their own products had.

At any rate, given the considerations, it doesn't pay to buy really cheap meters. You should own at least one good meter. You can save money by buying them "pre-owned" from the seminar circuit, or from leasing programs. So no matter what name is on the case, by one that is truly good. I would make a stronger statement that says to only buy good equipment. I didn't always, and they really hurt me on time and incorrect readings.

-Chris
 
After some thought I have decided it is not worth me investing in the equipment needed just to fix this thing if it's not a simple repair. Besides this is a nice boombox and I don't want to mess it up. So can somebody recommend someone who can fix this for me at a reasonable price? I live in the UK and will have to post it with lots of padding as the case is getting rather brittle and easily damaged. Although it is a Ferguson I believe it to be based on one of these GE 3-5286A | The Boombox Wiki
 
Administrator
Joined 2004
Paid Member
Hi soul music,
You know, sadly an expert service person can get by with very basic things and would be able to do that repair with your meter alone in most cases. But then, that person would also have thousands of hours invested into education and practical experience.

Don't feel badly. I guess you have made the correct decision for someone who only wants to fix the odd thing. I was hoping you would be starting your journey into this wonderful hobby. Given that thought, my advice would probably make more sense to you.

A good service person will have a lot of good equipment. His bench may be a mess or neat as a pin, what matters is his workmanship and how clean that is. I wish you luck in finding that service person. Stay in touch with anyone you find that is good as they will surely save you money over the years.

Best, Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.