Ok. A quick update before I go to bed. I ended up blowing both fuses on the amp so in order to try and get power to the right channel, I rigged a 680 ohm 5W resistor I had between each connection on the fuse holder. I know that power flowed because the resistor actually got quite hot.
Also, I went and added 4k ohms of resistance to that resistor I mentioned in my last post and replaced the cap with a 102 pF cap since that's all I had as well, which I think will be close enough for testing purposes...
Nevertheless, with the PCB in place as well as the resistor, I plugged the thing in and...nothing. Had my iPod on and a speaker hooked up and nothing, absolutely nothing....I'm running out of ideas...
I might add that I'm not entirely sure on how to test a transistor, other than to see if they pass any kind of voltage.
Also, I went and added 4k ohms of resistance to that resistor I mentioned in my last post and replaced the cap with a 102 pF cap since that's all I had as well, which I think will be close enough for testing purposes...
Nevertheless, with the PCB in place as well as the resistor, I plugged the thing in and...nothing. Had my iPod on and a speaker hooked up and nothing, absolutely nothing....I'm running out of ideas...
I might add that I'm not entirely sure on how to test a transistor, other than to see if they pass any kind of voltage.
Hi Paco,
I"m glad you are working away. A bit too fast for me. I'm a little under the weather, so bear with me please. It'll take some time to respond to your posts.
Have a look around here on testing transistors. There is the "diode test" which looks for catastrophic failures. Then there is a "beta test" looking at DC current gain, and "leakage tests" that is a good indicator for faults. I use my meter on "diode test" to check the junctions, and a Heathkit "IT-18" to test gain and leakage. The IT-18 should be $20 ~ $40 on Eeekbay, a wise investment. Most DVMs today do have a diode test. That you need at a very basic level.
-Chris
I"m glad you are working away. A bit too fast for me. I'm a little under the weather, so bear with me please. It'll take some time to respond to your posts.
Test - yes. Don't swap component between dead and working channels. Just being clear as I'm sure you know that.
That happens. You were rushing ...
No need.
No, they just reduced the tail current. Change the resistor to agree with the transistor change. They might be using more bulk emitter resistance in the transistors themselves, a design change.
All right. The first thing to do is to understand how to check a transistor and understand how it works. Until then, stop with your amp as you will need to retest the parts you think are good.
Have a look around here on testing transistors. There is the "diode test" which looks for catastrophic failures. Then there is a "beta test" looking at DC current gain, and "leakage tests" that is a good indicator for faults. I use my meter on "diode test" to check the junctions, and a Heathkit "IT-18" to test gain and leakage. The IT-18 should be $20 ~ $40 on Eeekbay, a wise investment. Most DVMs today do have a diode test. That you need at a very basic level.
-Chris
Chris, thanks for the responses. I'll hold off for now on doing much more. I've just gotten impatient since I just upgraded my speakers to some Infinity RSIIIBs .
I don't have anything that can test diodes right now, so I might end up getting one tonight. At the very least I need to go get some new fuses and a new resistor. If I can, I'll let you know tonight what I can about the transistors...
. I don't have anything that can test diodes right now, so I might end up getting one tonight. At the very least I need to go get some new fuses and a new resistor. If I can, I'll let you know tonight what I can about the transistors...
Hi Paco,
Please - don't rush out.
You will use a half decent meter for years and years. Buy a Fluke. Please. You will thank me for this. Especially if you do and try using another brand (except an Agilent). I used to calibrate multimeters, so I speak from a position of strong knowledge. Many other brands will not pass certification even new in the box!
Take your time with the transistor tester. You can use a jig you build if you buy a Fluke or Agilent meter. Otherwise, buy the Heathkit once one comes up. They make a few others, but you want Beta, and the two leakage modes. I use the Heathkit even with great meters on the bench. I do reconfirm the Heathkit reads properly now and again. For such a simple circuit, it's darned accurate.
-Chris
Please - don't rush out.
You will use a half decent meter for years and years. Buy a Fluke. Please. You will thank me for this. Especially if you do and try using another brand (except an Agilent). I used to calibrate multimeters, so I speak from a position of strong knowledge. Many other brands will not pass certification even new in the box!
Take your time with the transistor tester. You can use a jig you build if you buy a Fluke or Agilent meter. Otherwise, buy the Heathkit once one comes up. They make a few others, but you want Beta, and the two leakage modes. I use the Heathkit even with great meters on the bench. I do reconfirm the Heathkit reads properly now and again. For such a simple circuit, it's darned accurate.
-Chris
Chris,
Not rushing out/wasn't going to. Got stuff to do first. There's no guarantee I'll even be able to do anything tonight...
Just looking at prices for Fluke meters and junk, they aren't cheap. Probably a wise investment regardless, especially considering how much I've been using a meter lately...
I just have to ask; is there a recommended one. I was thinking the 115 would suit my needs well enough...
Also, considering that something fried completely on the amp (which I would assume would be filed under "catastrophic failure") getting something with a basic diode tester wouldn't be a bad idea right?
Not rushing out/wasn't going to. Got stuff to do first. There's no guarantee I'll even be able to do anything tonight...
Just looking at prices for Fluke meters and junk, they aren't cheap. Probably a wise investment regardless, especially considering how much I've been using a meter lately...
I just have to ask; is there a recommended one. I was thinking the 115 would suit my needs well enough...
Also, considering that something fried completely on the amp (which I would assume would be filed under "catastrophic failure") getting something with a basic diode tester wouldn't be a bad idea right?
Hi Paco,
The 115 might work well for you. Given how long these meters will last and how good they hold their calibration (Fluke is about the only one that does), buy the best one you can afford. A used one might be fine but you need to be a little more careful. The only hand held meter I would consider buying used is a Fluke. A "Demo" would be pretty safe too.
Some of these even have a pretty accurate capacitance function. Ohms is done with a kelvin connection on the older 87 and new equivalents! The jacks are split.
You will use this meter for probably more than 10 years, and the readings are trustworthy. It's one of the few that has very good AC performance. Yes, you can make accurate readings on an amplifier for gain and output power. Again, one of the very few meters you can trust.
I think you should have a good meter to continue troubleshooting. Even for basic voltage measurements and certainly for diode test and resistance readings. I think Fluke uses 1 mA for diode check and measures the voltage across the part. More valuable information.
-Chris
The 115 might work well for you. Given how long these meters will last and how good they hold their calibration (Fluke is about the only one that does), buy the best one you can afford. A used one might be fine but you need to be a little more careful. The only hand held meter I would consider buying used is a Fluke. A "Demo" would be pretty safe too.
Some of these even have a pretty accurate capacitance function. Ohms is done with a kelvin connection on the older 87 and new equivalents! The jacks are split.
You will use this meter for probably more than 10 years, and the readings are trustworthy. It's one of the few that has very good AC performance. Yes, you can make accurate readings on an amplifier for gain and output power. Again, one of the very few meters you can trust.
I think you should have a good meter to continue troubleshooting. Even for basic voltage measurements and certainly for diode test and resistance readings. I think Fluke uses 1 mA for diode check and measures the voltage across the part. More valuable information.
-Chris
Hi Paco,
A bad meter is your worst enemy! Any test equipment that lies will always cause far more trouble than it's worth to replace it with something good.
It sometimes blows me away when I see the junk some technicians try to use. This is gear they use to make a living! Ever notice that good technicians use good equipment?
So, I think that you are on the right track. You are also about to start enjoying this hobby a great deal more.
You can use the new meter to check the old one. Only use the old meter for basic voltage readings (like supply voltages) if it's accurate there. If the meter is not accurate on DC voltage (1.9 VDC scale), then everything else is out.
-Chris
A bad meter is your worst enemy! Any test equipment that lies will always cause far more trouble than it's worth to replace it with something good.
It sometimes blows me away when I see the junk some technicians try to use. This is gear they use to make a living! Ever notice that good technicians use good equipment?
So, I think that you are on the right track. You are also about to start enjoying this hobby a great deal more.
You can use the new meter to check the old one. Only use the old meter for basic voltage readings (like supply voltages) if it's accurate there. If the meter is not accurate on DC voltage (1.9 VDC scale), then everything else is out.
-Chris
"The best-laid plans of mice and men often go awry"
One of my all-time favorite quotes for a reason. My dad needed to go to Fry's to return some stuff I tagged along and asked him about a Fluke meter. He decided to take care of it while I went and found some other stuff I needed and we ended up with one of these: http://www.bkprecision.com/products/categories/sub_categories/models/?model=815
If it works as it says to any decent precision, it might actually be okay and if not, well the store has a return policy...
One of my all-time favorite quotes for a reason. My dad needed to go to Fry's to return some stuff I tagged along and asked him about a Fluke meter. He decided to take care of it while I went and found some other stuff I needed and we ended up with one of these: http://www.bkprecision.com/products/categories/sub_categories/models/?model=815
If it works as it says to any decent precision, it might actually be okay and if not, well the store has a return policy...
Hi Paco,
That was very nice of him.
Return it. You will understand once you try a Fluke. At least give this a go and test some diode junctions to get familiar with it.
I'm sorry to give you this advice, but your meter is one thing you absolutely have to count on. You will not calculate your error budget for each measurement and this meter will not hold cal as well. Nor will it have as good an AC performance as the Fluke (not even close). So long term, you can not depend on it.
I am sure it's probably a lot better than the one you have now. Does Fry's carry Fluke?
Now, why am I so positive about this? Simple. I worked for a company called Transcat in the calibration lab. We were doing electronic measuring devices for labs and the process control industry. Our products often were sent through our lab for calibration first. Most other meter makes either scrapped through certification, or failed to meet their published specs. When they came for there next calibration, most failed. Many could not be brought into tolerance due to the input voltage divider being out. Ac performance - forget it!
One the other hand, Fluke meters typically passed certifications for years and years. Out of tolerance conditions were caused by burnt components or bad jacks with the odd crush injury. Fluke uses a thick film module for the input divider with compensation capacitors to correct the AC parameters. Closed case calibration means that the higher frequency calibration will remain very close after the calibration. Open cover calibrations were very difficult as the case spacing does affect that parameter a lot. I made a jig for some of the Fluke meters to ease these problems. Anyway, because of the excellent attention paid to the design, Fluke meters tend to hold their calibration for long periods of time (years and years) without requiring optimization. This is what the home experimenter needs, because a calibration runs about $100 USD I think, with data would be closer to $150 USD. That's for a hand held multi meter.
The only other brand I would trust would be an Agilent, but these are newer and I didn't encounter any. I'm going on what I know of the brand.
Now, if you have got to get a less expensive meter, I would hesitatingly recommend an Escort. These are far better than B&K and the rest of the pack. B&K is still better than Escort and most other meters, so it's not total junk. It's just not going to give you the life and accuracy a Fluke will.
I do hope that's helpful for you. I do apologize to anyone I may have offended out there. Understand that for most of you, no matter what you have now, there is probably much worse out there. But if you are buying new, why not get the right meter?
-Chris
That was very nice of him.
Return it. You will understand once you try a Fluke. At least give this a go and test some diode junctions to get familiar with it.
I'm sorry to give you this advice, but your meter is one thing you absolutely have to count on. You will not calculate your error budget for each measurement and this meter will not hold cal as well. Nor will it have as good an AC performance as the Fluke (not even close). So long term, you can not depend on it.
I am sure it's probably a lot better than the one you have now. Does Fry's carry Fluke?
Now, why am I so positive about this? Simple. I worked for a company called Transcat in the calibration lab. We were doing electronic measuring devices for labs and the process control industry. Our products often were sent through our lab for calibration first. Most other meter makes either scrapped through certification, or failed to meet their published specs. When they came for there next calibration, most failed. Many could not be brought into tolerance due to the input voltage divider being out. Ac performance - forget it!
One the other hand, Fluke meters typically passed certifications for years and years. Out of tolerance conditions were caused by burnt components or bad jacks with the odd crush injury.
Fluke uses a thick film module for the input divider with compensation capacitors to correct the AC parameters. Closed case calibration means that the higher frequency calibration will remain very close after the calibration. Open cover calibrations were very difficult as the case spacing does affect that parameter a lot. I made a jig for some of the Fluke meters to ease these problems. Anyway, because of the excellent attention paid to the design, Fluke meters tend to hold their calibration for long periods of time (years and years) without requiring optimization. This is what the home experimenter needs, because a calibration runs about $100 USD I think, with data would be closer to $150 USD. That's for a hand held multi meter.The only other brand I would trust would be an Agilent, but these are newer and I didn't encounter any. I'm going on what I know of the brand.
Now, if you have got to get a less expensive meter, I would hesitatingly recommend an Escort. These are far better than B&K and the rest of the pack. B&K is still better than Escort and most other meters, so it's not total junk. It's just not going to give you the life and accuracy a Fluke will.
I do hope that's helpful for you. I do apologize to anyone I may have offended out there. Understand that for most of you, no matter what you have now, there is probably much worse out there. But if you are buying new, why not get the right meter?
-Chris
Thanks Chris. It's not that I wasn't going to go on what you say, it's just what my dad decided to go with. He was an electrician for many years and does know that quality of Fluke, but like I said, he decided to get that instead and if I ask about it, he'll get angry, so it's a lose-lose for me...whatever.
I ended up using the thing last night anyways to check the cap I thought was blown and it wasn't.
I also checked transistor Q707, which I also thought may be shot. When I checked the hFE, I got nothing and when I used the diode test to check between emitter and base and emitter and collector I got something like .1V and .2V respectively, or maybe it's the other way around. Nevertheless, close to zero. According to their documentation, "A shorted diode reads near 0 V".... maybe there's something in that???
I ended up using the thing last night anyways to check the cap I thought was blown and it wasn't.
I also checked transistor Q707, which I also thought may be shot. When I checked the hFE, I got nothing and when I used the diode test to check between emitter and base and emitter and collector I got something like .1V and .2V respectively, or maybe it's the other way around. Nevertheless, close to zero. According to their documentation, "A shorted diode reads near 0 V".... maybe there's something in that???
Hi Paco,
Well, you could have got much worse in the way of a meter. Just don't pay too much attention to the last digit on the display unless you are comparing like quantities close in time. Thank goodness for technology. Meter ICs have been getting better, so many meters are better (this one included) than they used to be. However, the input divider is still the weak point. You want to see a thick film assembly with very tight tolerance values used.
All hFE readings are variable on temperature. So handle the parts with pliers if you want even rough readings that are "more real". Your diode test is the very first test you should make. Often, you can do this in circuit, but beware of low value resistors across junctions (E-B normally). Your readings indicate a dead transistor. It's shorted. Junctions may also go open, but that is less common.
Hi Ragnwald,
-Chris
Well, you could have got much worse in the way of a meter. Just don't pay too much attention to the last digit on the display unless you are comparing like quantities close in time. Thank goodness for technology. Meter ICs have been getting better, so many meters are better (this one included) than they used to be. However, the input divider is still the weak point. You want to see a thick film assembly with very tight tolerance values used.
All hFE readings are variable on temperature. So handle the parts with pliers if you want even rough readings that are "more real". Your diode test is the very first test you should make. Often, you can do this in circuit, but beware of low value resistors across junctions (E-B normally). Your readings indicate a dead transistor. It's shorted. Junctions may also go open, but that is less common.
Hi Ragnwald,
Look for the features you will use. Pay attention to the frequency response as well. The capacitance function on the upper models is very accurate, much more than other brands are.
These days? A basic requirement if you measure AC voltages at all. A dB function (just a different scale and function on an AC wave form) would be extremely helpful. Always watch that the features you do get are reliable. I'd much rather use an accurate meter with no dB or record taking functions than a meter less accurate with those and more features. If you can't depend on your readings, the meter will mislead you.
-Chris
Chris,
The measurements taken on the transistor were done out of circuit, so I guess they're worth relying on.
If the transistor is fried, it's easy enough to get a new one, but what I want to know is what cause it to cook in the first place. It would seem to me that if some other faulty part caused the original transistor to bake and that part wasn't replaced, than it will cause a new transistor to fry as well right?
The measurements taken on the transistor were done out of circuit, so I guess they're worth relying on.
If the transistor is fried, it's easy enough to get a new one, but what I want to know is what cause it to cook in the first place. It would seem to me that if some other faulty part caused the original transistor to bake and that part wasn't replaced, than it will cause a new transistor to fry as well right?
Hi Paco,
Yes, since you measured out of circuit your readings will be more accurate. Mind you, it's a short. That you can see in circuit. Things are a little different at RF frequencies, but this is analog.
Most transistors fail due to over current. So, trace the path of current flow and test each component in line. Resistors may have been overheated or even opened without leaving a physical mark. Test everything so that you don't assume a part is okay.
You can also get a failure due to over voltage breakdown in some cases, but over current is often the actual failure mode.
Hi Ragnwald,
No problem. I hope that helps you out. The basic logic is that if you take the time and care to make a measurement, it should be accurate or you are wasting your time. Fluke hand held meters can often be more accurate than many other bench meters! For a bench meter, both Agilent and Keithley make the top products. Fluke makes good bench meters too, but here those other brands are better.
-Chris
Yes, since you measured out of circuit your readings will be more accurate. Mind you, it's a short. That you can see in circuit. Things are a little different at RF frequencies, but this is analog.
Absolutely!
Most transistors fail due to over current. So, trace the path of current flow and test each component in line. Resistors may have been overheated or even opened without leaving a physical mark. Test everything so that you don't assume a part is okay.
You can also get a failure due to over voltage breakdown in some cases, but over current is often the actual failure mode.
Now, there is one characteristic of a good technician. This type of thinking will serve you well. Look at your schematic. Follow the path of high current (assumption on my part) and test those parts. This is a very good time to sit down with the schematic and a cup of coffee. Just take your time and think logically.
Hi Ragnwald,
No problem. I hope that helps you out. The basic logic is that if you take the time and care to make a measurement, it should be accurate or you are wasting your time. Fluke hand held meters can often be more accurate than many other bench meters! For a bench meter, both Agilent and Keithley make the top products. Fluke makes good bench meters too, but here those other brands are better.
-Chris
Chris,
Will do. But that will have to wait until this weekend, or at least until Friday afternoon.
Changing the subject, I guess I should bring this up now that all the transistors (2N5087) I tested when matching them had an E-B voltage of zero. I still have more transistors that I can test, but is this a problem?
Will do. But that will have to wait until this weekend, or at least until Friday afternoon.
Changing the subject, I guess I should bring this up now that all the transistors (2N5087) I tested when matching them had an E-B voltage of zero. I still have more transistors that I can test, but is this a problem?
Hi Paco,
So to clarify, you were using the diode test there? I see there is a test called "Transistor leakage test". I would expect that to be in the low μA at the very most for a silicon transistor. Germanium types will show some leakage normally.
I had to download the manual for this meter for more information. Have you read the manual yet? It indicates that it does both leakage and beta (gain) testing for transistors.
So, to summarize transistor testing for you:
1.) Diode testing will show you junction failures (shorts and opens).
2.) Diode testing will help you identify the terminals, base, emitter and collector. Also the polarity.
3.) The transistor leakage test will help show damage without complete failure. Important tests for servicing!
4.) The beta test will give you an approximate gain for a transistor. This is very temperature sensitive.
The one thing that did surprise me (pleasantly) was the leakage test. It's only collector - emitter, but you may be able to turn the leads around so you can measure the collector - base leakage. This is the one that will cause the most trouble normally. Get in the habit of testing both leakage modes if you can. Beta is nice to know as well. Transistors that measure too high or too low should be viewed with suspicion. I replace them unless they are very expensive. In that case, I pay attention to them as I set the unit up.
I guess you should read that manual and figure out how to test all these parameters. You will find that the leakage tests will save you money if you do much service work.
-Chris
If that were true, they would all be duds. I haven't seen a score like that sine the 70s Radio Shack "grab bag" parts. Okay, they weren't that bad.
So to clarify, you were using the diode test there? I see there is a test called "Transistor leakage test". I would expect that to be in the low μA at the very most for a silicon transistor. Germanium types will show some leakage normally.
I had to download the manual for this meter for more information. Have you read the manual yet? It indicates that it does both leakage and beta (gain) testing for transistors.
So, to summarize transistor testing for you:
1.) Diode testing will show you junction failures (shorts and opens).
2.) Diode testing will help you identify the terminals, base, emitter and collector. Also the polarity.
3.) The transistor leakage test will help show damage without complete failure. Important tests for servicing!
4.) The beta test will give you an approximate gain for a transistor. This is very temperature sensitive.
The one thing that did surprise me (pleasantly) was the leakage test. It's only collector - emitter, but you may be able to turn the leads around so you can measure the collector - base leakage. This is the one that will cause the most trouble normally. Get in the habit of testing both leakage modes if you can. Beta is nice to know as well. Transistors that measure too high or too low should be viewed with suspicion. I replace them unless they are very expensive. In that case, I pay attention to them as I set the unit up.
I guess you should read that manual and figure out how to test all these parameters. You will find that the leakage tests will save you money if you do much service work.
-Chris
Chris,
I did actually read the manual. I do know about the leakage tests and I actually checked the bad transistor, but don't remember what it said.
The test numbers I mentioned above are from the diode test function, just so you know.
Later this week I can recheck the leakage test(s) and check more of the 2N5087s and get back to you with the results. Would the duds have been what caused the other transistor to fry?
I did actually read the manual. I do know about the leakage tests and I actually checked the bad transistor, but don't remember what it said.
The test numbers I mentioned above are from the diode test function, just so you know.
Later this week I can recheck the leakage test(s) and check more of the 2N5087s and get back to you with the results. Would the duds have been what caused the other transistor to fry?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.