In a earlier thread, I posted about working on my father's Heathkit AR-14, and how it was playing too quietly. I'd done some re-capping that brought it back from the dead, but the volume wasn't quite right. While experimenting, I blew up the transistor labeled Q100 on the schematic, so it was a lesson learned in caution.
This evening, I checked my solder joints and replaced Q100 with a locally obtained NTE85. Hooked everything up, turned on the tuner, and it was playing at a realistic volume. Plugged in some headphones and discovered that the sound might have the slightest bias to the left, but that it sounds pretty good for an elderly solid state stereo.
Something I couldn't help but notice is that it takes the right channel a second or two to start playing after the left channel comes on.
The trouble comes after about a half hour. Then, the left side starts cutting in and out, and when it's out, I can hear a bit of power supply hum before it comes back. After a period of time, it comes back less and less, and then when it comes back, it's quite crackly. I went back and re-flowed some of the old solder joints on the amp board, but this didn't make much difference.
Here's what I've done so far:
-Given all the pots a clean.
-Replaced C129 and 130 with 4700uF 50V They're the biggest caps on the board other than the big 4000uF filter cap.
-Replaced C124, 125, 126, and 127
-Replaced Q100
-Replaced C204, 205, 206, 207
-Replaced C202 and 203
What should I check next?
Thank you,
-William
This evening, I checked my solder joints and replaced Q100 with a locally obtained NTE85. Hooked everything up, turned on the tuner, and it was playing at a realistic volume. Plugged in some headphones and discovered that the sound might have the slightest bias to the left, but that it sounds pretty good for an elderly solid state stereo.
Something I couldn't help but notice is that it takes the right channel a second or two to start playing after the left channel comes on.
The trouble comes after about a half hour. Then, the left side starts cutting in and out, and when it's out, I can hear a bit of power supply hum before it comes back. After a period of time, it comes back less and less, and then when it comes back, it's quite crackly. I went back and re-flowed some of the old solder joints on the amp board, but this didn't make much difference.
Here's what I've done so far:
-Given all the pots a clean.
-Replaced C129 and 130 with 4700uF 50V They're the biggest caps on the board other than the big 4000uF filter cap.
-Replaced C124, 125, 126, and 127
-Replaced Q100
-Replaced C204, 205, 206, 207
-Replaced C202 and 203
What should I check next?
Thank you,
-William
The description still fits that of dried out caps or possibly another heat related issue if its fine for a half hour or so. As you say it's only on the left channel, isolate the inputs by turning volume down to zero. If this makes no difference to the crackling/cutting out and all caps have been replaced in that channel, you have to consider the semis themselves or even resistors may be breaking down due to heat and age. It's not correct method but I'd apply a gentle push or tap with a rubber handled screwdriver to these remaining parts and see if I couldn't coax a little change in the noise produced - perhaps narrowing down the fault location. It gets poo-poohed these days but this is situation where even freeze spray, carefully applied, could assist.
I don't know any local shop that has circuit cooler cans. The US EPA has a death wish on those products, the non-flammable ones anyway. Propane can light up your whole face.
I would take an AC scale of an analog VOM, and look to see where the signal disappears on the bad channel. First preamp stage, use the 2 VAC scale, after that use the 20 VAC scale. Put a .047 >100 v cap in series with the minus lead to inhibit a VOM from reading AC on DC voltages. I use a Simpson 260-6XLPM. DVM 's that I have owned, produce random numbers on music on the AC scale. They are for 50 hz or 60 hz, only, IMHO. Oscilloscopes that I have owned, broke after a couple of weeks. You can also use a "sound probe", an amplifier that has a blocking capacitor in the input and a variable gain, to probe through the circuit. Use a Y cable to put a FM radio into both channels at once, and compare volume of the good channel to the bad channel. Where the sound disappears, that stage has a problem requiring further analysis, perhaps DC voltage checks. Transistors should drop 0.6 v on the BE joint in the linear realm, and the CE voltage should be between 2 v and the rail voltage. Diodes should drop 0.6 v forward biased or be negative biased.
With my VOM, I found two bad solder joints in my most recent project, one made by me replacing an electrolytic coupler cap, one made at the factory in 1994. No wonder that amp kept blowing up the crowbar and output transistors on the right channel, the op amp kept whanging to +15 v output due to the input socket lead being badly soldered.
You could have a bad solder joint, you could have a component problem. Resistors frequently can be measured at the nominal or below due to leakage through some semiconductor.
BTW, my friend's AR15 receiver started having problems with the volume control, after two years. Cheap ****.
I would take an AC scale of an analog VOM, and look to see where the signal disappears on the bad channel. First preamp stage, use the 2 VAC scale, after that use the 20 VAC scale. Put a .047 >100 v cap in series with the minus lead to inhibit a VOM from reading AC on DC voltages. I use a Simpson 260-6XLPM. DVM 's that I have owned, produce random numbers on music on the AC scale. They are for 50 hz or 60 hz, only, IMHO. Oscilloscopes that I have owned, broke after a couple of weeks. You can also use a "sound probe", an amplifier that has a blocking capacitor in the input and a variable gain, to probe through the circuit. Use a Y cable to put a FM radio into both channels at once, and compare volume of the good channel to the bad channel. Where the sound disappears, that stage has a problem requiring further analysis, perhaps DC voltage checks. Transistors should drop 0.6 v on the BE joint in the linear realm, and the CE voltage should be between 2 v and the rail voltage. Diodes should drop 0.6 v forward biased or be negative biased.
With my VOM, I found two bad solder joints in my most recent project, one made by me replacing an electrolytic coupler cap, one made at the factory in 1994. No wonder that amp kept blowing up the crowbar and output transistors on the right channel, the op amp kept whanging to +15 v output due to the input socket lead being badly soldered.
You could have a bad solder joint, you could have a component problem. Resistors frequently can be measured at the nominal or below due to leakage through some semiconductor.
BTW, my friend's AR15 receiver started having problems with the volume control, after two years. Cheap ****.
Last edited:
I also find the local electronic parts retailers have clamped down stockholding and distribution due to transport restrictions but as techs find them useful on occasion, I guess trade suppliers are not averse to making profits. Newark and Mouser at least, in the US are offering at least 9 products each, starting around $10 (with freight restrictions.) If they come with a tube for a tiny drizzle application, all the better.
To be honest, I haven't used the stuff in more than a year so I have none on hand but I can always improvise if need be.
I'm beginning to think that the T03 transistors on the back of the chassis are part of the problem. Not cause they're failing, but instead because their heat-shedding is compromised. I took all those apart to run tests on the transistors. While the mica insulators were still present, any thermal gel that might have been there at one time was long gone.
Sent from my myTouch_4G_Slide using Tapatalk 2
Sent from my myTouch_4G_Slide using Tapatalk 2
That's certainly a problem but the fix is not difficult if you clean up well and apply the old white grease to both sides of the mica. That's the insulating electronics kind grease, not the computer CPU types which can actually be conductive. Check for shorts to the pins and case from the heatsink with a multimeter etc. before powering up.
It doesn't explain the crackling though. Have you tried the volume right down yet?
It doesn't explain the crackling though. Have you tried the volume right down yet?
Sorry, I have been away and busy with other projects.
Now, the the volume right down test... Would this work better with a dial-type multimeter? I have an inexpensive one. Rusty on how to use it best, but smart enough to re-learn.
Today, I removed all the T03 transistors from the back of the case, wiped their mounting areas and backs clean, wiped the mica insulators clean, and re-mounted everything with non-conductive heat sink paste in the appropriate places.
Something else I've done is replaced several more caps that are on the left channel, four in all. I turned the amp on and it sounded good. The only downside was that the left channel had more background hiss. The right was pretty quiet. Switching from radio to phono didn't change that. Other than the background hiss, the phono sounded fine.
But the same thing happened that's been happening all along, though it may have taken a little longer this time. Left signal starts humming and dropping out, and finally goes. Sometimes, it can be brought back, but is crackly while there.
Looking at the board, there is a resistor that appears to have gotten hot at some point. It even left a dark spot on the board under its mounting spot. I'll post a pic, but on the schematic, it looks like it could be R171, a 1ohm 2w resistor. Trouble is, the banding is a little ambiguous. I'm judging by the nearby resistor that could be identical.
Think that resistor could be part of the problem?
Thank you,
-William
Now, the the volume right down test... Would this work better with a dial-type multimeter? I have an inexpensive one. Rusty on how to use it best, but smart enough to re-learn.
Today, I removed all the T03 transistors from the back of the case, wiped their mounting areas and backs clean, wiped the mica insulators clean, and re-mounted everything with non-conductive heat sink paste in the appropriate places.
Something else I've done is replaced several more caps that are on the left channel, four in all. I turned the amp on and it sounded good. The only downside was that the left channel had more background hiss. The right was pretty quiet. Switching from radio to phono didn't change that. Other than the background hiss, the phono sounded fine.
But the same thing happened that's been happening all along, though it may have taken a little longer this time. Left signal starts humming and dropping out, and finally goes. Sometimes, it can be brought back, but is crackly while there.
Looking at the board, there is a resistor that appears to have gotten hot at some point. It even left a dark spot on the board under its mounting spot. I'll post a pic, but on the schematic, it looks like it could be R171, a 1ohm 2w resistor. Trouble is, the banding is a little ambiguous. I'm judging by the nearby resistor that could be identical.
Think that resistor could be part of the problem?
Thank you,
-William
An externally hosted image should be here but it was not working when we last tested it.
Sent from my myTouch_4G_Slide using Tapatalk 2
Can I suggest you check the channel voltage differences at circuit nodes like the Base, Emitter & Collector of the transistors in the power amplifiers. (ref. p23 in attached link)
Here's another copy of the schematic which is easy to print out and write voltages on for comparison of channels. you could use different colour pens to make comparisons simple. Heathkit Schematic and Manual Archive | Vintage Radio Info
A digital meter is best for precise measurements because a "dial" type, whilst great for seeing what fluctuating voltages are doing, can be way out on accuracy and repeatability unless it was originally a high quality instrument. They can be good for AC or audio measurements at high frequencies though - sometimes you can even see relative amounts of oscillation.
R171 and R169 are the emitter resistors for the output transistors and all power goes out through them, via C129. If both had gotten hot, you'd assume the speaker leads or speakers had been shorted for some time but if only 1 showed stress, I'd have to suspect oscillation and this may be consistent with the periodic bursts of noise from the bad channel. Amps do funny things like that when misbehaving. Test for heat in both resistors with a slightly moist finger and see if the problem persists with and without output to speakers. Keep one hand away from the live equipment and watch your safety. Moist skin invites current to flow.
Lets check those comparative voltages. As a crosscheck, base and emitter voltages on transistors should be around 0.65V different, so if this isn't the case, something is not quite right. If the amplifiers are working though, this is unlikely. Seeing your pic, that carbon resistor sure looks sick. Measure!
Here's another copy of the schematic which is easy to print out and write voltages on for comparison of channels. you could use different colour pens to make comparisons simple. Heathkit Schematic and Manual Archive | Vintage Radio Info
A digital meter is best for precise measurements because a "dial" type, whilst great for seeing what fluctuating voltages are doing, can be way out on accuracy and repeatability unless it was originally a high quality instrument. They can be good for AC or audio measurements at high frequencies though - sometimes you can even see relative amounts of oscillation.
R171 and R169 are the emitter resistors for the output transistors and all power goes out through them, via C129. If both had gotten hot, you'd assume the speaker leads or speakers had been shorted for some time but if only 1 showed stress, I'd have to suspect oscillation and this may be consistent with the periodic bursts of noise from the bad channel. Amps do funny things like that when misbehaving. Test for heat in both resistors with a slightly moist finger and see if the problem persists with and without output to speakers. Keep one hand away from the live equipment and watch your safety. Moist skin invites current to flow.
Lets check those comparative voltages. As a crosscheck, base and emitter voltages on transistors should be around 0.65V different, so if this isn't the case, something is not quite right. If the amplifiers are working though, this is unlikely. Seeing your pic, that carbon resistor sure looks sick. Measure!
I went ahead and changed the resistor seen in the pic earlier, along with its three companions. While that one had the worst appearance, they were all slightly misshapen. The one in the pic measured at 1.6ohms, which is pretty far out of spec. The new resistors have felt cool to the touch the whole time the amp operated.
Something I've noticed is that the old 150ohm 1w resistors nearby get really hot to the touch.
After all this was done, I was really expecting things to behave. Trouble is, they didn't...
Like usual, when I first turn the amp on from cold, the left channel seems to have more hiss in the background than the right, which is especially noticeable thru headphones.
After playing the radio thru speakers and headphones, after about 12 minutes, the left channel starts to hum and drop out. It'll come back a few times, but then it goes away.
Now, if I turn the volume up a bit, sometimes it comes in quite rough-sounding. Once it crosses that bad point, it never returns in such a way that sounds right, which is quite frustrating.
Now, when you say to check voltage differences in the power amplifier transistors, I assume that you mean the transistors that are mounted on the back of the amp itself?
Something I've noticed is that the old 150ohm 1w resistors nearby get really hot to the touch.
After all this was done, I was really expecting things to behave. Trouble is, they didn't...
Like usual, when I first turn the amp on from cold, the left channel seems to have more hiss in the background than the right, which is especially noticeable thru headphones.
After playing the radio thru speakers and headphones, after about 12 minutes, the left channel starts to hum and drop out. It'll come back a few times, but then it goes away.
Now, if I turn the volume up a bit, sometimes it comes in quite rough-sounding. Once it crosses that bad point, it never returns in such a way that sounds right, which is quite frustrating.
Now, when you say to check voltage differences in the power amplifier transistors, I assume that you mean the transistors that are mounted on the back of the amp itself?
In the manual I linked to above, the power amplifier schematic is on page 23. It is numbered only for the left channel parts in odd numbers. The right channel is numbered in the missing even numbers. So Q7 in the left channel is the same as Q8 in the right etc. I can't say without looking over your shoulder but I'm sure it's obvious what the sequence is from the copy you posted. All those 6 transistors Q5-16 in either channel need checking against their counterparts as a starting point after verifying that the voltages marked on the schematic agree with what you measure - within say, 5%. That assumes your local mains voltage is close to the rating of 110,115,120 AC etc or whatever is on the rear label.
The ground connection is your common reference , so clip the meter neg. lead of a multimeter to ground and measure the C,B,E, voltages on each transistor and compare with the same measurement in the other channel. It looks like 50-100V range will do. if you are using a DMM. Do it directly so you don't waste time writing down figures that are much the same. Obviously, they should be much the same and so you only need to note any clear differences. As said, you should find a 0.65V difference between B and E voltages on each transistor. Recheck by measuring across B&E if you aren't certain, This tells you the transistors at least are basically OK or not.
When it comes to the output transistors, (Q13-16 mounted on the back) we might look a little closer but let's get the preliminary checks out of the way.
I guess you are aware of the problems with using meters on live circuits but I'll say from personal experience that it's all too easy to slip with worn or rounded meter probes and wind up shorting parts and blowing them without realizing it. Take great care to steady your hand and limit the pressure to just get a steady reading. Keep one hand safely away from the amp. for your own safety.
The ground connection is your common reference , so clip the meter neg. lead of a multimeter to ground and measure the C,B,E, voltages on each transistor and compare with the same measurement in the other channel. It looks like 50-100V range will do. if you are using a DMM. Do it directly so you don't waste time writing down figures that are much the same. Obviously, they should be much the same and so you only need to note any clear differences. As said, you should find a 0.65V difference between B and E voltages on each transistor. Recheck by measuring across B&E if you aren't certain, This tells you the transistors at least are basically OK or not.
When it comes to the output transistors, (Q13-16 mounted on the back) we might look a little closer but let's get the preliminary checks out of the way.
I guess you are aware of the problems with using meters on live circuits but I'll say from personal experience that it's all too easy to slip with worn or rounded meter probes and wind up shorting parts and blowing them without realizing it. Take great care to steady your hand and limit the pressure to just get a steady reading. Keep one hand safely away from the amp. for your own safety.
Ok, finally getting back around to this. Here's what I have so far, and I'm just assuming that I did the measurements properly. They seem a little odd sometimes.
Q5 E=.002 C=4.49 B=.68
Q7 E=.002 C=.90 B=.665
Q9 E=.002 C=35.58 B=.90
Q11 E=.005 C=19.02 B=.645
Q6 E=.002 C=4.45 B=.684 hum was induced when I tested the base.
Q8 E=.002 C=1.08 B=.669
Q10 E=.465 C=31.5 B=1.09
Q12 E=.005 C=18.4 B=.68
The amp was on the whole time, tuned to FM stereo, though the antenna was not hooked up.
Q5 E=.002 C=4.49 B=.68
Q7 E=.002 C=.90 B=.665
Q9 E=.002 C=35.58 B=.90
Q11 E=.005 C=19.02 B=.645
Q6 E=.002 C=4.45 B=.684 hum was induced when I tested the base.
Q8 E=.002 C=1.08 B=.669
Q10 E=.465 C=31.5 B=1.09
Q12 E=.005 C=18.4 B=.68
The amp was on the whole time, tuned to FM stereo, though the antenna was not hooked up.
These numbers are mostly fine, as we might hope for an amp. that's idling. The only problem so far is Q9 where the voltage difference between Emitter and Base looks high as it might look when it's amplifying a signal. You could recheck without the tuner running next time. In fact speakers, inputs, everything should be removed or switched out and volume down when testing DC voltages. So far, so good with most parts anyway.
When you look at power supply connections to both channels, "D" is supposed to feed 28V to both. But down at Q9, it's measuring 36V and its counterpart Q10 is receiving 31V. 'Something odd there. Check what points A and D at the power supply read. The schematic shows you what all voltages are supposed to read. Check some more voltages but as said earlier, only those that show ~5% difference are much concern. Don't forget to include B,C,E readings of the output transistors Q13-16. Yes, those on the rear panel which they need as a heatsink.
When you look at power supply connections to both channels, "D" is supposed to feed 28V to both. But down at Q9, it's measuring 36V and its counterpart Q10 is receiving 31V. 'Something odd there. Check what points A and D at the power supply read. The schematic shows you what all voltages are supposed to read. Check some more voltages but as said earlier, only those that show ~5% difference are much concern. Don't forget to include B,C,E readings of the output transistors Q13-16. Yes, those on the rear panel which they need as a heatsink.
Ah, ok. I'll test them again in the next day or two. This time, I'll set the function switch to Aux Stereo, disconnect the speakers, and turn the volume all the way down. I wasn't aware that volume all the way down was a requirement.
Then, I'll check those other voltage points and the rear-mounted transistors.
Thank you,
-William
Then, I'll check those other voltage points and the rear-mounted transistors.
Thank you,
-William
- Status
- Not open for further replies.