JVC JA-S71 with odd protection circuit problem

I have a JVC JA-S71 with an odd reaction to audio input. I haven't found any posts that relate specifically about it, so I'll post mine here.

The amp seems to set off the speaker protection circuit when a strong bass passage is played through any of the aux or tape monitor inputs. It does it whether it is connected to the preamp or amp directly. To check further, I hooked up an audio signal generator and a dummy load with wattage meter to the amp and discovered that the signals below 80hz required little voltage to trip the protection, resulting in a 10w or less output in that frequency range. The lower the frequency, the less input voltage needed to trip the circuit, and the less wattage read at the output.
However, when inserting a signal above 200hz, the full 80 watt output is possible. I have tried all different frequencies up to 20khz and have achieved full 80w output.

So my question is... what the heck is wrong with this thing? I have used a DVM with a diode tester to check the output transistors and the 2 sets of four seem to be getting very close readings. None of them are obviously different than the others.

I have set the offset and bias voltages to the range indicated in the service manual. I have replaced all of the electrolytic caps, except for the large power caps and the ELNA low leak electrolytics.

I have only had this amp for a month and do not know its history, however, it will play all day long at the threshold of shutoff (about 30w output) without getting very warm or tripping the protection circuit.

Anybody have any ideas? Driver trannies, output trannies, etc.?

I would check the voltage 2 pre out/main in connections, as you run the low bass and get the protect to protect. Look for a DC voltage on the left one or right one. After looking at this, I would look at the DC balance by checking speaker out left, and speaker out right, looking for DC voltage right before the protect engages and shuts it down. Also leaky small electrolytics in the tone or control circuits can cause problems such as this. Look for the ones that ar tantalums usually inj the 2.2uf and less range. I have found elna's get leaky and lass DC. If you can't find DC passing on them, measure a few out of circuit. when they get leaky they usually increase in value a lot. like a 2.2uf reading 5 to 10 uf

Tom B said:

I would check the voltage 2 pre out/main in connections, as you run the low bass and get the protect to protect. Look for a DC voltage on the left one or right one. After looking at this, I would look at the DC balance by checking speaker out left, and speaker out right, looking for DC voltage right before the protect engages and shuts it down. Also leaky small electrolytics in the tone or control circuits can cause problems such as this. Look for the ones that ar tantalums usually inj the 2.2uf and less range. I have found elna's get leaky and lass DC. If you can't find DC passing on them, measure a few out of circuit. when they get leaky they usually increase in value a lot. like a 2.2uf reading 5 to 10 uf

Click to expand...

Thanks for the response. It gives me something to look for ,and I do have the problem just using the amp, without the preamp, so I feel it is a problem in the main amp itself.
The amp gets very good full frequency response at lower volumes and all of the loudness and tone controls have been sprayed with a cleaner/lube. I should add that my Sandisk E280 mp3 player was connected to the main amp input and , while its output is not very high and most of the tunes don't generate the low bass, a few of the country western and southern blues rock tunes do get low and can set off the protection circuit

Sounds like a bad cap - specifically in the protection circuit. Typical protection circuit uses a resistor coming from each channel to a Bipolar (or two polar back to back) electrolytic capacitor which feeds the DC protect circuit. Typical resistor is about 50k ohm, typical cap is about 300 uF. The cap function is to filter out signal so that only DC triggers the protect circuit, and if it were open, a strong low frequency signal would trigger protect. Look for UPC1237 or HA12002 IC spec sheets for typical circuits. A manual or schematic would help, if the circuit is discrete.

sregor said:

Sounds like a bad cap - specifically in the protection circuit. Typical protection circuit uses a resistor coming from each channel to a Bipolar (or two polar back to back) electrolytic capacitor which feeds the DC protect circuit. Typical resistor is about 50k ohm, typical cap is about 300 uF. The cap function is to filter out signal so that only DC triggers the protect circuit, and if it were open, a strong low frequency signal would trigger protect. Look for UPC1237 or HA12002 IC spec sheets for typical circuits. A manual or schematic would help, if the circuit is discrete.

Click to expand...

:thmbsp: Sounds like the Protection circuit's doing it's job. VLF will trigger the transistors to trip the relay.

If you have an oscilloscope, you can measure the frequency or DCV at the transistors.

sregor said:

Sounds like a bad cap - specifically in the protection circuit. Typical protection circuit uses a resistor coming from each channel to a Bipolar (or two polar back to back) electrolytic capacitor which feeds the DC protect circuit. Typical resistor is about 50k ohm, typical cap is about 300 uF. The cap function is to filter out signal so that only DC triggers the protect circuit, and if it were open, a strong low frequency signal would trigger protect. Look for UPC1237 or HA12002 IC spec sheets for typical circuits. A manual or schematic would help, if the circuit is discrete.

Click to expand...

That is a good idea, I did replace all electrolytic caps on the main amp board, which includes the 220uf and the 2 470uf caps servicing the relay. One thing I haven't done is check the mylar and ceramic caps for any value change, and the discrete transistors used in the circuit, as well. I have been curious about the possibility of the transistors in the protection circuit changing values from idle to higher volume level output, thereby possibly changing the sensitivity of the circuit. Their voltages in circuit are at reference level at each test point while idling.

jblmar said:

:thmbsp: Sounds like the Protection circuit's doing it's job. VLF will trigger the transistors to trip the relay.

If you have an oscilloscope, you can measure the frequency or DCV at the transistors.

Click to expand...

Are you referring to the output transistors, or to the protection circuit transistors? Or even to the main amp driver transistors?

Found manual at hifiengine.com. the capacitors I referred to were the 2 470 uF. When you replaced them, did they go in "back to back" i.e. they are in series and one is reversed from the other, so the 2 minuses should be connected together, and one + should be to ground and the other to the signal coming in (via 10K resistors). A quick check would be to jumper them to ground - this bypasses the DC sensing part of the protection and leaves everything else functioning.

One other question. Does it do it with either channel? helps to narrow down problem.

sregor said:

Found manual at hifiengine.com. the capacitors I referred to were the 2 470 uF. When you replaced them, did they go in "back to back" i.e. they are in series and one is reversed from the other, so the 2 minuses should be connected together, and one + should be to ground and the other to the signal coming in (via 10K resistors). A quick check would be to jumper them to ground - this bypasses the DC sensing part of the protection and leaves everything else functioning.

One other question. Does it do it with either channel? helps to narrow down problem.

Click to expand...

The replacement 470uf caps are in their proper position, with both negative poles connected via the pcb trace. I did find an anomaly in the protection circuit, a jumper wire was soldered in place of the R688 56ohm resistor at the factory. I replaced it with a resistor and saw no difference in behavior, other than it took a few seconds longer for the relay to come on at power up. I did find that the relay stayed functioning longer when either channel was operating by itself, but not with both of them operating together.

I did perform another signal test with input voltage at 180-190mv for all frequencies and observed a progressively higher output with each step.

10hz -.5w before relay shutoff
20 - 1w ""
30 -2.5w ""
40 -4.5w power output stops increasing a little bit before relay shuts down
50 -7 w ""
60 -10w ""
70 -13w ""
80 -18w ""
90 -21w ""
100 -30w relay doesn't shut off and power output doesn't increase
120 -41w ""
150 -70w ""
200 -95w relay doesn't shut off

As far a grounding the caps, I am unsure about doing it properly( don't want to see the 4th of July come too early)

Thanks for all the ideas and responses

In m opinion (and I think Sregor's), you really need to check the waveform at the top of the caps.

Your measurement show the the frequency sensitivity shows that the 235uF cap is not shunting enough low frequency signal to ground. .....allowing it to (prematurely) trigger the protection circuit.

For the -3dB frequency point of the protection circuit, I assumed 5kOhm series and a 235uF shunt C. The 5k is calculated by considering both (L&R) 10k resistors in parallel. The 235uF is the 2 470uF wired in series.

F(-3db) = 1/2*PI*R*C = 0.14Hz .....so, if everything is connected high-pass-wize, it shouldn't be so upper bass sensitive.

Jon_Logan said:

In m opinion (and I think Sregor's), you really need to check the waveform at the top of the caps.

Your measurement show the the frequency sensitivity shows that the 235uF cap is not shunting enough low frequency signal to ground. .....allowing it to (prematurely) trigger the protection circuit.

For the -3dB frequency point of the protection circuit, I assumed 5kOhm series and a 235uF shunt C. The 5k is calculated by considering both (L&R) 10k resistors in parallel. The 235uF is the 2 470uF wired in series.

F(-3db) = 1/2*PI*R*C = 0.14Hz .....so, if everything is connected high-pass-wize, it shouldn't be so upper bass sensitive.

Click to expand...

I agree with one comment - though the knee frequency is .14 hz, the trip point should be around 1 volt, which would would be (I'm too lazy to calculate it out like Jon did) 2-3 Hz - quite far from 200 hz. My opinion is one of the caps is bad or way out of spec. (do you have more, or the originals, or a cap tester) An alternative would be to use a 220 uF Bipolar type cap - most manufacturers use that instead of back to back caps. Good luck with it.

sregor said:

I agree with one comment - though the knee frequency is .14 hz, the trip point should be around 1 volt, which would would be (I'm too lazy to calculate it out like Jon did) 2-3 Hz - quite far from 200 hz. My opinion is one of the caps is bad or way out of spec. (do you have more, or the originals, or a cap tester) An alternative would be to use a 220 uF Bipolar type cap - most manufacturers use that instead of back to back caps. Good luck with it.

Click to expand...

I did replace all of the lytics on the main board, because of their age and the fact that there were so few to replace. I have attempted to use my esr meter to check them in circuit, but I doubt the readings. On a couple of caps, they read exactly what they're supposed to read. But the lytics in the relay circuit do not, so I shall remove them and check them out of circuit. Since the unit had the same problem with the original caps, placing them back in would not be very helpful. Using the formula, with parenthesis added - 1/(2Pi*R*C)= , showed that with my bad value readings from the esr meter, the caps should have kept the triggering point in the 2-3 hz range. There may be higher than normal resistance in the caps , or out of range value to the cap(s). If so, Digi-Key may hear from me ;-)

Thanks for your input. Every idea is useful and can spring new ones upon which to follow.

Logan, did I see you avatar upside down on a refrigerator somewhere?

Way out thought - are the 10k resistors 10k?

I wish to thank everyone for their suggestions and ideas. In one of my posts, I mentioned not trusting the esr meter value readings. Upon taking out the suspect caps and testing them out of circuit, I found the esr meter was correct and the caps all had a value of 48uf. Not exactly the 220uf and 470uf caps that were supposed to be in the circuit. Since I had a bunch of 470uf caps lying around , I replaced my replacement caps, using 2 470uf in series to get the 220uf range and, voila, This beast is operating as it should, or at least it is no longer sensitive to the low frequency shutoff problem. I will be ordering the proper replacement caps shortly, I'm trying to stick with the Elna Silmic II types. I have been listening to different sources, JT-V71 tuner with 750mv output, JVC XL-V211 cd player with about 2v output and my Sandisk mp3 play with about 100mv output, and the beast hasn't shutoff yet.
Now all I have to worry about is being hauled off for disturbing the peace and causing my neighbor's ears to ring.

Once again, thanks to all who posted their helpful ideas. Now on to other projects, 2 receivers the have FM tuner difficulties, not looking forward to them.

:thmbsp:

Nice. Good job! :huge: