Hello,
I was giving a KA-6100 a tune-up and while checking the BIAS found both channels were slowly cycling substantially above and below the 20mV spec. I checked for oscillations and there were none (.5V/div; .5mS/div). Also, the heat sink got warm but never hot and the unit sounds good to my ear and had no perceivable distortions on the scope. When I lifted a Thermistor (5TP41L) off the sink it still cycled And, when placed back on the sink it did change the BIAS (as it should). What do you think: bad thermistors?
Opinions solicited.
M.
I was giving a KA-6100 a tune-up and while checking the BIAS found both channels were slowly cycling substantially above and below the 20mV spec. I checked for oscillations and there were none (.5V/div; .5mS/div). Also, the heat sink got warm but never hot and the unit sounds good to my ear and had no perceivable distortions on the scope. When I lifted a Thermistor (5TP41L) off the sink it still cycled And, when placed back on the sink it did change the BIAS (as it should). What do you think: bad thermistors?
Opinions solicited.
M.
Thermal inertia or lag is one possible reason. The sensing element/s simply don't accurately follow the temperature change of the outputs in the short term and so the system behaves like a servo with a large time constant.
Given enough time... perhaps 30 minutes at idle, and it should settle to a stable value. If it doesn't then I would be more inclined to consider the possibility that's its just a characteristic of the design. Both channels behaving the same suggests that possibility.
Given enough time... perhaps 30 minutes at idle, and it should settle to a stable value. If it doesn't then I would be more inclined to consider the possibility that's its just a characteristic of the design. Both channels behaving the same suggests that possibility.
It may not be the actual reason for instability in this instance but just taking the cover off an amplifier is enough to change the thermal environment substantially. Some amplifiers are just hopeless to service without attaching test leads and closing the case over them before settling the amplifier and then testing.
It also takes forever to set bias precisely when this is necessitated by weak thermal coupling between output device and sensor or interference from driver dissipation. This is all too common, even with designs where the sensor relies on case air temperature or is only positioned near the output stage for easy assembly reasons. However, cycling up and down suggests overcoupling with an inappropriate tempco, to me. Within limits, this may still be normal and quite unnoticeable for the model, as Mooly suggests.
It also takes forever to set bias precisely when this is necessitated by weak thermal coupling between output device and sensor or interference from driver dissipation. This is all too common, even with designs where the sensor relies on case air temperature or is only positioned near the output stage for easy assembly reasons. However, cycling up and down suggests overcoupling with an inappropriate tempco, to me. Within limits, this may still be normal and quite unnoticeable for the model, as Mooly suggests.
bias and offset adjustment in an amplifier under repair will be the last thing to check but after the actual repair.
The unit needs a complete recaping given as a fact that is more than 30 years old
Tricky areas will be the secondary power supply and feedback and bootstrap capacitor
I would start from there
On the contrary to things said by Ian i could say that hardly ever consumer Japanese amplifiers had bias stability issues ... things like the above apply mostly to European or American designs and not really on the consumer side ...something like the more hi end side
Kind regards
Sakis
The unit needs a complete recaping given as a fact that is more than 30 years old
Tricky areas will be the secondary power supply and feedback and bootstrap capacitor
I would start from there
On the contrary to things said by Ian i could say that hardly ever consumer Japanese amplifiers had bias stability issues ... things like the above apply mostly to European or American designs and not really on the consumer side ...something like the more hi end side
Kind regards
Sakis
I would generally agree with your experience, Sakis but I have to take the OP word for the bias behaviour.
If the problem was due to a bad boostrap cap, I would expect some sign in the audio too and I assumed any necessary replacements had already been carried out.
There are 2 possibilities; an RC or a thermal time constant. It would be interesting to hear what the frequency of the cycling is, and to see if there is any correlation to VAS bootstap voltage with or without a steady tone applied.
If the problem was due to a bad boostrap cap, I would expect some sign in the audio too and I assumed any necessary replacements had already been carried out.
There are 2 possibilities; an RC or a thermal time constant. It would be interesting to hear what the frequency of the cycling is, and to see if there is any correlation to VAS bootstap voltage with or without a steady tone applied.
Thank you all for your replies.
I'll try again with the cover on and maybe apply some thermal grease to the underside of the Thermistors and see what happens. I might even change the location closer to the base of the sink. I have never seen bias swings like this in any Japanese amp of this pedigree. My inner engineer tells me it's the thermistors and their inability to respond fast enough to stay on top of thermal transients (cover notwithstanding). I'll post my results.
Thanks again,
Mike
I'll try again with the cover on and maybe apply some thermal grease to the underside of the Thermistors and see what happens. I might even change the location closer to the base of the sink. I have never seen bias swings like this in any Japanese amp of this pedigree. My inner engineer tells me it's the thermistors and their inability to respond fast enough to stay on top of thermal transients (cover notwithstanding). I'll post my results.
Thanks again,
Mike
always open to hear weird problems ....
Been doing almost 30 Sanyo Plus 55 in my life and most of them happily working till the year 2013 ...with in this year i had to replace the mains switch to almost half of them
weird don't you think ? But i am in for more if it comes !!!
Athens is getting pretty hot ...summertime ...how things are going down under ?
Kind regards
Sakis
Been doing almost 30 Sanyo Plus 55 in my life and most of them happily working till the year 2013 ...with in this year i had to replace the mains switch to almost half of them
weird don't you think ? But i am in for more if it comes !!!
Athens is getting pretty hot ...summertime ...how things are going down under ?
Kind regards
Sakis
well ... it is very easy to guide you in a wrong direction and this subject we talk about here might be a little more difficult than we thought .
I had a look at the 8100 schematic while repairing kept notes and then clicked the 6100 schematic to see the difference and see if its possible to help you ...To my very big surprize i realized that 8100 and 6100 have nothing to do with each other since your amplifier is an Sziklai or a CFP complementary feedback pair .
I actually dont know what this means to you and how deep is your knowledge on sziklai and the History on them .
In a few words
In this era all or 99% of Japanese amplifier was EFP one Harman Kardon i know and probably a few other that i miss was sziklai .
Now ...in an sziklai amplifier you don't monitor bias/ temperature on the outputs you do it on the drivers .... The reason is that the real outputs for his type of CFP amplifier is the drivers The big outputs only workhorses to provide current for the system and not actual audio signal .
Your attention should be on the drivers ...Now this may really sound very weird but there is a very good chance that the Kenwood people didn't know how to run the sziklai and that is why they installed the thermistor on the heatsink .
I can imagine that a quick acting thermistor will sense the heatsink temperature and try to drop the bias and then the circle will never close ...
Workable solution for such an amplifier is a small piece of metal that will common heatsink the drivers and the thermistor is to be attached there ...this will stabilize the bias but also alter the performance of the amp for the best .
Modification like that will increase safety also on this amplifier but still you may find that the ratings of the thermistor is wrong for the modification that can be changed and of course is up to you to find out how deep you would like to go .
In all cases it is important to notice bias behavior in a various range of temperature and that will include room temperature also .
There is capacitors Cm 57 and 58 that i am not familiar with the job they do ...you may as well try to remove them to see how it is going to work .
Kind regards
Sakis
I had a look at the 8100 schematic while repairing kept notes and then clicked the 6100 schematic to see the difference and see if its possible to help you ...To my very big surprize i realized that 8100 and 6100 have nothing to do with each other since your amplifier is an Sziklai or a CFP complementary feedback pair .
I actually dont know what this means to you and how deep is your knowledge on sziklai and the History on them .
In a few words
In this era all or 99% of Japanese amplifier was EFP one Harman Kardon i know and probably a few other that i miss was sziklai .
Now ...in an sziklai amplifier you don't monitor bias/ temperature on the outputs you do it on the drivers .... The reason is that the real outputs for his type of CFP amplifier is the drivers The big outputs only workhorses to provide current for the system and not actual audio signal .
Your attention should be on the drivers ...Now this may really sound very weird but there is a very good chance that the Kenwood people didn't know how to run the sziklai and that is why they installed the thermistor on the heatsink .
I can imagine that a quick acting thermistor will sense the heatsink temperature and try to drop the bias and then the circle will never close ...
Workable solution for such an amplifier is a small piece of metal that will common heatsink the drivers and the thermistor is to be attached there ...this will stabilize the bias but also alter the performance of the amp for the best .
Modification like that will increase safety also on this amplifier but still you may find that the ratings of the thermistor is wrong for the modification that can be changed and of course is up to you to find out how deep you would like to go .
In all cases it is important to notice bias behavior in a various range of temperature and that will include room temperature also .
There is capacitors Cm 57 and 58 that i am not familiar with the job they do ...you may as well try to remove them to see how it is going to work .
Kind regards
Sakis
That was my experience with a popular series of CFP Hitachi amplifiers and even a Sony model, I recall. I never heard complaints about the cycling affecting audio though, customers seemed happy with it. 4-5 mins cycling time sounds definitely a thermal matter where the control loop is overcompensated and either cools or heats for too long because the distance between driver transistor and sensor is too long or obstructed and otherwise, the tempco. could simply be wrong.
O/T - Well, it may be hot now for you so enjoy it, Sakis. Here it is getting cool at night and frosts are reported at higher locations. I have two pullovers, corduroy jeans and thermal sox to stay warm so it's no fun at the moment.
Thanks Sakis for the information about feedback. I have some understanding of feedback designs but like to learn more. I'm still trying to locate a datasheet for the thermistor so I can try a replacement or two. I did add a little thermal grease to each and the cycling dampened some what.
Thanks again,
M.
Thanks again,
M.
