Amplifier Distortion, DC-Offset, and You!
- Thread starter EchoWars
- Start date
OMI
Super Member
Hmmmmmmm
Like I said....ELECTRONICALLY CHALLENGED...
Looked at the manual again and they got the damn pages out of order.......
Yeppers..... There is a page with my beloved photos..... But it was stuck in the German section.... I glossed over that section as I sure dont read german
Have NO idea where the german picture page went.....
Now I can ADJUST
Back to the garage....
Got the DC offset down near 0. Noticed that there is a bit of creep. Have left the amp on and will see if the creep continues. It looks like next step is BIAS. I can see the pots, but the connect points are ?????. Will get a Bud from work who should know, to show me where to conncet the meter and then I can adjust BIAS....
Then the power meter....
Like I said....ELECTRONICALLY CHALLENGED...
Looked at the manual again and they got the damn pages out of order.......
Yeppers..... There is a page with my beloved photos..... But it was stuck in the German section.... I glossed over that section as I sure dont read german
Have NO idea where the german picture page went.....
Now I can ADJUST
Back to the garage....
Got the DC offset down near 0. Noticed that there is a bit of creep. Have left the amp on and will see if the creep continues. It looks like next step is BIAS. I can see the pots, but the connect points are ?????. Will get a Bud from work who should know, to show me where to conncet the meter and then I can adjust BIAS....
Then the power meter....
Last edited:
OMI said:Got the DC offset down near 0. Noticed that there is a bit of creep. Have left the amp on and will see if the creep continues. It looks like next step is BIAS. I can see the pots, but the connect points are ?????. Will get a Bud from work who should know, to show me where to conncet the meter and then I can adjust BIAS....
Then the power meter....
It is normal for the values to float or creep a little.... making adjustments in small increments and not overadjusting can make this a less frustrating exercise.
Allowing the unit to fully warm up at idle (tone controls neutral or defeated volume off, loudness off, no speakers attached) away from strong drafts so that the temperature can stabilize will make it easier to get a reading.
As for the idle current (bias), I found it to be even more volatile on my KA-7100 than the DC offset. I spent several hours tweaking the trimmer pots hoping to dial in the exact value, but it is a moving target....and eventually accepted the fact that it was never going to be "perfect".
As for where to locate the test points and trimmer pots for the bias/idle... this should also be clearly noted in the service manual... usually on the same page(s) as the DC offset... if such an adjustment is possible.
Note that the DC offset and idle current are somewhat interdependent AFAIK, and most manuals I have seen recommend an iterative process where you go back and forth making minor tweaks as required to bring both values in line.
Of course, I tend to be a bit on the anal side about this type of adjustment/tweak, and for many folks, close is good enough!
seadzz
Active Member
I suspect the creep you are seeing is directly related to any creep in your mains supply voltage. If your amp were connected to a regulated power supply then you should expect to see a rock steady number. However your local electrical company as everywhere produces a product that swings a great deal. Also the power supply in your amp is a good one but it is not really designed to produce perfect output without any sag. More important is that the power supply in the amp is quiet.hpsenicka said:
If you really think about it a good regulated power supply for lab or shop use would cost far more than all but the TOTL amps.
As long as your readings are close to spec and do not wander around too much there is little to worry about.
sdz
Nope... AFAIK it has nothing (or at least very little.. ) to do with fluctuating mains voltage... and everything to do with how the amp is designed to compensate for thermal fluctuations in the output transistors.
I am not an expert in this area, but EchoWars advised me that I was fighting a losing battle for the exact reason I stated above.... especially for the bias/idle current voltage.
I am not an expert in this area, but EchoWars advised me that I was fighting a losing battle for the exact reason I stated above.... especially for the bias/idle current voltage.
Last edited:
seadzz
Active Member
Maybe, but if the amp were to be turned on and left idle for several hours it would over time reach a temperature that would remain constant. It would be constant because there is no input signal.
If the amp is idle not drawing current because it is not being driven by a signal what then would explain a drift?
No question EW knows these amps far better than I and I would not bet against his opinion but a floating line voltage has to impact the stability of this (and every other) amp unless big money is spent line level regulation.
sdz
If the amp is idle not drawing current because it is not being driven by a signal what then would explain a drift?
No question EW knows these amps far better than I and I would not bet against his opinion but a floating line voltage has to impact the stability of this (and every other) amp unless big money is spent line level regulation.
sdz
Last edited:
Well, from my own observation I noticed that the degree of sensitivity to temperature variation is much higher than I would have imagined.
I had my voltmeters clipped on to the test points, with the amp cover loosely in place..... simply removing the cover and allowing cool air in caused an almost immediate change in the idle current voltage. Similarly, allowing the amp to stabilize with the cover completely removed, and them putting the cover back on caused an opposite fluctuation to occur.
On my Kenwood KA-7100 I was able to get down to deviations of only a few mV from the factory spec of 20mV, so it it not tremendously significant, but still present. I finally settled for a value that hovered mostly between 20 and 22mV.... but occasionally it would move outside that range, even at idle. This was a somewhat frustrating process, as sometimes it seemed like my tweaking was actually contributing to larger variances rather than dialing it in closer to the spec'd value.
Interestingly, my Pioneer SX-880 service manual specifies a value of 30mV, but indicates that a reading within the 20mV - 40mV range is acceptable. It is within that range now, and I dare not touch it without first upgrading the stock trimmer pots with Bourns multiturn pots.
I'm sure line voltage can be a factor as well, but at least in my case, thermal sensitivity seems to be a more significant factor.
I had my voltmeters clipped on to the test points, with the amp cover loosely in place..... simply removing the cover and allowing cool air in caused an almost immediate change in the idle current voltage. Similarly, allowing the amp to stabilize with the cover completely removed, and them putting the cover back on caused an opposite fluctuation to occur.
On my Kenwood KA-7100 I was able to get down to deviations of only a few mV from the factory spec of 20mV, so it it not tremendously significant, but still present. I finally settled for a value that hovered mostly between 20 and 22mV.... but occasionally it would move outside that range, even at idle. This was a somewhat frustrating process, as sometimes it seemed like my tweaking was actually contributing to larger variances rather than dialing it in closer to the spec'd value.
Interestingly, my Pioneer SX-880 service manual specifies a value of 30mV, but indicates that a reading within the 20mV - 40mV range is acceptable. It is within that range now, and I dare not touch it without first upgrading the stock trimmer pots with Bourns multiturn pots.
I'm sure line voltage can be a factor as well, but at least in my case, thermal sensitivity seems to be a more significant factor.
Phil H
Active Member
First I know very little about electronics. I recently got a Technics SE-9021 amplifier. I wanted to check the DC offset before connecting it to speakers. The amp has a button that reads 'operation' and has 2 positions 'DC' and 'low cut (2 hz)'. My mV readings are:
In low cut
L 16.1 R 15.6
In DC
L 13.2 R 176.6
Will it be OK to use this in the 'low cut' mode?
Also, the amp has 8 trim pots. Two are on a small board with 2 fuses, 12 diodes, 2 transistors and some resistors and caps. The others are on the main board. I haven't had any luck finding information on the amp.
I read EchoWars post that said most of the ss amps are DC. So I am guessing that the statement printed on top of the amp is marketing hype "Designed as a DC amplifier, reproduction characterisics are extended to the subsonic DC range and distortion is extremely low"
I did the test with the pre-amp (SU-9011) connected does this make any difference. I realize it wouldnt cause the big difference I am seeing. But, as a general rule, should the preamp be connected when testing offset?
To be honest, I only read the first 9 pages of this thread and then used the search function and couldn't find an answer. By the way, I bought this because I wanted the speakers that came with it. The system came with JBL L110 speakers, Technics tuner (ST-9031), turntable (SL-1600MK2) and the aforementioned amp, preamp.
In low cut
L 16.1 R 15.6
In DC
L 13.2 R 176.6
Will it be OK to use this in the 'low cut' mode?
Also, the amp has 8 trim pots. Two are on a small board with 2 fuses, 12 diodes, 2 transistors and some resistors and caps. The others are on the main board. I haven't had any luck finding information on the amp.
I read EchoWars post that said most of the ss amps are DC. So I am guessing that the statement printed on top of the amp is marketing hype "Designed as a DC amplifier, reproduction characterisics are extended to the subsonic DC range and distortion is extremely low"
I did the test with the pre-amp (SU-9011) connected does this make any difference. I realize it wouldnt cause the big difference I am seeing. But, as a general rule, should the preamp be connected when testing offset?
To be honest, I only read the first 9 pages of this thread and then used the search function and couldn't find an answer. By the way, I bought this because I wanted the speakers that came with it. The system came with JBL L110 speakers, Technics tuner (ST-9031), turntable (SL-1600MK2) and the aforementioned amp, preamp.
OMI
Super Member
O.K.
Thanks for the input...
I got the DC offset adjustment points located and have been playing around. I am using a Lowe's meter with lowest scale at 200mV. Seems to work O.K. on the offset pots. The pots seem to swing +/- 150 mV or so, and it is easy to zero them down towards +/- 0.
I think I figured out the points to hook the probe too for the BIAS. When I hook up and set the meter on the 200mV setting all I am reading is .2 to .3 or .4 on the dial when I rotate the pots????? Is this a case of needing a finer scale on the volt meter to get a good reading on the BIAS pots, or an indication of something screwed up???
Also can these littl pots be cleaned...I.E. A shot of Deoxit????
Thanks,
ELECTRONICALLY CHALLENGED
omi
Thanks for the input...
I got the DC offset adjustment points located and have been playing around. I am using a Lowe's meter with lowest scale at 200mV. Seems to work O.K. on the offset pots. The pots seem to swing +/- 150 mV or so, and it is easy to zero them down towards +/- 0.
I think I figured out the points to hook the probe too for the BIAS. When I hook up and set the meter on the 200mV setting all I am reading is .2 to .3 or .4 on the dial when I rotate the pots????? Is this a case of needing a finer scale on the volt meter to get a good reading on the BIAS pots, or an indication of something screwed up???
Also can these littl pots be cleaned...I.E. A shot of Deoxit????
Thanks,
ELECTRONICALLY CHALLENGED
omi
240 Volts
Strange, yet oddly normal
Phil H,
In a "true" DC coupled amplifier, DC anywhere in the signal path (including at the input terminals, e.g from a faulty CD player / Tuner / Tape Deck etc) will get through to the speaker outputs and appear as an offset voltage.
The fact that the low cut switch removes the large DC offset on the right channel indicates to me that there is a DC component somewhere in the signal path before the switch.
Just noticed that the measurements were taken with a pre-amp connected - remove this and repeat the measurements - this will tell you where the problem is, pre or power amp.
- Richard
In a "true" DC coupled amplifier, DC anywhere in the signal path (including at the input terminals, e.g from a faulty CD player / Tuner / Tape Deck etc) will get through to the speaker outputs and appear as an offset voltage.
The fact that the low cut switch removes the large DC offset on the right channel indicates to me that there is a DC component somewhere in the signal path before the switch.
Just noticed that the measurements were taken with a pre-amp connected - remove this and repeat the measurements - this will tell you where the problem is, pre or power amp.
- Richard
This is not something to guess at, and I'd suggest that you are not connecting your meter correctly, or you have a problem with the bias circuitry.
I would have suggested you leave it alone, but if you've been twiddling with the bias setting you have now commited yourself to finding a way to get it right.
Attachments
OMI
Super Member
Thanks EW…
I subscribe to the “learn by twiddling” school of electronic repair
If I don’t electrocute myself or burn the garage down I feel I have accomplished something
I got the manual and I think I have the probes connected to the correct points. There are 2 “U” shaped PC boards with the trim pots at the bottom. I can see the 3 points and am connecting the meter to the bottom terminals as per the picture. Connecting the + to the point indicated and so on. Now are we measuring the same type of voltage as DC offset? That is how I have the meter set up. I am using a Sperry DM-350A
One more question. The page in the manual with specs notes the voltage to be set at 20mV, while the picture show’s a measurement of 40mV??????
Noticed that the pic shows alligator clips on the test points. I am using the spring loaded long clips that have a small "U" shaped wire that protrudes when you push the spring. Could that have an impact?
I subscribe to the “learn by twiddling” school of electronic repair
If I don’t electrocute myself or burn the garage down I feel I have accomplished something
I got the manual and I think I have the probes connected to the correct points. There are 2 “U” shaped PC boards with the trim pots at the bottom. I can see the 3 points and am connecting the meter to the bottom terminals as per the picture. Connecting the + to the point indicated and so on. Now are we measuring the same type of voltage as DC offset? That is how I have the meter set up. I am using a Sperry DM-350A
One more question. The page in the manual with specs notes the voltage to be set at 20mV, while the picture show’s a measurement of 40mV??????
Noticed that the pic shows alligator clips on the test points. I am using the spring loaded long clips that have a small "U" shaped wire that protrudes when you push the spring. Could that have an impact?
Last edited:
Phil H
Active Member
Thanks EchoWars and 240 Volts!
Should I be using an analog meter or have kind of load across the speaker terminals? The one side is very flaking and sometimes the offset is high, and sometimes it is low. My meter is a 15-20 year old fluke 77. I don't think it is the meter because the right side is more consistent. Also I have a newer Fluke 322 amp clamp (not enough resolution only show 1/10 volt) that will some times show .2 V on the left channel.
This is the weird part. If the preamp is on and connected the offset is high. If the pre amp is disconected, the offset is too high. But with the preamp connected and turned off the the offset becomes reasonable.
Should I be using an analog meter or have kind of load across the speaker terminals? The one side is very flaking and sometimes the offset is high, and sometimes it is low. My meter is a 15-20 year old fluke 77. I don't think it is the meter because the right side is more consistent. Also I have a newer Fluke 322 amp clamp (not enough resolution only show 1/10 volt) that will some times show .2 V on the left channel.
This is the weird part. If the preamp is on and connected the offset is high. If the pre amp is disconected, the offset is too high. But with the preamp connected and turned off the the offset becomes reasonable.
The 77 is a fine meter. Change the battery if you haven't for a while.
RE: The preamp. Might see if the preamp is sending DC to the amp. As for the amp itself, get or make some shorting plugs for the amp input jacks, and check the offset again. You can also connect a load to the amp terminals and see if it helps.
RE: The preamp. Might see if the preamp is sending DC to the amp. As for the amp itself, get or make some shorting plugs for the amp input jacks, and check the offset again. You can also connect a load to the amp terminals and see if it helps.
OMI
Super Member
Houston… We have BIAS!!!!!
I went back to the scene of the crime….
I took out the heat sink with the BIAS boards when I was cleaning up the insides… So I retraced my steps… Went back and re-set the cables and made sure they were all connected and……..
THE POTS WORKED…….
Am tweaking my way to BIAS nirvana….
Now does one need to let the unit warm up and get to temp as you do in DC offset adjustments…????
I got both pots hovering around 20 mV, but still don’t understand the 2 different criteria.. 20mV VS the 40 mV on the drawing????
Will let the unit run for a bit and re check… Then re check the DC offset and call it quits…
Got into enough trouble on these 2
Will set ‘em to 20mV as that is what is on the spec page…
:banana2: :banana2: :banana2: :banana2: :banana2:
I went back to the scene of the crime….
I took out the heat sink with the BIAS boards when I was cleaning up the insides… So I retraced my steps… Went back and re-set the cables and made sure they were all connected and……..
THE POTS WORKED…….
Am tweaking my way to BIAS nirvana….
Now does one need to let the unit warm up and get to temp as you do in DC offset adjustments…????
I got both pots hovering around 20 mV, but still don’t understand the 2 different criteria.. 20mV VS the 40 mV on the drawing????
Will let the unit run for a bit and re check… Then re check the DC offset and call it quits…
Got into enough trouble on these 2
Will set ‘em to 20mV as that is what is on the spec page…
:banana2: :banana2: :banana2: :banana2: :banana2:
The emitter resistors are .47 ohms, so each millivolt you measure from emitter of NPN to emitter of PNP equals about 1mA of current. Therfore, 20mV equals about 20mA of current. For a Kenwood, this is awfully low. I'd use the 40mV setting (but in reality it probably makes little difference).
Yes, you should set it, and check it an half-hour later. Offset too. Repeat until it seems somewhat stable.
Yes, you should set it, and check it an half-hour later. Offset too. Repeat until it seems somewhat stable.
Phil H
Active Member
Thanks EchoWars,
Here is what I found with further investigation. I put a set of patch cords into the input and shorted the free end of the rca. Without the low cut
R: open -9; shorted +17
L: open -10: shorted +194
With the low cut, the offset remains fairly constant
R: 10
L: 21
The funny thing is with no patch cord inserted in the amp, it acts like it is shorted. Yesterday, I hooked up an old pair of boombox speakers (the detachable kind - boombox died years ago; I kept the speakers). I didn't hear anything awful; it actually sounded pretty good. Would these speakers be good to use to apply a load? If you have some suggestions, I am all ears. I have reached the limit of my ability.
I'd like to eventually get to the bottom of this. But, right now, I'd like to hook up a some decent speakers and not worry about them being fried. I have some JBLs with LE14C coxaxial drivers and parts aren't available for the mid-range/high. Also, I'm having surrounds replaced on some JBL L110's that I'd like to use with this. But, I am told if the voice coil go bad on a woofer, the cone kit cost about 200.00.
I am wondering if it would be good insurance for me to pay a shop to check this out. I have a set of 8" Yamaha speaker I got on eBay for 20.00 (nothing TOTL, that's probably all they are worth). I may try them, and see what happens.
Thanks
Phil
Here is what I found with further investigation. I put a set of patch cords into the input and shorted the free end of the rca. Without the low cut
R: open -9; shorted +17
L: open -10: shorted +194
With the low cut, the offset remains fairly constant
R: 10
L: 21
The funny thing is with no patch cord inserted in the amp, it acts like it is shorted. Yesterday, I hooked up an old pair of boombox speakers (the detachable kind - boombox died years ago; I kept the speakers). I didn't hear anything awful; it actually sounded pretty good. Would these speakers be good to use to apply a load? If you have some suggestions, I am all ears. I have reached the limit of my ability.
I'd like to eventually get to the bottom of this. But, right now, I'd like to hook up a some decent speakers and not worry about them being fried. I have some JBLs with LE14C coxaxial drivers and parts aren't available for the mid-range/high. Also, I'm having surrounds replaced on some JBL L110's that I'd like to use with this. But, I am told if the voice coil go bad on a woofer, the cone kit cost about 200.00.
I am wondering if it would be good insurance for me to pay a shop to check this out. I have a set of 8" Yamaha speaker I got on eBay for 20.00 (nothing TOTL, that's probably all they are worth). I may try them, and see what happens.
Thanks
Phil