Hi Mooly, it was me. I was measuring from one end of the diode to the other. Measured correctly now,(referenced to the chassis) I see 28v on D605 and 18v on D607.
Yes, you do have the right diagram. Straight ra-840. Everything matches the actual board. (Although it takes make a while to see it!).
Yes, you do have the right diagram. Straight ra-840. Everything matches the actual board. (Although it takes make a while to see it!).
No problem 🙂 I know its a steep a steep learning curve 😉 I hope this doesn't look to daunting but they are just simple voltage checks. We need to try and see where its falling down, where the loop is failing to correct the offset.
Lets elaborate a bit more on those voltages...
28 volts on D605 I'm assuming is the supply end of the diode
18 volts on D607 I'm assuming is the amplifier output end of the diode which is the same as measuring on the fuse. So that would be the DC offset (which is obviously at fault)
I've labelled these points on the diagram. We need to know what you have at each location as measured from chassis ground.
Point A is the positive supply which you have measured at +28 volts. That sounds OK.
Point B is the negative supply and should be - (minus) 28 volts.
(If the supplies are wildly different to -/+28 volts or they are unbalanced (i.e. the negative value is nothing like the positive value then we need to resolve that before continuing)
Point C is the amplifier output which I think you have measured at + 18 volts.
Point D (which is R617) should read approximately 1.2 volts more negative than whatever you see at point C. So if point C is +18v then point D should be about +16.8 volts.
Point E (which is R617) should read approximately 1.2 volts more positive than whatever you see at point C. So if point C is at +18v then point E should be about +19.2 volts.
Point F should read about the same as point A
Point G should read approx 0.6 volts more positive than point B. So if point B is -28v then point G should be around -27.6 It is not an easy one to measure accurately because the rail voltage can fluctuate a little. If you measure across across R605 you should see a steady 0.6 volts approx.
Lets elaborate a bit more on those voltages...
28 volts on D605 I'm assuming is the supply end of the diode
18 volts on D607 I'm assuming is the amplifier output end of the diode which is the same as measuring on the fuse. So that would be the DC offset (which is obviously at fault)
I've labelled these points on the diagram. We need to know what you have at each location as measured from chassis ground.
Point A is the positive supply which you have measured at +28 volts. That sounds OK.
Point B is the negative supply and should be - (minus) 28 volts.
(If the supplies are wildly different to -/+28 volts or they are unbalanced (i.e. the negative value is nothing like the positive value then we need to resolve that before continuing)
Point C is the amplifier output which I think you have measured at + 18 volts.
Point D (which is R617) should read approximately 1.2 volts more negative than whatever you see at point C. So if point C is +18v then point D should be about +16.8 volts.
Point E (which is R617) should read approximately 1.2 volts more positive than whatever you see at point C. So if point C is at +18v then point E should be about +19.2 volts.
Point F should read about the same as point A
Point G should read approx 0.6 volts more positive than point B. So if point B is -28v then point G should be around -27.6 It is not an easy one to measure accurately because the rail voltage can fluctuate a little. If you measure across across R605 you should see a steady 0.6 volts approx.
Hi Mooly, ok; Point A 29.6v
Point B -30.2v
Point C 18v
Point D 18.8v
Point E 16.6v
Point G -29.2v
Across R605 .7v
Point B -30.2v
Point C 18v
Point D 18.8v
Point E 16.6v
Point G -29.2v
Across R605 .7v
Basically OK I think. Supplies are good
I'm wondering if you have points D and E back to front. E should be the more positive one. If that's the case then the numbers look OK. Just check though.
Was point F OK? Should be around +29 volts.
I'm wondering if you have points D and E back to front. E should be the more positive one. If that's the case then the numbers look OK. Just check though.
Was point F OK? Should be around +29 volts.
I'll have a think later. What is odd is that the offset is at a voltage between the rail values. Often a catastrophic fault results in the offset going hard over to one or other rail.
In the meantime a couple more checks. Look at the circuit.
R603 that connects to the base of the first input transistor should have around hmmm.... it should be close to zero but I'm guessing you will see a highish voltage here. What do you see?
R607 should have around 0.7 volts across it.
The high offset you have (+18 volts) should be trying to turn OFF Q605 which in turn allows Q603 to conduct more. That 0.7 volts you measured across R605 should be turning ON Q611 which should be pulling the output back toward the negative rail.... and that is not happening. Q611 could be suspect but its far from conclusive at this point.
Sometimes quick and dirty tests can prove a lot. Of which there are a couple we can try...
In the meantime a couple more checks. Look at the circuit.
R603 that connects to the base of the first input transistor should have around hmmm.... it should be close to zero but I'm guessing you will see a highish voltage here. What do you see?
R607 should have around 0.7 volts across it.
The high offset you have (+18 volts) should be trying to turn OFF Q605 which in turn allows Q603 to conduct more. That 0.7 volts you measured across R605 should be turning ON Q611 which should be pulling the output back toward the negative rail.... and that is not happening. Q611 could be suspect but its far from conclusive at this point.
Sometimes quick and dirty tests can prove a lot. Of which there are a couple we can try...
Hi Mooly, so R603 across reads 0v and either end to ground reads 0v.
R607 across reads 0v (meter set to 200v and then 2v) but either end to ground reads 29.1v (plus and minus respectively).
I double checked these.
One question; when you said "Q611 which should be pulling the output back toward the negative rail", would I be correct
in thinking that the approximate goal of an amplifier circuit is to output a signal (waveform?) that is perfectly balanced between the positive and negative rails?And that "offset" is called such because it is an offset from the desired signal?
R607 across reads 0v (meter set to 200v and then 2v) but either end to ground reads 29.1v (plus and minus respectively).
I double checked these.
One question; when you said "Q611 which should be pulling the output back toward the negative rail", would I be correct
in thinking that the approximate goal of an amplifier circuit is to output a signal (waveform?) that is perfectly balanced between the positive and negative rails?And that "offset" is called such because it is an offset from the desired signal?
Your amplifier like many others uses a balanced supply meaning there is a positive and a negative rail. Those don't actually have to be equal in value....
The output point should be at zero volts DC to ensure that no voltage is laced across the speaker when there is no signal. You have a high DC output as a fault which we call an offset.
The upper and lower output transistors can be thought of as being a bit like a tap or valve controlling the flow of water. If we want the output to rise in a positive direction the upper tap has to be opened a little. The more we open it and the more water would be available to flow out of the amplifiers output. Close that valve and open the lower one and we control how much water flows from the negative side.
The output point is kept at zero by means of the negative feedback and the differential input pair of transistors which are like a set of scales or a balance. If the output rises positive or negative when it shouldn't then the balance is altered and more or less drive is delivered to Q611 which in turn alters the output transistors.
That indicates a problem. Look at the circuit diagram. D601 and D603 are in series and connected to ground via a 6.8k resistor. Each diode should have approx 0.6 volts as measured across it.
Check that first.
The transistor Q601 should have approx 0.6 volts across the base/emitter junction. Those are the pins that connect to R607 and R609.
Carefully check that as well. It will be easier measuring between the resistor points I think.
(Q601 forms a constant current source to feed the input stage)
The output point should be at zero volts DC to ensure that no voltage is laced across the speaker when there is no signal. You have a high DC output as a fault which we call an offset.
The upper and lower output transistors can be thought of as being a bit like a tap or valve controlling the flow of water. If we want the output to rise in a positive direction the upper tap has to be opened a little. The more we open it and the more water would be available to flow out of the amplifiers output. Close that valve and open the lower one and we control how much water flows from the negative side.
The output point is kept at zero by means of the negative feedback and the differential input pair of transistors which are like a set of scales or a balance. If the output rises positive or negative when it shouldn't then the balance is altered and more or less drive is delivered to Q611 which in turn alters the output transistors.
That indicates a problem. Look at the circuit diagram. D601 and D603 are in series and connected to ground via a 6.8k resistor. Each diode should have approx 0.6 volts as measured across it.
Check that first.
The transistor Q601 should have approx 0.6 volts across the base/emitter junction. Those are the pins that connect to R607 and R609.
Carefully check that as well. It will be easier measuring between the resistor points I think.
(Q601 forms a constant current source to feed the input stage)
I'm pleased that those results show a problem but its a bizarre one.
Is there a possibility anything got shorted in the past while working on this?
Anyhow...
I think the next step has to be a quick test in circuit of those diodes and the transistor. Use the diode range on your meter (amp off of course) and see if they read suspiciously low. Read across each diode and the B/E junction. You can compare against the other channel. Remember the reading will alter depending on probe polarity.
Assuming they read OK (and I suspect they won't but we'll see) then the only single component failure that I can see causing this would be R609, the 6k8 resistor that connects the diodes to ground.
Is there a possibility anything got shorted in the past while working on this?
Anyhow...
I think the next step has to be a quick test in circuit of those diodes and the transistor. Use the diode range on your meter (amp off of course) and see if they read suspiciously low. Read across each diode and the B/E junction. You can compare against the other channel. Remember the reading will alter depending on probe polarity.
Assuming they read OK (and I suspect they won't but we'll see) then the only single component failure that I can see causing this would be R609, the 6k8 resistor that connects the diodes to ground.
Please excuse my intrusiion.
Pictures can be very deceptive, but the R609 joint on right side (center), just above green edit mark, looks questionable. I suggest liquid rosin flux to get good reflow where necessary.
Agreed that Q601 current source is open. I suggest the following powered gremlin hunt: tie one meter lead to point F and probe your way down the D601, D603, R609 bias divider. There should be about 1.2V showing on meter at junction of D603 and R609. Any erroneous voltages should stick out and be easier to trace. Probe directly on component leads as necessary; integrity of parts, board traces, and solder joints are all suspects.
Assuming 1.2V on D603 the R609 junction, the same meter voltage should be present on base of Q601, and about 0.6V on emitter and its end of R607.
Another observation: R605 had 0.7V across in post 43. I surmise the current source must have been working at that point to deliver R605 bias current.
Good luck!
Pictures can be very deceptive, but the R609 joint on right side (center), just above green edit mark, looks questionable. I suggest liquid rosin flux to get good reflow where necessary.
Agreed that Q601 current source is open. I suggest the following powered gremlin hunt: tie one meter lead to point F and probe your way down the D601, D603, R609 bias divider. There should be about 1.2V showing on meter at junction of D603 and R609. Any erroneous voltages should stick out and be easier to trace. Probe directly on component leads as necessary; integrity of parts, board traces, and solder joints are all suspects.
Assuming 1.2V on D603 the R609 junction, the same meter voltage should be present on base of Q601, and about 0.6V on emitter and its end of R607.
Another observation: R605 had 0.7V across in post 43. I surmise the current source must have been working at that point to deliver R605 bias current.
Good luck!
Hi Mooly and Bsst. So, Mooly, re your last; D601= 838
D603=838
Q601=827 pos lead to e neg lead to b (over limit with leads swapped)
R609= 075 (meter set to 200k) Then 15 climbing to over limit when set to 20k
Bsst re your comments; Point F to D601=.5v
Point F to D603=0v
D603/R609 junction = 0v all measured on comp legs (which was very difficult).
D603=838
Q601=827 pos lead to e neg lead to b (over limit with leads swapped)
R609= 075 (meter set to 200k) Then 15 climbing to over limit when set to 20k
Bsst re your comments; Point F to D601=.5v
Point F to D603=0v
D603/R609 junction = 0v all measured on comp legs (which was very difficult).
The 838 sounds OK, particularly as both read the same. It is actually showing millivolts dropped across the diode when the meter passes a small test current through the diode. So we'll say OK for that at this point.
R609 will need one end isolating (unsoldering) to get a true reading.
I did a hatchet job 😀 on another amplifier simulation (sorry Doug) and built it up into the Rotel.
You can see the sort of voltages you should be getting and the very important approx 0.7 volts across the diode and base emitter junctions).
An open R609 would give the fault you have (second image).
R609 will need one end isolating (unsoldering) to get a true reading.
I did a hatchet job 😀 on another amplifier simulation (sorry Doug) and built it up into the Rotel.
You can see the sort of voltages you should be getting and the very important approx 0.7 volts across the diode and base emitter junctions).
An open R609 would give the fault you have (second image).