I have reached the limit of what i know (which is basically 0) and have resorted to randomly de soldering and testing various resistors and transistors in hopes of getting lucky but maybe somebody on here has some extra patience and would be willing to suggest at least a direction?
it's a Rolls RA200 amp and it's blowing all 4 mosfets and fuses. got it used, fuses and mosfets were blown, replaced all powered it on while everything was pulled apart on the bench and they all blew again (within maybe 30 seconds). replaced again and then reassembled with heatsink and this time it worked for about a month then it blew while i was out of the room sending 50v to my speakers taking one out.
At this point i started attempting to trace back on the schematic from the power switch to the mosfets assuming that whatever is failing must be something common to both channels? i've pulled a lot of transistors and other components to test and so far everything tests good.
the last time they blew i had a multimeter on the mosfets, measuring voltage and i noted there was 50v on both the gate and source and 3v on the drain. mosfet started getting warm to the touch and before i could power off they blew. (after they go then the fuse also goes).
hoping somebody might be able to give me a hint, i have attached the schematic below.
it's a Rolls RA200 amp and it's blowing all 4 mosfets and fuses. got it used, fuses and mosfets were blown, replaced all powered it on while everything was pulled apart on the bench and they all blew again (within maybe 30 seconds). replaced again and then reassembled with heatsink and this time it worked for about a month then it blew while i was out of the room sending 50v to my speakers taking one out.
At this point i started attempting to trace back on the schematic from the power switch to the mosfets assuming that whatever is failing must be something common to both channels? i've pulled a lot of transistors and other components to test and so far everything tests good.
the last time they blew i had a multimeter on the mosfets, measuring voltage and i noted there was 50v on both the gate and source and 3v on the drain. mosfet started getting warm to the touch and before i could power off they blew. (after they go then the fuse also goes).
hoping somebody might be able to give me a hint, i have attached the schematic below.
Hi . At first try without output mosfets , this design has no degeneration resistors for mosfets ,and therefore very sensitive to mosfet vgs voltage. And that voltage is very wide spread from transistor to transistor . Also potentiometers p2 and p102 may have bad contact and then idle current is maximal ,blowing out mosfets . Do not install mosfets first and measure voltage between G and S pins ,if all ok, it should be 2-3v i think , and change when adjusting potentiometer. Mark current position of potentiometer before adjusting . If mpsa transistors are shorted, you will get many volts g-s . This circuit is not self adjusting ,so very very sensitive to transistor replacement.
P.S. If you locate no bad parts ,can proceed to tests with mosfets installed .Replace all the fuses with small resistors ,like 0,25w 22 ohm .You will get them smoke if something is broken . But at first measure G-S voltages .Measuring voltage drop on those 22ohm gives you current consumption by that channel . Normally AB class amplifers output mosfet should have idle current (when amplifier has no signal) 20-100ma ,depending on schematic .So for 22 ohms and 100ma you should have 2,2volts drop .If you get then very hot ,high drop ,try carefylly adjust potentiometer for that channel and watch voltage drop. Those 22ohm resistors would limit current ,which mosfets can consume ,possibly protect them .
ximikas, thank you so much for taking some time to look through the schematic and reply!
without any mosfets installed i took some voltage readings between G and S pins..
Q6 = 0V
Q10 = 13.8V!
Q20 = 0V
Q23 = 13.8V!
adjusting the p2 and p102 pots didn't make any difference in voltage between G and S pins.
I removed mpsa transistors Q8 and Q21
These seem to be failed, tested as follows (tested for PNP transistors)
Positive to EMITTER - Negative to BASE = OL
Positive to COLLECTOR - Negative to BASE = OL
i guess my next question is since you noted in your reply that this circuit is very sensitive to transistor replacement how would i go about sourcing a replacement for these?
without any mosfets installed i took some voltage readings between G and S pins..
Q6 = 0V
Q10 = 13.8V!
Q20 = 0V
Q23 = 13.8V!
adjusting the p2 and p102 pots didn't make any difference in voltage between G and S pins.
I removed mpsa transistors Q8 and Q21
These seem to be failed, tested as follows (tested for PNP transistors)
Positive to EMITTER - Negative to BASE = OL
Positive to COLLECTOR - Negative to BASE = OL
i guess my next question is since you noted in your reply that this circuit is very sensitive to transistor replacement how would i go about sourcing a replacement for these?
At first , you need to repair all ,who is related to mosfets , until you get just few adjustable volts on each mosfet G-S .
But note , without mosfets installed output is not getting voltage ,so feedback is trying to make dc offset close to zero by opening a mosfet ,so then you get voltages like 13,8v G-S .So maybe only those mpsa were failed .But better recheck . For feedback to operate , you need to put mosfets ,but also a resistor in series with S pin ,do not connect pin directly,try 10 ohms . This will reduce idle current many times ,but feedback will operate .Also look at DC voltage at output ,should be close to zero ,maybe up to 0,1V . Use resistors in place of fuses also all the time as i have suggested earlier .
But note , without mosfets installed output is not getting voltage ,so feedback is trying to make dc offset close to zero by opening a mosfet ,so then you get voltages like 13,8v G-S .So maybe only those mpsa were failed .But better recheck . For feedback to operate , you need to put mosfets ,but also a resistor in series with S pin ,do not connect pin directly,try 10 ohms . This will reduce idle current many times ,but feedback will operate .Also look at DC voltage at output ,should be close to zero ,maybe up to 0,1V . Use resistors in place of fuses also all the time as i have suggested earlier .
Also ,about mosfet replacement . They must be original ,not from Al* or Eb** .Fakes may work too , but at very low power levels ,they mostly have small crystal die ,so can't dissipate needed amount of heat and blows ,sometimes they may be even lower voltages and current ones ,so better not risk and not waste money .Difficulty adjustment initial moment ,these mosfets are not linear ,they are vertical mosfets ,designed to be only in on and off conditions. So opening fully and heats up very quickly if you increase G-S voltage a little .
If you adjust P2 potentiometer to upper position according schematic ,it will decrease idle current , moving slider down will increase . First try with 10 ohms in S pins ,if nothing blows or smoke , decrease idle current ant try smaller resistor ,or direct connect S pins .Dont remove fuse resistors yet. Calculate idle current you get ,and try to adjust ,but very careful ,potentiometer may be old and very sensitive to position ,may lose contact . There are also risk of oscillation and blowing ,if idle current will be zero ,both mosfets off with no signal .Forget to mention - don't connect speaker . It may be difficult to set current ,you may replace those potentiometers to multi turn ones ,then its easier to adjust to some concrete current value . Also Q2 and Q16 should be close to heatsink or even touch it ,its thermal regulation ,should keep same current at different temperature ,but you may notice idle current decrease if heatsink will be hot and increased if just turned on cold .
If you adjust P2 potentiometer to upper position according schematic ,it will decrease idle current , moving slider down will increase . First try with 10 ohms in S pins ,if nothing blows or smoke , decrease idle current ant try smaller resistor ,or direct connect S pins .Dont remove fuse resistors yet. Calculate idle current you get ,and try to adjust ,but very careful ,potentiometer may be old and very sensitive to position ,may lose contact . There are also risk of oscillation and blowing ,if idle current will be zero ,both mosfets off with no signal .Forget to mention - don't connect speaker . It may be difficult to set current ,you may replace those potentiometers to multi turn ones ,then its easier to adjust to some concrete current value . Also Q2 and Q16 should be close to heatsink or even touch it ,its thermal regulation ,should keep same current at different temperature ,but you may notice idle current decrease if heatsink will be hot and increased if just turned on cold .
just an update, you were correct, when i re-tested the mpsa transistors they were okay.
i installed 22ohm resistors in place of the fuses and a 10ohm resistor in series with the S pin on the mosfets. powers up and nothing is blowing however i still have the same voltage readings... (i only am testing with the one channel for now.)
Q6 = 0V
Q10 = 13.8V
Changing the P2 Pot does not change the voltage between G - S at all.
I dropped the S pin resistors to 3ohm but still get the same voltages.
the mosfets came from digikey.
I am going through and cleaning up a number of components that were de soldered and the patched back in while i was testing just to make sure i don't have something on the board with a faulty connection.
i installed 22ohm resistors in place of the fuses and a 10ohm resistor in series with the S pin on the mosfets. powers up and nothing is blowing however i still have the same voltage readings... (i only am testing with the one channel for now.)
Q6 = 0V
Q10 = 13.8V
Changing the P2 Pot does not change the voltage between G - S at all.
I dropped the S pin resistors to 3ohm but still get the same voltages.
the mosfets came from digikey.
I am going through and cleaning up a number of components that were de soldered and the patched back in while i was testing just to make sure i don't have something on the board with a faulty connection.
I must have had a bad solder joint some where because after re-soldering all the components i touched and cleaning everything up i now have what looks like a more promising voltage across the G - S pins.
Q6 - 3.1v
Q10 - 2.7v
adjusting the P2 pot does change theses values
Q6 - 3.1v
Q10 - 2.7v
adjusting the P2 pot does change theses values
Good progress! So mosfets are installed ,s pin with 3ohms , fuses 22 ohms . Did you measured output voltage and voltage drop across these 3 and 22 ohms ? Or maybe calculated idle current ? No signal yet should be applied .
thanks for sticking with me.
There is no DC voltage on the output terminals. (a few mv)
now that i know it's not blowing mosfets i installed the same 3.3ohm S pin and 22 ohm fuse on the other channel as well so i can adjust idle current for both channels.
I carefully played around with the P2 and P102 pots
Voltage drop across the
3.3ohm resistor on Q6 is 1mv
3.3ohm resistor on Q10 is 2mv
3.3ohm resistor on Q20 is 1mv
3.3ohm resistor on Q23 is 2mv
the 22ohm resistors all have between 93mv - 97mv of drop.
(this was measured with no speakers and no input)
There is no DC voltage on the output terminals. (a few mv)
now that i know it's not blowing mosfets i installed the same 3.3ohm S pin and 22 ohm fuse on the other channel as well so i can adjust idle current for both channels.
I carefully played around with the P2 and P102 pots
Voltage drop across the
3.3ohm resistor on Q6 is 1mv
3.3ohm resistor on Q10 is 2mv
3.3ohm resistor on Q20 is 1mv
3.3ohm resistor on Q23 is 2mv
the 22ohm resistors all have between 93mv - 97mv of drop.
(this was measured with no speakers and no input)
So now idle current is too small ,less than 1ma ,maybe even dmm noise or measurement error. Drop on 22ohms means just 4ma ,so i think its mpsa load current ,not mosfets .Try adjust it first ,without removing resistors . 10mv on 3,3 ohm would mean 3ma , 100mv 33ma , 300mv 90ma .
I think for ensuring that nothing blows try adjusting to 100mv (~30ma) for test. Mosfets must be properly mounted to heatsink ,they begin to heat up with idle current .
When you get these 30ma idle current ,adjust again to smaller value , to 5ma ( you have to calculate current by voltage drop ,ohms law, I = U / R ).
Then change resistors from S pin to smaller ones , lets say 1 ohm or even less ,and check again current ,but now measure voltage drop only on 22ohms, more accurate .Current would increase while reducing these resistors on S pin.They reduce mosfet sensitivity , without resistors at all it may be too hard to adjust . So when you get small idle current with reduced S resistors , try with S pin directly soldered .Current will increase again ,you have to adjust it again ,now try to final value - to 30ma .If service manual is available, it may have more proper current values and maybe even simplier adjusting procedure .
Low dc value on speaker terminals is good sign , you can proceed with speaker and try listen on very low volume ,or 22ohms will smoke . Later change 22 ohms to 3,3ohms you have and ensure still nothing blows while playing not loud ,recheck current by voltage drop .Try with cold amplifier and leave it turned on for some time ,when heatsink will be warmer , may need adjust few times . Normally transistors like Q2 Q16 are placed near heatsink ,sometimes even have contact with it ,that used to stabilize idle current by temperature.Heating Q2 Q16 reduces idle current ,which sometimes have tendency to increase ,but depends on schematic . So continue step by step i think ,its more interesting ,than fireworks (for me at least ).
I think for ensuring that nothing blows try adjusting to 100mv (~30ma) for test. Mosfets must be properly mounted to heatsink ,they begin to heat up with idle current .
When you get these 30ma idle current ,adjust again to smaller value , to 5ma ( you have to calculate current by voltage drop ,ohms law, I = U / R ).
Then change resistors from S pin to smaller ones , lets say 1 ohm or even less ,and check again current ,but now measure voltage drop only on 22ohms, more accurate .Current would increase while reducing these resistors on S pin.They reduce mosfet sensitivity , without resistors at all it may be too hard to adjust . So when you get small idle current with reduced S resistors , try with S pin directly soldered .Current will increase again ,you have to adjust it again ,now try to final value - to 30ma .If service manual is available, it may have more proper current values and maybe even simplier adjusting procedure .
Low dc value on speaker terminals is good sign , you can proceed with speaker and try listen on very low volume ,or 22ohms will smoke . Later change 22 ohms to 3,3ohms you have and ensure still nothing blows while playing not loud ,recheck current by voltage drop .Try with cold amplifier and leave it turned on for some time ,when heatsink will be warmer , may need adjust few times . Normally transistors like Q2 Q16 are placed near heatsink ,sometimes even have contact with it ,that used to stabilize idle current by temperature.Heating Q2 Q16 reduces idle current ,which sometimes have tendency to increase ,but depends on schematic . So continue step by step i think ,its more interesting ,than fireworks (for me at least ).
woo hoo!, now i'm getting somewhere.
adjusted to 10mv and everything looked good, (you were right, the pots are very sporadic and difficult to adjust)
Next adjusted to 100mv and no issues. hooked up a speaker and confirmed i get audio
Changed 3.3ohm resistors to 1ohm then stared adjusting pots and measuring voltage drop on the 22ohm resistors. As you noted there is quite a bit a change as it warms up but i can get the voltage drop pretty close to .660v which if my math is right should be 30ma.
I then removed the 1ohm resistor and connected S pin directly. adjusted pots once again until around .660v drop across the 22ohm resistors. let it warm up, adjusted. powered off and when it's powered on cold that voltage drop measures around .400v but quickly climes up to around .650ish. the single turn pots are pretty finicky the smallest adjustment swings the voltage reading 200mv, i may look into getting mulit turn replacements.
i've been listening to it on low volume for about 20min. i am going to swap out the 22ohm resistors for 3.3 and test everything once more.
Thanks again.
adjusted to 10mv and everything looked good, (you were right, the pots are very sporadic and difficult to adjust)
Next adjusted to 100mv and no issues. hooked up a speaker and confirmed i get audio
Changed 3.3ohm resistors to 1ohm then stared adjusting pots and measuring voltage drop on the 22ohm resistors. As you noted there is quite a bit a change as it warms up but i can get the voltage drop pretty close to .660v which if my math is right should be 30ma.
I then removed the 1ohm resistor and connected S pin directly. adjusted pots once again until around .660v drop across the 22ohm resistors. let it warm up, adjusted. powered off and when it's powered on cold that voltage drop measures around .400v but quickly climes up to around .650ish. the single turn pots are pretty finicky the smallest adjustment swings the voltage reading 200mv, i may look into getting mulit turn replacements.
i've been listening to it on low volume for about 20min. i am going to swap out the 22ohm resistors for 3.3 and test everything once more.
Thanks again.
So far it's working great. There is one thing I can't figure out. In the schematics posted above there is a note... "Set p2 and p102 for 1-2mv across r11 and r111" I found resistor r11 but voltage across is 10v and I don't see r111 anywhere.
Good progress! I looked at schematic, R11 can't have voltage drop like 1mv ,maybe model is different ,or version . I don't see r111 too. Maybe its meant 0,1 ohm resistors voltage drop , R18 R31 R61 R77 .1mv means 10ma idle . Also to note , transistors around these 0,1 ohm is current limiter, check them also . About 5 ampere limit .
yeah .1mv at the .1ohm resistors makes more sense. i reached out to Rolls to see if they might be able to provide a service manual that would confirm the correct idle current, until then i dropped it a bit to about 15ma idle. (no distortion that i can tell).
I have ordered some new multiturn pots as i am concerned this might have all been caused by a bad contact in one or both of the pots as you pointed out above. once those arrive i will hopefully have a more exact idle current to adjust to.
I have ordered some new multiturn pots as i am concerned this might have all been caused by a bad contact in one or both of the pots as you pointed out above. once those arrive i will hopefully have a more exact idle current to adjust to.
Also you can extend pots range , to easier adjustment . You can add resistor like 1-2 kiloohms in parallel to potentiometer end legs , if will work best if slider is about middle position. If its close to one of ends, then this will not work, you may change surrounding resistors to slightly bigger values ,but it may be not easy , in example next to 6,8k is 7,5 k , huge difference. Multiturn pots is best solution , just they may have not 1:1 pinout / footprint,so need to check for that before ordering , value may be 470 ohms .