Just to test the module if everything working fine,
Power up with 22-0-22 Vdc instead of 53-0-53Vdc (not changing the R6 & R17 stick to 6k8 and 27k). Manage to adjust the voltage across 100r to -9Vdc (Neg) and 5.35 (Pos). Previous voltage before adjusting VR2 are -11.46 and +6.24.
Adjust the output then connect the LOAD. Everything working fine (test with slow music volume). At higher volume, sound is distorted (due to 22Vdc used). All 3 pairs FETs working as the Heatsink becoming warm and the MJEs heatsink much more hotter then the main heatsink.
Will try to connect / adjust with 53Vdc once back in the evening.
Power up with 22-0-22 Vdc instead of 53-0-53Vdc (not changing the R6 & R17 stick to 6k8 and 27k). Manage to adjust the voltage across 100r to -9Vdc (Neg) and 5.35 (Pos). Previous voltage before adjusting VR2 are -11.46 and +6.24.
Adjust the output then connect the LOAD. Everything working fine (test with slow music volume). At higher volume, sound is distorted (due to 22Vdc used). All 3 pairs FETs working as the Heatsink becoming warm and the MJEs heatsink much more hotter then the main heatsink.
Will try to connect / adjust with 53Vdc once back in the evening.
Vivek
I'm also waiting for any guidance on how to troubleshoot.. since power up with 53-0-53 Vdc still given problem.
Getting high Vdc across 100r for Neg Rail when adjusting Vr2 (-16.4v to -14v).
However Vdc across 100r at Pos Rail still looks OK which fluctuated @5v-6v when adjusting Vr2.
Both measurement was done with one of the multimeter probe lead clipped to Gnd and the other end clipped to the fuse holder after the 100r/5W.
R22@220r which located near T10/MJE350 producing smoke with burning stripe however it still working as i measure the resistance across R22. I can only play around with Vr2 less then a minute before R22 start to burn.
Seems this module working fine with 22Vdc even though this not the voltage the module design to work with but given problem when powering up with 53Vdc which this the voltage the amp supposed to work with the combination of 6k8 (R6) and 27k (R17).
MJEs heatsink temperature only warmer a little bit.
Anybody are kind enough to guide where's to start troubleshooting with this kind of symptom
Looks like is more then challenge to DIY this amp.. imagining dust bin rite now...
I'm also waiting for any guidance on how to troubleshoot.. since power up with 53-0-53 Vdc still given problem.
Getting high Vdc across 100r for Neg Rail when adjusting Vr2 (-16.4v to -14v).
However Vdc across 100r at Pos Rail still looks OK which fluctuated @5v-6v when adjusting Vr2.
Both measurement was done with one of the multimeter probe lead clipped to Gnd and the other end clipped to the fuse holder after the 100r/5W.
R22@220r which located near T10/MJE350 producing smoke with burning stripe however it still working as i measure the resistance across R22. I can only play around with Vr2 less then a minute before R22 start to burn.
Seems this module working fine with 22Vdc even though this not the voltage the module design to work with but given problem when powering up with 53Vdc which this the voltage the amp supposed to work with the combination of 6k8 (R6) and 27k (R17).
MJEs heatsink temperature only warmer a little bit.
Anybody are kind enough to guide where's to start troubleshooting with this kind of symptom
Looks like is more then challenge to DIY this amp.. imagining dust bin rite now...
Vivek,
In my amp modules, I noticed that the voltage starts to rise only when the the trimmer is at 3 O'clock CW. You can follow bawang's or samuel's method - that is to measure the voltage across t he 0.33R resistors.
Take note also to refer to latest quasi pcb tracks. There are revisions on the T8 area.
Joey
In my amp modules, I noticed that the voltage starts to rise only when the the trimmer is at 3 O'clock CW. You can follow bawang's or samuel's method - that is to measure the voltage across t he 0.33R resistors.
Take note also to refer to latest quasi pcb tracks. There are revisions on the T8 area.
Joey
My PCB designed based on latest track which published at the Quasi web only different on R29, R32, R33, R30, R32, R34 value which supposed on the schematic using 0.33r/5W but mine using 0.47r/5W (followed the value on the board layout).
Can somebody attach the up to date schematic / pcb track layout here if there is changes on the one that published on the quasi web.
Many Thanks
Can somebody attach the up to date schematic / pcb track layout here if there is changes on the one that published on the quasi web.
Many Thanks
Hello everyone.
I have 200 pcs more 2SK2698 in my box so I decided to use it in this quasi creation but no luck I cant bring the bias current higher than 5mA/pcs if more than that one by one tha transistor will dead. I use 94 volt AC CT trafo I also get trouble with the DC protection it seem can not work with high level it will oscilate open ans shut very quickly. any body can shed a light to me ?
I have 200 pcs more 2SK2698 in my box so I decided to use it in this quasi creation but no luck I cant bring the bias current higher than 5mA/pcs if more than that one by one tha transistor will dead. I use 94 volt AC CT trafo I also get trouble with the DC protection it seem can not work with high level it will oscilate open ans shut very quickly. any body can shed a light to me ?
I think the reason why we can't get to the correct biasing is due to the mismatching of the mosfets. And one way to find this out is to measure every voltage across the 0R33 (0R47) resistors and compare.
IMHO, the 100R fuse-replacement procedure is not accurate way to bias the mosfets, unless these resistors are fixed on the pcb.The correct way is to read the current across every mosfet. You may need to increase/decrease the 27R resistors to "force" matching.
IMHO, the 100R fuse-replacement procedure is not accurate way to bias the mosfets, unless these resistors are fixed on the pcb.The correct way is to read the current across every mosfet. You may need to increase/decrease the 27R resistors to "force" matching.
The temporary 100r is a damage limitation component.
It is only used to get the amp running at near working voltage but with zero output bias current.
These 100r resistors allow voltage checks to be made around the circuit to check correct operation of the voltage amp stage.
10r must be fitted to start the biasing procedure.
Quasi says one can complete the bias setting with these 10r in place, but I recommend changing to 1r and eventually to fuse for the final bias setting.
This has been said and repeated recently.
Somebody is not reading!!!!
It is only used to get the amp running at near working voltage but with zero output bias current.
These 100r resistors allow voltage checks to be made around the circuit to check correct operation of the voltage amp stage.
10r must be fitted to start the biasing procedure.
Quasi says one can complete the bias setting with these 10r in place, but I recommend changing to 1r and eventually to fuse for the final bias setting.
This has been said and repeated recently.
Somebody is not reading!!!!
A bit of ground to cover
The Nmos series amplifiers were specifically designed to use power mosfets with characteristics similar to IRFP450s or IRF840s. If other FETs are selected then modifications may be required to get the amp to operate correctly.
Whilst the Nmos series can be run over a wider voltage range than many amps they are still limited. The Nmos200 voltage range is 40 to 50 volts. The Nmos350/500 voltage range is 40 to 85 volts. When selecting your operating voltage, consideration must be given to heatsinking and the SOA of all stages.
One of the downsides of using mosfets in the output stage of this type of AB class amplifier is lower efficiency when compared to the same topology using a transistor output stage. This means that a large main heatsink is required. Look at my build to get the idea. If your heatsink is smaller than this then your amp output stage will run hot (read bad).
The schematic shown on my web site indicates the current flows through a correctly operating amp module. These should be used to calculate the voltages across the various components. If you do not know how to do this then stop building and find out how. For example a 40 volt drop across a 10 ohm setup resistor means a 4 amp current flow (I = V/R) and this equates to a power dissipation of 160 watts (P=IxV). Something is wrong with this amp module.
Take great care with the orientation of the smaller transistors. They are not all the same and the incorrect insertion of just one is likely to damage of few of its friends. This means that if one is damaged then all of the transistors should be tested or replaced. If you have an “in-circuit” transistor tester well and good, if not you will need to remove every transistor to be tested.
I have modified the setup procedure in the Nmos350/500 construction guide and this is presented here;
5. If everything seems ok adjust VR2 to set the output stage bias current, by measuring the
voltage across the positive rail resistor. Adjust for a reading of 3 volts per output FET pair. I.e.
For a 6 FET board set for a voltage of 9 volts. This equates to a bias current of 30mA per
FET pair or 90 mA total. For a 10 FET board set for a voltage of 15 volts.
6. If everything seems ok, check the output offset voltage and adjust VR1 to achieve an offset of
less than 10 mV.
7. All being well switch off, back off the bias control trimmer (VR2) and replace the 100 ohm
resistors with 10 ohm 1 watt resistors. Switch on again and re-adjust VR2 to get 0.3 volts per
FET pair.
There will be a difference in the voltages across the setup resistors. This is because the fuse in the negative rail sits inside the gate voltage reference loop. When a resistor is inserted here it upsets the reference. This is why the final setup is done with 10 ohm resistors, or as Mr T has suggested with 1 ohm resistors.
The fuse location will be changed in the Nmos350 MkII (yes it’s on the way!).
Finally and with sincere respect to all, if much of the above does not make sense to you then please build something simpler (chip-amp?) and gain experience before building this level of amplifier.
Oh a sneak peak at the Nmos350 MkII is presented on condition that it is not yet attempted.
Cheers friends
Quasi
The Nmos series amplifiers were specifically designed to use power mosfets with characteristics similar to IRFP450s or IRF840s. If other FETs are selected then modifications may be required to get the amp to operate correctly.
Whilst the Nmos series can be run over a wider voltage range than many amps they are still limited. The Nmos200 voltage range is 40 to 50 volts. The Nmos350/500 voltage range is 40 to 85 volts. When selecting your operating voltage, consideration must be given to heatsinking and the SOA of all stages.
One of the downsides of using mosfets in the output stage of this type of AB class amplifier is lower efficiency when compared to the same topology using a transistor output stage. This means that a large main heatsink is required. Look at my build to get the idea. If your heatsink is smaller than this then your amp output stage will run hot (read bad).
The schematic shown on my web site indicates the current flows through a correctly operating amp module. These should be used to calculate the voltages across the various components. If you do not know how to do this then stop building and find out how. For example a 40 volt drop across a 10 ohm setup resistor means a 4 amp current flow (I = V/R) and this equates to a power dissipation of 160 watts (P=IxV). Something is wrong with this amp module.
Take great care with the orientation of the smaller transistors. They are not all the same and the incorrect insertion of just one is likely to damage of few of its friends. This means that if one is damaged then all of the transistors should be tested or replaced. If you have an “in-circuit” transistor tester well and good, if not you will need to remove every transistor to be tested.
I have modified the setup procedure in the Nmos350/500 construction guide and this is presented here;
5. If everything seems ok adjust VR2 to set the output stage bias current, by measuring the
voltage across the positive rail resistor. Adjust for a reading of 3 volts per output FET pair. I.e.
For a 6 FET board set for a voltage of 9 volts. This equates to a bias current of 30mA per
FET pair or 90 mA total. For a 10 FET board set for a voltage of 15 volts.
6. If everything seems ok, check the output offset voltage and adjust VR1 to achieve an offset of
less than 10 mV.
7. All being well switch off, back off the bias control trimmer (VR2) and replace the 100 ohm
resistors with 10 ohm 1 watt resistors. Switch on again and re-adjust VR2 to get 0.3 volts per
FET pair.
There will be a difference in the voltages across the setup resistors. This is because the fuse in the negative rail sits inside the gate voltage reference loop. When a resistor is inserted here it upsets the reference. This is why the final setup is done with 10 ohm resistors, or as Mr T has suggested with 1 ohm resistors.
The fuse location will be changed in the Nmos350 MkII (yes it’s on the way!).
Finally and with sincere respect to all, if much of the above does not make sense to you then please build something simpler (chip-amp?) and gain experience before building this level of amplifier.
Oh a sneak peak at the Nmos350 MkII is presented on condition that it is not yet attempted.
Cheers friends
Quasi