Vivek said:
When the amp is run at high power the emitter of T10 will be at +46v for brief moments when the output approaches the +ve rail. The collector of T10 is tied to the negative rail via a 220 ohm resistor and this is locked at the 50v rail voltage. So we have a voltage differential of 94 volts. The same applies for T9 only in reverse.
Although only brief, this condition is way beyond the Vce of the BD139 / 140 device.
Cheers
Q
Thanks guys. I will try to find them in the local market again.
PS: Quasi, do you watch cricket?
PS: Quasi, do you watch cricket?
ashok said:
Ashok
Is there any good source where i can buy quality cabinets for amp.(not in bulk)
Regards
Sony Soman
Rated Ripple Vs Capacitance
Hi All,
I know there are various suggested formulas for mF per Ipk or Irms. Is the rated ripple or the capacitance (mF) of an electrolytic capacitor more important in determining how many electrolytics are used in parallel for the PSU?
I also know there is a preference to using two or more electrolytic capacitors in parallel to reduce ESR. I have read mixed opinions on the use of small film capacitors to bypass electrolytics to reduce ESR. For now I like to keep the question simple and just in the context of the question posed.
Regards,
John L. Males
Willowdale, Ontario
06 October 2007 16:10
Official Quasi Thread Researcher
Hi All,
I know there are various suggested formulas for mF per Ipk or Irms. Is the rated ripple or the capacitance (mF) of an electrolytic capacitor more important in determining how many electrolytics are used in parallel for the PSU?
I also know there is a preference to using two or more electrolytic capacitors in parallel to reduce ESR. I have read mixed opinions on the use of small film capacitors to bypass electrolytics to reduce ESR. For now I like to keep the question simple and just in the context of the question posed.
Regards,
John L. Males
Willowdale, Ontario
06 October 2007 16:10
Official Quasi Thread Researcher
Hi Sony,
This is for most city based DIY'ers . There are lots of small shops that work with sheet metal. You will need someone with a shear to cut sheets , a bending machine and drilling machine ( unless you do that yourself ).
Most work with steel sheets and might not have aluminum sheets that they will cut up to your size but there are many small shops that will sell small pieces of metal of various thickness . Aluminum from about 0.5 mm to about 4mm is what I have seen in small sheet metal shops. You can buy pieces as small as 2sq feet !
If you need enough the machine shop might order the sheet directly from the suppler and get your job done.
I would say you should find out where your small scale industries are located and visit the place and walk into a few small factories and ask someone to direct you around. It always works and it does take some time. Every city is the same over here. Only thing is that in places like Delhi and Mumbai the factories might be a bit far away from the center of town.
You can carry in your own sheet metal but that is not possible in some cities unless you use your own transport ( not 2 wheeler ).
There are small scale factory directories and also the Yellow Pages directory . You can search through that to get a starting point . A phone call is generally not good enough . You might need to visit the place.
You can get it done but you need to find out where unless some helpful soul tells you exactly where to go. Have any friends who own or work in small scale industries ?
Cheers.
This is for most city based DIY'ers . There are lots of small shops that work with sheet metal. You will need someone with a shear to cut sheets , a bending machine and drilling machine ( unless you do that yourself ).
Most work with steel sheets and might not have aluminum sheets that they will cut up to your size but there are many small shops that will sell small pieces of metal of various thickness . Aluminum from about 0.5 mm to about 4mm is what I have seen in small sheet metal shops. You can buy pieces as small as 2sq feet !
If you need enough the machine shop might order the sheet directly from the suppler and get your job done.
I would say you should find out where your small scale industries are located and visit the place and walk into a few small factories and ask someone to direct you around. It always works and it does take some time. Every city is the same over here. Only thing is that in places like Delhi and Mumbai the factories might be a bit far away from the center of town.
You can carry in your own sheet metal but that is not possible in some cities unless you use your own transport ( not 2 wheeler ).
There are small scale factory directories and also the Yellow Pages directory . You can search through that to get a starting point . A phone call is generally not good enough . You might need to visit the place.
You can get it done but you need to find out where unless some helpful soul tells you exactly where to go. Have any friends who own or work in small scale industries ?
Cheers.
Quasi,
For Nmos500 should i use IRFP450 like for Nmos350 ?
I have doubts (i don't know how to calculate ) about the gate capacitance that T9,T10 must drive ... If i use IRFP450 is 26nF but for IRFP460 if 42nF ...
Do you take care of this ? I just want to be sure that everything it's ok ... and I want to learn. Sone formulas will be apreciated 🙂
And please tell me if i'm correct if I say : in the first and second stage the signal si aplified in class A and from the third to the output in complementary AB class. 🙄
--------
PS : I have done my preamp with TL072 for my Nmos200 version but it adds some noise. Should i make a single transistor preamp with low noise 2SC1845 or should i raise the gain of Nmos200 ... i don't know what to do.
I tested the Nmos200 with sinusoidal. signal
On 8 ohms i got 100W and the amp was working werry well, no distorsion 😀 Cheers !
For Nmos500 should i use IRFP450 like for Nmos350 ?
I have doubts (i don't know how to calculate ) about the gate capacitance that T9,T10 must drive ... If i use IRFP450 is 26nF but for IRFP460 if 42nF ...
Do you take care of this ? I just want to be sure that everything it's ok ... and I want to learn. Sone formulas will be apreciated 🙂
And please tell me if i'm correct if I say : in the first and second stage the signal si aplified in class A and from the third to the output in complementary AB class. 🙄
--------
PS : I have done my preamp with TL072 for my Nmos200 version but it adds some noise. Should i make a single transistor preamp with low noise 2SC1845 or should i raise the gain of Nmos200 ... i don't know what to do.
I tested the Nmos200 with sinusoidal. signal
On 8 ohms i got 100W and the amp was working werry well, no distorsion 😀 Cheers !Hi Ashok
Thanks for the info. There are some guys from whom i used to get inverter cabinets made long time back. Let me find out if they can make one or two amp cabinets
Cheers
Sony Soman
Thanks for the info. There are some guys from whom i used to get inverter cabinets made long time back. Let me find out if they can make one or two amp cabinets
Cheers
Sony Soman
For Marus.
"...I have done my preamp with TL072 for my Nmos200 version but it adds some noise..."
Use a different opamp. If the FET input is not required and if the resistors are reasonably low value ( 20 K region , to minimise offset ) there are a whole lot of dual opamps that you can use .
Easiest one to get would probably be the NE5532 . The other popular one is the OPA2134 .
Remove your TL072 and put a DIP socket in its place then you can experiment with different opamps easily.
"...I have done my preamp with TL072 for my Nmos200 version but it adds some noise..."
Use a different opamp. If the FET input is not required and if the resistors are reasonably low value ( 20 K region , to minimise offset ) there are a whole lot of dual opamps that you can use .
Easiest one to get would probably be the NE5532 . The other popular one is the OPA2134 .
Remove your TL072 and put a DIP socket in its place then you can experiment with different opamps easily.
I looked around in the local market but could not find the MJE340/350. The shops seemed to have run out of stock. Any other transistor I could use for the NMOS200?
Hi Marus
For the intended application (domestic hi-fi) you should use 10 x IRFP450 or similar. This should drive comfortably into 4 ohms. If you intend to use the amp for PA then you should use IRFP460, but note that I'm not supporting this and you will have to make your own judgement. The drivers should be able to handle the extra capacitive load but I'm worried that in PA applications the whole thing will just run too hot. Use big heatsinks.
That is correct.
Ashok's got the right idea here......
Cool ...well done. This is becoming a popular module. It'll be great to see them in a 5.1 or 7.1 setup.
Cheers friends
Quasi
marus said:
For the intended application (domestic hi-fi) you should use 10 x IRFP450 or similar. This should drive comfortably into 4 ohms. If you intend to use the amp for PA then you should use IRFP460, but note that I'm not supporting this and you will have to make your own judgement. The drivers should be able to handle the extra capacitive load but I'm worried that in PA applications the whole thing will just run too hot. Use big heatsinks.
marus said:
That is correct.
marus said:
Ashok's got the right idea here......
marus said:
I tested the Nmos200 with sinusoidal. signal
Cool ...well done. This is becoming a popular module. It'll be great to see them in a 5.1 or 7.1 setup.
Cheers friends
Quasi
Hi All,
I am curious if anyone has done a SPICE simulation of the NMOS350?
I tried using a spice program and the negative side of the waveform either does not exist or is -12V, -9V, 0V or such when it should be -25V. Often the negative side of the waveform is jut chopped off. I have tried a variety of "creative" values for the LTP/Current mirror with generally limited improvement and odd time success. The instant I have success and change the rail voltages even 10V less I end up with the same chopped of negative output waveform.
Probing about the source of the problem might be the LTP stage, but not certain yet.
I have to assume the problem is either the models or some error in my inputing the schematic into the GUI interface of the spice program. Any thoughts where I need to look right off start for such a basic problem?
Regards,
John L. Males
Willowdale, Ontario
Canada
09 October 2007 13:27
09 October 2007 13:43 Typo correction. jlm
Official Quasi Thread Researcher
I am curious if anyone has done a SPICE simulation of the NMOS350?
I tried using a spice program and the negative side of the waveform either does not exist or is -12V, -9V, 0V or such when it should be -25V. Often the negative side of the waveform is jut chopped off. I have tried a variety of "creative" values for the LTP/Current mirror with generally limited improvement and odd time success. The instant I have success and change the rail voltages even 10V less I end up with the same chopped of negative output waveform.
Probing about the source of the problem might be the LTP stage, but not certain yet.
I have to assume the problem is either the models or some error in my inputing the schematic into the GUI interface of the spice program. Any thoughts where I need to look right off start for such a basic problem?
Regards,
John L. Males
Willowdale, Ontario
Canada
09 October 2007 13:27
09 October 2007 13:43 Typo correction. jlm
Official Quasi Thread Researcher
Hi Keypunch,
I do a SPICE simulation. And it's woking verry well. 🙂
I'm not sure why do you say it should be -25V. In my Nmos300/8ohms it's about 70V at full power (with +/-80V rails).
This is how it looks (10Khz sine wave)
What program do you use ? If it's Orcad PSpice I can give you my simulation.
I do a SPICE simulation. And it's woking verry well. 🙂
I'm not sure why do you say it should be -25V. In my Nmos300/8ohms it's about 70V at full power (with +/-80V rails).
This is how it looks (10Khz sine wave)
What program do you use ? If it's Orcad PSpice I can give you my simulation.
Thanks Hari. But I was thinking instead of you or someone else taking the trouble of sending it all the way to me, if there was some other transistor I could possibly use things would be simpler.
marus said:
Hi Marus,
I suspected from your postings and images you had done a simulation, but I was not sure with what. I used LTSpice, which definitely showed Class B crossover distorition when I am able to get it to show a proper +-. The reason I said should be -25 was based on the simulation input signal I should have a 25V P-P output signal. I was seeing mostly a +25 with a generall a clipped -12V, -9V, -3V or 0V negative rail output signal. The odd time the negative rail signal was just much smaller, but a sine wave. When I make some gross changes to a number of resistor values in the current mirror and LTP stage I can get the expected P-P output signal. The problem is as soon as I change the rail voltage in the LTspice simulation I the negative side of the P-P output signal clips, even when I increase the rail voltage. The extent of the value changes, when I can find them and the overall problem suggests to me either LTspice is the problem or the models I have used to do the simulation with.
I believe the Orcad demo will not alow me to simulate the NMOS350. PSpice I am not sure as it is an old program if I recall and I am not sure if the PSpice demo will have restrictions that will apply to the NMOS350. I certainly would be interested in your file if you know what I can use, even if demo version.
If you can export the spice models for all of the components you used to start with I will try to import them to LTSpice and see if that improves the LTspice simulation results to me more in line with all of the builders real world experiences. I know if there was crossover distortion builders would have heard and seen it. Of course if the negative rail based output signal was being clipped as I have seen in the LTspice simulation there would of been all sorts of builder comments, not to mention quasi never had this problem since he posted the design.
I used the Toshiba spice model for the 2SC2240. I could not find a 2SC1845 spice model. I used the Philips spice files for the BC546B and BC556B. I even used the BC546B for the LTP pair just in case the 2SC2240 spice model was suspect. I used the Motorola and Failchild MJE340/350 spice files. I used the IRFP450 subck file for the output devices.
Regards,
John L. Males
Willowdale, Ontario
Canada
10 October 2007 (11:38 -) 12:09
Official Quasi Thread Researcher
Keypunch,
I have the simulation in full version of Orcad PSpice and it don't work in demo version, i've tryed. I don't know how can i help you...
Quasi,
In Nmos200 it's a 0,1uF capacitor in parallel with T8. In the Nmos350 schematics it isn't but it have a place on the PCB.
So ... what should i do with it for Nmos350 ?
I have the simulation in full version of Orcad PSpice and it don't work in demo version, i've tryed. I don't know how can i help you...
Quasi,
In Nmos200 it's a 0,1uF capacitor in parallel with T8. In the Nmos350 schematics it isn't but it have a place on the PCB.
So ... what should i do with it for Nmos350 ?
ltspice simulation
Hi John,
I have done simulation in Ltspice. I am posting the files here, maybe it can be of some help to you somehow.
Cheers
Biren
Hi John,
I have done simulation in Ltspice. I am posting the files here, maybe it can be of some help to you somehow.
Cheers
Biren
marus said:
Hi Marus,
I see if I can get access to a full version of Orcad. Thanks so much for taking the time to try your simulation with the demo version of Orcad. Orcad just says demo has limits based on "complexity". That is so vague. At least other Spice demo articulate a pin or componebt limit. I know quasi's amp is simple, but most demos will not handle either the pin count, component count or have the devices needed for the amp.
Marus,
If I may jump in here. In the early versions of the NMOS350 schematic the 0.1uF capacitor in parallel with T8 and also in series with C4 was missing from the NMOS350 schematic, but was on the PCB. That omission was corrected sometime ago by quasi. The 0.1uF capacitor in parallel with T8 is known on the NMOS350 schematic as C14. Please note that the NMOS350 schematic with the current flows does not have the 0.1uF capacitor in parallel with T8 as it predated the correction to the schematic.
Regards,
John L. Males
Willowdale, Ontario
Canada
12 October 2007 (11:05 -) 11:53
Official Quasi Thread Researcher