I'll guess then, that IC1, T2, T2' are a DC servo.
And if I didn't miss spotting another gain stage, voltage gain is about 80. So, really, no NFB. Interesting concept.
Regards,
Peter Jacobi
And if I didn't miss spotting another gain stage, voltage gain is about 80. So, really, no NFB. Interesting concept.
Regards,
Peter Jacobi
Re: Re: Re: Re: Commentable Thoughts
Really? Vbe changes -2.2mV per deg C. Vgs threshold voltage changes at about -5.5mV per deg C. That means you are going to have more than double the thermal stability problem with a mosfet as with a bipolar. Please explain ampman...???amp_man_1 said:
hi.
circlotron are you taking amp_man serious?
(i didnt make that mistake , not when he was factor or whatever.....)
rgds karsten madsen - www.cadaudio.dk
>Dont woory friend , Ampman will help u in resolving the amplifier situation.>
🙂 sure thing.....
circlotron are you taking amp_man serious?
(i didnt make that mistake , not when he was factor or whatever.....)
rgds karsten madsen - www.cadaudio.dk
>Dont woory friend , Ampman will help u in resolving the amplifier situation.>
🙂 sure thing.....
Re: Re: Re: Re: Re: Commentable Thoughts
Hi Circlotron,
I'm pretty clueless WRT amp design, but I'm pretty sure the description of my series 200 amp (from EA 1986) mentions something about this, I'll have a look when I get home.
edit: Frisbee,
Don't worry too much, with 1000W you will have a hell of a lot of headroom, hopefully at 10W or so the THD will still be very low.
Tony.
Circlotron said:
Hi Circlotron,
I'm pretty clueless WRT amp design, but I'm pretty sure the description of my series 200 amp (from EA 1986) mentions something about this, I'll have a look when I get home.
edit: Frisbee,
Don't worry too much, with 1000W you will have a hell of a lot of headroom, hopefully at 10W or so the THD will still be very low.
Tony.
Re: Commentable Thoughts
Seriously Ampman,
I'm curious how any of your amps with more than just a single pair of HEXFET(if you have any, let us say an Amp of yours with 3 pair), which are not using any kind of source resistor or other thermal protection(as I understood you don't need in your amps with IRF FET's, right?) would perform if you use sinus (not square!) signal and drive the output voltage to 2/3 of the rail voltage (note, not 2/3 of max output) into a 8 or 4 ohm load?
I just wan't to say that I really take you seriously now, would appreciate verry much to hear what your results are when performed in the way I asked stated above.
Reffering also to: http://www.diyaudio.com/forums/showthread.php?postid=304119#post304119
Thanks in advance! 🙂
amp_man_1 said:
Seriously Ampman,
I'm curious how any of your amps with more than just a single pair of HEXFET(if you have any, let us say an Amp of yours with 3 pair), which are not using any kind of source resistor or other thermal protection(as I understood you don't need in your amps with IRF FET's, right?) would perform if you use sinus (not square!) signal and drive the output voltage to 2/3 of the rail voltage (note, not 2/3 of max output) into a 8 or 4 ohm load?
I just wan't to say that I really take you seriously now, would appreciate verry much to hear what your results are when performed in the way I asked stated above.
Reffering also to: http://www.diyaudio.com/forums/showthread.php?postid=304119#post304119
Thanks in advance! 🙂
Public apology to ampman
One thing I did forget about was that despite mosfets having a much greater thermal sensitivity of gate voltage threshold than bipolars with Vbe, they also have a much lower transconductance; i.e. 1mV change in bipolar base voltage makes a much bigger difference than 1mV change in fet gate voltage. So, I wasn't really completely right.
/Circlotron - must be having a bad week, but no excuse for biting ampman.
One thing I did forget about was that despite mosfets having a much greater thermal sensitivity of gate voltage threshold than bipolars with Vbe, they also have a much lower transconductance; i.e. 1mV change in bipolar base voltage makes a much bigger difference than 1mV change in fet gate voltage. So, I wasn't really completely right.
/Circlotron - must be having a bad week, but no excuse for biting ampman.
Also, when the MOSFET heats up, you will generaly have an increase in Ron which will act as a sort of negative feed back for the amout of curent flowing through the device.
But in any case you may find, as long as you keep thing cool enough in the first place you should be fairly safe with either typ of output devices. Right?
But in any case you may find, as long as you keep thing cool enough in the first place you should be fairly safe with either typ of output devices. Right?
Commentable Thoughts
Mr Ultima Thule I have Performed the test on my amp which u have asked me to perform.
Here are some specifications:
VCC=+-58VDC
Filter Capacitance=22000MFd/63V per rail.
Load=2 Ohms PURE RESISTIVE LOAD
Mosfets per rail = 3 IRFP250N with Vgs Matching to +-50mV
Composite Chamber Turbulance Heatsink with 120mm AC220V Fan Cooling.
Input Signal=20Hz SINEWAVE Signal with amplitude adjusted to give 2/3 of rail voltage i.e. 38VAC RMS
Time Duration =2Hours Continous.
#################################
Now the results
HEATSINK Temperature=55degree CELCIUS
Thermal Runaway=NIL
VGS change = 3.7V at cold To 3.9V at Full Drive
SO Whats ur opinion Now!
Regards
AmpMan
Ultima Thule said:
Mr Ultima Thule I have Performed the test on my amp which u have asked me to perform.
Here are some specifications:
VCC=+-58VDC
Filter Capacitance=22000MFd/63V per rail.
Load=2 Ohms PURE RESISTIVE LOAD
Mosfets per rail = 3 IRFP250N with Vgs Matching to +-50mV
Composite Chamber Turbulance Heatsink with 120mm AC220V Fan Cooling.
Input Signal=20Hz SINEWAVE Signal with amplitude adjusted to give 2/3 of rail voltage i.e. 38VAC RMS
Time Duration =2Hours Continous.
#################################
Now the results
HEATSINK Temperature=55degree CELCIUS
Thermal Runaway=NIL
VGS change = 3.7V at cold To 3.9V at Full Drive
SO Whats ur opinion Now!
Regards
AmpMan
I think this is the relevent bit.......
quote from Electronics Australia January 1985 Page 27.
Sorry if I've completely mis-interpreted this and it isn't what AmpMan was talking about.
Tony.
quote from Electronics Australia January 1985 Page 27.
Sorry if I've completely mis-interpreted this and it isn't what AmpMan was talking about.
Tony.
Commentable Thoughts
Mr.Frisbee: The Thermal compensation is inadequate because
Their is no special compensation technique [apart from diodes and resistors] used in this amp to prevent thermal runaway. The bias is setup at fixed value and doesn't have any means to alter it.
Regards
Ampman
Frisbee78 said:
Mr.Frisbee: The Thermal compensation is inadequate because
Their is no special compensation technique [apart from diodes and resistors] used in this amp to prevent thermal runaway. The bias is setup at fixed value and doesn't have any means to alter it.
Regards
Ampman
Amp_man,
Apprently, and if I understand right, there is thermal compensation on this design.
First, T3 & T4 are used as diodes to compensate the Vbe of T5 & T6. To ensure stabilty between the two pairs, T3 & T5 are physically linked together and so are T4 & T6.
Elsewhere, we have D4 linked with T11 & D7 with T14 for the same purpose.
Same again for D5 & T15 and D6 & T16.
And finally, T12 & T13 are mounted on the same heatsink with the power transistors T17, T18, T19, T20, T21 & T22.
Your thought?
Apprently, and if I understand right, there is thermal compensation on this design.
First, T3 & T4 are used as diodes to compensate the Vbe of T5 & T6. To ensure stabilty between the two pairs, T3 & T5 are physically linked together and so are T4 & T6.
Elsewhere, we have D4 linked with T11 & D7 with T14 for the same purpose.
Same again for D5 & T15 and D6 & T16.
And finally, T12 & T13 are mounted on the same heatsink with the power transistors T17, T18, T19, T20, T21 & T22.
Your thought?
Re: Commentable Thoughts
Hi Ampman,
I think it's great, 38VAC RMS into 2 Ohm is lot of power!
Did you check btw the case temperature for each FET?
You say you have 3,9 Vgs at full drive, I don't know if I'm just unable to interpret the datasheet of IRFP250N but the treshold is in Fig 6 around 5 volt??
http://www.irf.com/product-info/datasheets/data/irfp250n.pdf
Could you give the theory why IRF FET's can be parallelled without the Source resistor?
In Fig 3 the Vds is 50 V and for the lowest Vgs pictured is 4,5 Vgs and the current varies from around 1,5A@Tj25C to 7A@Tj175C, I can't recall for the moment what the Vgs use to be for Vds lower than 50V...
What do you think?
amp_man_1 said:
Hi Ampman,
I think it's great, 38VAC RMS into 2 Ohm is lot of power!
Did you check btw the case temperature for each FET?
You say you have 3,9 Vgs at full drive, I don't know if I'm just unable to interpret the datasheet of IRFP250N but the treshold is in Fig 6 around 5 volt??
http://www.irf.com/product-info/datasheets/data/irfp250n.pdf
Could you give the theory why IRF FET's can be parallelled without the Source resistor?
In Fig 3 the Vds is 50 V and for the lowest Vgs pictured is 4,5 Vgs and the current varies from around 1,5A@Tj25C to 7A@Tj175C, I can't recall for the moment what the Vgs use to be for Vds lower than 50V...
What do you think?
Re: Re: Commentable Thoughts
Listen Man The theory is simple . The positive temp coeff. plays its role very well, whenever a mosfet gets hot its Rds increases , this inturns limit the current flow through it and helps in current sharing. But CONDITIONS APPLY
1 Source resistors can only be eliminated with Vcc around +-50VDC with MOSFET IRFP250N and above series.
2 Heatsinking must be efficient such as force cooling.
3 Matching of Mosfets to +-50mV tolerance of Vgs
4 They must be in good thermal coupling i.e. they must be mounted together as close as possible.
Secondly, When u apply VCC around +-95VDC , then the situation becomes worst and u have to implement source resistors to compensate for thermal stability.
The case Temp was around 62-65degree Celcius
And regarding the comparision of ratings with datasheet is that the test is basically performed with pulsed signal rather than continous signal.
Regards
Ampman
Ultima Thule said:
Listen Man The theory is simple . The positive temp coeff. plays its role very well, whenever a mosfet gets hot its Rds increases , this inturns limit the current flow through it and helps in current sharing. But CONDITIONS APPLY
1 Source resistors can only be eliminated with Vcc around +-50VDC with MOSFET IRFP250N and above series.
2 Heatsinking must be efficient such as force cooling.
3 Matching of Mosfets to +-50mV tolerance of Vgs
4 They must be in good thermal coupling i.e. they must be mounted together as close as possible.
Secondly, When u apply VCC around +-95VDC , then the situation becomes worst and u have to implement source resistors to compensate for thermal stability.
The case Temp was around 62-65degree Celcius
And regarding the comparision of ratings with datasheet is that the test is basically performed with pulsed signal rather than continous signal.
Regards
Ampman
Re: Re: Re: Commentable Thoughts
that's true. But that million dollar question is "will the increase in Rds be enough?".
I can assure you that it will not be for at least IRF mosfets. Yes, you can and will blow them.
amp_man_1 said:
that's true. But that million dollar question is "will the increase in Rds be enough?".
I can assure you that it will not be for at least IRF mosfets. Yes, you can and will blow them.
Thanks Ampman, I appreciate your explanatory answer!
I just getting confused when interpreting the datasheet in this case for IRFP250N.
That the Rds on increase with rising temperature is known for me but I didn't know it would compensate also for the increasing Id for a given Vgs which is at least around 4,5 times for Vgs 4,5 under the conditions mentioned in Fig 3 datasheet.
As you mentioned a tight physical mounting of the FET's and a proper heatsink with forced cooling will surely help the situation, you seem to have found a sollution that works anyhow.
You mentioned that source resistors can be eliminated for IRFP250N and above working with rail voltage of around +-50 V, I have some IRFP630N and probably some older IRFP250 around, would it work for them too?
You mentioned also about 95 V rail voltage would be worst, what parameter in the fet does change in the negative direction, is it the change of Vgs or Rds on or else, how much?
Thanks for your input and Regards!
I just getting confused when interpreting the datasheet in this case for IRFP250N.
That the Rds on increase with rising temperature is known for me but I didn't know it would compensate also for the increasing Id for a given Vgs which is at least around 4,5 times for Vgs 4,5 under the conditions mentioned in Fig 3 datasheet.
As you mentioned a tight physical mounting of the FET's and a proper heatsink with forced cooling will surely help the situation, you seem to have found a sollution that works anyhow.
You mentioned that source resistors can be eliminated for IRFP250N and above working with rail voltage of around +-50 V, I have some IRFP630N and probably some older IRFP250 around, would it work for them too?
You mentioned also about 95 V rail voltage would be worst, what parameter in the fet does change in the negative direction, is it the change of Vgs or Rds on or else, how much?
Thanks for your input and Regards!
hi.
>I can assure you that it will not be for at least IRF mosfets. Yes, you can and will blow them.>
so true , we have built power amplifiers for years with irfp power fets and i can only agree.
we also built amplifiers with hitachi fets (2sk and 2sj) and thats another story.
if you like make one and see for yourself 🙂 (buy extra fets)
rgds karsten madsen - www.cadaudio.dk
ps. amp_man , again you dont know what you are talking about , build one first , talk about it later , ok 🙂
>I can assure you that it will not be for at least IRF mosfets. Yes, you can and will blow them.>
so true , we have built power amplifiers for years with irfp power fets and i can only agree.
we also built amplifiers with hitachi fets (2sk and 2sj) and thats another story.
if you like make one and see for yourself 🙂 (buy extra fets)
rgds karsten madsen - www.cadaudio.dk
ps. amp_man , again you dont know what you are talking about , build one first , talk about it later , ok 🙂
Re: Re: Re: Commentable Thoughts
Of course, this all presumes that the Rds *increase* with temperature has a greater effect than the Vgs threshold *decrease* with temperature =at the same time=.
In a saturated switching circuit, e.g. a SMPS it certainly will. In a linear amp cct...
not so certain.
I have often thought (but never tried it) that paralleled mosfets would be better off with separate heatsinks rather than thermally tied to their neighbours. If a mosfet gets hotter because of hogging the current, it's temperature increase will be greater if it is thermally isolated from the others. This greater temperature increase will result in a greater reduction of current back toward what it's neighbours are carrying.amp_man_1 said:
Of course, this all presumes that the Rds *increase* with temperature has a greater effect than the Vgs threshold *decrease* with temperature =at the same time=.
In a saturated switching circuit, e.g. a SMPS it certainly will. In a linear amp cct...
not so certain.
km said:
Just for the record, the quote I did from EA earlier in the thread was using Hitachi 2SK134 and 2SJ49...... didn't realise the discussion was WRT a particular brand.
Tony.
Commentable Thoughts
LISTEN MR KARSTEN 'MAD'SEN
I dont want to indulge in any controversy , U have ur own experience and Ampman has its own experience with IRF mosfets.
Secondly, I also have a disscussion with IRF's Technical Team a year ago about this type of implementation, and they have responded positively in my favour with full technical reference.
So Mr 'Mad'sen thinks that I talk about the amp first and builds it later.
So What about all the amplifiers with Mosfets which i have built in past years and which are still working in excellent conditions at various DJ's and Discotheques without any single breakdown. So u think that All these amplifiers are just blown away at their first startup and only their GHOST or SPIRIT is performing amplification!
Mr MAD'SEN u might seems to get in thermal Runaway as ur mind is overbiased.So kindly use Source resistors.[u said it]
Regards
Ampman
Ampman Respects the MODERATORS AND RULES AND REGULATION OF THIS FORUM
km said:
LISTEN MR KARSTEN 'MAD'SEN
I dont want to indulge in any controversy , U have ur own experience and Ampman has its own experience with IRF mosfets.
Secondly, I also have a disscussion with IRF's Technical Team a year ago about this type of implementation, and they have responded positively in my favour with full technical reference.
So Mr 'Mad'sen thinks that I talk about the amp first and builds it later.
So What about all the amplifiers with Mosfets which i have built in past years and which are still working in excellent conditions at various DJ's and Discotheques without any single breakdown. So u think that All these amplifiers are just blown away at their first startup and only their GHOST or SPIRIT is performing amplification!
Mr MAD'SEN u might seems to get in thermal Runaway as ur mind is overbiased.So kindly use Source resistors.[u said it]
Regards
Ampman
Ampman Respects the MODERATORS AND RULES AND REGULATION OF THIS FORUM
- Status
- Not open for further replies.