medogrizli said:
Once again, look at the datasheet, in fact, the diagram posted above. The current is PER MOSFET, and is in fact about 15A (look at where 25C and 125C curves intersect).
WHY on earth would anyone consider producing 700W of heat as an acceptable way avoid using a simple Vgs multiplier??? But of course, what do I know - maybe you had some divine revelation why thermally compensating the bias current is 'bad'? You have now started several threads and in all of them we get back to what kind of temperature coefficient VMOS has, when it is clearly documented in the relevant datasheets. besides, Prof. Biljanovic is still alive and well, so you can ask him about this. He will probably tell you that ALL MOSFETs have a region of positive and a region of negative coefficient, the only question is what drain current gets you in what region. This is strongly dependant on channel geometry.
Finally, if you still don't believe what a multitude of us (some of which have used MOSFETs by the kg in many amps), you can always ignore what we are saying and make your hexfet amp with no bias compensation, and see for yourself. In the eman time, we might be looking to the horison for that plume of smoke
ilimzn said:
At +-70V rails, yes I am sure he wants single ended
Why not?
I used IRF mosfets loaded by a BJT current sources, with a servo opamp, and powered it from a bipolar source. Nice single ended A class heater.
It scared me once, when a waterfall run on CD. I jumped and turned around, but fortunately found no water in the room.
It uses some IRFPs, I don't remember exactly. Tubes are 12J5 and 12L6 in triode connection, directly coupled through a voltage shift (bi-polar B+ source) and have a global feedback including DC. Also, I used TV compactrons in such design, sound well. However, every tube needs own current to work with minimal distortions. Instead of IRFPs I used IXFKs, the same good result.
Nearly any active element may be used with good results, if used properly.
ilimzn said:
Once again, look at the datasheet, in fact, the diagram posted above. The current is PER MOSFET, and is in fact about 15A (look at where 25C and 125C curves intersect).
WHY on earth would anyone consider producing 700W of heat as an acceptable way avoid using a simple Vgs multiplier??? But of course, what do I know - maybe you had some divine revelation why thermally compensating the bias current is 'bad'? You have now started several threads and in all of them we get back to what kind of temperature coefficient VMOS has, when it is clearly documented in the relevant datasheets. besides, Prof. Biljanovic is still alive and well, so you can ask him about this. He will probably tell you that ALL MOSFETs have a region of positive and a region of negative coefficient, the only question is what drain current gets you in what region. This is strongly dependant on channel geometry.
Finally, if you still don't believe what a multitude of us (some of which have used MOSFETs by the kg in many amps), you can always ignore what we are saying and make your hexfet amp with no bias compensation, and see for yourself. In the eman time, we might be looking to the horison for that plume of smoke
No no not at all for sure termaliy COMEPNSATION IS NOT BAD : I just wanted to discuss theoretically
medogrizli said:
No no not at all for sure termaliy COMEPNSATION IS NOT BAD : I just wanted to discuss theoretically
Practically, this amp has no thermal compensation, but its VAS draws a lot of idle current, and OpAmp needs to be very fast to push crossover distortions above the audible band. It is a 3-step approximation amp, and output FETs and BJTs that work in parallel must be selected such a way FETs provide full current on voltages below B-E break-down of output BJTs.
1'st step: driver supplies output (up to 0.2A), then 2'nd step: BJTs (up to 5A), then 3'rd step: FETs (up to 20A peak).
bogdan_borko said:
Actually, there is a pretty good translation, and it is 'spite'
medogrizli said:
It is of course possible - pretty much everything is, but the question is, is it practical? You have to look at this with a bit of lateral thinking. It's not really about wether you can make a HEXFET amp without thermal compensation, but what kind of topology does not need thermal compensation.
You may also explore what MOSFET you would need to practically remove thermal compensation. One of the types is a lateral MOSFET, but, you also need to look at what other properties of the MOSFET will change.
thanks for translation, but spite is closer to "prkos"...
MEDOGRIZLI, sorry, I was nervous.
You are new in this forum, so when you want something to learn: FIRST search the previous threads, becouse a lot of new members, ask the same questions again and again and that could be iritating sometime...
Cheers brother!
MEDOGRIZLI, sorry, I was nervous.
You are new in this forum, so when you want something to learn: FIRST search the previous threads, becouse a lot of new members, ask the same questions again and again and that could be iritating sometime...
Cheers brother!
So pleae if you have GOOD pdf tutorual book about AB class amplifier design
So usualy books can not be fond on the net freely
or good web links: please send it here or email velikigrizli@gmail.com
thanks
So usualy books can not be fond on the net freely
or good web links: please send it here or email velikigrizli@gmail.com
thanks
Wavebourn said:
Why not?
I used IRF mosfets loaded by a BJT current sources, with a servo opamp, and powered it from a bipolar source. Nice single ended A class heater.
It scared me once, when a waterfall run on CD. I jumped and turned around, but fortunately found no water in the room.
It uses some IRFPs, I don't remember exactly. Tubes are 12J5 and 12L6 in triode connection, directly coupled through a voltage shift (bi-polar B+ source) and have a global feedback including DC. Also, I used TV compactrons in such design, sound well. However, every tube needs own current to work with minimal distortions. Instead of IRFPs I used IXFKs, the same good result.
Nearly any active element may be used with good results, if used properly.
That's an interesting design. Pretty simple. Good PSRR too? The tube running off the common power supply rails? You might get some interest posted on the tube forum.
Sheldon
Check out VZachenko's VHex amps. Excellent sounding amplifiers. http://www.diyaudio.com/forums/solid-state/294632-sons-vhex.html
Laterals are easier to bias as current flow drops through them as they heat up, but they are getting expensive and harder to find. Verticals are cheap and easy to find. Both are very robust devices.
Boards and kits for all the VHex amplifiers are available, along with boards and kits for supplies and protection systems. Virtual Zero Audio - power amplifier products
Boards and kits for all the VHex amplifiers are available, along with boards and kits for supplies and protection systems. Virtual Zero Audio - power amplifier products
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.