Workhorse said:
Am I missing something? what posititive tempco? RDS-on has positive tempco wich is usefull in SMPS switching stuff(where we really dont want to use triode operating area) but gate treshold voltage has negative tempco(around -3...-5mV/C) wich may cause biasing problems.
Lateral mosfets are the easy ones, but availlability and price....
Hi Mikeks,
I have to, as a practical designer, disagree with you on zener clamping of MOSFET gate voltage.
If a MOSFET output is only required to pass 4A per device it is absolutely valid to use a zener that clamps just above the Vgs that achieves the design goals. By allowing 30A or whatever simply because the MOSFET is at very low voltage and can take it is inviting disaster under a meltdown amplifier fault condition - where most every internal current limit is welcome.
Cheers,
Greg
I have to, as a practical designer, disagree with you on zener clamping of MOSFET gate voltage.
If a MOSFET output is only required to pass 4A per device it is absolutely valid to use a zener that clamps just above the Vgs that achieves the design goals. By allowing 30A or whatever simply because the MOSFET is at very low voltage and can take it is inviting disaster under a meltdown amplifier fault condition - where most every internal current limit is welcome.
Cheers,
Greg
Yep.
Almost as bad as the clipping which is the main occasion it would be triggered. And we'd be very thankful for it under a fault condition.
It is unlikely to have any effect on sound quality under normal operating conditions and, in fact, from the market response to the product which was considered very highly, I would say your fears fall on deaf ears.
Cheers,
greg
Almost as bad as the clipping which is the main occasion it would be triggered. And we'd be very thankful for it under a fault condition.
It is unlikely to have any effect on sound quality under normal operating conditions and, in fact, from the market response to the product which was considered very highly, I would say your fears fall on deaf ears.
Cheers,
greg
Hi Anatech,
In that case the driver was a symmetrical Vas running 10ma and current limited under overload to 20mA. No bad effect.
My new design has EF drive but the current to the zeners comes from effective current source resistors not the driver transistors.
I will publish this soon I am just finessing the design.
cheers,
greg
In that case the driver was a symmetrical Vas running 10ma and current limited under overload to 20mA. No bad effect.
My new design has EF drive but the current to the zeners comes from effective current source resistors not the driver transistors.
I will publish this soon I am just finessing the design.
cheers,
greg
The gate drive voltage is not needless curtailed...it is limited...when driving the normal maximum permissible loads the gate voltage seldom exceeds 6V and the current through each of the device in parallel bank of 10 Mosfets is thus maximum of 5A with rails voltages are at +-120VDC ...with a total peak current of 5 X 10 = 50 amperes which is more than enough to drive a 4 ohm load ! .....
P.S: Driving the 4 ohm load with 84volts[120Vpeak] max rms resulted in power output of 1700Watts with RMS current = 21 and thus peak current requirement is 31Amperes max and in practice we could be able to get as much as 50A peak very easily!
K a n w a r
P.S: Driving the 4 ohm load with 84volts[120Vpeak] max rms resulted in power output of 1700Watts with RMS current = 21 and thus peak current requirement is 31Amperes max and in practice we could be able to get as much as 50A peak very easily!
K a n w a r
An amp output relay in combination with fuses in the amps supply lines worked for me for over 20 years, never had a real contact problem and if so replaced the relay. The problem would be a bad contact surface. But these days the relays are so refined it is no issue. Would not dare to use an amp. without relay protection.
Frankly speaking, a good amplifier, when used correctly (e.g. not for welding) will never break. I had some examples when the PSU capacitors dried out resolving in more noise (after 10-15 years of work), but never a full-blown output stage with a 50 V DC output.
I of course realize, that "things may happen", but a "relay+fuse" protection is IMHO much cheaper and easier to implement, than the "power mosfet+current sensor" type.
I of course realize, that "things may happen", but a "relay+fuse" protection is IMHO much cheaper and easier to implement, than the "power mosfet+current sensor" type.
Hi Kanwar,
There was no worry about total o/p current. Paralleled outputs are fairly common and the concept is well understood. I have designed amps with FETs, I just don't care for them as much. My preference.
I also agree with fuses and relays, again, my preference. I've beat that one to death.
-Chris
There was no worry about total o/p current. Paralleled outputs are fairly common and the concept is well understood. I have designed amps with FETs, I just don't care for them as much. My preference.
I also agree with fuses and relays, again, my preference. I've beat that one to death.
-Chris
anatech said:
Hi Chris,
Sorry friend I failed to understand what you are saying...
Every high power amplifier has parallel output devices and relays and fuses although but i donot able to grasp the context of your saying ..please clarify a little bit more...
Or do you want say that the zener SOA protection used by us doesnt worth, If yes its only your opinion buddy...
K a n w a r
Hi Kanwar,
I was responding to your previous post ...
If you want to limit the gate current, the zener works. Especially if you don't want to bother detecting the actual current and heatsink temperature. I really don't have a problem with this.
-Chris
(edit because I hit the wrong key to post after Hi Kanwar)
I was responding to your previous post ...
It seemed like you didn't think I understood the concept. I am responding to that.
If you want to limit the gate current, the zener works. Especially if you don't want to bother detecting the actual current and heatsink temperature. I really don't have a problem with this.
-Chris
(edit because I hit the wrong key to post after Hi Kanwar)
Hi Chris,
Thanks for your reply...
In the event of Short circuit , the output current is also detected .. when its crosses the reference value the protection circuit activates and mutes the input audio signal and also the power supply rails are turned off...
regarding heatsink temp measurement, yes it is of concern in BJT's but in mosfets this doesnot count to that extent..we do have overtemp protection which is another way of safety measure...
K a n w a r
Thanks for your reply...
In the event of Short circuit , the output current is also detected .. when its crosses the reference value the protection circuit activates and mutes the input audio signal and also the power supply rails are turned off...
regarding heatsink temp measurement, yes it is of concern in BJT's but in mosfets this doesnot count to that extent..we do have overtemp protection which is another way of safety measure...
K a n w a r
Re: All eyes on me!!!!!
This is totaly untrue...
Elementary...really...
When a 'short circuit is encountered' you cannot then have 'feedback action'.....
....assuming of course that your short-circuit is to ground, then by definition your feedback network is in parallel with zero ohms!
Which necesarily means feedback is disabled....
viz: there is no voltage to feedback.....
Workhorse said:
This is totaly untrue...
Elementary...really...
When a 'short circuit is encountered' you cannot then have 'feedback action'.....
....assuming of course that your short-circuit is to ground, then by definition your feedback network is in parallel with zero ohms!
Which necesarily means feedback is disabled....
viz: there is no voltage to feedback.....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.