If you are using k170/j74 from the lower end of the bl grade be careful that the bias does not exceed the Idss.
No one suggests that the bias of a jFET should exceed the Idss.
Some experts have said and repeated the message that operation (implying AC duty cycle) can take them into the Idpk>Idss area.
No one suggests that the bias of a jFET should exceed the Idss.
Some experts have said and repeated the message that operation (implying AC duty cycle) can take them into the Idpk>Idss area.
Did you try your alternative N-Channels FETs from Fairchild on the board?
Yes, I am going to have transfer curves of the FQA16n25C vs FQA19n20C in next day or so. I am also going to include the FQP 19n20C in the mix. I have had both FQP and FQN versions on the board, and both have functioned perfectly. The F5/F5T is a REALLY easy build and a great project for first timers IMO. I will be playing music through both channels tonight with bias at 400mV. If all is ok, I am going to hotbox them and see if that changes anything, heeding Jacco's wisdom of real life amp temps. I
that numbers on a datasheet are average numbers.
I measure voltage drop on LEDs, zeners and diodes. Matching MOSFETs to 10mV Vgs spread, then hooking it up to a random pick diode bit strange imo, better safe than sorry.
(i compensate lack of wisdom, by thinking up excuses for still having close to 1k in power diode duallies burried in the back yard)
So i have some disappointing news. I hooked up both versions of the F5T V3 to parallel set of these.
http://www.creativesound.ca/pdf/CSS-EL70-data-080909.pdf
2ohm and 86dB doesnt make for an efficient load, or so I thought. Both versions with just a single pair of output devices could easily make for an uncomfortable listetning session yet neither version even raised the heatsink temperatures. This reminded me of one of the really great assets of the F5, GRUNT. I am not sure i have enough of a difficult load to push these things. I cannot imagine what the 4 pair version is capable of. I guess i will have to hook both channels up and give a listen tomorrow. A little critique for the group Anybody got any big subwoofers they want to send my way. I guess i can try a pari of parallel ca22rnx's that i have. If that doesnt work, i could try powering my neighbors Prius. Enough screwing around, hope to have more info tomorrow. Interesting point, the freestanding diode version showed no signs of sweating.
http://www.creativesound.ca/pdf/CSS-EL70-data-080909.pdf
2ohm and 86dB doesnt make for an efficient load, or so I thought. Both versions with just a single pair of output devices could easily make for an uncomfortable listetning session yet neither version even raised the heatsink temperatures. This reminded me of one of the really great assets of the F5, GRUNT. I am not sure i have enough of a difficult load to push these things. I cannot imagine what the 4 pair version is capable of. I guess i will have to hook both channels up and give a listen tomorrow. A little critique for the group
Anybody got any big subwoofers they want to send my way. I guess i can try a pari of parallel ca22rnx's that i have. If that doesnt work, i could try powering my neighbors Prius. Enough screwing around, hope to have more info tomorrow. Interesting point, the freestanding diode version showed no signs of sweating.
What the diodes do for you is get the full available power from a given output fet. So if your thing is to hear the diodes, then you should build an amp that may seem borderline power wise under number F5 conditions, and what you will get is more power simply as a result of the diodes present. Under full power, a single output pair, can source about 12a off current due to the Vdrop os Rs. With the diodes, you get the full 28A or therabouts. I am still not sure about the whole Class A vs A/B figure in this scenario, but you can see that you double your avaiable power just by adding the diodes. If I am wrong, I hope someone will correct me.
You also delve into the questions of whether or not an amp sounds better having more power than needed. Some folks like having a 200W amp even though they will never use more than 50W. Perhaps you can hear a difference. Nelson has suggested that his customers say as much as you move through the progression of his Pass stuff. I don't know personally. Forgive me Nelson if I am remebering this wrong.
You also delve into the questions of whether or not an amp sounds better having more power than needed. Some folks like having a 200W amp even though they will never use more than 50W. Perhaps you can hear a difference. Nelson has suggested that his customers say as much as you move through the progression of his Pass stuff. I don't know personally. Forgive me Nelson if I am remebering this wrong.
this is confusing.
The F5 will give full power without the diode bypassing Rs.
The diodes allows more peak current to flow if the speaker demands that.
That extra current demand does not coincide with a nominal speaker impedance. The speaker at that transient demand moment has an effective impedance much lower than nominal. That is not the same as more power delivery. It is avoidance of current clipping. An amplifier that current clips is just as bad as an amplifier that voltage clips. Both are incapable of meeting the demand set by the signal and the load.
86dB/W/m for 4 ohm is 89dB/2.83V/m
With 0.6V across the source resistors, diode die at 25C, output current total would be at ~1.5A.
Half of that going through one driver makes 89.5dB continuous.
(in a "normal" room, me find that a rather uncomfortable SPL)
Ask yourself which mechanism would stop diode temperature to creep up higher and higher once it starts to sweat.
IMO once they start to sweat they conduct earlier and start to sweat a bit more
With no heath sink they will stop to conduct when temperature go hi enough for diode to pop. Luky that most time they pop open circuit not like mosfets, one may ewen ear them pop if music is not to loud but not big worry about that the smell is prety nasty.
but if you miss both clues most probably one will find out next time he goes near F5T with DM
With heath sink it all change especialy if on main one that has litle thermistors taking care of mosfet temperature.
those shuld be prety constant at around 50 C few wats trough diodes won't make much difference for sink designed for say 250W.
And as temperature won't change much diodes wil be pre conditioned to conduct at almost a constant place so no difference between low and loud transients and laud and laud transients
Or not?
IMO once they start to sweat they conduct earlier and start to sweat a bit more
With no heath sink they will stop to conduct when temperature go hi enough for diode to pop. Luky that most time they pop open circuit not like mosfets, one may ewen ear them pop if music is not to loud but not big worry about that the smell is prety nasty.
but if you miss both clues most probably one will find out next time he goes near F5T with DM
With heath sink it all change especialy if on main one that has litle thermistors taking care of mosfet temperature.
those shuld be prety constant at around 50 C few wats trough diodes won't make much difference for sink designed for say 250W.
And as temperature won't change much diodes wil be pre conditioned to conduct at almost a constant place so no difference between low and loud transients and laud and laud transients
Or not?
Not saying that they should not be on heat-sink, just that they do not generate much heat themselves. Most of the heat they will be subjected to will be that of the Mos-fets. Just because they are on the heat sink does not change this. Any situation that would change their conduction, and as a result their temperature in a rapid fashion, would result in similar outcome even on heat sink. We are talking instantaneous current here.
I think that Jacco is suggesting that pushing the bias point too far in relation to the diode conduction simply negates the stability that the source resistor brings to the mos-fet and increases the chance that you will have some sort of fault occur. If i am thinking correctly (probably not), it is still about the mosfet and how it maintains its control over the circuit. As Andrew has stated, that the ideal scenario is one where the diode never does anything other than handle extreme momentary transients.
Let the corrections begin!
Not saying that they should not be on heat-sink, just that they do not generate much heat themselves. Most of the heat they will be subjected to will be that of the Mos-fets. Just because they are on the heat sink does not change this. Any situation that would change their conduction, and as a result their temperature in a rapid fashion, would result in similar outcome even on heat sink. We are talking instantaneous current here.
I think that Jacco is suggesting that pushing the bias point too far in relation to the diode conduction simply negates the stability that the source resistor brings to the mos-fet and increases the chance that you will have some sort of fault occur. If i am thinking correctly (probably not), it is still about the mosfet and how it maintains its control over the circuit. As Andrew has stated, that the ideal scenario is one where the diode never does anything other than handle extreme momentary transients.
Let the corrections begin!
I think you got that right too! In other words, I agree with what andrewT has said recently. I'm not sure many of the other posters were seeing it this way??? My opinion basically is; Get all the classA you can afford(HeatSinks, $$$, etc.), Take care of those continuos, reoccuring common transients with the ClassAB capability. Take care of the real headrtoom issues with the diodes. Something like 25Wrms ClassA, 50Wrms ClassAB, 100wrms ClassABw/Diodes...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.