bocka said:
If you use a DC feedback circuit like the original Zen, the
marginal resistance in the inductor will help. Certainly any
additional resistance from Source to ground will also stabilize
the circuit.
Apart from those considerations, if you are running enough
current through the Mosfet (and with an inductive load you will
run current greater than the peak output) you will start getting
into the zero or even negative temperature coefficient region,
where increased temperature does not appreciably increase
the Drain current for a given Vgs. This assumes that you
have lots of heat sinking, as do all Zen designs. (By this I mean
4 or 5 times what you think you need)
So you bias up the Gate with a pot and adjust it until you get
stable operation at a given temperature and so on. If you find
it wandering too much, maybe put a thermistor in parallel with
the Gate to ground to compensate it a bit.
Better yet, put a knob and a current meter on the front panel
and watch it all day long. 😉
ok, instead of the pot a DAC for the bias, and a microcontroller with temperature sensor will do the job....
4 or 5 times what i think means we run here about 10A bias, and need a heatsink with about 200W at 30 K difference hot/cold side?
Steve Eddy said:
I was thinking of driving it from the differential outputs (600 Ohms) of my Behringer DCX2496, which has differential AK4393 DAC outputs. Differenntial to single ended conversion & power amplification with just one active device, it has an attractive ring to it...
Jan Didden
janneman said:
Well you don't need additional voltage amplification in order to get power amplification. What's the maximum output of your Behringer? If it's high enough voltage-wise, you can dispense with the step-up transformer.
Those 600 ohm outputs are rather on the high side though and you might get a bit more high frequency rolloff than you want with the 1:5 step-up.
But if you've got enough voltage output from the Behringer, you could go with a 1:2 or a 1:1.
se
Steve Eddy said:
Output voltage is about 1Vrms. At this point, I have 30dB attenuators and then into power amps with 28dB gain for normal listening levels.
So a stepup of 5x for those cases when I want it really loud would be enough.
I still need volume control, but a custom passive switched attenuator between the Behringer and the 1-clap could do that.
Jan Didden
So,
janneman said:
But a stepup of 5x is only 14dB. And if you're starting out with 1 volt RMS, then you're not going to get any more than 5 volts RMS at the amp's output. Into 8 ohms, you're only talking about 3 watts. Can you get by with just 3 watts?
The 600 ohm outputs of the Behringer are already on the high side. A passive attenuator will only make that worse.
As I said, this isn't an amp without caveats. Low source impedance is desirable in most cases, but even moreso in this case.
When I first thought of doing an amplifier using the 13K7-A as a voltage gain stage feeding an output follower, I'd intended to mate it with a preamp of similar design in order to deal with the unusually low input impedance and to provide a very low source impedance.
Ultimately what I ended up with was the 13K7-A driving Sziklai pair follower using BJTs. But that uses two transistors which is one more than what Fuling was calling for. 🙂
I'd never got 'round to doing the preamp because I've just been using the variable output from my Theta Miles. But what I was planning to do was using the Jensen JT-11P4-1 which has just a wee bit of voltage gain (1.5dB) and is ideally loaded resistively with a 20k load which was going to be a nice 20k attenuator.
Should be able to do something like that with a single JFET and maybe get the output impedance down to a couple hundred ohms.
se
I just realized that if Fuling wants only ONE transistor, then power MOSFETs won't fit the bill. I seem to recall that a power MOSFET is basically just a gob of MOSFET cells all wired in parallel (which I guess would explain why the input capacitance is so high). It's like a whole crowd of tiny transistors clapping.
So an amp using a single MOSFET device would only be one transistor clapping in the same sense that the Gaincard has only 9 parts. 🙂
Hmmm...
Fuling, how literal do you want to be here?
se
So an amp using a single MOSFET device would only be one transistor clapping in the same sense that the Gaincard has only 9 parts. 🙂
Hmmm...
Fuling, how literal do you want to be here?
se
Steve Eddy said:
Right now having 1Vrms at the speaker is good enough for normal listening - actually a bit too loud if you want to do reading as well.
My speakers have each 4 x Jordan 50mm for mid/high (>280Hz) and 2 x Jordan 150mm for low. Must be over 90dB sensitivity.
And the stepped atten will work, don't worry. leave it to me....
Jan Didden
Puhh could be a little bit boring build a ZEN and watching this amp all day long 😉 A litte bit music at this time wasn't bad...
First I looked to a BUZ11AL MOSFET datasheet. The crossing point between negative and positive coefficient was at 30 amps, a little bit too much although for the ZEN I think.
A IRF 530 should work fine at 3 amps (and I dont' have to look at the amp meter all the time).
Yes, I remember. In the former days not everything is done with semiconductors. Good idea, I'm satisfied with this solution😎
So here it is, a new thread dealing with "Overshooting, ringing and blowing up mosfets if you build stuff based on irresponsible advice" http://www.diyaudio.com/forums/showthread.php?s=&threadid=18365
If you feel that I overlooked anything, let me know.
If you feel that I overlooked anything, let me know.
As for the question of current control:
Surely the point is that there is just one device in the signal path. Any regulation of current or voltage has nothing to do with the concept. After all, there might be active devices controlling the mains supply; you can't say that they're part of the circuit, even though they may affect it.
So I'd say go for a servo circuit, and be damned. It'll probably have less detrimental effect than a thermistor hung on the gate.
Oh, and ESD protection is optional....
Surely the point is that there is just one device in the signal path. Any regulation of current or voltage has nothing to do with the concept. After all, there might be active devices controlling the mains supply; you can't say that they're part of the circuit, even though they may affect it.
So I'd say go for a servo circuit, and be damned. It'll probably have less detrimental effect than a thermistor hung on the gate.
Oh, and ESD protection is optional....
Hi,
In the case of a SE topology such as this the entire PS is in the signal path.
Conceptually we tend to not mention this for clarity's sake.
Yuk.
Cheers,😉
In the case of a SE topology such as this the entire PS is in the signal path.
Conceptually we tend to not mention this for clarity's sake.
Yuk.
Cheers,😉
As Nelson has posted it is not necessary to use a thermistor, servo unit or son on. On a constant Vgs the current for a IRF 530 for example rises only from 3 amps @ 25° to about 4,5 amps @ 175°. My fault was that I first have looked to any MOSFET I can get and that was a BUZ11AL.
A IRF 530 on a large heatsink should be stable without servo loop or thermistor.
A IRF 530 on a large heatsink should be stable without servo loop or thermistor.
But control of the current is necessary for the best cost /performance of the choke.
So from that angle, it might be better not to use such a large heatsink, and allow the MOSFET to limit it's own current.
So from that angle, it might be better not to use such a large heatsink, and allow the MOSFET to limit it's own current.
MIXED FEELINGS...
Hi,
A bit of both, I'm afraid... but if you know of one that's utterly reliable and doesn't degrade the sound without being too complex...I'm willing to listen...Errr to you, not the servo.
Servos, like most heavy feedback circuits are like dogs chasing their own tail.
Cheers,😉
Hi,
A bit of both, I'm afraid... but if you know of one that's utterly reliable and doesn't degrade the sound without being too complex...I'm willing to listen...Errr to you, not the servo.
Servos, like most heavy feedback circuits are like dogs chasing their own tail.
Cheers,😉
Ok, i have a large heatsink, and some diy copper coils. And some IRF510 etc.
To prevent me fool from frying them all - will anyone of you post a complete inductor - Zen circuit?
To prevent me fool from frying them all - will anyone of you post a complete inductor - Zen circuit?
Till,
I think it's this one:
The trouble is someone has drawn diodes all over it 🙄
Cheers,
edit: I thought I embed an image from another thread, but apparentl not 🙁
I think it's this one:
The trouble is someone has drawn diodes all over it 🙄
Cheers,
edit: I thought I embed an image from another thread, but apparentl not 🙁
- Status
- Not open for further replies.