I agree for the lateral sound but Hitachi or Exicon is the same. Indeed High Damping factor (>100) has no real effect on a home system but a DF lower than 20 (and less than 10 for Zenotron) has a real big effect with many speakers.
Fab
I was referring to the Zenotron and not the Zenquito. The Zenquito being a common drain output has a much higher damping factor thus lower Zout.
But it is indeed quite revealing the effect of this 10K resistor value (open loop gain) for the Zenquito (or other amplifiers) on the overall sound signature....
Fab
Last edited:
Hello,
Extract from Homecinema.fr
<<Ampli Mosquito et ***Kito a base de mosfets SK1058>> - 29715000 - sur le forum <<Amplis et Preamplis>> - 1056 - du site Homecinema-fr.com
picatchumm a écrit:
Bonsoir,
Quels est l'identiter sonnore des ampli a base de 1058 sont plutôt crisalin ou plutôt chaud (~ tube) ???
Qu'est qui fait l'identité sonnore d'un ampli mosfet , ses condos ???
Quel difference (sonnore ) y a t'il entre les differents Mos-fet :
2SJ77/2SK214
BUZ900DP/BUZ905DP
2SJ162/2SK1058
ou autre transistors ...
@Merci
Entre les BUZ900DP/BUZ905DP et les 2SJ162/2SK1058 la seule différence est que le premier contient 2 des seconds dans un seul boitier. Le brevet est le même et donc les puces sont quasi equivalente (les buz assurent 16 A alors que les SJ SK assurent 7 A (1 A de mieux X2 pour les BUZ).
Le son chaud typé tubes des mosfets vient du fait que contrairement aux bipolaires qui écrêtent brutalement à cause de leur VCE sat, donc harmoniques impaires), les mosfet de puissance écrêtent progressivement (sorte de soft clipping naturel) car leur caractéristique Id (courant drain source, donc côté puissance) en fonction de Vgs (tension de commande) est à peu près une droite, la limitation en courant de sortie arrive parce que l'on à plus assez de tension de commande (du moins dans un ampli à une seule alimentation pour les drivers et pour les mosfets) et ce phénomène est progressif, donc moins violent.
Philippe
In terms of paralleling up devices, I just noticed Exicon makes
an ECW20n20 and its partner an ECW20p20.
These have double the |yfs| and current rating of the ECX10n20 and ECX10p20 's.
Exicon Lateral MOSFETS - Our Product Range
So for the Zenquito, could 1 set of the ECW's be employed in place of 2 sets
of the ECX 's ??? That would make things quite a bit easier.
Unfortunately, I've already bought the ECX's . I'm not happy at this particular moment.
.
Please don`t parallel up the output devices, especially not when driven by a common emitter amplifier. Double die may be somewhat better, but are you sure that the extra current handling will be necessary? The ECF10N20 / ECF10P20 in TO-3 metal can package would be my choice if it was practically feasible.
Thanks for posting the the link to the 2sk1058 / 2sj162 discussion.
This application note by ti and posted by Jackinnj is really useful.
https://www.ti.com/lit/an/snaa045a/snaa045a.pdf?ts=1607920358635
OK so to calculate the gate resistors, just rearrange the following equation posted in the ti application note, page 14 .
Rgate = 1 / (2 x Pi x Ciss x f -3dB )
Hopefully, this is correct
Looking at the graphs of the datasheet for the ECX10n20
At Vgs = 0.9Vdc , Ibias = 300 mA and Ciss = 500pF
But Vgs could rise typically to something in the order of 1.5 Vdc when music is playing , and at that point Ciss ~ 650 pF.
Which I think would be a good compromise.
But selecting an f -3dB = 100 KHz , gives Rgate = 2.45K , which is a magnitude out.
So select f -3dB = 1MHz , which gives Rgate = 245R , which is a believable answer.
For the ECX10p20 , the Ciss curve is much easier.
At Vgs = 1.55 Vdc , Ibias = 300 mA and Ciss ~ 1,100 pF
and pretty well stays flat after that.
So using f -3dB = 1MHz , gives Rgate = 145R
I think ...
.
Last edited:
> Please don't parallel up the output devices
Many people do in practice with MOSFETs, including Nelson, the late Charles Hansen, .....
How to Parallel Lateral MOSFETs...
How to Parallel Lateral MOSFETs...
Especially the lateral MOSFETs have a negative tempco when biased > 100mA.
As long as you have done some matching upfront, the equal current sharing is self regulating due to the -ve tempco.
The gate stopper values in the TI application note applies to that particular circuit topology.
It is also affected by the PCB or wiring layout.
The only way to be certain is to test in real circuit.
But one can start with a value of 220~470R, as close to the MOSFET gate as possible.
Cheers,
Patrick
Many people do in practice with MOSFETs, including Nelson, the late Charles Hansen, .....
How to Parallel Lateral MOSFETs...
How to Parallel Lateral MOSFETs...
Especially the lateral MOSFETs have a negative tempco when biased > 100mA.
As long as you have done some matching upfront, the equal current sharing is self regulating due to the -ve tempco.
The gate stopper values in the TI application note applies to that particular circuit topology.
It is also affected by the PCB or wiring layout.
The only way to be certain is to test in real circuit.
But one can start with a value of 220~470R, as close to the MOSFET gate as possible.
Cheers,
Patrick
I consider my self as a hobbyist, at certainly not at the level of the designers here, but IMO the real purpose of paralleling output gain devices
is to lower Zout of the amp, so that more current can get from the power supply to the load (ie speakers).
Recall, a voice coil needs current to build up a magnetic field.
Also, Zout of the power supply should be taken into consideration when calculating Zout of an Amp.
As explained to me, the power supply is 1/2 the amp.
Some speaker loads are far more difficult to drive than others.
Apparently, wild swings in the phase angle cause issues.
I was planning to use the Zenquito to drive a pair of B&W 603 s2 's
in which the magnitude of the impedance drops down to 3.9 ohms.
Which I think is a medium difficult load.
.
Last edited:
Hello Éboissier
,Yes, after 2000, I gave up electronics a bit to devote myself to a house that took up a lot of my time.
Then, as I am in IT, from 2009 I became more passionate about and occupied with PC, troubleshooting and Windows installation for which I specialized. However, retirement is fast approaching, 15 months, phew ... It will be necessary to find work. The electronics workshop froze a little after 2000 ... but current projects will resume.
Greetings JF
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.