nope
this isn't my playground
( so I'll not tell that already proposed are probably best easy available
)
this isn't my playground
( so I'll not tell that already proposed are probably best easy available
)
> naah ...... never mind
Well, if you do not want to have an open discussion, I cannot force you.
As far as I am concerned, the pass device for a MOSFET cap multiplier should have the following characteristics :
1. High transconductance at bias -> low Zout
2. Low saturation voltage and flat Id vs Vds after saturation -> allows low dropout, low dissipation and good PSRR
3. Low Vgs at bias -> allows low dropout
4. Low Cgd at those low Vds values -> quick response to Vin ripples and good PSRR
5. +ve tempco -> when heated up Vgs drops for the same bias -> less dissipation -> self stabilising
Have I miss anything ?
Patrick
.
Well, if you do not want to have an open discussion, I cannot force you.
As far as I am concerned, the pass device for a MOSFET cap multiplier should have the following characteristics :
1. High transconductance at bias -> low Zout
2. Low saturation voltage and flat Id vs Vds after saturation -> allows low dropout, low dissipation and good PSRR
3. Low Vgs at bias -> allows low dropout
4. Low Cgd at those low Vds values -> quick response to Vin ripples and good PSRR
5. +ve tempco -> when heated up Vgs drops for the same bias -> less dissipation -> self stabilising
Have I miss anything ?
Patrick
.
Last edited:
Patrick
we started talking about output devices , not cap multi ones ..... so I just commented on difference between 201 and 618 , in this context .
anyway - splitting a hair isn't my fave way of killing time and I hate when open discussion is leading anywhere else than to better knowledge of all of us .
splitting a hair is ego trip thing .......... and I'm pretty much used to be court jester/fool and avoiding ego trips is sort of my occupation
as far as cap multi devices are in question , I have no objection and/or comment
It was just a casual comment that if we were getting them for Cap Multiplier anyhow, there would be no harm putting them in the amp for a trial.
Wasn't dead serious and certainly wasn't suggesting that it would have to be better.
In amplifier applications, such devices have a few issues, e.g. bias stability because of the huge tempco.
And they are in general better suited for follower applications than in circuits such as the F5.
Having said that, Kaneda has used them successfully in quasi push pull circuits, so those problems can be mastered.
There are many such UHC-MOSFETs, most of them from Hitachi. I have tried a few.
Their high transconductance does offer some advantages, even in F5 topology.
Patrick
Wasn't dead serious and certainly wasn't suggesting that it would have to be better.
In amplifier applications, such devices have a few issues, e.g. bias stability because of the huge tempco.
And they are in general better suited for follower applications than in circuits such as the F5.
Having said that, Kaneda has used them successfully in quasi push pull circuits, so those problems can be mastered.
There are many such UHC-MOSFETs, most of them from Hitachi. I have tried a few.
Their high transconductance does offer some advantages, even in F5 topology.
Patrick
Last edited:
> I was willing to do real work, no-one asked me.
I asked a couple of times openly for volunteers.
I also asked you once per PM about PCB.
Anyhow, the followings are things that you (in general and not necessarily ashaw) can consider organising :
1) choose a make for the Tx and maybe gauge interest for GB
2) consider GB & group match for MPC74 0R22
3) consider GB & group match for 2SK3497 / 2SJ618 (I did ask Zhoufang for availability, no reply.)
4) start discussing what protection and control features you want and think about circuits for those
These things take time. Other components are easily available, so no need to GB now.
As I promised, I shall release schematics and BoM when PCB Gerbers are ready to go to Variac.
No point releasing work in progress.
Patrick
I asked a couple of times openly for volunteers.
I also asked you once per PM about PCB.
Anyhow, the followings are things that you (in general and not necessarily ashaw) can consider organising :
1) choose a make for the Tx and maybe gauge interest for GB
2) consider GB & group match for MPC74 0R22
3) consider GB & group match for 2SK3497 / 2SJ618 (I did ask Zhoufang for availability, no reply.)
4) start discussing what protection and control features you want and think about circuits for those
These things take time. Other components are easily available, so no need to GB now.
As I promised, I shall release schematics and BoM when PCB Gerbers are ready to go to Variac.
No point releasing work in progress.
Patrick
Last edited:
No issue.
If you want to help, consider starting a new thread on those 4 points above.
But it is really up to you all whether you want to do this, independent of what I am doing here.
The component choice from me is only a recommendation.
You can always choose your own according to your imagination, as if you would buy a SE F5 PCB from DIYA, or a chipamp 3886 PCB, or whatever else.
Patrick
If you want to help, consider starting a new thread on those 4 points above.
But it is really up to you all whether you want to do this, independent of what I am doing here.
The component choice from me is only a recommendation.
You can always choose your own according to your imagination, as if you would buy a SE F5 PCB from DIYA, or a chipamp 3886 PCB, or whatever else.
Patrick
I would not presume to be able to do any 'real work' towards the design of this PCB. I have no special requests that need fulfilling, I will help where I can, but I do not think that will go very far here in an area that is not all that familiar to me.
I would like pushbutton, I would like regulated supply, I would like spacing for Rifa low Z caps, but most of all I would like those with real skill and knowledge in here to be able to go wild and come up with what makes them happy and then come along for the ride.
if there is some way I can help that you can think of (after I have moved, so 2-3 weeks) I will do so
I would like pushbutton, I would like regulated supply, I would like spacing for Rifa low Z caps, but most of all I would like those with real skill and knowledge in here to be able to go wild and come up with what makes them happy and then come along for the ride.
if there is some way I can help that you can think of (after I have moved, so 2-3 weeks) I will do so
Last edited:
I have enough PCB layout assistance for now. Thx.
The rest (GB or not) is up to the public to decide.
Patrick
The rest (GB or not) is up to the public to decide.
Patrick
I can't wait to see the final product. A big thanks to you Patrick and everyone who is helping out.
My limited understanding of the F5 was that it was a third harmonic amp. I guess you could possibly eliminate second harmonic altogther. Seems power is the main driving force behind this design. I like the idea of better immunity to noise, but this seems to be a source problem. I fully concede to your knowledge, expertise and opinion. It is vastly superior to mine. I am trying to go beyond figuring out how and asking why. I want to build an F5 or f-something to drive full range speakers and i am tryng to ascertain if this is an unnecessary step for me. Hell, i don't even know if i like 2nd or 3rd harmonics better😱
Nothing is necessary, and you can get low 2nd harmionics if you know what you are doing wit the SE F5.
There are people who have already built the balanced F5 at the F5 thread.
Take a look there and see what they have to say.
I only build balanced amplifiers since I built my AJX100, so I do not know how the single ended F5 sounds.
Maybe it sounds very nice as well with some second harmonics.
😉
Patrick
There are people who have already built the balanced F5 at the F5 thread.
Take a look there and see what they have to say.
I only build balanced amplifiers since I built my AJX100, so I do not know how the single ended F5 sounds.
Maybe it sounds very nice as well with some second harmonics.
😉
Patrick
Last edited:
I wish to add that the F5X PCB currently being converted to Gerber supports both floating and grounded X.
That means you could, if you so wish, also use the PCB as 2-channel single ended F5.
e.g. if you want to use active crossover to drive a 2-way speaker system.
In such applications, you need to use a higher rail voltage (e.g. +/-24V), lower bias (1.3A), and a larger source resistor (0R47).
The rest can stay essentially as is.
BUT two separate SE F5 PCBs placed further apart would be a better solution.
Patrick
.
That means you could, if you so wish, also use the PCB as 2-channel single ended F5.
e.g. if you want to use active crossover to drive a 2-way speaker system.
In such applications, you need to use a higher rail voltage (e.g. +/-24V), lower bias (1.3A), and a larger source resistor (0R47).
The rest can stay essentially as is.
BUT two separate SE F5 PCBs placed further apart would be a better solution.
Patrick
.
Last edited:
Since I am out of town next week for the entire week, and probably would not be posting regularly, here is an update of current progress:
1. The F5X PCB is basically finished, waiting for proof check. Only then will schematics & bill of material be released.
2. The design of the case and all detailed drawings are finished. We shall order a prototype next week.
3. There is a default layout of using the case for 2 channels of F5X. This includes :
-- 2 channel F5X PCB at optimal position on heatsink
-- 2 pieces 500VA toroidal transformer 150 diameter and 80 high, or smaller
-- PCBs for holding 8x capacitors per channel, 30mm dia x 50 high, 10mm snap-in (e.g. EPCOS 22000uF 40V), placed next to the F5X PCB (left & right)
-- an additional internal heat sink for 4x bridge rectifiers (16x TO220) & 2 sets of MOSFET follower regulators (using any TO247 devices)
-- a control unit consisting of 2 independent speaker DC detection, amplifier status control (off, standby, operate), and associated relay drivers
-- an axillary PSU (+/-12V, 10VA) for speaker protection & control circuitry
4. All miscellaneous mechanical parts not part of the standard case have detail drawings finished.
5. Layout of all other PCBs are currently in progress.
Patrick
.
1. The F5X PCB is basically finished, waiting for proof check. Only then will schematics & bill of material be released.
2. The design of the case and all detailed drawings are finished. We shall order a prototype next week.
3. There is a default layout of using the case for 2 channels of F5X. This includes :
-- 2 channel F5X PCB at optimal position on heatsink
-- 2 pieces 500VA toroidal transformer 150 diameter and 80 high, or smaller
-- PCBs for holding 8x capacitors per channel, 30mm dia x 50 high, 10mm snap-in (e.g. EPCOS 22000uF 40V), placed next to the F5X PCB (left & right)
-- an additional internal heat sink for 4x bridge rectifiers (16x TO220) & 2 sets of MOSFET follower regulators (using any TO247 devices)
-- a control unit consisting of 2 independent speaker DC detection, amplifier status control (off, standby, operate), and associated relay drivers
-- an axillary PSU (+/-12V, 10VA) for speaker protection & control circuitry
4. All miscellaneous mechanical parts not part of the standard case have detail drawings finished.
5. Layout of all other PCBs are currently in progress.
Patrick
.
Last edited:
My thanks also to thanh, who has done really a lot behind the scene, to bksabath, fitzfish and ashaw, for relieving me of the PCB work, to MarkLai for getting the prototypes made, and Variac for supporting the whole thing in the first place. Hi share of the work has yet to come.
I only hope you would like what we do, in the end.
😉
Patrick
I only hope you would like what we do, in the end.
😉
Patrick
- Status
- Not open for further replies.