As for those who might be tempted to use separate power supplies for the front end and output stages, here is a friendly piece of advice :
Similar to the Aleph & Aleph-X, the output MOSFETs (both N & P) are driven at their gates by the drain resistor of the front end JFETs. These drain resistors are both connected to the top & bottom rails. If you split the rails between the front end and the output stage, then any difference (whether DC drift, or HF noise, or whatever else) between those rails will appear as changes in Vgs of the output MOSFETs, and hence in output current.
Of course, as in any circuit with global negative feedback, the feedback resistor network will attenuate this by a certain amount, in the case of the F5 about 30+ dB (if I remember correctly).
So the sound will change for certain, for better or for worse.
I personally would stick to tightly coupled rails between front end and output stage, and try to use a cap multiplier or Zen V3 type of regulated supply for the entire circuit.
But, by all means, feel free to experiment.
Patrick
Similar to the Aleph & Aleph-X, the output MOSFETs (both N & P) are driven at their gates by the drain resistor of the front end JFETs. These drain resistors are both connected to the top & bottom rails. If you split the rails between the front end and the output stage, then any difference (whether DC drift, or HF noise, or whatever else) between those rails will appear as changes in Vgs of the output MOSFETs, and hence in output current.
Of course, as in any circuit with global negative feedback, the feedback resistor network will attenuate this by a certain amount, in the case of the F5 about 30+ dB (if I remember correctly).
So the sound will change for certain, for better or for worse.
I personally would stick to tightly coupled rails between front end and output stage, and try to use a cap multiplier or Zen V3 type of regulated supply for the entire circuit.
But, by all means, feel free to experiment.
Patrick
Last edited:
That's what I thought.
I only asked because I was wondering how vitalica managed to build his amp exactly like yours... Just occured to me, it must be your boards in his amp?
Nevermind me...
Hey Vitali,
> After several days, I finished two F5 balanced monoblock.
That PCB looks familiar !!
I am glad that they are now sitting in a housing that is worthy of them.
Very well designed.
And I recognise that transformer as well.
But I still see Fairchild MOSFETs in your photos.
When did you change them to Toshiba's ?
> without revealing too many "secret details" over this way a number of experiments have been run regarding the power supply rails to the front end of the F5.
> It is daunting.
> Changes in that path appear at this point to radically effect the sound presentation!
You mean you are running separate rails for the input and output stages ?
Good luck.
Patrick
Hi Patrick ,
But I still see Fairchild MOSFETs in your photos.
When did you change them to Toshiba's ?
No, I changed nothing ,that was the beginning .
Do you think it would be better to change with TOSHIBA 1530/201?
Anyway, I am very happy as can be heard!
Thanks Patrick, great work!!
I feel happy with them
Best regards,
Vitali
> In other words, don't even think about it.
Glad to have your support, Nelson.
Vitali,
Changing to Toshiba MOSFETs involves more changes than just the MOSFETs themselves. So I suggest you listen to your current F5s first. If they are good enough for Nelson, they are not by any means bad.
When after a few months you absolutely want to change to Toshibas, then contact me again by PM.
As I already said, I like your housing. Worthy home for those boards.
Patrick
Glad to have your support, Nelson.
Vitali,
Changing to Toshiba MOSFETs involves more changes than just the MOSFETs themselves. So I suggest you listen to your current F5s first. If they are good enough for Nelson, they are not by any means bad.
When after a few months you absolutely want to change to Toshibas, then contact me again by PM.
As I already said, I like your housing. Worthy home for those boards.
Patrick
> In other words, don't even think about it.
Glad to have your support, Nelson.
Vitali,
Changing to Toshiba MOSFETs involves more changes than just the MOSFETs themselves. So I suggest you listen to your current F5s first. If they are good enough for Nelson, they are not by any means bad.
When after a few months you absolutely want to change to Toshibas, then contact me again by PM.
As I already said, I like your housing. Worthy home for those boards.
Patrick
Thanks Patrick ,
I'm sure I will want to make Toshiba 1530/201.
I come back in a few months on PM
Regards,
Vitali
After several days, I finished two F5 balanced monoblock.
I am very very happy sound!
We compared with F5, 5 Aleph, Aleph X.
F5 balanced - is my favorite!
I very much Nelson Pass and Patrick (EUVL) for everything!
Sorry for my English
Vitali
Hi Vitali, nice design, very original. Did you make the boards?
In other words, don't even think about it.
Blasphemy!!
Not only that, we muck about with it!!
_-_-bear
Blasphemy!!
Not only that, we muck about with it!!
_-_-bear
Is it possible to blaspheme oneself???
He is after all the creator, no?
Ron
We have been around that issue a while ago, partly from me
The advice was the same
Due to the exstreme close coupling of Jfet and output, a seperate supply on Jfets may cause problematic imballance
I remember one member having problems just from mounting exstra local decoupling caps
I have always thought about different ways to connect a power supply
Looking at a PASS Lab proto, shown here a short while ago, it "seemed" like there was multiple supply connections to output devices
But Im not certain about it
And anyway, its only interesting with a BA-1 and 2, with multiple output devices
The advice was the same
Due to the exstreme close coupling of Jfet and output, a seperate supply on Jfets may cause problematic imballance
I remember one member having problems just from mounting exstra local decoupling caps
I have always thought about different ways to connect a power supply
Looking at a PASS Lab proto, shown here a short while ago, it "seemed" like there was multiple supply connections to output devices
But Im not certain about it
And anyway, its only interesting with a BA-1 and 2, with multiple output devices
Hi there,
first i built F5 followed by original Pass schematics with all limiters and toshibas... Result was sweet but highs were dead,too smooth (some kind of distortion?).
Next,a friend of mine who has ESR measurment equipment,said that the last cap in PSU should be as close as possible to amp pcb.Other caps distance are not so important.So idealy it should be on pcb i guess.Size 2.2-50uf mkp ,tried 50uf with hiraga amp and result was stuning,well not so stunning it was with another type 50uf mkp.4 example Zen Mod suggested Icar made in Italy motor caps.
This time i choose 2.2uf obbligato aluminum.Result as expected was clearly noticable,but givven another day burn-in didn't gave "this is the amp" impression.
So the last hope - changed toshibas to irfp240-9240 and with first turn on- trebles are here,the air ,at last! saxs shines now and instrument separation is superb but still slightly "laid back" sound (mids are not extremely expressive as i wanted) so since it's very very clean amp technically i guess it's time to remove thermistors,put some silver solder,caddock resistors which in poor schematics ,in my opinion, are wasting time-money,this time i fell it may be worth the investment.
So thank you Pass
first i built F5 followed by original Pass schematics with all limiters and toshibas... Result was sweet but highs were dead,too smooth (some kind of distortion?).
Next,a friend of mine who has ESR measurment equipment,said that the last cap in PSU should be as close as possible to amp pcb.Other caps distance are not so important.So idealy it should be on pcb i guess.Size 2.2-50uf mkp ,tried 50uf with hiraga amp and result was stuning,well not so stunning it was with another type 50uf mkp.4 example Zen Mod suggested Icar made in Italy motor caps.
This time i choose 2.2uf obbligato aluminum.Result as expected was clearly noticable,but givven another day burn-in didn't gave "this is the amp" impression.
So the last hope - changed toshibas to irfp240-9240 and with first turn on- trebles are here,the air ,at last!
saxs shines now and instrument separation is superb but still slightly "laid back" sound (mids are not extremely expressive as i wanted) so since it's very very clean amp technically i guess it's time to remove thermistors,put some silver solder,caddock resistors which in poor schematics ,in my opinion, are wasting time-money,this time i fell it may be worth the investment.So thank you Pass
Well this is just my experience,maybe i should change something in schematics? i asked nobody answered...Some said toshibas magic,maybe it sounded sweet wide,but ir doesnt matter if its has no top end,in this schematics,with my equipment,maybe who wants them,dirty cheap?
btw some hours ago found a thread (not english) the experienced man who tried many transistors put toshibas and irfp also in the bottom in his list .And wrote that toshibas has dirty trebles.Of course do not take it too serious.
1)ONS MJ15024\15025,
2) ONS MJL21193\21194
3) Sanken 2SA1215\2SC2921
4) Sanken 2SA1216\2SC2922
5) ONS MJL1302\3281
6) Toshiba 2SA1943\2SC5200
7) Magnatec BUZ900DP\BUZ905DP
8) Hitachi 2SJ49\2SK134
9) Toshiba 2SA1942\2SC5199
10) Toshiba 2SJ201\2SK1530
11) Sanken 2SA1186\2SC2837
12) Hitachi 2SJ162\2S1058
13) Toshiba 2SJ115\2SK405
14) IGBT Toshiba GT20D101\GT20D201
15) IR IRF240\IRF9240
btw some hours ago found a thread (not english) the experienced man who tried many transistors put toshibas and irfp also
in the bottom in his list .And wrote that toshibas has dirty trebles.Of course do not take it too serious.1)ONS MJ15024\15025,
2) ONS MJL21193\21194
3) Sanken 2SA1215\2SC2921
4) Sanken 2SA1216\2SC2922
5) ONS MJL1302\3281
6) Toshiba 2SA1943\2SC5200
7) Magnatec BUZ900DP\BUZ905DP
8) Hitachi 2SJ49\2SK134
9) Toshiba 2SA1942\2SC5199
10) Toshiba 2SJ201\2SK1530
11) Sanken 2SA1186\2SC2837
12) Hitachi 2SJ162\2S1058
13) Toshiba 2SJ115\2SK405
14) IGBT Toshiba GT20D101\GT20D201
15) IR IRF240\IRF9240
Hi there,
first i built F5 followed by original Pass schematics with all limiters and toshibas... Result was sweet but highs were dead,too smooth (some kind of distortion?).
Next,a friend of mine who has ESR measurment equipment,said that the last cap in PSU should be as close as possible to amp pcb.Other caps distance are not so important.So idealy it should be on pcb i guess.Size 2.2-50uf mkp ,tried 50uf with hiraga amp and result was stuning,well not so stunning it was with another type 50uf mkp.4 example Zen Mod suggested Icar made in Italy motor caps.
This time i choose 2.2uf obbligato aluminum.Result as expected was clearly noticable,but givven another day burn-in didn't gave "this is the amp" impression.
So the last hope - changed toshibas to irfp240-9240 and with first turn on- trebles are here,the air ,at last!
So thank you Pass
If you want to use toshibas successfully, you need to play around with the values. Sounds like you have created a low pass filter somehow - you will never change that by any amount of silver you put in the circuit. Try to pull out the limiter transistors and see if it helps with the IRF's - if you go back to toshiba, you select different resistor values.
No there was no low pass i believe,just muffled-not dynamic mellow highs.
Well next step is to remove all limiters and see what happens next,this thing will never end
But when things about improving schematics i am starting to be not so creative and happy
1. No there was no low pass i believe,just muffled-not dynamic mellow highs.
2. Well next step is to remove all limiters and see what happens next,this thing will never end
3. But when things about improving schematics i am starting to be not so creative and happy
1. How do you think a low-pass sounds?
2. Good call.
3. Try to get the Fairchild devices. IRF has a N/P dis-symmetry, which gives you more distortion. This might be what you don't like, but believe me, the amp is plenty detailed - in fact if you have speakers with very high clarity, then this amp might be too much for the highs.