F5 Turbo Builders Thread

Jump to Latest
#3,003
6L6 posted that he didn't feel fancy rectifiers were necessary for our class A amps, and standard bridges were fine. I don't remember that he thought they degraded the sound. Fancy rectifiers make more difference in AB amps according to Jim. Feel free to experiment and draw your own conclusions. This is DIY after all. 😉
 
#3,004
yes, you are right: page 12, "The same Vishay rectifiers are used here" Nelson wrote....

I remember a very old article from Nelson where he describes that slow bridges are best for class A amps. But later on, he says they are using fast recovery diodes at Pass Labs, to meet some RFI requirements....

So it doesn't really matter, I guess...
There are people who say they hear a difference between rectifier diodes, but as usual I don't hear any ....😀

Walter
 
#3,005
What I was getting at is that Nelson has an entire article on PSU, found here -- https://www.passdiy.com/project/articles/power-supplies (yes, the order of the article is messed up a bit)

...and he mentions that slow diodes are better for our class-A PSU.

I've used slow, fast, and very fast diodes and honestly can't tell any difference. Because of that I find that I prefer the 4-pin monolithic diode bridges, as they are very compact, noticeable less expensive than discrete diodes+heatsinks, have little rejected heat and are quite easy to mount.
I would not go out of my way to use high-speed diodes thinking that they are better, but if I have them on hand, I'll gladly use them. The PSUv3 has holes for discrete and heatsinks and if nothing else, they look fantastic. 🙂 And it is very nice to have one PCB with the entire PSU on it.

The short answer is that there probably isn't a wrong answer.

Where the fast diodes seem to make the most difference isn't in AB amps, but in line-level projects. After a long conversation with Wayne at Burning Amp '11, he explained all the reasons why they are worth using for preamp and phono applications.

BobEllis said:
Feel free to experiment and draw your own conclusions. This is DIY after all. 😉
Click to expand...

No better advice could be given. 😀
 
Last edited:
#3,009
No, you can not use it, because it is of silicon nitride passivated ion-implanted epitaxial planar construction according to datasheet 😀


of course you can if used as rectifier diode. Or do you want to use it in the output section?


 
Last edited:
#3,014
Anyone tried These MOFETS?

Has anyone on the thread tried these devices: FQA28N15 / FQA36P15.
If so, what were the results in terms of stability, biasing, flatness, or just sound?

Did you also have the chance to compare against the IRFP240/9240?

Thanks.
 
#3,015
WalterW said:
I just read this, in article about the new .8 Pass amps: " The power supplies still use paralleled fast/soft rectifiers and large toroidal transformers from Plitron, but the AC primary circuits now incorporate additional RF filtering."
Click to expand...

I wonder how much of that part is driven by sound quality and how much be regulations about how much noise the amps can impart on the line. Not to impugn the products' quality at all, I've just been beaten about the head and shoulders by too many regulations lately.
 
#3,016
BobEllis said:
I wonder how much of that part is driven by sound quality and how much be regulations about how much noise the amps can impart on the line. Not to impugn the products' quality at all, I've just been beaten about the head and shoulders by too many regulations lately.
Click to expand...

eeeeee...the head, I can understand. The shoulders...once they're gone, no more head to hold and beat about... :smash:
 
#3,017
sangram said:
The latest version has thermistor on only one pair. Which was my exact question.

The initial F5 had the thermistor routed to the supply rail, this meant that even if you use 4 or 5 output pairs the bias on all of them would start going down as the thermistor operated as it would regulate the total Vgs of all devices.

In the latest F5 schematic which was published in the F5T pdf, the thermistor was now routed to the source of one of the output pairs and I assume thermally coupled to that one, hence 'pinching' the Vgs of that device as it got hotter. I am not sure how the unmonitored devices behave when the thermistor operates, which is really what I want to know. I guess I'll simply use one on each pair, safer that way I guess.

FWIW I ran my F5 without thermistors as they seemed to degrade the sound quality quite significantly, or so I imagined at the time. This meant that I had to readjust the bias every week, as the ambient temperature changed.

I grew intimately familiar with the amp, to the point that I knew how many turns would result in how much bias. I have seen up to 1.8A on the output devices, but they lived and the amp never went into runaway. The hottest bits in the amp were the source resistors, I measured 80 degrees C on those. At this time, the sinks - 10" single lengths of these ones - were at 55 degrees.
Click to expand...

Dear sangram,
I just can't understand the thermistor move. The first version makes much more sense to me, since it acts on the root of the bias mechanism, the jfet drain resistance. The second version, besides your point, confuses me in the sense that even if the thernistor is, as you said, "pinching off" the device Vgs, the voltage generated by the drop on the drain resistor is going to be over the source resistor, increasing the current. Perhaps Mr Pass could enlighten us?

Thanks
Daniel
 
#3,018
buzzforb said:
Multiple manufacturers make that particular diode. I believe yours is a discontinued Fairchild version. You can use it just fine, just check the Vf in the datasheet to confirm it is the same.
Click to expand...


Any opinions on the ON MUR3020WTG compared to the Vishay MUR3020WTPBF; I ask especially since the price for the ON version is very attractive at $1.69 at Arrow. To my novice eye the data seem similar though the charts look slightly different.
 

Attachments

#3,019
Imo, the key parameter is the Vf and the associated graph. This tells you the cut on point of the diode and its subsequent response and this is relevant to biasing the output stage. Other than that, I don't know that you could tell any difference at all in the parts. On a couple of occasions, I have seen where Nelson uses slight differences in different parts manufacturers parameters to achieve a final goal. A faster/lower Vf means an inherent decrease in bias(maybe not significant) and slower/higher Vf, the opposite.
 
#3,020
Finished building my F5T V2 yesterday. Powered on one channel with the bulb tester and no smoke! So for biasing I should shoot for 0.6vdc across the test points on the output boards? I did hook it up to a speaker and only a little sound came out with the full 2v of the source. Thanks! I am preparing the close up pictures for the inevitable issues.
 
You must log in or register to reply here.
Top Bottom