Increasing Fetboy output drive...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Silvas

Well-known member
Joined
Aug 15, 2006
Messages
727
Location
Chia, Colombia
Hi!

I´ve been playing with the fetboy circuit for a while and love it. I´m getting about 7.2 volts on the output (unbalanced and transformerless) before the output begin to clip. i sticked to the original 47k resistor value as higer values clipped one side of the waveform too early. Less than 42k is playing with fire i think, with nothing to gain. I also used the original darlington but for the npn used a standard bc547 with approx. beta 130, and worked great.

I would like to increase the output drive capability of the Fetboy and would like to know the options. I´ve been messing with supply voltages, up to 30 volts and custom-biasing the fets for these voltage, and this way i can squeeze a couple more volts out of it at the expense of higher dissipation...any other transformerless options?

I noticed that at 30 volts my Fetboy is quieter, i mean, a different kind of noise, less noticeable for the same gain amount.

Something odd: there´s some noise with the gain totally ccw with no input, the noise gradually dissapears until i reach 9:30 o´clock and starts to increase normally from here...but this noise isn´t noticeable with normal operation. But this bothers me, any ideas?

Thanks!
 
Do you want to be able to drive more (heavier) loads or do you want more signal swing ?

The latter could be accomplished by a step-up TX, but you said you wanted TX-less.
How about leaving the orig. circuit like it was (making colour) and adding a transparant circuit after it to do the level- or muscle-thing ? It won't be hard to make a circuit that adds less colour than the FETboy :wink:
 
[quote author="MartyMart"]How about adding the THAT balanced out driver chip ? ( or SSM2142 ? )
Isn't that a 6dB boost in signal .... would that work ?
MM.[/quote]
Yes & no... depending on amount of boost needed. Not unlikely the requirements surpass the allowed rails for these chips (+/-18V for the '2142). Maybe an opamp+discretes-combination need to be used (see circuits as posted by Wayne/Mediatechnology). Or reconsider the need for those kiloVolt-swings :wink:
 
[quote author="MartyMart"]Ah, yes he said supply up to 30v so that's a "no-no" , it was just a thought.
MM.[/quote]
All fine. Or use two: one inverting driver and one non-inverting, both sniffing the same FETboy-output. Same supply, but double the swing. The FETboy might need to get some dummyload now.
 
Thanks for the replies

In fact, what i want to do is increase output swing...mantaining the circuit as is, all discrete class A.

I started lowering the resistor from 68k to 47k and lower, so the waveform doesn´t clip on one side...and i was able to get around 7.2 volts on the output without distortion. Next thing i done was increasing the voltage and rebiasing the fet so this time i got 9.1 volts and also got a little more dissipation. For my application i would like a bit more...my audio interfaces and all gear is calibrated for 18dB headroom above +4, and i don´t want the pre clipping before the interface reaches 0dBFS. Yeah, i know, "just recalibrate everything for -14dB headroom"....
 
[quote author="Silvas"]Thanks for the replies

In fact, what i want to do is increase output swing...mantaining the circuit as is, all discrete class A.

I started lowering the resistor from 68k to 47k and lower, so the waveform doesn´t clip on one side...and i was able to get around 7.2 volts on the output without distortion. Next thing i done was increasing the voltage and rebiasing the fet so this time i got 9.1 volts and also got a little more dissipation. For my application i would like a bit more...my audio interfaces and all gear is calibrated for 18dB headroom above +4, and i don´t want the pre clipping before the interface reaches 0dBFS. Yeah, i know, "just recalibrate everything for 14dB headroom"....[/quote]
 
You need to think about the fet specs at 30VDC what is the Drain to Gate voltage? The look at the fet spec for gate current leakage. Maybe that has to do with the noise and volume control.

I just looked at the schematic again, as I have posted in the past I don't like the output current sink? setup.

I would redesign the output for a little more current and use a 1:2 transformer as a first option, I did read you don't want that.

You are going to need a different design gain stage or to change a few parts if you don't want to add a higher power supply voltage gain of say two amp after the fetboy. The issue will be it might not sound the same.

Note the direct connection off the drain to the EF with a current sink. The drain is about two base emitter drops above the center. If you want more signal swing (higher voltage supply)you might need to find a different FET. What are your drain and source VDC to ground readings at the different power supply voltages.

What FET number did you use?
 
I just did some simple math for the fetboy gain stage
The Q7,3 BC550c 400-900hfe at 2ma, NS P04 die, TO92 from a NS book

Rounding current in the output to 40ma looks like under 1ma for the fet and not much thru Q2,6.

40ma at 12v, .04A X 12VDC =.48watts
then 40 ma at 24v , .04A X 24VDC =.96watts

TO92 cases are 600mw, .6watts IIRC.

How hot are the transistors Q3,7?
 
My fet is a 2n5457

The pot noise problem i was talking about is present no matter which supply voltage is used...

The transistors are hot at 30 volts, using a heatsink. Extremely hot without it. At 24 volts they run cooler but still need the heatsink. The 100R resistor is quite hot as well.

In my first try i used BC550c (leftovers from the green pre), but it was just too hot, i had to put a giant heatsink and the current consumption was about 80mA per gain stage, it worked fine until i blew the power transformer.

I still have to test if i can reach 0dBFS in the 192 interface with the new voltage...

Any recommendations for a 1:2 output trafo, suitable for this circuit?
 
or...

any suggestion for a suitable FET that yields more voltage swing?

Stupid question...The swing isn't limited by the output devices?
 
Its 7.2 RMS

I need an insane level on my Digi 192 i/o using the "A" trim, which is calibrated for +22db input=0dBFS...i know, i can recalibrate or switch to trim "B", which is +18dB input=0dBFS.

BUT

If possible, i want to reach the +22dB mark...not an absolute necessity, just something i want to do...

1:2 trafo on the output

or

different fet?
 

Latest posts

Back
Top