Hamptone JFP trouble

when you have a current consumption of 160mA you can use the 10VA at 24V.

i am currently building on my single channel jfp preamp. my first issue got
a problem on high level inputs, sounds like motor boating.
next i will try a different supply for the two stages and a feedback resistor between output (before output cap) and input, maybe 100k.
for input and output i use OEPs.
i had a toriod with 15VA. its a little bit oversized for one channel.
what board or layout do you use?

Usually, the motorboarding is lack of input trafo shielding....

I´m using my own layout, you could see a pic of the populated boards of my last prototype board at www.myspace.com/audioeng80 .The PCB is big, but have all i need and works great. The pcb is for 1 channel. I also left a pot for "harmonics" following Chrissugar´s advice but still working on that. I mounted multiturn trimmers for biasing the jfets, which worked great for me (but probably not the most correct thing to do) the polys are mounted on the copper side due to its size, but on the prototype i´m using in the studio i used tantalums and i´m quite happy with the sound.

10VA at 24 volts, how many mA? i´m not sure about how to convert VA to mA.

i use shielding cans for both trafos. with the trims i do the same way.
can you explain how sounds the different between poly and tantal?
your link does not show a picture, only your myspace account.
my prototype can you see here:
http://gasman.lima-city.de/diy/fetboy_print_klein.JPG
http://gasman.lima-city.de/diy/fetboyinside.JPG

the cable wiring on the second picture is not the best, i know. but
not a problem.
for my new board i must route a new layout with some enhancement.

In the myspace account, click on my neve photo, then click on my DIY photo section....

Tantalum vs poly...i haven´t compared side by side but i think tantalums sound more "midrangey" and forward than poly.

Nice layout!

have a question about the adjust of the half supply voltage on the output (DC)
i did not match the FETs, so i become values there was out of the 12V range (supply = 24V)
so i put a trim instead of the 1k82 resitor and trimmed it to 12V DC output voltage.
btw, i adjust the bias from the transistor to have 4V at the 100 ohm resistor. (=40mA current source)

is this a right way to adjust the correct working point of 12Vdc?

i dont´t know if it is correct or not, but the trimmer instead the 1k82 resistor works fine for me and sounds nice...but i think you should match your fets to similar voltages if you´re using the 2 stages.

What kind of wall wart should be used for the mnats version? Mako's page doesn't give specs for the wall wart.

I know Scott Hampton's original version called for a 24v wallwart, per this page: tapeop2004diy.htm but mnats design calls for a 7824 regulator.

Should I use a >24v (30v?) wall wart, and then regulate it down to 24v?

Or is it better to use a 24v wall wart and skip the regulator altogether?


----------
edited to fix link

in this circuit a 24v unregulated wallwart puts out about 28v so you are using the regulator to knock it back down

[quote author="buschfsu"]in this circuit a 24v unregulated wallwart puts out about 28v so you are using the regulator to knock it back down[/quote]

Thanks Buschfsu.

Be sure to check the output of the wallwart under load. Often times, the wallwart will approach its rated voltage when a portion of the rated current is drawn. They use a slightly higher voltage secondary to ensure that it will still deliver the rated voltage when in use.

If the voltage gets nearer to 24V when connected to the amp, the regulator will not have enough headroom to work with and can start generating noise and doing all sorts of f-ed up things.

[quote author="skipwave"]Be sure to check the output of the wallwart under load. Often times, the wallwart will approach its rated voltage when a portion of the rated current is drawn. They use a slightly higher voltage secondary to ensure that it will still deliver the rated voltage when in use.

If the voltage gets nearer to 24V when connected to the amp, the regulator will not have enough headroom to work with and can start generating noise and doing all sorts of f-ed up things.[/quote]

Thanks for thinking it through with me.

This complicates things a little because the fetboy will have to be built before I can really test the real-world wallwart output. Unless there's some way to emulate a load.

I guess I will start with the 7824 installed and check the voltage before and after the regulator. That should give me a good starting place - and will most likely (it sounds like) work just fine.

What happens if a regulator doesn't get enough voltage to work with?
Shuts down? Or lets the voltage pass through as it receives it?

Let's see, in the original Hampton article, he says each JFP gain block draws 40mA. Two per channel times two channels, 160mA. The prototype used a 500mA wallwart that output 31VDC unloaded.

You can use a resistor, with value selected using Ohm's law to draw the same current as the preamps will. E/I=R or 24V/.160A=150ohms at 3.84W dissipation. Better make that a 5W resistor minimum. RadioShack has those 50ohm 10W cement resistors, could string three together.

http://www.radioshack.com/product/index.jsp?productId=2062292&cp=&sr=1&origkw=resistor&kw=resistor&parentPage=search

Ah, I see this was kinda covered earlier in this thread, and in the other "Fet Boy" thread that's up in the Lab again. Also, Scott discusses the regulator stuff on his page. A little reiteration is alright, especially when it comes to electronics.

This is a rather fussy little preamp to build, you trade simple, low part count design for time spent selecting parts and tweaking. It's definitely worth the trouble.

Hey, thanks for teaching me Skipwave.

Mnats ver. is a 1 channel, so I'll use 300R/5w resistor.

[quote author="skipwave"]
This is a rather fussy little preamp to build, you trade simple, low part count design for time spent selecting parts and tweaking. It's definitely worth the trouble. [/quote]

I'm finding myself more attracted to the fiddley projects because there's inevitable more learning / thinking involved.

I remember first reading the Hamptone article and thinking "I'll never be able to build that." But 5 years later, it seems quite do-able with all the helpful resources here. Thanks again.

Hi,

i know this is old thread but it makes no sense opening another.
So i'm building another pair of HJFP preamp from original schematic and i'm having some problems. When i put unbal. line signal through it (one module, 150k r series to the input and no 10M r to ground), sound is very thin and is not amplified at all, actually comes out even lower.
I'm wondering what would the problem be. When i turned ZTX635 the wrong way i got much stronger and nicer signal, but probably no amplification.

Can anyone help?

rotation said:

no 10M r to ground

Click to expand...

Put one in. You can't do without it.

rotation said:

150k r series to the input

Click to expand...

Why?

Hi,

i needed some time to reply because i moved to another apartment (finally mine
So, problem is the same, no gain, thin sound, very little output..
I measured some voltages:
2N5457
drain-13,8V
source-1,1V
output before 10u cap-12,8V
ZTX653
base-3,3V
emitter-2,8V

I added 10k pot at the input, series 150k resistor after it makes no change. I added it first because it is on Hamptone schematic, where this op amp is used as a line mixer. I have 10M to ground resistor.
Schematic is here:
http://rockstudio.com/jpg/JFPsmall.jpg

The 150k resistors in the line mixer schematic are the summing resistors. They are there to present the right impedances to each channel as well the summing amp, the JFP. That is a gross over-simplification, but no concern for now.

The summing resistor is not needed for a single preamp. Get it out of there.

You should be able to simply feed a line level signal to the JFP input, the 10k pot is fine for level control.

Your voltage readings look normal. I can open mine up later tonight and verify.

Strange that you say you got a stronger signal with the ZTX the wrong way. It is only functioning as a current source for the output darlington. Try picking the signal off at the drain of the 2N5457 - through a small non-polar (most any poly-film will do) blocking capacitor - and take that to the input of your working JFP. This way you are signal tracing to check whether the problem lies in the first stage (the voltage gain stage) or the second (the current gain stage). If you don't get a nicely amplified signal, try a different FET.

Strange thing with ZTX turned around was happening with the other card which is working fine, but is somewhere in mountain of my stuff (i moved to new apartment). I would like to get this one to life, i stuffed only one op amp for now..
Btw, my source is cd player, i will try another fet now. Thanks.

Miha

Well, it works fine now, i plugged it directly into amplifier. It looks like the problem was that HJFP and hi-fi preamp didn't agree about something, i have another problem now   Hi-fi preamp is based on Elektor ECC82 preamp (i was told so, it's not built by me) and has very sensitive inputs. Btw, HJFP sounds much, much better than this tube preamp..
Thanks a lot for helping me out!

Miha