Hamptone jpf gain stage current consumption too high

74 ma per stage...2 stages per mic preamp

it should be about 40ma for each gain stage right??

Output transistors heat up a lot! but i have some really big heatsinks on them so there's not much heat.

Using everything as in the original hamptone schematics, but added bias trimmer for the prototype and idss control. I'm using the fairchild 2n5457 fets. The 1/2W 100R resistor were getting too hot, changed it for a 1W. My modules are built in my own designed pcb's and biased at exactly 12v. I should use a SOT resistor for biasing instead of the trimmer but for now it's working fine.

The preamp sounds stunning but i'm really concerned about too much current draw if i plan to build 4 4ch 1RU unit...i blew my little transformer so in the meanwhile i'm just getting a bigger one.

I've heard that the farichild fets can cause some trouble. But cannot get the ON semi fets for now...

Any help on this??

Thanks!

Look around some of the older threads about this.

Fets have a range of specs in the same part number. BJTs do as well

IMO I don't think you parts have any issues.

I have not built this but maybe some measured the specs of parts that worked and sounded good That is maybe what you are looking for.

This is by design. Check out John Roberts' comment about that current source: http://www.groupdiy.com/index.php?topic=296609#296609

I wonder if the one extra $0.05 part for a reliably working current source would kill the sound... :twisted:

;Matthias

This was the controversy---whether the sound was dependent on that "terrible" I source design. High beta parts with that arrangement will give high output currents, and to make matters worse beta goes up about 0.5% per degree C.

I'd do some A/B listening tests between two designs, one with decent temperature compensation as suggested by Gus in the other thread, or similar, and one without (but maybe the nominal 68k R raised to reduce the high dissipation).

It's going to be a highly colored preamp in any event, with gobs of second.

so, you think 74 ma is common draw for each jpf module?

no 40ma is normal, but the actual current draw is dependent on having exactly the right devices -- you have to sort the FETs and select the BJTs for the correct hfe. also, in the commercial version of the HJFP2, the 68k resistor has been changed to 127k.

the details are all in the other threads about the hamptone jfet pre. keep reading, there's lots of good info.

ed

iirc, i just picked up random 2n5457 (i wanted to know if the biasing with a trimmer will work with "any" 2n5457) and bipolars...and the current draw is exactly 74 mA in both modules. i dont remember if i used 68k...maybe i used 47k as described in the original article so i´m raising them to 68k to see what happens.

Thanks for the replies!

Success!

changed the 47k to 120k, current draw for both modules now is 82ma. I need a new power transformer to replace the one i fried due to high current draw, need to use my pre in a session i was recording with it.

Thanks.

i was wondering...what if i raise the value of the 47/68/127k resistor? what will change? the gain of the module will change? today i built a couple more modules, fired them and have about 54 mA each, with current draw raising as temperature increases on the output transistors...which are cooler now but quite hot.

Gain is not a sensitive function of current sink current. I'd listen, and also take the resistor up to where the current is not so large. And listen again.

The voltage gain is mostly set by the first stage. The second stage is roughly unity gain.

It really is a very unstable, ill-advisedly-unmanufacturable circuit. But maybe it is fun.

[quote author="bcarso"]It really is a very unstable, ill-advisedly-unmanufacturable circuit.[/quote]
In a magazine I forgot the name of there was once an article about "50 sounds that must die", with entries like that & that preset on this & that synthesizer.

If we'd make a "50 circuits that must die"-list I assume quite a few people here would vote for this circuit here, not ? :twisted: :wink:

BTW, if that list indeed emerges let's not forget to add all TR-808 circuits :!:

Yeah, maybe is a difficult circuit, i am aware that it needs some tweaking to get it working right...but the sonics...i work in the top studio in Colombia and i have a little nice choice of preamps here including a 4 ch hardy, amek DMA, hhb radius 10, neve 31102 (28 in the console) and the stock mxp3000 pres which i like as a clean sound. The prototype i built constantly surprises me and the clients as a very colorful and magical sound, i mean, the neves are great, but mine is sweeter and the top is super defined but silky smooth. maybe is the iron...anyway i dont have any reason to be wanting to kill this circuit, i cannot agree with that. Surely it is ineficcient but THE SOUND! i´m in love with the sound, and want to build and tweak a bunch of them and some of my clients want one too.

Thanks for the replies!

The circuit-killing just for fun of course, this one is a nice example of 'sound' vs 'technical (in-)elegance.

[quote author="Silvas"] maybe is the iron...[/quote]
Just curious, which input- & output-iron did you use?

The 1:6.45 OEP for input here and I have some 600:600 TXs to try (Altec 15356A & Edcor) for the output.

Regards

would some api style output trasnformers would work on this??

[quote author="3nity"]would some api style output trasnformers would work on this??[/quote]

This design can barely drive 600 ohms through a 1:1 turns ratio transformer, so you could use a 2503-style step up output only if you are running into hi-Z loads.

It's typical of simple emitter follower output stages, and is why we see a lot of push-pull outputs in the discrete opamps that can perform heavy lifting. Take a look at the 2520 schematic, or 990, FETBloak, etc.

Silvas - How many volts at the source of your FET? The idea behind selecting that part is to idle at 12V (half the supply rail) with the stock values. That will yield the voltage gain described in the article. You could adjust the source R to match your FETs, but compromise voltage gain. Most of the current is drawn by the output device, but don't overlook the importance of an appropriate FET.

Input iron is a rare peerless 1153, from a 1950´s ampex tape recorder, no output iron. Voltage at source is 12.4, trimmer adjusted for approx. 12v at the output of the module.