freestompboxes.org

MoonWatcher wrote:

briggs wrote:I found that when I started sorting jfets around about 50% didn't meet my tolerance levels...

I seem to remember Brian Marshall going thru all of that with the Liquid Sunshine, until he finally found a source (or sorting method) for consistent ones. But in that case, they are 2N5458's, right?

I personally don't care for doinking around with j201's because of the sorting part, even though they could all probably be sub-grouped for different projects.

But a tip of the hat to those who DO take the time to sort those things. I wonder how it ranks compared to picking fly shit out of pepper with boxing gloves on?

The tolerances of Jfets do still amaze me! The tolerances of some devices are akin to old germanium transistors....

briggs wrote:The tolerances of Jfets do still amaze me! The tolerances of some devices are akin to old germanium transistors....

Yeah, and unfortunately I oftentimes have little patience for either! I'll just use a jFET as an input buffer here and there, and be done with it. Silicon BJT's with a small cap across the B & C are a lot more appealing than dealing with temperature and leakage issues with germs, too.

Again, a tip of the hat to anyone who will take the time to sort germ trannies, or any other component with tolerances all over the place.

it's easy. just buy non-leaky, consistent Ge transistors. they are extremely easy to find. really.

RnFR wrote:it's easy. just buy non-leaky, consistent Ge transistors. they are extremely easy to find. really.

How are they priced? How are they affected by temperature changes?

Leakage and consistent gain are just the first two hurdles. And until I win the lottery, price is also important. If I can get decent germanium transistors for a nickel a pop, I'll give it a whirl. Otherwise, I'm out...

Russian transistors all the way!

Finally. A amp sim trying to capture power tube distortion that is made with the tone stack BEFORE some of the clipping gain stages!

did i miss something here?

You can tell in the youtube clips. Burgerman or Gearmandude turn up the bass and the bass mostly just gets more distorted, not really louder. Just like when a real amp is turned up enough to get the power tubes and/or phase inverter clipping. The tonestack at that point is more for sculpting the distortion than changing frequency response.

In these "amp model" jfet designs you usually find the tonestack is in the same position as it is in the actual amp they're trying to model - usually after the first/second gain stage before the "power amp" (final gain stage)...

Would love to build this one, the '57 Tweed by Wampler is gonna be even awesomer! Anyone that can take some gut shots?

I would love to hear a YouTube video not using a tele. Not that those tones are not great. But, specifically, I'd love to hear someone try for the early Santana tone.

Well.. traced this for my own purpose and since I made such a wonderful work, I'd like to share it (since I could'nt find one myself when looking for it).
If someone is interested in the original DSN files, PM me for them.
Nir

A little follow up on this.
The design mimicks either a 5F6A Bassman or a plexy Marshall. The EQ values/gain stage construction is 5F6A (which the Marshall is of course based on). The resistors/pots values were reduced x10 while caps increased x10 to get approx the same EQ curves. This is not a "Black 65" as the name might hint (Twin Reverb/Deluxe Reverb etc). To get a Twin EQ the value of C12 needs to be 1uF, while C14 needs to be 0.47uF. The 5.6K contour resistor (connected to C12 and C14) should be increased to 10K and the bass pot should be changed to 25K.
The output rolloff (C7/C11) might need to be removed as well for that "glassy" Twin sound.

Cool, thanks for the share! I may bread board this one soon!

I think that Wampler matches the FET's, i'll be nice to get them measured to be sure, and the bias voltages.

Thanks!!

Hi Sweetalk.. Still have the pedal here (belongs to a friend). I will measure the bias point of all 4 FETs tomorrow. I have seen test points on the PCB for all 4 bias points but not sure they do match..
Also not sure it's worth matching - assymetrical clipping always sounds nicer...

OK - here are the bias points:
First FET - 4.08V at the drain.
Second FET 4.00V
Third FET - 3.95V
Last FET - 3.93V
VCC set at precisely 9.00V.
Doesn't look like either hand picked/matched set or really good bias point - I'd bias them for 5V @ drain so I get more dynamic range before clipping but I guess it doesn't really make such a hugh difference anyway.

dnts wrote:OK - here are the bias points:
First FET - 4.08V at the drain.
Second FET 4.00V
Third FET - 3.95V
Last FET - 3.93V
VCC set at precisely 9.00V.
Doesn't look like either hand picked/matched set or really good bias point - I'd bias them for 5V @ drain so I get more dynamic range before clipping but I guess it doesn't really make such a hugh difference anyway.

Thanks a lot, grat job!!!. Maybe they're matched to have a Vp around some place to get a 4V bias with that fixed resistors.

Okay.. so I traced the schematics for this pedal.. but placed them in the wrong forum (duh):
viewtopic.php?f=1&t=18561
Anyway - matched JFETS indeed (all ~4V @ drain). J201. Straight forward design. Gain, Gain, tone, output buffer. A few things noted:
The tone stack is Bassman 59 (5F6A) and NOT blackface. (resistors are x10 smaller, caps x10 bigger). The output buffer has ~10KHz roll off @ 40dB/decade filter. The 10pF caps used (metalized film) are probably useless (give Ft @ over 100KHz). No real voodoo in this one so I guess that whatever keeps it ticking is pure hype... decent sounding preamp that is no different from dozens of other preamps about.
Anyone interested can probably breadboard it to see what it sounds like (I'd skip the extra gain stage as it's not pleasant sounding to my ears).

About these FETs: I breadboarded the balck 65 from the schermatic above and got a fine clean sound but no overdrive. I found all the J201s needed bigger resistors to bring that drain voltage down to 4v, the quoted measured value from a production unit. Then I got OD.

This brings up the whole issue of sorting FETs. At least one expert has said that putting trimmers on the drains is amateur stuff and hinted that there is a better way. Perhaps this better way is buying a big lot of FETs and sorting them?
I had about 25. I made a test rig with a 1k between source and grnd, a 1m between gate and ground (the parallel caps didn't change the results), and a trimmer in series with a 3k between drain and 9v, mimicking the set up of the FETs in the black 65 schematic. I understand this is crude, that the real method involves tracing a curve of different voltage/current relationships.
I tweaked the trimmer until I got 4v on each FETs drain, then measured the resistance. The group came in between 15.8k and 52k. I also notices that the FETs from Fairchild all came in at the bottom of the group, all 20k or below.
The others just say J201, but appear to be hand color-coded on the back: three little stripes. This implies I have bought previously sorted and then re-mixed transistors. I don't remember who I bought them from, but it was one of the well known suppliers to pedal boutiquers/hobbyers.
Is it possible that makers buy big lots, color code, then resell their trash?
It would seem the FETs in the original pedal all were chosen to be happy with 5.6k resistors, but that with my set I need higher values. I had a couple of groups of more than 5, one requiring 40k, the other 52k.
I tried the 40k group of FETs in the circuit with 39k drain resistors. They now all read around 4v on the drains. The circuit has too much drive, gain on minimum doesn't clean it up. The sound is clear out of the first stage but pretty fuzzy out of the second...I realize now I built this without the boost switch but with all four stages always on. I took out the first stage and now its almost clean... still some work to do.