Dead Astronaut FET Bassman Preamp

[patrick398]

Hello all,

Today i finally got round to building the FET bassman preamp. I'm using the dead astronaut schem:

I'm thinking that it's breaking up more than it's supposed to though. I don't have a reference point so it's just a feeling but i assumed this thing would have a lot of clean headroom. As it is it's breaking up into mild overdrive when i hit a chord hard on my tele. It sounds really nice but i'd rather it performed as it's supposed to at first.

Here is the vero layout i drew up, i buggered up the pinout for the fets so i've had to bend the legs around. On the build i've got drain at the top, gate in the middle, and source at the bottom.

Any ideas what could be causing it to break up? or is it supposed to be like that?

[Danich_ivanov]

It doesn't have a capacitor on the input which is unusual, i would put 100nf right before 68k, and it might just be the reason for a breakup. The other thing are these 100k pots on the drain which do adjust biasing and gain, realistically i wouldn't use anything above 10k with two 2,2k resistors, one from drain to pot, the other one from pot to "+".

[idy]

Plus 1. Did you try to bias the FETs? You can even do it by ear, although 100k trimmers willl make it pretty hard to find the sweet spot.

[patrick398]

It doesn't have a capacitor on the input which is unusual, i would put 100nf right before 68k, and it might just be the reason for a breakup. The other thing are these 100k pots on the drain which do adjust biasing and gain, realistically i wouldn't use anything above 10k with two 2,2k resistors, one from drain to pot, the other one from pot to "+".

I never thought about there not being an input cap but thinking about it i don't think i've ever built anything without one. I guess because it's designed for bass? Without an input cap all frequencies are being allowed to pass meaning there's more signal to drive the fets?
Both fets are biased to 4.5v and then adjusted a lot trying to find a spot with the least break up. At either extreme of the trimpots rotation it's fizzy and splattery as you'd expect from mis-biasing, i'm pretty sure i'm in the sweet spot, its a very pleasant distortion i'm getting...i would like to clean it up though.

[Danich_ivanov]

It doesn't have a capacitor on the input which is unusual, i would put 100nf right before 68k, and it might just be the reason for a breakup. The other thing are these 100k pots on the drain which do adjust biasing and gain, realistically i wouldn't use anything above 10k with two 2,2k resistors, one from drain to pot, the other one from pot to "+".

I never thought about there not being an input cap but thinking about it i don't think i've ever built anything without one. I guess because it's designed for bass? Without an input cap all frequencies are being allowed to pass meaning there's more signal to drive the fets?
Both fets are biased to 4.5v and then adjusted a lot trying to find a spot with the least break up. At either extreme of the trimpots rotation it's fizzy and splattery as you'd expect from mis-biasing, i'm pretty sure i'm in the sweet spot, its a very pleasant distortion i'm getting...i would like to clean it up though.

Yes, i assume it is there to pass more frequencies, which may also result in overdriving fets harder. I don't think that adding 100nf - 1uf will result in being less usable for bass, but certainly more stable performance. With 1k on the souce (+ cap) i doubt you will need more than 4,7k - 10k (max) on the drain for "clean'ish" effect.

[patrick398]

Ok i'll try a 100nf on in input and see if that tames it a little bit.
How much variation is there likely to be between fets? I got a bag of 10 so could try some others

[anotherjim]

It doesn't absolutely need to have an input cap, but if you plugged something in that had some DC on its output, then it would misbias the 1st FET.
With 1M input impedance, a 100nF input cap isn't going to be cutting any music frequencies. It will still be full-range for bass & guitar if you make it 10nF.

I'm not sure if it could have enough gain to clip easily. If you do want to cut the gain, then removing C2 does it in one hit. If you want less bottom end push, reduce C2 and/or C6. A lot of "fenderish" circuits have 22uF in those positions.

As suggested, check FET bias with the 100k trimmers. Setting for 4.5v on the drains is a guide. Get 4.5v (half of the battery volts) first, then tweak by ear for best clarity.

[Jeema]

The problem may just be that the circuit has too much gain for 9V of headroom.  The second stage alone is gonna be about 50x gain with the trimmer halfway by my calculations.  That alone might cause breakup.

Removing one or both of the source bypass caps (C2/C6) would for sure reduce gain, though.  Maybe try that if nothing else.

[patrick398]

Ok i'll try and reduce/remove the source caps tomorrow when i get home. I'll double check those 1k Rs too, if i've accidentally put in 10k would that result in more gain or less gain? It's a long shot but worth double checking

[Danich_ivanov]

Ok i'll try and reduce/remove the source caps tomorrow when i get home. I'll double check those 1k Rs too, if i've accidentally put in 10k would that result in more gain or less gain? It's a long shot but worth double checking

If you mean source resistor, it shouldn't affect gain much, since there are those giant 47uf caps that greatly diminish the effect of that resistor on gain, it will affect the tone (and even then not so much) more so than gain. But if you decide to get this caps out, it will affect gain depending on where drain pot is set, the less source resistor is in relation to drain resistor, the more gain, untill you reach a dead zone. But i personally wouldn't touch this caps, since they sculp the tone of the whole thing, it most likely won't sound (tone/mojo) as it intended to without those caps.

[patrick398]

Thanks so much for all the help. I don't want to mess with the tone/mojo of the circuit as it does sound fantastic. There must be a reason i'm experiencing more clipping than there's supposed to be though and i'm keen to found out why.
I'll double check all values tomorrow and maybe record a sound clip so you can judge for yourselves

[Marcos - Munky]

The problem may just be that the circuit has too much gain for 9V of headroom.

I built something similar to this circuit (a fet version of the Alembic F2B) and it breaks up at 9V. I tried it at 19V (a regular notebook power supply) and got way less distortion than at 9V.

[anotherjim]

Bearing in mind that...
The second FET is fed from a volume control, so we can turn that right down. If that doesn't reduce the distortion, something is clipping in the first stage.
With that Mids pot value, the tone stack itself loses quite a bit of the power out of the signal from the 1st stage. If you disconnect the tone pot, you will hear just how much louder it gets.
On 9v supply, the JFET's are often working with drain resistors of around 12k to 18k.

Playing a Tele into a proper Fender amp can be done using the second "Low sensitivity" input for a different clean flavour. That input has a lower impedance and also cuts the input level by half. You can simulate it exactly by fitting another 68k resistor across the 1M input resistor.

[patrick398]

Bearing in mind that...
The second FET is fed from a volume control, so we can turn that right down. If that doesn't reduce the distortion, something is clipping in the first stage.
With that Mids pot value, the tone stack itself loses quite a bit of the power out of the signal from the 1st stage. If you disconnect the tone pot, you will hear just how much louder it gets.
On 9v supply, the JFET's are often working with drain resistors of around 12k to 18k.

Playing a Tele into a proper Fender amp can be done using the second "Low sensitivity" input for a different clean flavour. That input has a lower impedance and also cuts the input level by half. You can simulate it exactly by fitting another 68k resistor across the 1M input resistor.

So that 'volume' pot is really more of a gain pot, reducing the signal level going into Q2? I did consider making R1 (68k) a pot, a sort of 'pre-gain' control.
When i probed the circuit there was no clipping at Q1 so first stage should be ok.
With regards to the fitting another 68k R do you mean in paralell with the 1M to ground? I thought that was just a pulldown resistor to avoid popping?

[EDIT] Just did some reading, makes sense. 68k in parallel with the 1M decreases signal level by allowing my signal to bleed to ground yes? That is going to decrease the input impedance though right?

[thermionix]

I know nothing about converting tube circuits to JFETs, some changes at the input may have been necessary.  But to copy the #2 (Low) input of a typical Fender amp, change R2 to 68K.  To copy the #1 (High) input, change R1 to 33K (34K) and connect R2 (1M) to the input side of R1.

[anotherjim]

Yes, that "other" input does have a lower impedance. It's 68k + 68k (with 1M in parallel which makes little difference) = 136k. However, the standard single coil Telecaster doesn't seem to mind working into less than the standard guitar 1M load. It's actually easy to be plugged into that input without realising it.

Strictly, a "Gain" control directly affects the amplification factor of a circuit. Gain controls often cannot turn to silence, just minimum gain to maximum gain. Here the gain is set by the drain resistors of each FET and there is no panel knob for it. A volume control placed before an amplifying section can be considered a "Drive" control.

[thermionix]

I'm not seeing where the 1M is ever in parallel with a 68K.  Plug into #1 and  the 68Ks are in parallel with each other (34K), the 1M from tip to ground.  Plug into #2 and the switch in the #1 jack shorts the 1M, so 68K in series to the grid, 68K from grid to ground.

[anotherjim]

Yep, my memory failed me there. The 1M is shorted out in the Fender scheme. Although it's still of no significant consequence to the impedance.

[patrick398]

So I dropped the 1M R to ground down to a 68k, dropped the 47uf (C1) down to 22uf and removed C6 and it has definitely cleaned it up a bit. The 'hi/lo' switch seems to change volume though. The hi setting is a fair bit louder than the lo setting so I'm starting to think something else is going on here. Is there an attenuation of frequencies significant enough in the lo setting that I should expect a drop in volume?
As expected the volume pot does effect gain, pretty clean and chimney at lower settings but when it's cranked its full on distortion, especially in the hi setting. I feel something is amiss

[deadastronaut]

edit: cant believe nobody spotted this...

scrap your layout...the fets are wired wrong...

the drain goes to v+ trim..

the source goes to r/c to ground.

the 68k goes to gate.....



watch your pinouts.