Why biasing of some transistors in Sandspur/Sunflower PCBs doesn’t work?

[harmaes]

I’m trying to better understand why biasing sometimes works and sometimes doesn’t work.
I have a practical question about the Sandspur / Sunflower PCBs. I’m trying to use 2n1308 NPN germaniums in the Sandspur and noticed that the collector voltage of Q2 can’t get below around 5,5V. Bias and sundial to minimum. Also the range of max/min is just over 1v with somewhere in the 6-7V on the upper end of the trim and sundial.
These 2n1308s are around 200 HFE, with a VBE of 0,358V, Iceo 84,3uA, Ib 4,58mA.
The silicon BC546A has a similar HFE with VBE around 0,76V, Ib 4,1mA. These bias correctly.

So I’m trying to understand why the 2n1308 can’t bias correctly without modification to the circuit? How to calculate and explain this and what needs to be changed and why? Is there a formula on how to calculate this?

The Sunflower is identical to the Sandspur, only the inverter is in the Sunflower.
2n1305, 2n1309 PNP germaniums etc bias correctly in the Sunflower. Is this related to the NPN germaniums?

 

[giovanni]

My guess is the leakage current and different BE voltage have a significant effect on biasing. This article discusses some of these differences.

 

[Harry Klippton]

Have you tried the amplified parts bias calculator?

https://www.amplifiedparts.com/tech-articles/fuzz-face-bias-calculator

 

[harmaes]

Have you tried the amplified parts bias calculator?

https://www.amplifiedparts.com/tech-articles/fuzz-face-bias-calculator

I filled in the details of the calculatur for the 2n1309s and get 5.1v on the collector of Q2 with 8,2k resistance.
I measure 5.5v with full resistance of 11k. I can replace the sundial with a 25k to be able to raise the resistance enough to lower the voltage.

EDIT: replacing the 5k sundial with a 25k solved the problem.

 

[SillyOctpuss]

What is the hfe and leakage of the 2n3108 that you can't get to bias? I've only ever had trouble with transistors that are too leaky. Also are you letting them settle after fitting them before trying to bias? The gain and leakage (especially leakage ) will shoot right up as you handle them when putting them into sockets, especially if if takes a while to line the legs up with sockets. The leakage can easily double.

 

[harmaes]

What is the hfe and leakage of the 2n3108 that you can't get to bias? I've only ever had trouble with transistors that are too leaky. Also are you letting them settle after fitting them before trying to bias? The gain and leakage (especially leakage ) will shoot right up as you handle them when putting them into sockets, especially if if takes a while to line the legs up with sockets. The leakage can easily double.

Mentioned above what the HFE and leakage is. I’m aware on how to handle germanium transistors.

 

[SillyOctpuss]

Mentioned above what the HFE and leakage is. I’m aware on how to handle germanium transistors.

I only asked because I could only see one leakage figure and wasn't trying to talk down to you about handling germanium transistors.

 

[harmaes]

I only asked because I could only see one leakage figure and wasn't trying to talk down to you about handling germanium transistors.

Cool, which other leakage figures are you looking for?

 

[Chuck D. Bones]

If you use a Ge tranny for Q1 in the Sandspur, the biasing will definitely be thrown off. Q1's Vbe and leakage have a major impact on the overall bias. Ge tranny Vbe is between 60mV and 170mV. Si Vbe is around 650mV. Huge difference. Try using Si for Q1, Ge for Q2 and see how you like it. I should also add that 2N1308 is not my favorite Ge tranny in a FuzzFace. The high HFE and bandwidth results in a harsh tone. You might try AC127, 104NU71 or MP38A for Q2.

 

[harmaes]

If you use a Ge tranny for Q1 in the Sandspur, the biasing will definitely be thrown off. Q1's Vbe and leakage have a major impact on the overall bias. Ge tranny Vbe is between 60mV and 170mV. Si Vbe is around 650mV. Huge difference. Try using Si for Q1, Ge for Q2 and see how you like it. I should also add that 2N1308 is not my favorite Ge tranny in a FuzzFace. The high HFE and bandwidth results in a harsh tone. You might try AC127, 104NU71 or MP38A for Q2.

Good to know, I didn’t really like the response of the 2n1308 in the fuzz face either. I will try some silicone in Q1 with a NPN germanium in Q2.

For Si I like 2x BC183C around HFE 500 and a combination of BC183B and C (250/500). Not a fan of BC108C and BC109C but a 2n2369a with a BC108C sounds much better. Also like 2n3903 and 2n3904 and 2xBC546A.

In the Sunflower I’ve used PNP germaniums and I like different combinations. The GT308 with lower HFE 60-70 and 80-90 sounds great, liked AC125 with HFE 80-90 and 120-130 as well. Also 2SB171s are nice but I can’t find 2Sb171 or 2SB175 with higher HFE for Q2.

 

[Chuck D. Bones]

I don't think high HFE is beneficial for Q2, but try them and see what you like.

 

[harmaes]

I don't think high HFE is beneficial for Q2, but try them and see what you like.

The Thorpy Tacit Blue (Tactical Fuzz PCB) uses a set of CV7351 with 120 and 180 HFE (2n1306) which sounds quite nice in their slightly modded fuzz face circuit. I could get them around 114 and 164 HFE which worked quite well IMO.
Only thing is that all fuzz faces seem to produce quite some noise.
Cutting down the volume and fuzz seems to help a bit and a noise gate too.

 

[Chuck D. Bones]

I think it's more than a "slightly modded fuzz face." There is a lot of filtering in there and the SMOOTH pot, none of which you'd find in a standard FF.

With gain comes noise. Ge transistors are leaky and leakage gives rise to noise.

 

[harmaes]

I think it's more than a "slightly modded fuzz face." There is a lot of filtering in there and the SMOOTH pot, none of which you'd find in a standard FF.

With gain comes noise. Ge transistors are leaky and leakage gives rise to noise.

I don’t think it’s a heavy mod but it does make it a nice sounding fuzz, noise is quite prominent though.