a problem with biasing a BS250 mosfet (obsidian)

[SpheriX]

I've just built an Joe Davisson's Obsidian (BS250 version). The problem is that drain resistor value (originally 3.9k) doesn't provide the required voltage at the drain of the BS250. The voltage is -8.2V (the battery measures 8.5V) so the voltage drop is only 0.3V,but should be about 4.5V as stated in schematics. When I put in a 100k R, drain voltage is somewhere about 6V.

The question: Could it be that the leads from the BS250 are not what the datasheet says (D-G-S),and  what should I try to do?  :shock:

[Joe Davisson]

The pinout is right, looking at the flat side with the leads down. It ends up like a 2N3904 put in backwards.

Is the battery flipped around for -9 volts? If so, then it sounds like a blown (or partially blown) MOSFET. The differences between parts could be .5v off or so, but not that bad to need 100k+.

The only other (and typical) answer is a misread resistor value or wiring mistake.

[SpheriX]

I've doublechecked the wireing and it seems ok. The interesting thing is that all three BS250 act the same way (could all three be blown); the sound is close to the soundsample I've found on the web (maybe even better). BTW the MOSFETS reads BS250 (no manufacturer logo or whatsoever), so if I flip it the other side round to check if the D-S-G pins are inverted could i damage them?

[bassmeister]

I have about the same problem right now. I hooked up the whole circuit, and it worked immediately. It sounded very thin though, just like a misbiased transistor. Drain voltage on each transistor was ~ -8,3 to -9 V, power supply ~ -9,3 V, so misbiasing confirmed. First question, should I vary the drain resistor or the biasing resistors to get Vsupply/2 at each drain? If I vary the drain resistor, don't I change the sound characteristics of the circuit as a result of messing with the gain in each stage?

Anyway, I wanted to experiment with the biasing, so I hooked up one single stage, like a "one transistor Obsidian" (with 0,1u output cap and 100k volume control). When biasing like the schematic suggests, I again read ~ -8,3 V to -9 V at the drain, depending on which transistor I use. So, I tried to bias using a 100k trimmer hooked up as a voltage divider with one end of a 3.3M resistor connected to the wiper and the other end connected to the gate. By doing this, I had no trouble trimming for Vsupply/2 (in my case, -4,65) V at the drain. This happened, with all transistors, when the trimmer was set very close to the middle.

Conclusive questions: I understand that the biasing resistors need to be tweaked to get Vsupply/2 at the drain of each transistor, but the ratio between the biasing resistors in the original schematic differs quite much from what I used to get it working (1.5M/(1.5M+3.3M) vs. 50% - 5% perhaps). Why is that? Does it solely depend on the characteristics of the transistor? I prefer using a trimmer to adjust for inconsistencies when designing with FETs because I get a good result quickly. But, is there something I could think of when using a two resistor divider to bias to make designing easier?

Bonus question: When measuring the gate voltage using an ordinary two resistor voltage divider as biasing network, I read -1,4 V. That should mean Ohm was wrong, since I use 1.5M/3.3M and ~ -9.3 V as power supply. Could the problem lie in the DMM having a too small input impedance, effectively parallelling the 1.5M gate-to-ground resistor?

(Did anyone get through to here? )

[R.G.]

Has everyone read "Designing MOSFET Boosters" at GEO? Good place to start.

[bassmeister]

Has everyone read "Designing MOSFET Boosters" at GEO? Good place to start.

I've read it. I learned from that document how biasing of a MOSFET works when using a trimmer and a resistor from the wiper to the gate rather than using a two resistor voltage divider for biasing, as the original Obsidian schematic suggests. I would've opted for the first solution from the beginning, but I wanted to try in order to learn. If I haven't read the document, I would probably have given up when it didn't work. You need to have knowledge when troubleshooting, and that article provided that for me.

Thanks R.G.

However, it takes a while and also some practice to REALLY learn. Just reading it through doesn't help very much, but the document serves as a very good help when practicing the theory. Read, build, make mistakes, read again, rebuild. In that way, "Designing MOSFET boosters" has helped me a lot.

[Khas Evets]

Are there any plans to complete the "Designing MOSFET Boosters" article at Geo, R.G. I know I'd be interested.

[R.G.]

Quote:
Has everyone read "Designing MOSFET Boosters" at GEO? Good place to start.


I've read it.

OK. From your description, you either can't get the MOSFET to turn on enough to pull the drain resistor down, or the source resistor (if any) is too high. That suggests to me that there is either a connection that has not been made correctly or one of your resistors is not the value you think it is. Check them with your ohmmeter.

Are there any plans to complete the "Designing MOSFET Boosters" article at Geo, R.G. I know I'd be interested.

Yes, there are. There's one coming up in the next couple of months, depending on my schedule.

[Khas Evets]

[Dave_B]

See: http://www.diystompboxes.com/smfforum/index.php?topic=39974.0