|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
70.117.113.124
This is my first attempt at diode biasing, and I know little or nothing about solid state devices. I bought some Cree 1 amp 600V Schottky SiC diodes. The specs at Mouser said that they have a forward voltage drop of 1.8V, so thought two in a series string on the cathodes would give me 3.6V bias, plus .2V for my 10 ohm sensing resistor (tube running at 20ma.)for a total bias of 3.8V. When I fired her up I find I only have about 1.67V bias, so allowing for the .2V supplied by my 10 ohm sensing resistor, looks like each diode is only dropping about .735V each, at 20ma. Am I correct that at 20ma. a string of 5 should give me ~ 3.65 bias, and with the 10 ohm sensing resistor give me about 3.85V total bias?
In case it matters, this is on a simple triode gain stage parafeed preamp, using a 6N6P tube, with the K&K cascode CCS set at 20ma for the plate load. The plate is also directly connected to one side of a pair of Magnequest nickel core 15K B-7 OPTs, the other side of the primaries goes to a parafeed cap, which returns to the top of the cathodes of the 6N6P. Am I on the right track here?
twystd
Follow Ups:
A few years ago I went to the trouble of measuring the forward voltage of those diodes at lower currents than the spec sheet. It turns out that the dynamic impedance is still very low at lower currents, much less than most LEDs, for example, especially once you get above 15 mA or so. I use them with great success to bias the 6E5P in my 833C amps to ~28mA.Here's the curve:
Edits: 09/12/14
Thanks for the curve, seems as if my diodes (TO220s) are dropping slightly less voltage than the curve indicates. BTW, those "Mid Life Crisis" amps are absolutely awesome! Maybe one day I can afford to finish my 300BC PP class A boat anchor project. Not on the same level as your amps, about 25-30 watts, at least 2 chassis, one for the signal, and the other for the 3 power supplies, lots of big iron. I started it when I had money...now not so much. :-(
twystd
Am I correct that at 20ma. a string of 5
Yep, you will have to string several together.
To bias my 6N6P's cathode, I used six Crees (csd01060E-TR) with an STPS2150 to get just where I wanted to be, around -5.4V
If you go this route, buy a few extra.:-)
Those Crees are the TO252-2 package. They're rather tedious to work with due to their size. Once done, it's a tight compact string of diodes.
I found stringing that many TO220s was not realistic and they wouldn't stand up.
Hope this helps.
Cheers,
W
Thanks wheezer, the next time I diode bias a project I'll try those. Since I already bought 15 of the TO220 diodes, I'll try and use them. Plan is to mount 2 strings of 5 on perfboard, using silicone. I can also mount the sensing resistors, and allow for test points on the board. I'll mount the board as one unit on standoffs. Should work, what you think?
twystd
That will work fine. A problem arises when stringing those 220 cases w/o adequately securing them.
My choice of silicone is Permatex sensor safe or GC electronics grade silicone adhesive.
Cheers,
W
Nylon studding is the way to do it
Al
Just finished, all worked out fine, right on the expected voltage, 3.9V. Kind of a hassle doing two strings of 5 diodes, two sensing resistors (to measure current through tube) and 4 test points, but well worth it. Line stage sounds better now, even more detail yet smoother.
Don't know if it was the difference between battery bias and diode bias, or whether it was changing the parafeed topology from the conventional (parafeed cap between plate and transformer and grounding the other side of the primary) to directly wiring the plate to one side of the primary, and using the parafeed cap between the other side of the primary and returning to the top of the cathode.
Maybe it was both things working together, never know for sure, as I did two things at one time. I know better, but the underside of the tube socket is hard to get to the way I built the thing, and didn't want to take it apart twice. Very pleased with the sound, just the ticket if you need an active line stage, about 10dB of gain, about 80 ohms output impedance. Maybe one day I can save up for those AVCs, then I'll put it in a pretty chassis and call it a done deal.
twystd
The 1.8v forward voltage spec is the maximum at 1 amp current (typical is 1.6v).
The plot of forward voltage against current is unreadable at 20mA, other that to say it's "less than 0.8v".
If you look carefully at those curves, you will see that they are essentially straight above about 200mA, meaning the forward voltage drop at these currents is dominated by resistance, not by the semiconductor behavior.
Since the tube is running a lot less than 1 amp, the voltage drop across the diodes is less too.
If you had measured the voltage drop across the 10 ohm resistor, you would have sorted that out right away!
Below 200 mA, "semiconductor behavior" means that the voltage increases as the logarithm of the (forward-bias) current.
That all sounds right to me.Maybe there is some threshold voltage on the diodes and below that the drop will not be as great????
Edit, Current threshold, not voltage threshold.
Thanks Paul!
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/11/14
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: