I can see how this constant bias fantasy would work better with a cascode constant current source, and the output taken from the "mu" connection.
Bear in mind that a simple triode stage with a CCS load and a bypassed cathode resistor will have constant voltage across the resistor (and hence no capacitor distortion) until you apply a load to the anode. Real circuits are always more complicated in operation than ideal ones, although to understand real ones you first have to understand ideal ones.
Ah, yes, Dave - the pesky real world. It ruins many a fine idea!
Fred, I think that you could connect things pretty much as shown here:
With the A connect on top of each chip in the image, K on the bottom and G the side - but I must confess that I've never done it. One of the nice things about this board is that a guy like me who slaps up a post just to fish for comments gets to chat with real experts! One of them may be able to tell us if the above is likely to work or not.
This still untested schematic, you will notice, has evolved quite a bit from the initial concept, based on the advice given in the posts thus far.
Bill
Fred, I think that you could connect things pretty much as shown here:
With the A connect on top of each chip in the image, K on the bottom and G the side - but I must confess that I've never done it. One of the nice things about this board is that a guy like me who slaps up a post just to fish for comments gets to chat with real experts! One of them may be able to tell us if the above is likely to work or not.
This still untested schematic, you will notice, has evolved quite a bit from the initial concept, based on the advice given in the posts thus far.
Bill
Thank you Bill. Only reason I ask is I have a couple of 10m45s regulators and some time to play with them. I got them several years ago and I put them away when I received a shock from a heat sink off of one and decided SS was not for me. Now a bit wiser I have them out again. They're kind of a black box to me. I imagine there is a pair of fets inside. -Fred
Here's a design I found that uses the 10m45 and shows high and low impedance outputs. It's near the bottom of the page.
link
I read "k" on the regulator as cathode so it seemed to take the output there would be similar to a cathode follower. It worked on my breadboard project O.K. and worked well as a driver.
I like getting advice here too 🙂. Thanks -Fred
link
I read "k" on the regulator as cathode so it seemed to take the output there would be similar to a cathode follower. It worked on my breadboard project O.K. and worked well as a driver.
I like getting advice here too 🙂. Thanks -Fred
Thanks for the link!
Sam Ochi of IXYS Corp. commented in an article that, in these units:
1) the transistors are high voltage MOSFETS
2) the internal voltage reference is 3V
3) current is controlled by the voltage drop in the current set resistor, which (by feedback) will always be held to equal the internal voltage reference of 3V
Sam Ochi of IXYS Corp. commented in an article that, in these units:
1) the transistors are high voltage MOSFETS
2) the internal voltage reference is 3V
3) current is controlled by the voltage drop in the current set resistor, which (by feedback) will always be held to equal the internal voltage reference of 3V
Which may just be another way of saying that the Vgs is 3V.
You know, I've been looking at what we were calling a mu connection - will it work on these depletion mode chips, or do you need a more elaborate electronics, as shown on Gary Pimm's website?
You know, I've been looking at what we were calling a mu connection - will it work on these depletion mode chips, or do you need a more elaborate electronics, as shown on Gary Pimm's website?
Using the source terminal of the upper MOSFET as the MU output works fine with the Ixys and Supertex parts.
You may need to have a 47 to 100 ohm resistor in series with the MU output if driving a capacitive load. I've had oscillation issues when the MU output was used in a preamp to drive interconnect cables. A small resistor in series with the MU output takes care of the oscillation. Pretty much the same as you find in almost any opamp circuit with a capacitive load.
You may need to have a 47 to 100 ohm resistor in series with the MU output if driving a capacitive load. I've had oscillation issues when the MU output was used in a preamp to drive interconnect cables. A small resistor in series with the MU output takes care of the oscillation. Pretty much the same as you find in almost any opamp circuit with a capacitive load.
Great! Did you mean connecting to the source of the lower MOSFET, just above the sense resistor?
Nice information to have, as the lower impedance mu output should greatly simplify driving the next stage!
I ran across this post while trying to wrap my mind around CCS operation:
http://www.waltjung.org/PDFs/Sources_101_Letter_Revisit_0409.pdf
Mr. Jung connected the gate of the upper MOSFET to a voltage divider across the lower transistor's Vds to boost the upper transistor's bias, and found a small but measureable performance gain. Good idea?
Again, thanks for sharing your knowledge.
Nice information to have, as the lower impedance mu output should greatly simplify driving the next stage!
I ran across this post while trying to wrap my mind around CCS operation:
http://www.waltjung.org/PDFs/Sources_101_Letter_Revisit_0409.pdf
Mr. Jung connected the gate of the upper MOSFET to a voltage divider across the lower transistor's Vds to boost the upper transistor's bias, and found a small but measureable performance gain. Good idea?
Again, thanks for sharing your knowledge.
Yep, It's the source of the lower MOSFET that is the MU output. That's what I get for not looking at the schematics...
Increasing the voltage across the lower MOSFET does improve circuit performance.
I've found that having a higher bias voltage available for the lower MOSFET also increases performance as you can use larger resistor values to set the output current. The higher value current set resistor provides better feedback in the CCS circuit and drifts less with temperature changes. The higher value current set resistor also reduces the noise gain of the CCS/load circuit.
Increasing the voltage across the lower MOSFET does improve circuit performance.
I've found that having a higher bias voltage available for the lower MOSFET also increases performance as you can use larger resistor values to set the output current. The higher value current set resistor provides better feedback in the CCS circuit and drifts less with temperature changes. The higher value current set resistor also reduces the noise gain of the CCS/load circuit.
- Status
- Not open for further replies.