They are not far from 220uF. Make sure its stable on the scope, bcs ZLH is very low impedance. Don't use C6, its wrongly added in that drawing.
I don't have a scope alas Salas. I guess that being the case I should wait and order a pair of 220uFs then? Any brands/types recommended which would be suitable in terms of impedence?
I've left C6 free. Is C1 ok to populate? I have an Arcotronics 0.1uF Polypropylene I could use here.
Cheers.
I would leave C1 out.
Or, add 1r in series to create a damped R+C. That 1R will probably need some or much adjusting to optimise it.
Don't waste a low ESR polypropylene in a R+C snubber on the PSU side.
Or, add 1r in series to create a damped R+C. That 1R will probably need some or much adjusting to optimise it.
Don't waste a low ESR polypropylene in a R+C snubber on the PSU side.
I don't have a scope alas Salas. I guess that being the case I should wait and order a pair of 220uFs then? Any brands/types recommended which would be suitable in terms of impedence?
I've left C6 free. Is C1 ok to populate? I have an Arcotronics 0.1uF Polypropylene I could use here.
Cheers.
C1 is the main filter electrolytic on top section drawing, maybe you meant C2 its bypass? I don't recommend random bypassing especially without instruments, avoid. Electrochemical caps alone will do fine. Its just in many drawings 0.1uF get sprinkled around ideally. But things aren't ideal, can resonate. C1 can be a Nichicon KG.
Use output cap 220uF Nichicon, Elna, Panasonic, not LowZ series any you may get easily. 105C are welcome if near the sinks. C4 Vref filter cap can be Silmic II, Nichicon Muse or Fine Gold.
On a different subject, has anyone tried to use(one section) of these regs for the Pass JFet BOZ? Since it requires less than 20ma, would you really need the power Mosfets? It is such a simple circuit, it really shows up the character of the p/s reg used. So far I keep going back to a 18v battery supply, everything else "gets in the way" of the sound.
C1 is the main filter electrolytic on top section drawing, maybe you meant C2 its bypass? I don't recommend random bypassing especially without instruments, avoid. Electrochemical caps alone will do fine. Its just in many drawings 0.1uF get sprinkled around ideally. But things aren't ideal, can resonate. C1 can be a Nichicon KG.
Use output cap 220uF Nichicon, Elna, Panasonic, not LowZ series any you may get easily. 105C are welcome if near the sinks. C4 Vref filter cap can be Silmic II, Nichicon Muse or Fine Gold.
Thanks for all that Salas - much appreciated! And yes sorry I meant C2 (which I shall remove). I'll order the other caps now.
Many thanks 🙂
HI Salas, any guide on how far the reg (Reflektor) is allowed to be from it's load? I'd prefer to keep the ac (pre rectifiers) short and run up to 30cm DC lines (force/sense). Is that ok?
Salas, on R5 (value of 82R in 5V version), how critical is value? Could 75R work?
On the snubber (1R, 0.47u), what is design guide? Could 0.4R or 2.3R work with 0.22uF?
ON the 150R's (lead to IRF9610 and gate of IRF9610) how critical is value? Could 100-200R work?
On the snubber (1R, 0.47u), what is design guide? Could 0.4R or 2.3R work with 0.22uF?
ON the 150R's (lead to IRF9610 and gate of IRF9610) how critical is value? Could 100-200R work?
Depends on Idss but 82R should have about he same effect as 75R on dropping it. So use it.
Don't change the R on the Zobel, its the damping that can change the phase margin a lot. With the cap is the frequency, you may wander a bit when still stable.
Yes, 100-200R can work but if you have to deviate better go towards 100R.
Don't change the R on the Zobel, its the damping that can change the phase margin a lot. With the cap is the frequency, you may wander a bit when still stable.
Yes, 100-200R can work but if you have to deviate better go towards 100R.
Thx Salas, Idss of 7.5mA on the K170's
on the Zobel R, could that be 0.8 as 2x 0.4R?
And the 560R on the current mirror. Would 470R suffice?
on the Zobel R, could that be 0.8 as 2x 0.4R?
And the 560R on the current mirror. Would 470R suffice?
Last edited:
I'll get the required components tomorrow, LOL.
What determines the range of the reg? I have built the 5V version, using 1K pot in series with a green LED. I have 18V input and yet my output range is just between 5V and 6V...
What determines the range of the reg? I have built the 5V version, using 1K pot in series with a green LED. I have 18V input and yet my output range is just between 5V and 6V...
Thx Salas, saw that. Still not sure why the 1k pot would only give me a range from 5 to 6V. Even the formula would suggest a broader range.
The setting resistor should not be allowed to fall bellow 1/3 of its opposite 1K resistor, the one between the Mosfet's gate stopper and ground. Thus 330R run by about 4Vgs/1K=4mA mirrored, gives 1.32V. Add 2V from the LED and two 0.7Vbe from the right branch mirror BJTs. That gives you the near 5V base. I would expect 4.7-7.4V with 1K VR. Tolerances or some problem may have squeezed that to 5-6V? We should know the practical Vgs, LedVf and Vbe in a build to see if it computes.
Hello,
If placing V1.2R LV sensing part near load circuit, big in size C1 (4.7uf) becomes an issue.
Can it value be greatly lowered, to 10n, and for low output impedance R10 also lowered, from 27k to 2k, and R9 adjusted? See attachment for 5v output. It simulates ok, near DC impedance only about 1.5-2 times bigger than with big C1. Does it has negative effects?
Also, what limiting use of 2sk170 as q8 and q3? (For output 5-15v, input 12-23 v)
If placing V1.2R LV sensing part near load circuit, big in size C1 (4.7uf) becomes an issue.
Can it value be greatly lowered, to 10n, and for low output impedance R10 also lowered, from 27k to 2k, and R9 adjusted? See attachment for 5v output. It simulates ok, near DC impedance only about 1.5-2 times bigger than with big C1. Does it has negative effects?
Also, what limiting use of 2sk170 as q8 and q3? (For output 5-15v, input 12-23 v)
Attachments
On which schematic the designations you talk refer to? If you will run a noise simulation you will see the problem with 10n. Its not only about impedance. Use a compact electrolytic 16V with very low ESR & HF impedance 1000uF if you plan for digital. That can filter deeply and have good impedance on low VrefR also. You can find a multitude of computer motherboard capacitors offered cheaply with those characteristics. For J3 in the schematic you show here K170BL can do because in low voltage its high idss will not overheat Q18. Although I would prefer you cut its idss down to 2mA with a resistor for Q18 to work in a less noisy manner. If the other position you refer to is some common gate top half of a cascode CCS to another K170 , forget it, you need double the Vgs and up, for top half. So to have some Vds for bottom. Try a J310 or BF245C.
Although I would prefer you cut its idss down to 2mA with a resistor for Q18 to work in a less noisy manner.
Any degeneration resistor will increase noise not vice versa. Moreover, jfets are less noisier at higher currents
If he will use 100 Ohm between g&s on an average BL to go ~2-3mA, the contribution is going to be 0.18uV for 20kHz BW.The original calls for a low noise VHF 2-3mA Idss Jfet for Q18's load anyway. No resistors, low capacitance semi. Q18 is a bjt. Current noise also.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Power Supplies
- The simplistic Salas low voltage shunt regulator