EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: stefan_o on June 11, 2017, 10:27:39 pm
-
Hey,
does someone have an good idea how to achieve high-voltage current limiting? The voltage can vary up to 160 Volts, but I need a constant current limiting at 0.2mA (=200µA). Initial precision does not need to be very high. Output voltage should not (less than 500mV) change below the threshold.
Any good idea how to implement that? It is to limit the current through a avalanche photo diode.
Best regards
Stefan
-
Simplest unqualified solution: depletion mode MOSFET e.g. LND150, DN2540, etc.
A BJT (PNP) or PMOS, current source or mirror, would also be an option, but may be less desirable for bias reasons.
If high frequency response, capacitance, is an issue, find the smallest transistor you can tolerate (i.e., Rds(on) as high as possible -- presumably, Rds(on) <= (0.5V) / (0.0002A)), with Idss > 0.2mA if depletion mode. Also consider a cascode configuration, and a damped LC filter (in series with the collector/drain) to increase the impedance at high frequencies.
Tim
-
A depletion mode MOSFET seems to be a great idea, didn't thought of that. The LND150 also seems to be the most suitable one I've found, RDS(on) of 1kOhms is the highest I could find (using parametric search of Mouser/Digikey). For the resistor I probably need a trimmer to tune to the individual MOSFET, correct?
I don't think high frequency response will be an issue.
Thank you very much,
Stefan
-
If accuracy "isn't much of an issue", you might just as well leave it as-is -- otherwise, yes, a trimmer can do better. It will still be temperature sensitive, so keep that in mind.
Should be easy enough to SPICE up the right resistor range and tempco required to keep it very stable, if you like.
Tim
-
You could use one or two of those LR8 high voltage linear regulators from microchip : http://www.farnell.com/datasheets/2138212.pdf (http://www.farnell.com/datasheets/2138212.pdf)
(https://www.eevblog.com/forum/projects/high-voltage-(160v)-low-current-limiting-(0-2ma)/?action=dlattach;attach=323297;image)
ps. though note that datasheet says it minimum output current typically 0.3mA max 0.5mA including resistor and load current... but you could probably add a resistor in parallel with the load to eat 0.1-0.2mA
-
Problem with the LR8: The operating voltage may be different, so I cannot just add a resistor, current would be different. But the LR8 looks like a nice part for voltage regulation, much better suited than the TL783 I had in mind.
If accuracy "isn't much of an issue", you might just as well leave it as-is -- otherwise, yes, a trimmer can do better. It will still be temperature sensitive, so keep that in mind.
Should be easy enough to SPICE up the right resistor range and tempco required to keep it very stable, if you like.
Tim
Can you explain how? By reading the datasheet of the LND150 (http://ww1.microchip.com/downloads/en/DeviceDoc/LND150%20C041114.pdf) I assumed that I need about 4kOhm as source resistor and according to the graph on page 4 (top right) it is quite temperature stable. How would you optimize it?
Best regards,
Stefan
-
Because transconductance is so low at 200 uA I'm not sure the corner is going to be sharp enough to meet that 0.5V transition between constant voltage and constant current.
PS. why do you care about that though? Do you want to experiment with both constant voltage and constant current biasing?
-
If accuracy "isn't much of an issue", you might just as well leave it as-is -- otherwise, yes, a trimmer can do better. It will still be temperature sensitive, so keep that in mind.
Should be easy enough to SPICE up the right resistor range and tempco required to keep it very stable, if you like.
Tim
Can you explain how? By reading the datasheet of the LND150 (http://ww1.microchip.com/downloads/en/DeviceDoc/LND150%20C041114.pdf) I assumed that I need about 4kOhm as source resistor and according to the graph on page 4 (top right) it is quite temperature stable. How would you optimize it?
For 200uA you need 4.3K with that depletion mosfet. The lower the current the more temp stable it is. Keep in mind that it is not a current limiter but a constant current source, ie it will try to keep that current flowing and for that it will raise the output voltage as much as it can.
-
With such a low output current, can you not use a cheap 300v Rail-Rail output opamp & design a simple voltage/current regulated supply output?
-
Ah, LND150 does indeed have quite good temp stability at low currents. Nice.
Tim
-
.. the TL783 I had in mind.
I don't think TL783 (http://www.ti.com/lit/ds/symlink/tl783.pdf) will work, it has a high min. output current spec. of 15mA
I'd use discretes ala 2SC2690 CRT drivers or audio power amp VAS transistors for good linearity.
-
For 200uA you need 4.3K with that depletion mosfet. The lower the current the more temp stable it is. Keep in mind that it is not a current limiter but a constant current source, ie it will try to keep that current flowing and for that it will raise the output voltage as much as it can.
? There is no switching, nothing, isn't it just Ohm's law? If there is 160 V for example at the input and a current of 0.05mA output will be 160V-(RDS(on)+RSource)*I, in this case 160V-(1000Ohm+4300Ohm)*0.00005A=159.735V?
Because transconductance is so low at 200 uA I'm not sure the corner is going to be sharp enough to meet that 0.5V transition between constant voltage and constant current.
PS. why do you care about that though? Do you want to experiment with both constant voltage and constant current biasing?
Ah, you mean RDS(on) depends on current? And with low current RDS(on) will be a lot higher? In fact, it would be prefect if a higher resistance would compensate the lower current for the voltage drop to be constant. If the voltage drop for lower currents is higher, that is bad.
The easiest I think is to try it out, I already ordered a few LND150, all I need is a few resistors and see how it works.With such a low output current, can you not use a cheap 300v Rail-Rail output opamp & design a simple voltage/current regulated supply output?
Cheap 300v Opamp? I looked at Digi-Key, all are above 100 €/US$. A LND150 is less than 1 €...
.. the TL783 I had in mind.
I don't think TL783 (http://www.ti.com/lit/ds/symlink/tl783.pdf) will work, it has a high min. output current spec. of 15mA
I'd use discretes ala 2SC2690 CRT drivers or audio power amp VAS transistors for good linearity.
I had the TL783 for voltage regulation in mind. For voltage I can choose the resistor low enough for it to be stable. But the LR8 has two advantages: Less power losses (resistor divider with higher values) and I can regulate down to almost 0. But this has nothing to do with the current limiting, that's the voltage one step before.
-
Exactly, and when the current goes up it will try to restrict that increase by taking up voltage and lowering the output voltage
-
I had the TL783 for voltage regulation in mind.
Ok, it's just that 15mA load at 160V is a big fat resistor making 2.4W of heat and also heating up the TL783, to supply 0.2mA
This is why I say it's not going to work well here.
-
With such a low output current, can you not use a cheap 300v Rail-Rail output opamp & design a simple voltage/current regulated supply output?
Cheap 300v Opamp? I looked at Digi-Key, all are above 100 €/US$. A LND150 is less than 1 €...
I thought the 5$ LTC6090 was +/-140v, but it's actually 140v total.
Digikey does have this 15$ 350v device, but, it is way out of budget and at 60ma, overkill on power output:
https://www.digikey.com/product-detail/en/apex-microtechnology/PA340CC/598-1917-ND/2700701 (https://www.digikey.com/product-detail/en/apex-microtechnology/PA340CC/598-1917-ND/2700701)