Hi,
I am using the LS4117 JFET in diode mode as clamps, because they have ultra-low leakage current. They work like a charm for this application - a well-kept secret ;-)
Looking up the max allowed Ids, I found the data sheet specs 50mA as max gate current, but there's no max Ids spec, wtf?
Thinking I found a typo in the LS data sheet, I checked the Calogic and Interfet data sheets for the 4117, and lo and behold - they also spec 50mA max gate current and NO max Ids current!
Is this a typo that has been propagating since day one or am I missing something?
Jan
I am using the LS4117 JFET in diode mode as clamps, because they have ultra-low leakage current. They work like a charm for this application - a well-kept secret ;-)
Looking up the max allowed Ids, I found the data sheet specs 50mA as max gate current, but there's no max Ids spec, wtf?
Thinking I found a typo in the LS data sheet, I checked the Calogic and Interfet data sheets for the 4117, and lo and behold - they also spec 50mA max gate current and NO max Ids current!
Is this a typo that has been propagating since day one or am I missing something?
Jan
50mA must be the bond wire limit. Idss < 0.09mA so the only way to get 50mA is via the gate diode junction
Maybe my use of Ids is confusing, I mean the drain current as compared to BJT collector current. The is always max Ic beyond which the device gets damaged. What is the comparable max Id of 4117 beyond which it gets damaged, the abs max Id?
Jan
Jan
How do you plan to get Id much beyond Idss? A small -ve voltage on the gate before the gate junction diode actually conducts is only going to raise Id to about 1mA
Maybe my use of Ids is confusing, I mean the drain current as compared to BJT collector current. The is always max Ic beyond which the device gets damaged. What is the comparable max Id of 4117 beyond which it gets damaged, the abs max Id?
Jan
There's no way to force a lot of channel current without forward biasing the gate, the current flows in the drain and/or the source in any case, it has no where else to go. The 50 mA specified max gate current has to flow in the channel anyway (I assume you tie the drain and source together and use the gate as anode). These are typically 20fA leakage as diodes, BTW the BAV199's might be cheaper and back to back diodes in one SOT23.
Last edited:
There's no way to force a lot of channel current without forward biasing the gate, the current flows in the drain and/or the source in any case, it has no where else to go. The 50 mA specified max gate current has to flow in the channel anyway (I assume you tie the drain and source together and use the gate as anode). These are typically 20fA leakage as diodes, BTW the BAV199's might be cheaper and back to back diodes in one SOT23.
Scott thanks I think I have to plug a leak in my understanding o these devices. 😯
I'll check out the BAV199.
Jan
I'll second the idea of a BAV99L for use as a diode clamp. You get a series pair in an SOT-23, they're readily available, very inexpensive, and, at least in my application, add extremely little distortion when this is tested for in simulation. Not sure how they differ from the BAV199, but both seem to have similar characteristics, and I know the BAV99 is extremely clean in my simulations.
I'll second the idea of a BAV99L for use as a diode clamp. You get a series pair in an SOT-23, they're readily available, very inexpensive, and, at least in my application, add extremely little distortion when this is tested for in simulation. Not sure how they differ from the BAV199, but both seem to have similar characteristics, and I know the BAV99 is extremely clean in my simulations.
The BAV199 is the 20fA typical leakage version. Linear Systems has a replacement for what we used 40yr ago in modules.
Those (and other) diodes are pretty good but the reverse current depends strongly on temp and reverse voltage. For a typical 15V analog supply it is best to go for 100V reverse blocking diodes.
Still, max guaranteed values can easily be 1000 x typical values, so there is some uncertainty there.
I am working on an auto ranger for soundcard inputs and 1N4148 clamps worsened the THD by at least 6dB, while the 1N4117 made no difference. The 4117A is even better. Only drawback is the relatively high price.
But I am going to order some of those low leakage diodes and try them out.
Jan
Still, max guaranteed values can easily be 1000 x typical values, so there is some uncertainty there.
I am working on an auto ranger for soundcard inputs and 1N4148 clamps worsened the THD by at least 6dB, while the 1N4117 made no difference. The 4117A is even better. Only drawback is the relatively high price.
But I am going to order some of those low leakage diodes and try them out.
Jan
And the jfet gate will behave exactly the sameThose (and other) diodes are pretty good but the reverse current depends strongly on temp and reverse voltage.....
Jan
BTW I was wrong: reverse current does NOT depend on reverse voltage - it is also called reverse saturation current.
BTW2 I have seen a reference somewhere to a 'bootstrapped clamp' as a means to use a diode clamp that has extremely low leakage until it reaches the clamp threshold. But I can't find any info on it. Anybody has a pointer for me?
Jan
BTW2 I have seen a reference somewhere to a 'bootstrapped clamp' as a means to use a diode clamp that has extremely low leakage until it reaches the clamp threshold. But I can't find any info on it. Anybody has a pointer for me?
Jan
- Status
- Not open for further replies.
- Home
- Design & Build
- Parts
- 2N/LS4117 max current?