In normal operation the diode will be reverse biased (switched off) and the depletion layer between the P and N type semiconductors forms the dielectric of a capacitor (just like the varactor diodes used for tuning your television). Varactor diodes work by changing the thickness of the depletion layer in inverse proportion to the diode voltage, so their capacitance is inversely proportional to voltage. Power diodes need bigger bits of silicon so their depletion layer has a larger area, increasing capacitance. You don't really want to add shunt capacitance that changes significantly with applied voltage, so a small-signal diode such as 1N4148 seems the best bet.
I agree that shunt capacitance is a bad thing, but probably the ratio of change of capacitance with voltage is worse. This makes me cautious about advocating the smallest signal types, whose ratio I tested some years ago and found a big enough change to make them useful as cheap varactors.
If you must use a clamp, how about trying one of the many small signal thermionic diodes Eg EB91 (2.5pF at any voltage)?
Of course all the arguments hinge on the source impedance of the driver stage, which in a good design should squash a few pf's here and there.
If you must use a clamp, how about trying one of the many small signal thermionic diodes Eg EB91 (2.5pF at any voltage)?
Of course all the arguments hinge on the source impedance of the driver stage, which in a good design should squash a few pf's here and there.
Thanks for the replies.
The preceding stages can handle a few pfs, but there is going to be a 4-5Vp GK voltage swing across one set of diodes.
dhaen, how much capacitance change were you seeing?
neons are the perfect solution. are there any in current production that would be suitable for this?
The preceding stages can handle a few pfs, but there is going to be a 4-5Vp GK voltage swing across one set of diodes.
dhaen, how much capacitance change were you seeing?
neons are the perfect solution. are there any in current production that would be suitable for this?
dhaen said:
It looks as though it was for both. The 555 oscilloscope had neons between grid and cathode of the cathode followers driving the horizontal deflection plates, and also between the deflection plates as deflection indicators.
Runeight: I can't imagine that modern CRTs are immune to flashover, so protection must still be necessary. However, very few oscilloscopes now use CRTs, so those particular neons might not be in production. Telephones use neons etc for transient protection, and they will definitely still be in production.
runeight said:
Don't panic! Now I've looked back, we're talking really low pf's here. Something in the range of:
IN4148: 5p to 1p for 10v change or so,
1n4001: 25p to 10p for 10v change or so.
You can see the 1n4148 is the best varactor, but probably the best clamp too, as the capacitance is so small.
1N4148's are 'ok', no doubt, but neons approach the theoretical ideal capacitance wise, plus they give visual feedback
. There's companies that still make them - Newark Electronics stocks them - I buy the type with the lead wires.
You can get small neons rated for under a milliamp to at least several milliamps continuous, but they'll clamp significantly higher currents for shorter periods of time - it's a matter of heat dissipation with them mostly.
. There's companies that still make them - Newark Electronics stocks them - I buy the type with the lead wires.You can get small neons rated for under a milliamp to at least several milliamps continuous, but they'll clamp significantly higher currents for shorter periods of time - it's a matter of heat dissipation with them mostly.
Thanks John! Panic was beginning to set in.
thoriated, which kind do you buy. i've searched several mail order places and most of the neons are high voltage (65-230V). I haven't found any that are not in some kind of lamp assembly that lite at less than 10V.
Much thanks to everyone for the help. At worst, I'll use the diodes and probably not even notice their presence in the amp except when they save a tube or two.
thoriated, which kind do you buy. i've searched several mail order places and most of the neons are high voltage (65-230V). I haven't found any that are not in some kind of lamp assembly that lite at less than 10V.
Much thanks to everyone for the help. At worst, I'll use the diodes and probably not even notice their presence in the amp except when they save a tube or two.
Hi, runeight -
I believe that 60 volts or so is the standard breakdown voltage for most available neon lamps, presumably so that they can be easily operated off of 117 VAC with a simple series resistor. This is low enough to protect against tube grid to cathode breakdown and the neon lamps have the advantage of working for both polarities, but if a clamp voltage of much less than that is needed for other reasons, a diode would be more suitable.
I believe that 60 volts or so is the standard breakdown voltage for most available neon lamps, presumably so that they can be easily operated off of 117 VAC with a simple series resistor. This is low enough to protect against tube grid to cathode breakdown and the neon lamps have the advantage of working for both polarities, but if a clamp voltage of much less than that is needed for other reasons, a diode would be more suitable.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.