As others have pointed out, turn down the AC test voltage amplitude and you can measure the diode junction capacitance at lower DC bias voltages. For instance, if you use an AC test voltage of 0.1 volt RMS, you can probably measure the junction capacitance all the way down to zero DC bias, as there will not be enough peak voltage to cause any significant conduction current. Be sure to pay attention to the D-value. When the D-value, dissipation factor, (reciprocal of Q, the quality factor) exceeds one, the instrument is saying that the device under test is behaving more like a resistor than a capacitor. In other words, the device under test is dissipating more than it is storing. Good capacitors have a very low dissipation factor. Note that for the diode, the dissipation factor drops very rapidly as the DC reverse bias increases until the diode is no longer experiencing any forward conduction.
I think the reason you need to have at least 1 volt of reverse bias on the diode is so when the machine applies its AC signal the AC signal does not ever forward bias the diode. I can imagine when the diode gets forward biased during the capacitance measurement it would mess up the bridge etc...
Nice piece of equipment ... maybe you could do part 3 and show us the capacitance vs bias voltage difference between a 1N4001 and a 1N4007. Always something interesting and thought-provoking to see on the channel!
Nice, I designed my own impedance analyzer some years ago for a few $100s, it doesn't do DC bias on it's own, but you can use an external power supply to measure both capacitors and inductors with DC bias (voltage and current). It covers from low ~1Hz to 20MHz and performs very close to one of the high end Hioki units.
You might be able to measure capacitances at below 1 V bias, if you can reduce the test signal AMPLITUDE. If I understood, you have had 1 V amplitude, which means -0.5 V to +0.5 V peaks. Even more, if the instrument signal is stated as RMS. The common assumption of some 0.6 or 0.7 V forward drop of silicon junction is not valid with currents down to micro- or nano-amperes. There is a formula for the voltage versus current, as well as for temperature. You can see an exponential term there. In any case, you cannot measure your capacitances below 1 V bias, because the forward current changes your bias! Another note regards manufacturer standards on their diode capacitance measurements -- their graphs mention that the measurement was done at 1 MHz.
True stories from real life: “I have a couple of them around in the garage somewhere but I could not find them.” Sometimes it is easier to get some fresh ones, the old ones will surely come up just when the new ones arrive.
BTW: If you put a 1N400X diode in series with an inductor and connect it to your signal generator, you can get chaos to happen. You need an inductor that will get the resonance to be less than about 5MHz for it to work. You set the generator to add a small reverse bias and then you tune to find resonance. somewhere near resonance the following starts to happen. The resonant circuit rings up until the voltage gets big enough to forward bias the diode. When the 1N400X is forward biased, a bunch of carriers are injected into the junction area. Those carriers massively change the capacitance and also kill the Q The amplitude crashes to near zero The diode has to go through its recovery time. Then the ring up happens again. The recovery time depends on the amplitude of the current just before it. When the recovery time ends, vs the phase of the generator controls the current you will get before it crashes
Why wouldn't you use Zener diodes instead. I found their voltage VS capacitance much more repeatable among the same version zener voltage and wattage between multiple manufacturers. Especially for 36v through 75v zeners which usually can replace authentic varicaps. The 1N400x series just differ too much between manufacturers.
Hans Summers did a useful investigation about different types of diodes with dc bias in his artcle "Rectifier diodes as varicaps (varactors) ". He found the most "variable" common diodes, and draw graphs "Capacitance / Bias voltage". Also he measured "an ordinary LED " as a varactor. So, it will be interesting to see how act different LEDs from different ages.
Neat toy you got there! When I was working on modifying CB radios to work on 10m, we would use a diode that a friend acquired that was made by Bell Telephone Labs back in the late 1970's. We replaced the stock varactor in a Cobra 148GTL and were able to get the "Voice Lock" to shift the frequency +7KHz and -15KHZ, which was more than the spacing between the channels that the radio used. STOCK was only able to do +3.5 and -6KHz.
Hi! Great video. Is there any possibility that you share the values with us? In a google doc or something. If I were to guess, I would say that junction cap is proportional to 1 / V_r
If my memory is not playing tricks on me, the capacitance goes down with ~1/sqrt(U_reverse), and as others said: if you reduce the ac amplitude, you could measure below 1 V reverse voltage, cheers.
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
As others have pointed out, turn down the AC test voltage amplitude and you can measure the diode junction capacitance at lower DC bias voltages. For instance, if you use an AC test voltage of 0.1 volt RMS, you can probably measure the junction capacitance all the way down to zero DC bias, as there will not be enough peak voltage to cause any significant conduction current. Be sure to pay attention to the D-value. When the D-value, dissipation factor, (reciprocal of Q, the quality factor) exceeds one, the instrument is saying that the device under test is behaving more like a resistor than a capacitor. In other words, the device under test is dissipating more than it is storing. Good capacitors have a very low dissipation factor. Note that for the diode, the dissipation factor drops very rapidly as the DC reverse bias increases until the diode is no longer experiencing any forward conduction.
I think the reason you need to have at least 1 volt of reverse bias on the diode is so when the machine applies its AC signal the AC signal does not ever forward bias the diode. I can imagine when the diode gets forward biased during the capacitance measurement it would mess up the bridge etc...
Nice piece of equipment ... maybe you could do part 3 and show us the capacitance vs bias voltage difference between a 1N4001 and a 1N4007. Always something interesting and thought-provoking to see on the channel!
Nice, I designed my own impedance analyzer some years ago for a few $100s, it doesn't do DC bias on it's own, but you can use an external power supply to measure both capacitors and inductors with DC bias (voltage and current). It covers from low ~1Hz to 20MHz and performs very close to one of the high end Hioki units.
You might be able to measure capacitances at below 1 V bias, if you can reduce the test signal AMPLITUDE. If I understood, you have had 1 V amplitude, which means -0.5 V to +0.5 V peaks. Even more, if the instrument signal is stated as RMS. The common assumption of some 0.6 or 0.7 V forward drop of silicon junction is not valid with currents down to micro- or nano-amperes. There is a formula for the voltage versus current, as well as for temperature. You can see an exponential term there. In any case, you cannot measure your capacitances below 1 V bias, because the forward current changes your bias! Another note regards manufacturer standards on their diode capacitance measurements -- their graphs mention that the measurement was done at 1 MHz.
"And I'm not gonna do that"
D'aaw! That would have been pretty fun :D
What about capacitance of ceramic capacitors vs. voltage?
redrok
AD0TJ
True stories from real life: “I have a couple of them around in the garage somewhere but I could not find them.” Sometimes it is easier to get some fresh ones, the old ones will surely come up just when the new ones arrive.
EVERYTHING always appears when you stop looking for them, lol.
BTW: If you put a 1N400X diode in series with an inductor and connect it to your signal generator, you can get chaos to happen. You need an inductor that will get the resonance to be less than about 5MHz for it to work. You set the generator to add a small reverse bias and then you tune to find resonance. somewhere near resonance the following starts to happen.
The resonant circuit rings up until the voltage gets big enough to forward bias the diode.
When the 1N400X is forward biased, a bunch of carriers are injected into the junction area.
Those carriers massively change the capacitance and also kill the Q
The amplitude crashes to near zero
The diode has to go through its recovery time.
Then the ring up happens again.
The recovery time depends on the amplitude of the current just before it.
When the recovery time ends, vs the phase of the generator controls the current you will get before it crashes
Why wouldn't you use Zener diodes instead. I found their voltage VS capacitance much more repeatable among the same version zener voltage and wattage between multiple manufacturers. Especially for 36v through 75v zeners which usually can replace authentic varicaps. The 1N400x series just differ too much between manufacturers.
Hans Summers did a useful investigation about different types of diodes with dc bias in his artcle "Rectifier diodes as varicaps (varactors) ". He found the most "variable" common diodes, and draw graphs "Capacitance / Bias voltage". Also he measured "an ordinary LED " as a varactor. So, it will be interesting to see how act different LEDs from different ages.
There was an internal bias option card (opt 001 or 002) for those and the 16023b bias controller add on, I want to find the board for my 4274a.
watch tomorrow
Neat toy you got there! When I was working on modifying CB radios to work on 10m, we would use a diode that a friend acquired that was made by Bell Telephone Labs back in the late 1970's.
We replaced the stock varactor in a Cobra 148GTL and were able to get the "Voice Lock" to shift the frequency +7KHz and -15KHZ, which was more than the spacing between the channels that the radio used. STOCK was only able to do +3.5 and -6KHz.
Funny, that's where I remember messing with varactors too.
Very nice!
Hi! Great video.
Is there any possibility that you share the values with us? In a google doc or something.
If I were to guess, I would say that junction cap is proportional to 1 / V_r
If my memory is not playing tricks on me, the capacitance goes down with
~1/sqrt(U_reverse), and as others said: if you reduce the ac amplitude, you could measure below 1 V reverse voltage, cheers.
SRC is Series Resonant Converter.
Would a 1N4007 provide more capacitance than a 1N4001?
I heard that once, but have not confirmed. maybe larger range not value?
wow