All seem to be correct, except pin 6 maybe.
Try to replace the 40106, or check for shorts or wrong value.
Try to replace the 40106, or check for shorts or wrong value.
Getting capsule polarization voltage from phantom power in a Schoeps-style circuit
- Thread starter skyy7
- Start date
Help Support GroupDIY Audio Forum:
rogs
Well-known member
I can see the logic for your response, but I think the DC supply to pin 14 should be higher - with other readings around half that supply voltage...... It really does sound as if the circuit is drawing too much current.
Last edited:
maosante
Active member
I followed rog's advice to check the IC soldering since hand soldering is a bit tricky specially for ICs. After resoldering all IC's pin this are the new readings;
Pin 2: 2.8v
Pin 4: 3.2v
Pin 6: 2.7v
Pin 8: 2.8v
Pin 10: 3.5v
Pin 12: 3.7v
Pin 14: 4.3v
D1 cathode: 12.53v
I've configure it for 60v at the output with a 3.3k resistor following Mic Scharf's R15 value.
The pulse is now gone, randomly it can happen but is not more an often sound, sometimes there’s a little flicker sound in the audio signal, can it be because the 1nf ceramic capacitor?
The C1 could be a 63v? 100v has been difficult to find.
With 12.5V VDD, the output voltage is bound to be about 72V, which exceeds the capacitor rating, so it may go into brekdown and result in occasional noise.
If you can't find a 100V, put two 22nF 50 or 63V in series.
rogs
Well-known member
I'm afraid I don't know whether the Mic Scharf version of my project uses the same component idents ?
(I'm not able to download his BOM directly for some reason).
There is no need for C1 on the original schematic (posted on THIS PAGE ) to be rated at 100v.
The oscillator runs at the supply voltage of the IC. That should be typically be around 9v to 11v depending on the value R5 ( on the original schematic ).
You've measured the IC supply as 4.3v, which is low. It does suggest something is still drawing too much current.
I'm assuming you are measuring 12v at the cathode of the zener diode ?
Couple of extra points....
• You will see that the original circuit specifies a CD40106. On occasion, people have used alternative versions of the 40106 - like the HEF40106. Functionally these are identical devices, but the Schmitt trigger thresholds are a little different, so you are likely to find the multiplier and output vottages will be a bit higher with th HEF device. The oscillator will run faster too (c.180KHz as I recall ). Not a problem normally, but something to be aware of.
• You do need to check your IC supply voltage. It seems that the IC is drawing around 2.5mA, if the dropper resistor value is 3k3.
That seems a bit high for a CD40106 ? .... You could try a lower resistor value - say 1k - and see how that affects the figures.
• If you're using the multiplier in conjunction with a standard OPIC LDC circuit, don't forget to add the value of the op-amp 'half rail' supply (typically 10 to 12v) to the selected multiplier output voltage, to get the actual required capsule polarisation voltage.
maosante
Active member
That’s right C1 is for the chip supply but the bom suggests 100v for all capacitors I guess just for not complicating things. But according to your schematics it shouldn’t reach that voltage.
By the other hand there's a resistor R4 1M in your design 10k (R14) in Scharf's. Can this value affect current?
The Mic I'm using consumes about 3.5ma for the preamp of the capsule.
rogs
Well-known member
I mention using 100v capacitors for simplification in my circuit description HERE . It's ony the latter multiplier stages that actually need that voltage rating
There is essentially no current being drawn through R4 - it's load is only the capsule, which is just a capacitor.
I have no idea why that value has been reduced to 10k in Mic Scharf's version? ...... He simply made contact and asked if he could use my circuit for his PCB version of the project.....
From your measurement of the voltage on pin 14 of the IC, it would appear that you have 7.7 v (12v - 4.3 v ) being dropped across the 3k3 resistor you have fitted for R15 (R5 on the original circuit). That suggest a current of c.2.3mA being drawn by the IC, which is more than twice as much as I would have expected.
Are you using a CD40106 or a HEF40106?
maosante
Active member
The website I’ve bought it from says it’s a CD40106 but the actual data sheet is this one. https://www.sigmaelectronica.net/manuals/1718770.pdf which is a HEF. I’ve installed sockets for R15 (R5 in your circuit) for easy testing.
The Capsule is a Dachman audio K47 and I want to polarize it to achieve a hypercardiod pattern. I don’t really know which is this voltage but according this document 60v is cardiod and above that is the hypercardiod and then it becomes figure 8 at 120v. So I guess Hypercardiod is in between at 80v.
First I’ve got to have a stable circuit, any advice for the HEF version? I’ll be changing the 1nf 100v ceramic for a 1nf 63v MLCC as a first step then.
I did ask in PCB way the reason for the 1M to 10k change. By the way thanks for your support and design.
Last edited:
rogs
Well-known member
I have built a version using HEF40106:
• With a 12v zener diode, and R15 fitted as 2k2, I measure 10.6v at pin 14 of the IC. .....
( So the IC and it's associated circuitry is drawing c. 0.64mA ...... Probably slightly less than a CD40106 in the same circuit)
• The voltages on all IC pins (except pins 14 and 7) measure at c. 5.2v ( +/- 3%).
• The output voltage at the cathode of D7 measures as 72.2v. (Slightly higher than a CD40106 in the same configuration).
That 72v provides me with 60v of capsule polarisation, once the op-amp 'half rail' bias voltage has been subtracted.
With regard to your points on variable pattern generation, I'm guessing you are describing the technique described by Abbey in post#2 of this thread ?
I've not ventured into the world of generating voltages in excess of 100v, although providing the components are appropriately rated, it should be fine.
I have seen a hobby page describing this 'multi-voltage' pattern option HERE but I've not ventured down that route myself.
For multi-pattern use I took a different approach which works well, but it doesn't include options for more 'exotic' patterns.
I did briefly try out a dual polarity version of a voltage multiplier, to see how it compared to my op-amp selection multipattern approach, but decided I'd stick with the op-amp selection version.
maosante
Active member
I have sockets on R15 (Mic Scharf board) so I can change the voltage easily by swapping resistors. Then the HEF version outputs 2.2v more than the regular CD chip. With this info I will try to match your measurements.
Thanks!
Do you realize that in order to produce the hypercardioid pattern, you need to have two different bias voltages? One positive for the front capsule and one negative for the rear capsule?
Actually, with a non-grounded backplate, you need two positive voltages. One for the backplate and one for the front diaphragm, with the rear diaphragm grounded.
Since you can't use a resistive voltage divider, you have to tap the backplate voltage from one of the diodes, probably the second or third one, via a 1 Meg resistor and with a 100nF cap to ground.
Then the bias voltage for the front diaphragm needs to be close to 100V.
There is another possibility.
You can derive both positive and negative voltages from the multiplier.The positive output is about 7X the supply voltage, that's because you have 6 stages of multiplier added to VDD.
You can add a negative multiplier chain to provide X times the VDD voltage in reverse polarity.
Tapping D12 from nodes I or J results in -12 and -24V respectively.
Actually output is not exactly X times VDD, it's a little less, as one can see.
It depends on the exact part for the hex inverter. As rogs mentioned the HEF part gives a slightly different result than the CD part.
You can derive both positive and negative voltages from the multiplier.The positive output is about 7X the supply voltage, that's because you have 6 stages of multiplier added to VDD.
You can add a negative multiplier chain to provide X times the VDD voltage in reverse polarity.
Tapping D12 from nodes I or J results in -12 and -24V respectively.
Actually output is not exactly X times VDD, it's a little less, as one can see.
It depends on the exact part for the hex inverter. As rogs mentioned the HEF part gives a slightly different result than the CD part.
Last edited:
