Dual Polarity Voltage Multiplier...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
If you cannot get the simulation running in Tina try LTspice. It works there. Though instead of actual logic gates, I use two voltage sources with square wave outputs of opposite polarity. Maybe that will also fix your problem?

Jan
 
Ltspice sim with logic gates worked fine for me, while tweaking my own version, for what it's worth.
 
I use two voltage sources with square wave outputs of opposite polarity

Yes, I had the simulation working with that, I wanted to provide a simulation file with actual logic gates so that someone wanting to build could try different resistor/capacitor combinations in case they wanted to minimize number of different components in the BOM, or wanted to try different clocking frequencies.
 
I've been tinkering with diverse variations of your OPIC circuits, and the more i play with them the more i get irritated by phantom power limitations. If all those tube mics have clunky PSUs that follow, why wouldn't we have dedicated PSUs for opamp based mics? This is definitely my next step. No more hustle with voltage multipliers, no limitation of how many opamps in a mic we get to use for various purposes, and the benefit of basically unlimited headroom is priceless.

I just want a mic where capsule's SPL limitation is the only limiting factor and i believe this is the way to go. Sorry for this rant, just had to share some frustration after several hours of LTspice desperation.
 
the more i get irritated by phantom power limitations

You could take Thor's approach and modify your pre-amps with lower value resistors and increase the phantom voltage to 60v so the phantom can supply more power.

The nice thing about modern power supplies is if you make an external supply it can be very compact and lightweight compared to the old tube mic supplies.
 
I've been tinkering with diverse variations of your OPIC circuits, and the more i play with them the more i get irritated by phantom power limitations. If all those tube mics have clunky PSUs that follow, why wouldn't we have dedicated PSUs for opamp based mics? This is definitely my next step. No more hustle with voltage multipliers, no limitation of how many opamps in a mic we get to use for various purposes, and the benefit of basically unlimited headroom is priceless.

I just want a mic where capsule's SPL limitation is the only limiting factor and i believe this is the way to go. Sorry for this rant, just had to share some frustration after several hours of LTspice desperation.
As I've mentioned before, I'm a great one for 'simple' inside the mic, and to move everything else (eq etc) further down the signal path.
So the simplicity of the original OPIC LDC wth a single op-amp appealed to me.
My first OPC multipattern mic uses 3 opamps, and a single polarity VM. ...Obviously that draws more current - so a lower DC supply is available for the op-amps - and each extra op-amp and associated signal path ressitors will add some extra noise as well. (Not a lot - but some! ).

The original idea of this dual VM was to enable a multi-pattern mic to be created, using just one op-amp again.

Allowing for 1mA of current to power a VM, adjusting the values of the op-amp supply feed resistors (R4 and R5 on the OPIC LDC schematic ) to be 2k2 instead of 10k should make the available DC supply to the op-amp should be aroud 36v.
Close to the upper limit for many op-amps - and certainly for the OPA164* series.
There will obviously be some attenuation of the output signal, working into this lower impedance load, but that should be less than 0.5dB, so not that significant.

That should give a maximum signal rail to rail output swing of c.36volts p-p. Probably slightly less than that, to allow for a finite impedance of the phantom power supply, but not a lot less I suspect.

I'm not sure that you'd need any more headroom than that ? ... and that's still using a standard phantom power supply.

Of course if you want to include more op-amps internally, for other functions, then those figures would no longer compute.
As I mentioned before, I'm not a fan of doing any 'non-linear' signal path modifications at mic level, so I quite like the idea of a single OPA1641 being fed from a 36v DC supply as being the simplest impedance converter, with the maximum output signal that an op-amp can provide.....
And that would still only need standard phantom power..
 
Of course if you want to include more op-amps internally, for other functions, then those figures would no longer compute.
My post was certainly not a critique towards your original design which is marvelous. It led me to further experimentation for which i just need more juice.
 
My post was certainly not a critique towards your original design which is marvelous. It led me to further experimentation for which i just need more juice.
No, I didn't read you post as being critical .... I was just intrigued to know why you needed more power.....
I think maybe I'm taking the idea of doing everything apart from the impedance converter outside the mic a bit too seriously ? :)
 
To get the double power (or more) you could also use a passive external splitter box and connected the mic output to two or more mic pre input channels (increases the load a bit though). On the other hand you would like to record both front and back diaphragm outputs anyway (so a XLR5->2x XLR3 adapter cable is needed anyway). Also the efficiency of 48V to whatever voltage op amp circuits like to use could be improved using DC/DC converters instead of zeners.
 
I have one of Ari's Flat 47's fitted into a multi-pattern OPIC. That is fitted with the original VM, and I have R5 fitted as 1k.
There are 3 op-amps in that circuit, and the circuitry draws c.7mA in total from the phantom power supply.
That provides a DC supply to the op-amps of c.16v.

R5 is fitted as 1k, providing an output polarisation voltage of 70 volts - measured at the cathode of D7 - so that the actual voltage applied across the capsule is 70v, minus the op-amp 'half rail' bias voltage - which is 8 volts.

So the capsule actually has 63v across it.

I think the 'R5' Khron is describing refers to the output series resistor in the newer 'adjustable' version, which is 1M and, as he suggests, is not a critical value.

One final point. I have specified a CD40106 as the VM project hex inverter.
It is of course possible to use alternative types of 40106 - like the HEF40106 - but the thresholds are a bit different, so the output voltages may vary a little from the values listed in this circuit. For example, you would probably need to fit R5 as 2k2 to get a 72v output from a VM fitted with a an HEF40106 intead of a CD40106.
For the newer 'variable' type, this is not a concern of course....
I'm testing R5 with a value of 10R, wich should be supplying the flat 47 capsules with around 72V, after deducting the opamo half rail bias voltage. Ari has mentioned the capsules should be able to handle 75V. Well, capsules appear to handle 72V well, no issues so far in my testings. I wonder what value could be choosed to get exactly around 70V, somewhere around 180-200 ohm?
Rogs, Khron is there a simple way to establish the position of the trimmer to match more precisely the output on both diaphragms?... I have a DMM, an average signal generator, and a not so good scope. Right now I'm doing it by ear, omni, talk on both sides while I adjust the trimmer and try to match them, but with the body tube removed I obviously get a great deal of hum too so....I have roughly matched them like this so so at the moment, but would like to make it precise... Prior to that I was using fig of 8 and try to find the best spot to get the most rejection at the null sides by ear, but again its somewhat difficult too with the tube removed and hum all over....and then when I tought I'd find the spot, I switch to Omni and one side is always little hotter than the other, so I don't trust this idea much... Its not like I can inject a sine tone, because the whole idea is to match differences between the actual capsule sides so...? Forgive me If this is obvious and thanks in advance.
On another note, I'm really loving these mics, the vintage sound of the Flat 47, and Rog's straight "wire" like circuit to deliver the sound as it is from the capsule. I know Ari has mentioned this capsule is better served with the more open 47 style grill, but I actually like the transparency of the HL77 bodies' 87 style basket I'm using on these. The only thing I've noticed, is that the Flat 47 has some hefty low end reach, and with no filtering going on in the circuit, its usually to much "boomy" for close micing in directional pattern. Almost always I'm using them with low cut on the preamp, and sounding great like that. Of course for kick, padded, that low end reach is welcomed, so I'm not saying I dislike that, just some observations on this combo. Anybody notice this too?
 
Last edited:
I'm afraid that selecting the value of R5 is the only way of easily changing the DC output level with the original OPIC voltage multiplier ... which is one of the reasons I went on to do the variable version for my OPIC43 version... ( lower part of that PDF)
Easy (and cheap) enough to replace the VM if you're using stripboard - not so easy if you're using Mic Scharf's PCBs!

Not that it's really necessary to be that precise. A difference of 5v - or possibly even up to 10v? - in the polarisation voltage - within the 65v to 75v range) - is only likely to affect the sensitivity by about 1 dB or less ...

When it comes to using the trim to calibrate the output levels, I've found that, with the head backet still in place, the level of 50Hz hum encountered by removing the body sleeve was low enough to still carry out useful measurements.

I would suggest a simple direct level test might be sufficient. Use a headphone transducer or small speaker in direct contact with one side of the capsule and apply a 1KHz signal at a resonably high level. Observe the output on a preamp or mixer level meter..
Turn the mic over and apply the same transducer and signal to the other side of the capsule..... Adjust the trim pot to match the level indications on the preamp or mixer level meter.

I the case of Ari's 'flat 47' you probaby won't need to move the trim pot far from the centre position... the two sides should already be matched within about 1 dB or so...

For a less well matched capsule you should be able to adjust for up to about 6dB of difference in output levels between the two sides.... I would think that you should be able to use this 'meter level' calibration to be reasonably accurate.

It may be possible to use the figure of 8 'side null' point to match the two sides..... If they are significantly different, you might expect the null not to be at the 90 degree point......
No actually tried that, but it might provide some useful results?
 
[...] I know Ari has mentioned this capsule is better served with the more open 47 style grill, but I actually like the transparency of the HL77 bodies' 87 style basket I'm using on these. The only thing I've noticed, is that the Flat 47 has some hefty low end reach, and with no filtering going on in the circuit, its usually to much "boomy" for close micing in directional pattern. Almost always I'm using them with low cut on the preamp, and sounding great like that. Of course for kick, padded, that low end reach is welcomed, so I'm not saying I dislike that, just some observations on this combo. Anybody notice this too?

I also appreciate the sound of the K47 Flat in a U87-style basket. It produces a slightly "fatter" tone and a more neutral high-frequency response compared to the U87 clone I built using a K87 capsule in the same body.

While using the K47 Flat in the U87 ORS flat circuit, I noticed that even a slight bump on the grill caused the mic preamp to clip. Engaging the low-cut filter on the preamp didn't resolve the issue, as the microphone signal was already too strong (requiring about +37 dB of gain to achieve -18 dBFS from a 15 cm distance).

To address this, I kept C9 in place, removed C6, and added a 4pF capacitor between the gate and drain.

For testing both the K87 and K47 Flat capsules, I used the microphone bodies available at this link:
https://www.aliexpress.com/item/100....order_list.order_list_main.40.21ef1802ugPGja

I believe these are the same or similar to the ones found here (where you get two for a lower price):
https://www.aliexpress.com/item/1005004825669344.html?spm=a2g0o.cart.0.0.5fc138da86Jnv9&mp=1

I’ve read that the HL-77 body is prone to ringing when tapped. The body I’m using doesn’t exhibit much ringing. However, the full-size HL-95, has significantly more ringing (it sounds almost like a bell), although it does have better metalwork.
https://www.aliexpress.com/item/100....order_list.order_list_main.17.21ef18020sGtMg
 
Last edited:
You could use adhesive tar inside the body : it's quit heavy, not so thick and damps vibrations very well. I think you can find it in most hardware shops.
I use it to wrap bass reflex vents in my speakers (for example) or to damps electronics chassis (enclosure)
 
I am interested in polarizing U87 ORS with higher voltage and have been looking around for suitable and small converter enough to fit in. So just +60V needed.

Variants such as from midwayfair, Fuchs and Sparklebear comes to mind (Schoeps?).
Using CMOS circuits looks easier as done in the OPIC project(s) than choosing (and placing) inductors to get it to work.

But also if there might be noise/radiation concerns to address with different oscillator solutions.

When I look at U87AI schematics, the voltage regulation uses double the amount of components than different DIY I found here.
That won't probably fit in a smaller body.

I know SMD might not be so DIY friendly but I wonder if this IC would require least physical space and be good enough?
Totally 8 components: https://www.analog.com/media/en/technical-documentation/data-sheets/LTC3290.pdf
 
Last edited:
You could use adhesive tar inside the body : it's quit heavy, not so thick and damps vibrations very well. I think you can find it in most hardware shops.
I use it to wrap bass reflex vents in my speakers (for example) or to damps electronics chassis (enclosure)
Yes, I have some adhesive tar from way back in my "audiophile" DIY days. :).

Sorbothane comes to mind too but expensive. I read long time ago that human flesh is even better to dampen vibrations than Sorbothane.

Where is Shylock when you need him?
 
The problem with building a dual polarity voltage multiplier using charge pumps is the large number of components required.
I decided to have another look at Schoep's simpler Hartley oscillator concept -- knowing that I might encounter problems with inductor placings, and reliable oscillator start up?.. :)
In the event, I decided to place the two inductors laying flat next to each other with a 0.1" spacing, and experimented with the orientation.
To my surprise it appears to make very little difference - although one orientation did perform slightly better than the other (less loading of the oscillator output from the attached voltage multipliers).

In the end, I have a mic that has fewer components and is simpler to construct than most of my other mics!
Ari's 'flat47' is a lovely capsule, and having pretty well matched outputs is easy to adapt to switched voltage pattern changing.
I also think the triple layer mesh, and the shape of the U87 headbasket help contribute to what I think is both the simplest - and the best sounding? - mic I've built so far.

I've called it 'OPIC.45'. Some notes here: https://www.jp137.com/lts/OPIC.45.pdf
 
Last edited:
I decided to place the two inductors laying flat next to each other

How have manufacturers implemented that in volume production? Same style of two inductors side by side, or a tapped inductor with dual windings on the same core? Or a transformer style of winding where the two windings are on top of one another on the same core? I thought I had seen at least one schematic which appeared to show a single inductor with tapped winding, but I don't remember which model that was, I will have to look through my schematic library to see if I can find it again.
 
How have manufacturers implemented that in volume production? Same style of two inductors side by side, or a tapped inductor with dual windings on the same core? Or a transformer style of winding where the two windings are on top of one another on the same core? I thought I had seen at least one schematic which appeared to show a single inductor with tapped winding, but I don't remember which model that was, I will have to look through my schematic library to see if I can find it again.
Some of the various inductor options are mentioned in the page HERE I linked to in my text. It seems to be the 'mutual inductance' question that can cause problems with the reliable running of a Hartley oscillator. Probably one of the reasons charge pump designs get chosen in preference by some maunfacturers? No running 'iffyness' there!....
But when it comes to a building a dual polarity version- especially as a hobby project - the extra component count of two charge pumps can be a bit of a pain......

One original Hartley layout I did try had the inductors laid end to end..... It worked OK - sometimes !
Fitting Bourns style 78F inductors - side by side - seems to work more reliably - perhaps obviously?
Whether that would still be true with other stye of inductors, I'm not sure?

For mass production I'm guessing some manufacturers have concluded that the charge pump style of multiplier is more predictable, and have gone for that approach... For this hobby project, I thought the simpler Hartley oscillator, driving two voltage doublers might be worth another look?
Certainly uses fewer components! :)
 
Back
Top