EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: willeye on May 16, 2020, 02:46:58 am
-
Hi All, Long time viewer, first time contributor. I am planning to build the circuit attached to power and use with the Grundig GCM3 Microphone. My question is how can i generate the 100v from the 9v without using 33 battery's. I am new to electronics and i have built a few kits and could do with some help with this. I read about a boost converter using a 555 but i am not sure if that wold be suitable.
The microphone was a gift and as i am stuck inside (imunocompromised) i would like a nice project. :)[attach=1]
-
Welcome to the forum :)
A common circuit for generating condenser mic capsule HV bias is a one-transistor Hartley oscillator with voltage multiplier afterwards. Early Schoeps CMC3 or CMC5 used it. Output is usually around 60VDC (7.5V power) but with 9V you can get more. Very little current is required, uA. The disadvantage compared to using a stack of coin batteries is the oscillator's output is not regulated unless the input is regulated - so it will drop as the 9V battery ages. It also has some RF noise but this is filtered and seems to work fine.
These are circuits used by many mic manufacturers. MXL mic pic shows it as the second board on the far right, pic from http://recordinghacks.com/microphones (http://recordinghacks.com/microphones) (you should send pics to them, they have no Grundig). The other mic schematics shows the bias voltage circuit.
I was working on a small SMT Hartley oscillator pc board for mics but Mouser sold out of inductors so I have to find substitutes.
Another method is a voltage multiplier using CMOS CD4049.
Most important is how much room do you have?
-
If space is not a problem you could just join a whole lot of small 9 volt batteries in series. They clip together quite well. Get it going very quickly, and if it’s not a success you can always use the batteries for something else. Probably lower noise than an inverter too.
https://m.youtube.com/watch?v=ousUTivJoaM (https://m.youtube.com/watch?v=ousUTivJoaM)
-
Have no limit on size but am hoping to put it all in a guitar pedal size enclosure. but tbh it could be any size. I know stacking batteries will work but there must be an way without.
-
I laughed at "33 x CR2032"; that sure is safe.
Any oscillator circuit with a high voltage winding on the resonating inductor can work and results in low noise. A transformer based inverter is feasible as well. But since the inductor or transformer is not common, a micropower boost switching regulator can directly drive a capacitance multiplier to easy get to 100 volts.
-
A dc-dc converter, something like this https://uk.rs-online.com/web/p/isolated-dc-dc-converters/7331102/ (https://uk.rs-online.com/web/p/isolated-dc-dc-converters/7331102/) If you go down this route you may want to go for a medically approved converter for safety reasons, low leakage current. EDIT: Sorry had my safety hat on, was thinking about leakage current.
-
Does it stringently need 100V, or would it work with the usual 48V phantom power over XLR as well? (Then it were powered from the mixer/amp, and no need for batteries at all.)
-
Apparently it doesnt work with 48v and only produces nice results up at round 100v. I have seen cap/diode doublers but how can i simply produce the ac? i cant design my own circuits yet so any pointers to online resources. I will look at rs but wanted to build the whole system myself if i can. it is my isolation challenge to get this mic to work.
-
A dc-dc converter, something like this https://uk.rs-online.com/web/p/isolated-dc-dc-converters/7331102/ (https://uk.rs-online.com/web/p/isolated-dc-dc-converters/7331102/) If you go down this route you may want to go for a medically approved converter for safety reasons, low leakage current.
This looks a promising option but it is expensive and I would like to know how it works rather than just using sealed unit. but if all else fails it would be a solution :)
-
If it's for the polarising voltage for a condenser microphone then the 33 lithium coin cells actually sounds like a relatively compact and electrically very quiet solution. The only load will be insulation resistance (as indicated by the 500k series resistor in the schematic), so they should last for pretty much shelf-life.
Just stack them in suitable insulated tube (or maybe two half sized stacks side by side) with a single compression spring and a suitably large series resistor. They don't need to be expensive cells for this application, just look for cheap ones on Amazon. CR2025s would save you some stack height too (2.5mm vs 3.2mm), 82.5mm total.
The microphone might well have used a 90V dry cell radio HT battery originally.
Edit: Qty 10 for £1.89 on ebay (with 5% multi-buy discount).
Edit1: CR1616?
-
If it's for the polarising voltage for a condenser microphone then the 33 lithium coin cells actually sounds like a relatively compact and electrically very quiet solution. The only load will be insulation resistance (as indicated by the 500k series resistor in the schematic), so they should last for pretty much shelf-life.
Just stack them in suitable insulated tube (or maybe two half sized stacks side by side) with a single compression spring and a suitably large series resistor. They don't need to be expensive cells for this application, just look for cheap ones on Amazon. CR2025s would save you some stack height too (2.5mm vs 3.2mm), 82.5mm total.
The microphone might well have used a 90V dry cell radio HT battery originally.
I am starting to think this is getting to be my only option, What do you mean by "suitably large series resistor" isnt that what the 500k is for. I know the £1 shop sell packs of 20 but think might go little better quality
-
Yes, it is, sorry, I wrote that bit before looking at the schematic (again).
If you want really compact you could maybe try CR1616s but they might be more expensive as they are less common.
P.S. Old Zamboni would have given his eye teeth for lithium coin cells! ;D https://en.wikipedia.org/wiki/Zamboni_pile (https://en.wikipedia.org/wiki/Zamboni_pile)
-
I have seen cap/diode doublers but how can i simply produce the ac? i cant design my own circuits yet so any pointers to online resources.
Check out figure 27 on page 18 of Linear Technology application note 45 for an example of how a boost converter can drive a capacitive voltage multiplier. The virtue of this particular implementation is that it can run on a single 3 or even 1.5 volt cell:
https://www.analog.com/media/en/technical-documentation/application-notes/an45f.pdf (https://www.analog.com/media/en/technical-documentation/application-notes/an45f.pdf)
R3 is the "suitably large series resistor" mentioned by Gyro and is there for safety as it limits the output current if the high voltage is applied to something delicate, like the user.
-
If it's for the polarising voltage for a condenser microphone then the 33 lithium coin cells actually sounds like a relatively compact and electrically very quiet solution. The only load will be insulation resistance (as indicated by the 500k series resistor in the schematic), so they should last for pretty much shelf-life.
Just stack them in suitable insulated tube (or maybe two half sized stacks side by side) with a single compression spring and a suitably large series resistor. They don't need to be expensive cells for this application, just look for cheap ones on Amazon. CR2025s would save you some stack height too (2.5mm vs 3.2mm), 82.5mm total.
The microphone might well have used a 90V dry cell radio HT battery originally.
Edit: Qty 10 for £1.89 on ebay (with 5% multi-buy discount).
Edit1: CR1616?
Well i have ordered 40 cr1616 from ebay, that way i can build the circuit then i can test while I learn about doublers. So where does the 1m resistor go in this case?
-
I have seen cap/diode doublers but how can i simply produce the ac? i cant design my own circuits yet so any pointers to online resources.
Check out figure 27 on page 18 of Linear Technology application note 45 for an example of how a boost converter can drive a capacitive voltage multiplier. The virtue of this particular implementation is that it can run on a single 3 or even 1.5 volt cell:
https://www.analog.com/media/en/technical-documentation/application-notes/an45f.pdf (https://www.analog.com/media/en/technical-documentation/application-notes/an45f.pdf)
R3 is the "suitably large series resistor" mentioned by Gyro and is there for safety as it limits the output current if the high voltage is applied to something delicate, like the user.
So figure 27 cold be connected to a cap/diode doubler via the out but what is the +90v Bias mean
-
You could buy 8 A23 batteries : https://uk.farnell.com/gp-batteries/gp23a/battery-12v-38mah/dp/300408 (https://uk.farnell.com/gp-batteries/gp23a/battery-12v-38mah/dp/300408)
10 pcs for 4.5 pounds : https://www.amazon.co.uk/GP-Super-Voltage-Alkaline-Battery-Chrome-Black-Red/dp/B072FBK3P9/ (https://www.amazon.co.uk/GP-Super-Voltage-Alkaline-Battery-Chrome-Black-Red/dp/B072FBK3P9/)
Should take up little space, but kinda lousy battery life, i don't know how many hours the mic is gonna run on a set of 8.
-
Safety might be a secondary aim. But primarily, the series resistor between voltage source and condenser microphone capsule is supposed to ensure that (once the capacitor is charged) audio frequency variations of the capacity do not alter the charge, but rather lead to a changing voltage across the capacitor, which can be picked-up and amplified. So I'm surprised about the low value of only 500k. For a cut-off freqency of say 20Hz, this imples that the capsule capacitance would need to be >= 1/(500000*2*pi*20), i.e. >= ~15nF. Admitedly it's a mic with a large diaphragm, but is the capacitance really that high? Otherwise low-frequency performance suffers. I've seen other condenser mic amp schematics which rather used dozens of MOhm up to 1GOhm for the charging resistor, and similarly high values for the gate biasing resistor, too. Nevertheless it indeed seems that these old Grundig tape recorders feed the mic input via only 500kOhm, e.g. R10 in http://www.hifi-archiv.info/Grundig/TK5/tk5-16.JPG (http://www.hifi-archiv.info/Grundig/TK5/tk5-16.JPG)
So figure 27 cold be connected to a cap/diode doubler via the out but what is the +90v Bias mean
The left part of figure 27 already is the boost converter, including the doubler.
You can likely change the divider R1/R2 a bit to get 100V instead of 90V.
(The pulse generator at the right hand side of figure 27 is not part of the converter, you don't need it, it is just powered by the 90V generated by the converter)
-
I really like the solution posted by David Hess, AN15 fig. 27. An external high voltage switch could also be used with the LT1073 as shown in fig. 29 which uses the LT1072. And thanks to gf post above pointing out the 500k series resistor the maximum available current is only 200uA.
-
And thanks to gf post above pointing out the 500k series resistor the maximum available current is only 200uA.
But DC current is not an issue. Once the mic capsule capacitor is charged, there won't flow DC current any more. It is rather the converter's quescent current which drains the battery then, not its load.
EDIT: When the mic capsule were biased directly from a battery, then this battery would last forever (almost - i.e. rather limited by self-discharge and leakage currents only), while the no-load quiescent current of the LT1073 is still 135..250µA. This relativizes of course, when we consider that the amp needs some power anyway.
-
Thank you all for the help! I am learning so much. So if i understand I only need this part of the [attach=1] Figure 27. Do i need to do anything to smooth the output voltage. Would noise be transferred to the signal from mic. I will draw up a final design and order a lt1073. The batteries I have ordered can be a way of testing and a fall back. Does the inductor value have to be exact. I have some i have removed from other recycled boards. Is there a simple way to test an inductor.
Would this https://www.ebay.co.uk/itm/10pcs-LT1073CS8-Micropower-DC-DC-Converter-Ajustable/261185316852?hash=item3ccfdba7f4:g:XggAAMXQLs5Rp0sL (https://www.ebay.co.uk/itm/10pcs-LT1073CS8-Micropower-DC-DC-Converter-Ajustable/261185316852?hash=item3ccfdba7f4:g:XggAAMXQLs5Rp0sL) be correct. It aslso says in the specs it can run upto 30v so could i run the whole thing off a 9v square battery
-
Are you really sure that the mic won't work with 48V, too? (besides giving a ~6dB lower signal)
Looking around in the web, I found the following two circuits interesting.
Both pretty simple, still just a few components, directly powered over XLR from the mixer, no battery, no voltage converter.
http://www.audioimprov.com/AudioImprov/Mics/Entries/2016/3/4_Johnny_Zhivagos_Lika.html (http://www.audioimprov.com/AudioImprov/Mics/Entries/2016/3/4_Johnny_Zhivagos_Lika.html)
http://www.audioimprov.com/AudioImprov/Mics/Entries/2017/5/7_Schoctava%2C_a_simple_mic_circuit.html (http://www.audioimprov.com/AudioImprov/Mics/Entries/2017/5/7_Schoctava%2C_a_simple_mic_circuit.html)
(The second one may not be an option if one side of the capsule is undetachably connected to ground.)
-
Are you really sure that the mic won't work with 48V, too? (besides giving a ~6dB lower signal)
Looking around in the web, I found the following two circuits interesting.
Both pretty simple, still just a few components, directly powered over XLR from the mixer, no battery, no voltage converter.
http://www.audioimprov.com/AudioImprov/Mics/Entries/2016/3/4_Johnny_Zhivagos_Lika.html (http://www.audioimprov.com/AudioImprov/Mics/Entries/2016/3/4_Johnny_Zhivagos_Lika.html)
http://www.audioimprov.com/AudioImprov/Mics/Entries/2017/5/7_Schoctava%2C_a_simple_mic_circuit.html (http://www.audioimprov.com/AudioImprov/Mics/Entries/2017/5/7_Schoctava%2C_a_simple_mic_circuit.html)
(The second one may not be an option if one side of the capsule is undetachably connected to ground.)
Everything I have read has said 48v works but does not allow the mic sing to full potential. As this is a project for fun i am going to keep trying to get as good a result as i can. There are only one wire from the mic and the shield. Obviously all i can get is anecdotal info so i think if i use the original schismatics as a start i can hopefully make something nice.
-
Thank you all for the help! I am learning so much. So if i understand I only need this part of the (Attachment Link) Figure 27.
That is correct.
Do i need to do anything to smooth the output voltage. Would noise be transferred to the signal from mic.
Noise may be a problem so I would add an RC filter before the load by dividing the high value current limiting resistor into two parts and placing a capacitor between them to ground.
Does the inductor value have to be exact. I have some i have removed from other recycled boards. Is there a simple way to test an inductor.
The inductor needs to be able to support the full switching current from the LT1073 without saturating but is otherwise not critical. The LT1073 datasheet has details.
-
@willeye, here is yet another low-current voltage multplier from CMOS invertes:
http://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf (http://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf)
Btw, did you watch the EEVblog microphone video series?
https://www.youtube.com/watch?v=ihAG6cMpUlY&list=PLHoGZ7nuFTo8VmQBhzZxl-qDEIN7T5Vt8 (https://www.youtube.com/watch?v=ihAG6cMpUlY&list=PLHoGZ7nuFTo8VmQBhzZxl-qDEIN7T5Vt8)
Quite a couple of (proven) circuit ideas are sketched here.
[ It's actually a playlist of seven videos, IIRC, but the board software inserts only the first video into the message. At YouTube you can search for "EEVBlog Microphone Series" ]
As this is a project for fun i am going to keep trying to get as good a result as i can. There are only one wire from the mic and the shield. Obviously all i can get is anecdotal info so i think if i use the original schismatics as a start i can hopefully make something nice.
As "original schematic" I'd consider the tube preamp in the Grundig tape recorders, to which these microphone models are supposed to be connected. But I'm not sure that Grundig's aim was the best result one can get from this microphone capsule. My feeling is that they rather had a different design goal in mind: Unlike other high quality microphone manufacturers they wanted to renounce a (tube?) preamp inside the microphone enclosure (I guess for cost reasons), so they rather did make some compromises in favor of this goal. Note, for best results one would rather not have a long cable from an condenser microphone capsule to the preamp (but the preamp would reside directly in the mic enclosure, close to the capsule), and one would not charge/load the capsule with such a low impedance either (500k is low for a condenser microphone capsule).
EDIT:
One fundamental question is of course whether to keep the gadget itself untouched, for historic preservation reasons, or whether to modernize it.
-
@willeye, here is yet another low-current voltage multplier from CMOS invertes:
http://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf (http://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf)
Btw, did you watch the EEVblog microphone video series?
https://www.youtube.com/watch?v=ihAG6cMpUlY&list=PLHoGZ7nuFTo8VmQBhzZxl-qDEIN7T5Vt8 (https://www.youtube.com/watch?v=ihAG6cMpUlY&list=PLHoGZ7nuFTo8VmQBhzZxl-qDEIN7T5Vt8)
Quite a couple of (proven) circuit ideas are sketched here.
[ It's actually a playlist of seven videos, IIRC, but the board software inserts only the first video into the message. At YouTube you can search for "EEVBlog Microphone Series" ]
As this is a project for fun i am going to keep trying to get as good a result as i can. There are only one wire from the mic and the shield. Obviously all i can get is anecdotal info so i think if i use the original schismatics as a start i can hopefully make something nice.
As "original schematic" I'd consider the tube preamp in the Grundig tape recorders, to which these microphone models are supposed to be connected. But I'm not sure that Grundig's aim was the best result one can get from this microphone capsule. My feeling is that they rather had a different design goal in mind: Unlike other high quality microphone manufacturers they wanted to renounce a (tube?) preamp inside the microphone enclosure (I guess for cost reasons), so they rather did make some compromises in favor of this goal. Note, for best results one would rather not have a long cable from an condenser microphone capsule to the preamp (but the preamp would reside directly in the mic enclosure, close to the capsule), and one would not charge/load the capsule with such a low impedance either (500k is low for a condenser microphone capsule).
EDIT:
One fundamental question is of course whether to keep the gadget itself untouched, for historic preservation reasons, or whether to modernize it.
I have been watching those videos over and over and I am learning so much and am starting to understand the parts of the circuit. Parts have started to arrive so I will start to breadboard it.
I can see that Grundigs goals might not have been to get the best out of it but mine are, It is such a nice looking thing and there are quite a lot of them about so it feels like a good project.
I will experiment with the 500k resistor once it is breadboarded to see if I can improve the response after the info in the episodes above. Looking at some Grundig schematics it always has the 500k but it often is accross a 3m /2m resistor divider so the final resistance on the load looks like 2.5m on some but maybe i havent read them right.
I have a din socket on the way so I should be able to build this as a module that has no impact on the mic and wont require modification.
One video that helped a lot was https://www.youtube.com/watch?v=SToBPCajwc0 (https://www.youtube.com/watch?v=SToBPCajwc0) seeing it step by step and explaining every part of the circuit is amazing.
-
it always has the 500k but it often is accross a 3m/2m resistor divider so the final resistance on the load looks like 2.5m on some but maybe i havent read them right.
In the TK5 schematic I see a 3M/5M divider feeding the 500k resistor, but connected to the center tap there is also a bypass capacitor (C35=50nF), shorting AC to ground (cut-off ~6Hz). So the divider resistors don't play a role for AC at audio frequencies.
EDIT:
Where do you actually want to connect the mic? To home audio / PC sound card, or rather to a mixer with XLR microphone inputs?
-
it always has the 500k but it often is accross a 3m/2m resistor divider so the final resistance on the load looks like 2.5m on some but maybe i havent read them right.
In the TK5 schematic I see a 3M/5M divider feeding the 500k resistor, but connected to the center tap there is also a bypass capacitor (C35=50nF), shorting AC to ground (cut-off ~6Hz). So the divider resistors don't play a role for AC at audio frequencies.
EDIT:
Where do you actually want to connect the mic? To home audio / PC sound card, or rather to a mixer with XLR microphone inputs?
I am planning to use both on my mixer and on my pc via either a di box or a xlr to 3.5 cable, have both. Most of the parts have now arrived and I have put together then main circuit [attach=1] and just waiting for some 16mm ID tube to build the battery and the 400v cap. First time I have worked with jfets.