Apex 460 capacitor change?

[LDT2]

Phil,

Apex actually has the wrong power supply schematic with the 460 (which has been mentioned I think at Prodigy pro). The "correct" P.S. schematic is with the 450 mic here , which is a bit blurry.

I actually opened both power supplies and wrote down the info from the actual parts.

Lucio

 

[PhilGood]

Ha!!

Nice detective work!! It hadn't even dawned on me when looking at the "460" power supply that the pattern control isn't even there! D'oh!

I spent so much time looking at just the mic schematic.

Looking at both of these I'm wondering why the zeners are even there! The voltage could be adjusted with the resistors and caps alone. That's how Neumann and AKG did it in the past.

 

[mshilarious]

Looking at both of these I'm wondering why the zeners are even there! The voltage could be adjusted with the resistors and caps alone. That's how Neumann and AKG did it in the past.

If you left that rail unregulated, it would be subject to fluctuations in AC line voltage. Since there is very little current required on the +200V rail, if C1/R1 in the mic schematic is not providing enough filtration, perhaps the easiest thing to do would be to increase both values. Alternatively, you can route the zener reference voltage into a transistor with a high impedance input (filtered as in C1, but with a much larger R1), and take the voltage off the emitter (again filtered). It might be a trick to find an appropriate transistor . . .

BTW it doesn't work to stack low voltage zeners as a method to avoid the noise; zener noise is a function of voltage (so it also isn't true that only 6.8V zeners are quiet), so if you stack them, so will you get the same noise signal as if you used a single higher voltage zener.

Finally, zeners at the same rating vary by brand, batch, and individual diode. It's not always possible to say select this voltage, or this part number, and take the next zener on the tape. I recommend testing the individual diodes IN THE CIRCUIT because you need the zener to be conducting the same current as the actual circuit when you test it for noise. You could drop an extra 22uF/400V at pin 7 of the PSU and route that to an AC voltmeter, o-scope, whatever you got (I just use my converters!) to test & select for noise.

 

[LDT2]

Phil,

Yeh, modding these mics is quite an adventure.

So, what about C9 and C10 having such drastic differences in size? The schematic says 10,000uf, but the actual supply has 1000uf.

Also, what would the difference in size do with C7? Just wondering if it makes a big difference in the sound of the mic.

Thanks,

Lucio

 

[Phil Townsend]

Those darn Zeners....

I removed my zeners...but I put a switch on the heaters and B+.
Turn on the heaters first then apply the B+... Small cap across the B+ switch
Not idiot proof.
Yes that does sound better.
A relay with a timer would work but I didnt have one and did have a couple of switches.
Phil
Santa Fe

 

[PhilGood]

Phil,

Yeh, modding these mics is quite an adventure.

So, what about C9 and C10 having such drastic differences in size? The schematic says 10,000uf, but the actual supply has 1000uf.

Also, what would the difference in size do with C7? Just wondering if it makes a big difference in the sound of the mic.

Thanks,

Lucio

Well, 10,000 would be much better than 1000. I once built a power supply according to a schematic but found that there was a considerable buzz coming into the mic. After disussing things with Marik he suggested I either increase the size of the caps for filament side or add another. Sure enough it did the trick! I would think a larger size would be worth doing if they'll fit.

Phil T or mshilarious could probably give you more input on C7, or Marik if he's lurking nearby.

 

[mshilarious]

I wouldn't expect a larger C7 would do much . . . compared to R1/C1 (mic schematic), which is 10K/1uF, R6/C7 is 150K/1uF, which in theory filters noise an extra 23dB. Keep in mind there is a noise floor from the self-noise of the capsule, and reducing the noise is the rest of the circuit will have diminishing returns. Also, the tube itself will spit out some noise, they aren't the quietest beasts in the forest.

Oh hey! In my confusion over the conflicting schematics yesterday, I didn't notice that the zener noise isn't filtered to the second plate! That could be bad . . . again, probably I wouldn't remove the zeners, but a little something would be nice. Maybe it doesn't matter as much in the cathode follower stage, I'm not too familiar with that. But try another 6.8K/22uF in the PSU after the zeners (that would filter to both plates and the capsule, although the capsule looks OK), and a 1uF close to the plate in the mic, if there's room. If there is not room in the PSU, you could move the zeners before C5/R6, bump up their value accordingly, and maybe go to higher value caps.

Finally, is there a heatsink on the 7806? There really ought to be; if the AC line voltage gets high, that puppy can get hot.

 

[LDT2]

PhilGood,

There's plenty of room for the 10,000uf caps, although I don't have any hum or anything with the 1000uf caps. I think I'll try plugging the 460 into the Nady PSU and see if it sounds different/better. Thanks.


mshilarious,

7806 does have a heatsink in both PSU's.
Would there be much difference with C7 being a electrolytic vs mylar cap? The Apex has an electro and the Nady has a mylar.
Man, I wish I understood this stuff better. I apologize for what I'm sure is a stupid question, but, when you say,

If there is not room in the PSU, you could move the zeners before C5/R6, bump up their value accordingly, and maybe go to higher value caps.

Are you talking about C5/R6 in the PSU or in the mic?

Bump their value to what?

Higher value caps for which caps?

I understand if this is too involved, but it would be great for those of us just getting into this. Half the information on the schematics is either missing (like voltage on some caps or wattage on some resistors) or unreadable because it's so blurry.

Thanks, it's much appreciated.

Lucio

 

[mshilarious]

Would there be much difference with C7 being a electrolytic vs mylar cap? The Apex has an electro and the Nady has a mylar.

My rule is to use poly film caps (mylar is metallized poly film, also good, usually smaller) wherever they fit. You can get 1uF in a fairly small poly film box cap, that is what I would use.

Are you talking about C5/R6 in the PSU or in the mic?

The PSU. The series of 20K/22uF caps, then 6.8K/22uF, are smoothing caps to filter the rectified AC off the bridge to DC. Each stage drops the voltage a little, too. Finally, the pair of zener diodes regulates the voltage to a fixed figure; however, zeners have self-noise that needs to be filtered or it will bleed through to the audio.

Now, the capsule and the first stage plate of the tube have their own filters to help get rid of this noise. The second stage plate does not, but that tube stage is configured as cathode follower, and I don't know from personal experience whether or not the zener's self-noise would hurt the audio there. But it doesn't hurt to filter out to the noise, so why not?

So, if you move the zeners to before C5/R6, then C5 will filter the zener noise . But you'd also reduce plate/capsule voltage a little, because R6 would drop the voltage a little from the zener voltage. To compensate, you could use slightly larger zeners. What value, I don't know, but you should be able to figure that out by metering. It could be a small enough change not to matter.

A word about zener removal and audio effect: there are a few effects there, the least of which may be less noise. The plate and capsule voltages will increase, which will alter the tube parameters, and possibly change the capsule's sensitivity and frequency response. So one should be careful about attributing various effects to zener diodes rather than their underlying function in the circuit. Again, if there are AC line voltage fluctuations, the tube and capsule would be subject to them, so I would not contemplate removal unless the PSU is plugged into regulated AC voltage.

Higher value caps for which caps?

The 22uFs in the 200V rail in the PSU.

(like voltage on some caps or wattage on some resistors) or unreadable because it's so blurry.

The cap voltage should be according to the power rail; measure the voltage at the transformer, and make sure the cap ratings in that rail are higher than that. You can usually tell wattage ratings simply by looking at the size of a resistor; or you can calculate it.

For example, in the 200V rail, figure it starts at 240V, so short-circuit current across that rail (with 73.6K supply resistance) is 3mA, which gives a little less than 1W. Probably they used 1W resistors for that rail, just to be conservative; they may have used less. The zeners should be rated 1/2W each, although they might have used larger ones.

The filament rail, I'd expect R13 to be a big fat resistor, since 12AT7 draws 300mA, that's a 9V drop across a 30 ohm resistor, for 2.7W, so R13 should be 5W or more. I have a hard time understanding why the transformer selected would have voltage high enough to need that kind of drop. Usually you only want to feed a regulator 1 or 2V higher than the output voltage, otherwise you just waste power and generate heat. Maybe they have a big bag of them

 

[LDT2]

mshilarious,

Wow, thanks!!! That cleared up alot!!!

But when you refer to poly film caps, do you mean polyester or polypropylene?

Also, I get that the voltage rating should stay above the rail voltage, but, for instance, in the 460 mic schematics, the cathode bypass caps have one at 63v and the other at 25v. In the ELAM M251 it's 6v. In the C12 and C24 schematics there are no voltage ratings except for the tranny coupling cap. And forget about resistor wattage except for the M251 and U47 schematics.

I guess I'm just wondering if it'll do damage to change these parameters. The 460 has a couple resistors rated at 1/2w and another at 2w. Others unrated.

I've experimented with some different values that reflect the original mic that I'm trying to emulate and it makes quite a difference in the sound. For instance, in my Nady TCM1050 I changed the 1000pf/630v (.001uf) to a .01uf/250 (which is what is used in the U47) and it made a huge difference in the fullness of the sound. I also changed the 1KM resistor to a 100M (also used in the U47) and again, it thickened the lower mids. But I don't want to damage the capsules, etc. Is this safe to do?

Thanks again for your help.

Lucio

 

[Phil Townsend]

DC for heater caps or battery

You want pure dc at the heater.
Well... maybe...I use ac heaters for all my playback amps cause it sounds better to me. But that wont work in this app. Can you imagine AC running up inside the power cord right next to the sig. output...YIKES!
I use Black Gate caps before and after the voltage regulator.
If you will measure the AC that rides on the DC at the heater pins on the tube you will get a baseline. No AC is best. Size the cap so that you get no AC (Bigger is better here).
I love to use battery power for heats in this app. Sealed Lead Acid (panasonic) Put the battery in the Power supply and size them so they will last all day...Then recharge at night.
Nicads work well too. But its good to discharge them down to 1 volt per cell to avoid the memory issues.
If you do use SLA batteries put some caps across the the plus and neg wires...Batteries do make some noise as they do their chemical dance...The BGs will filter 99% of that noise. 470 mikes or better.

Rebuilding the whole PS for this Apex has really proved to be worth the time and cost. I'm shocked how good this mike is...The PS wont fit in the plastic case that the mic came in but thats no problemo.


The Duncan Amp Power Supply simulator is right on.
You guys have got to download that app and learn how to use it.

Phil
Santa Fe

 

[Phil Townsend]

Duncan Amp Links

Read the info on Duncan Amp PS design at :
http://basaudio.net/diymag.htm
and download the power supply sim at:
http://www.duncanamps.com/software.html

Both free!
Phil T
Santa Fe

 

[mshilarious]

mshilarious,

Wow, thanks!!! That cleared up alot!!!

But when you refer to poly film caps, do you mean polyester or polypropylene?

Generally I used metallized polyester, the Panasonic ECQE series.

Also, I get that the voltage rating should stay above the rail voltage, but, for instance, in the 460 mic schematics, the cathode bypass caps have one at 63v and the other at 25v. In the ELAM M251 it's 6v.

63V seems excessive, 25V is kinda the stock value for 12A_7 cathodes. But some of that is just the nature of the type of cap; the 0.1uF cap is a bypass cap to improve the high frequency response of C6, and poly caps often don't come in ratings less than 63V or 50V.

I guess I'm just wondering if it'll do damage to change these parameters. The 460 has a couple resistors rated at 1/2w and another at 2w. Others unrated.

Well as long as you're not underrated, it will make little difference. It's a good idea not to massively overrate caps, especially electros. I recall that the higher voltage rating at a given value will increase ESR, which is bad for fidelity, but I could be making that up . . .

I've experimented with some different values that reflect the original mic that I'm trying to emulate and it makes quite a difference in the sound. For instance, in my Nady TCM1050 I changed the 1000pf/630v (.001uf) to a .01uf/250 (which is what is used in the U47) and it made a huge difference in the fullness of the sound.

What was the original, ceramic? I think either value is OK, probably the quality of the cap made the difference.

I also changed the 1KM resistor to a 100M (also used in the U47) and again, it thickened the lower mids.

Probably the smaller resistor will shed some highs. Also there might be some impact with respect to the voltage supply, but I'm really not a true condenser expert, I'm more of an electret guy But I wouldn't expect any damage from those changes.

 

[LDT2]

mshilarious,

Wow again!!!! Oh, the 1000pf/630v cap was a polystyrene. I used a WIMA polypropylene.

Thanks, I really appreciate it.

Lucio

 

[LDT2]

But in the end I don't know that I'll even use it that much more than say, an AT 4047. We'll see. It's still too damn bright.

Even w/ Peluso CEK347 (if that's what I even have, who knows?) a Peluso 47-style Xformer (who knows what this is? It's probably just an el cheapo Chinese tranny. Looks surprisingly close to the original. I'm beginning to wonder. ) Auricap for C8, 56V zeners, GE 6072 tube. I upgraded all resistors, and other caps, even replaced the el cheapo lytics in the p.s. w/ Nichicons, removed the cathode follower. Next step is to put a black gate back in for C6 which I don't have in there right now. I'll give it the full run-through, but in the end I don't know if it'll really be worth all the hassle.

If you really want to get that "U47" vibe with the 1050 I've discovered a couple things that will work really well.

First replace C4 with a .01uf/250v cap (stock is 1000pf or .001uf). This will really fill out the bottom of the mic.

Second change R10 (1KM) to a 100M/1w planar. This will bring out the mids more.

These are the same values as the U47 and it really works well.

Also, I've found that, for some reason, lowering the zeners in the 1050 makes the mic thinner sounding. My PSU came with one 1N4764 (100v) and one 1N4762 (82v). I just replaced the 82v with a 1N4758 (56v) to bring it down to 156v and it works pretty well. But for some reason, the higher the voltage, the fuller the mic sounds. Just the opposite of the Apex, which got fuller sounding when dropping the zeners. The only problem is that the gain drops with the voltage, too.

Anyway, try at least those 2 changes, you'll be surprised. Very noticeable.

Lucio

 

[LDT2]

I forgot to mention changing R2 from 100k/2w to 100k 1w. This also takes some of the edge off. The U47 actually used 100k/.25w, but I haven't tried that.

Lucio

 

[crazydoc]

I forgot to mention changing R2 from 100k/2w to 100k 1w. This also takes some of the edge off. The U47 actually used 100k/.25w, but I haven't tried that.

Lucio

This has gone far enough - I just have to say that changing resistor wattage is the Emporor's new clothes, no doubt about it in my mind.

 

[PhilGood]

This has gone far enough - I just have to say that changing resistor wattage is the Emporor's new clothes, no doubt about it in my mind.

This is called splitting hairs and I agree.

 

[LDT2]

Yeh, but it makes an audible difference. I went up to 3w and the mic got brighter. And I'm not talking something only dogs can hear.

Lucio

 

[Phil Townsend]

Parallel resistors

I almost always use parallel resistors at the plate and cathode. To me it sounds better. NOS Allen Bradley carbons 2 watters. Nice sound. The films have an edge kind of sound.
Phil
Santa Fe