octavia mk 319 dumb question 1 and smart question 2

That is very, very cool. I might actually measure the current operating conditions to get an idea of what's going on (and locate some 5840 triode curves) to see if there is any room for improving things. Looks like you've done a great job though. The grid bias resistor value may be one to look at as I can't remember how the 5840's miller capacitance compares with the VF14's. Dale posted some great info. regarding the U47's use of a low value grid resistor and how it manages to work well with a relatively low value.

Oh, here you go:

http://www.groupdiy.com/index.php?topic=10980.msg129657#msg129657

Thanks for the link! I was (and still am) looking for good info on choosing the right size grid/gate to ground resistor value. As I said, I basically just copied values from the U47 schematic. There is another 500M resistor on the original MK-319 board, so that would be another inexpensive option to try (but then again the ELA M251 uses only 33M!). Since the capsule is floating with 60V on one side and the tube grid on the other side, I figured that the 60M version would be closer to 0V than with a 500M resistor. Although the polarization voltage could be raised to make up for that.

My bias points are actually quite similar to what's written on the U47 schematic, despite the different tube and the very different heating/cathode resistor arrangement.

If I can find the time tonight, I'll make comparative frequency plots of the (modded) FET mic and the "tubified" MK-319. But the tube version doesen't sound horrendously different, just a little more present and direct. I'm curious as to how that shows in the frequency domain. We'll have to keep in mind, though, that there is some variation in sound from capsule to capsule, and the two 319s I have aren't even close in serial numbers.

What I find interesting is that previously I liked the modded FET model better than the stock FET. Ever since I modded the stock FET mic to this tube circuit, I prefer the now "tubified" mic over the modded FET model. So it does seem to be a step up. On the other hand, this is a pretty fresh piece of work, so this preference might be a case of "initial pride". Sometimes you see things differently later on...

Some tests with an audio test set could be quite cool if you have one. There are many things which you could test, especially at different frequencies. I'm guessing though, that the main cause of distortion at higher levels will be due to the transformer, at least in the case of the valve version. Worth checking though! The usual THD tests at nominal levels might be quite cool too, say driving the mic output at -20dBu or so....

When I get around to studying the U47 more, I'll take a Lindos test set home from work and a 'scope and see what I can establish from using different valves too.

I think Zebra50 may have posted the primary inductance values for the MK319 transformer around somewhere which may be of interest to you. If they are missing, I could re-measure if you don't have a meter with a good L range.

Roddy

Exellent stuff chaps wish I understood all you are saying Im still learning tho .....Its making me ich to try the 319 capsule with Elam circuit & T14  as I dont understand the theory I have to try stuff ....I totally understand the pride thing Rossi Iv kinda talked my ears into hearing what I was aiming for in the past...Im try to be more honest with my self these days thats why i these latest 319 questions in this old thread, Im over the moon with my Diy Ela m251 but didnt expect the Octava to sound as good as it does against it....As I said I have to admit I dont have classic expensive mics to compare to but I think the 319 sounds good I get reminded of that old story about the king whos has no cloths on & no one has the bottle to tell him....it seems some poeple say Octavas are crap because they are cheap & not a neuman or whatever.

Anyway In rant on ....the 319 has made me wonder or doubt my violet capsule a little Im tryin to be honest with myself but Im just not sure which mic I like best at the moment maybe I could make some short recording clips couple a line from pop songs with no eq & so on .......Anyway Rossi you now have two flavours that you like so thats great.....Cheers.

Unfortunately I don't have any convenient way of measuring distortion in a mic. What I can do at home is comparative frequency plots. I also have a measurement mic with an accurately measured frequency plot, so I can judge the absolute frequency response as well. For distortion I usually use a very loud tambourine in close proximity. That doesn't cover low frequency distortion, of course, but otherwise it gives a good impression of how much SPL the mic can handle.

I found Zebra50's Oktava tranny explorations:
http://www.groupdiy.com/index.php?topic=25887.msg308409#msg308409

It's pretty much what I remember: ratio between 7.5 and 8:1
Inductance on the secondary about 1 H, on the primary >40H (Zebra's meter maxed out?)
IIRC, inductance on the primary should be inductance on the secondary multiplied by ratio^2 so it ought to somewhere between 56 and 64 H.

Should do for a tube mic, I think. I don't have the impression the tube version is lacking in bass response. But we'll see, once I get to doing some frequency plots.

Gary, I have an assload of mics (no U47, though) so this is not primarily about making a cheapo mic for myself. It is more like an exploration and a challenge for myself and the sake of "microphonehood". Also I've always been a fan of Henry David Thoreau, so there's also an element of trying to with as little as possible and not wasting stuff (i.e. re-using parts in the mic). I certainly wish I had this mic when I didn't have as many mics as I have now. That said I have a feeling it sounds better than some of the more expensive mics I own. I'm pretty sure my tubified MK-319 would sound better than my Gefell UM92 on many singers. It certainly is lower noise. And yes, the MK-319 sounds pretty decent stock, and it's pretty easy to improve it somewhat with a few simple mods. I wrote an article about modding the MK-219/319 for Sound & Recording, Germany, last year. I might do a follow-up with a tube version, if I keep liking the result.

Some frequency plots. Here we go; the first one is the best I can do in terms of absolute frequency response. These are measurements against my Behringer measurement mic Meaning the ECM8000s frequency response represents the 0db-line. Level differences are evened out. The tube version is about 6 dB louder than the FET version. BTW. the ECM8000 is pretty flat, when you use it 90 degrees off axis, i.e. aimed at the ceiling (which is what I did).

Next is a difference plot of Tube version vs. the FET version. As you see there is a slicht bass loss, but not very much: about 3 dB down at 40 Hz. I'm not 100% sure this is because of the circuit, but it probably is.

Remember those are two different MK-319s with very different serial numbers. Surprisingly the frequency plots are almost congruent. The "tubified" MK-319 has a 1 dB rise in the upper Frequencies (which, to my ears, is more apparent than the slight bass loss).

To illustrate the dimension of the LF loss, I also measured the low cut on the FET version (my tube version doesn't have a low cut or pad). Note that the FET one, too, is a modded MK-319; the low cut frequency on my FET version is lower than on the stock mic. On many sources you won't even notice this low cut because there is a slight resonance above the low cut, which in a way makes up for what is cut off in the lower regions. That's a feature, not a bug 

Anyway. The FET low cut plot is just there to show you how small the bass loss actually is in the tube version. Right now I'm not sure if this slight bass loss is due to a lack in primary inductance, tube bias or size of the cathode bypass cap (22u). As I said, it could also be the different capsule.

Rossi,
Can you share a little bit on Oktava FET mod?
thanks

Can you please explain thoroughly how you did that? i would LOVE to do that, it is brilliant. Any help you can offer would be amazing.