ELA M 251 (non export) build with a 5718

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

OneRoomStudio

Well-known member
GDIY Supporter
Joined
Jul 8, 2004
Messages
590
Location
Minneapolis, MN, USA
Hi,

Thought I would share my experience building up a non-export version of the 251 with poctop's 251 boards.

I used a 5718, which sounds great and I highly recommend it. If anyone else wants to do the same, I ended up using a 2k74 cathode resistor, and that seems to do the trick. I also chose to under-heat the tube as in the original (~5.9V).

251_front.jpg

251_rear.jpg


I'm still having some trouble fitting the body tube around the relays, so I think I'm going to have to file the corners down.

I also recorded a little test so you can all hear what it sounds like (we don't do that enough around here).
https://soundcloud.com/peter-bregman/ela-m-251-test/s-yKwJX
 
try to make sure the pcb is as straight as possible in the body tube , this will greatly help fitting the mic sleeve over ,
you can fiddle a little wit bending the rail or the screw stamped parts,  this is indeed a tight fit :)

thanks for doing that,
Best,
DAn,
 
i had great results as well, with 5718,  in my Neumann M367
now i have installed a new AC701k, but it worked a long time with a 5718
the sound was a bit different but closer than anything else i've tried
that tube has nearly the same internal construction
i just noticed that it was a bit difficult to pick a quiet one, underheating was needed also...
cheers

 
Just curious as to why the under-heating, and what does this mean for the charts which assume a heater voltage of x.  Does under-heating make the charts no longer valid, and if yes then how do you plot?
 
There are a number of reasons to under-heat a tube in a microphone. Under-heating decreases the grid current, which in turn increases the input impedance and presents less of a load on the capsule. It can also increase the life of the tube (which is nice with a sub-mini since it has to be soldered in place). Under-heating can also lower the current noise of the gain stage which is beneficial in a microphone.

On the other hand, not all tubes react well to under-heating. In the tubes that can be under-heated, the linearity of the tube can be altered. So to answer your second question, yes, the chart would look slightly different. For a cathode-biased tube though, we're only in the ballpark anyway (we're stuck with standard resistor values with a 1% or 5% tolerance, plus the charts aren't exactly precise tools). In general, the same plate and cathode resistors should work fine as long at you're within 10% of the original heater voltage.
 
Awesome!! I've been thinking about doing just this. I've got a couple of GE 5star 5718s (dunno if they're quiet) and a couple of cinemag 2840s lying around, and the poctop 251 board was the first thing that came to mind.

You mentioned using a transformer with a more appropriate ratio. Are you thinking lower ratio like a 5:1?

I've managed to convince my wife to build a mic along with me so we'll end up with a stereo pair of 251s. Yummy!!



 
Lowfreq - I'm jealous, my wife thinks the tinkering I do with circuits is roughly equivalent to playing with legos - and she has no interest in joining me  :p

I just threw in an RCA 5718 I had laying around, and it seems to be plenty quiet. Hopefully the GE's will be the same.

As for the transformer ratio - the 5718 has a low plate resistance and coupled with a 100K plate resistor, the output impedance is only about 4K. I knew that would let me get away with using a BV-8 style transformer (6.5:1) without loading down the tube.

While I still might experiment with that at some point, after more testing I'm getting plenty of level with the CM-2480 (10.5:1), and it seems to have plenty of low-end too.  It's a nice sounding transformer, and I think it's better than a lot of the T14/1 style transformers out there, so I may stick with it.

Now I just have to save up for one of Tim's CK12's or get one of tskguy's HK12's  ;D
 
OneRoomStudios said:
As for the transformer ratio - the 5718 has a low plate resistance and coupled with a 100K plate resistor, the output impedance is only about 4K. I knew that would let me get away with using a BV-8 style transformer (6.5:1) without loading down the tube.

How did you arrive at this 4K figure? Also, FWIW, underheating tends to result in increased output impedance
 
Melodeath00 said:
How did you arrive at this 4K figure? Also, FWIW, underheating tends to result in increased output impedance

As far as I know (aside from measuring it) there are two standard equations when it comes to calculating output impedance of a triode stage. In this case, they end up pretty close:

1) Zout = Rp/mu

In this case, Rp=100K and the mu=27, so

Zout = 3703

and

2) Zout = Rp || ra

In this case, Rp still equals 100K, and ra equals 4650, so

Zout = 4443


The average of the two is 4073, so 4K.


As for under-heating increasing the output impedance, that's true, although I doubt it would be much of a change. Let's say that a 10% change in heater voltage has a directly proportional effect on ra. That would result in a Zout of 4886...not much different.
 
OneRoomStudios said:
Melodeath00 said:
How did you arrive at this 4K figure? Also, FWIW, underheating tends to result in increased output impedance

As far as I know (aside from measuring it)

Sorry for the stupid question, but how could one measure output impendance, with ohmmeter ? I have a few tube and I would like to build some test circuit to measure that. To find then a suitable output transformer ...

THanx.
 
lampas said:
Sorry for the stupid question, but how could one measure output impendance, with ohmmeter ? I have a few tube and I would like to build some test circuit to measure that. To find then a suitable output transformer ...

Unfortunately, you can't measure impedance with an ohmmeter. Impedance is an AC characteristic, whereas ohmmeters only measure the DC characteristic of resistance. Here's a good description I found with a quick Google. Suffice it to say measuring impedance isn't the simplest thing in the world - that's why we use math and theory.

http://www.learnabout-electronics.org/ac_theory/impedance73.php
 
That's right. A tiny T14/1.
The Siemens SM204's I posted were the only C12's I've ever seen personally with Henry Radio transformers. They are extremely rare.
 
micaddict said:
I'm not Tim, but Hiller and Henry Radio trannies would be in (early) C12s, not in Ela Ms.
Indeed, the ELAM only ever had the T14/1. That's why it would be so nice if somebody would recreate that transformer!
 
What did you use for your grid to ground resistor? Did you use 8M like the ac701 schem or did you go for the 30M like the 6072 schem?

I've breadboarded a simple cardioid only version so far to see how the tubes are, and I've used a 50M grid to ground, as that's what I had lying around.  It's not working enough as a low cut enough to compensate for my Tim Cambell capsule, which has oodles of low frequencies.

Although, you did say that you're using a Chinese ck12 so maybe you don't have so much low end to try to tame.
 
Hi all,

I am putting together a ELM 251 at the moment and like where this thread is going and talking about.
I have a spare AC701 tube which I was going to save for using in my other m269c so I can have a stereo pair, but if I've read this right the AC701 can be used in this elam 251?
If so I was going to adapt the AC701  to a 9 pin connector from an old broken 12ax7 so it would slot into the 9 pin connector I have in there.
Would both the AC701 and the 12AY7 be swappable with the same circuit config?
Also I am using 33Meg resistor would this still be ok?
Have also used 5000pf rather than 4700pf
8megs are 8.2megs
My 500k resistors are 0.25w so might need to change these
Also I'm electing to use 12v relays and then a 30R resistor for H+

Regards

Spence.
 
Lowfreq - I used a 30M, and I haven't noticed any strange low frequency issues. I would be surprised if the grid resistor was the issue in your case. The first thing I would do if I had too much low end would be to decrease C1. The C12 had a 0.5mF cap on the output, so a 3u3 might be overkill in your case.

Spencerleehorton - Unfortunately, you won't be able to swap out a 6072/12AY7 and an AC071 in the same mic. These tubes have to be biased differently. In the original AC701-based 251, the plate resistor was 200K and the cathode resistor was 3k6. For an Export (6072) version of the 251, you'd want a 100K plate resistor and a 1K8 cathode resistor.
 
Back
Top