API 312 Thread!

[Rybow]

Ok. So if I jumper my 75101 in series for 600ohms on the primary, I will have a 1:5 ratio. A 25k resistor will give me a 1K impedance on the primary. So, is this in addition to the transformers impedance? If I left the load out, would the primary impedance be just 600 ohms? Transformers are one thing I still don't fully understand.

Is 1k the optimal impedance? It seems to me that messing around with the load resistor gives you different impedance matching values for different mics. You could throw a couple resistors on a switch and then be able to change impedance depending on the mic. Not that I want to, just trying to understand.

Thanks for all your help! I'll get to reading this thread again.

 

[bobschwenkler]

I'm not sure what the winding impedances of these transformers are, but they've got to be negligible next to the load impedances.

From what I've read 1k or 1.5k is a pretty standard primary ratio. You probably could just use a switch to get different impedances at the input. A shorting switch might be preferable here.

 

[Rybow]

Alright. So I have been doing some reading. I found on article on the great river site that says that the input impedance of a pre amp should be 5 to 6 times that of the output impedance of the microphone. So I went through my mics/mics I will soon own (all ribbons and dynamics) and came up with some figures. I will either go for broke on a 1.5K impedance on both channels, which if I read correctly is the input impedance of an actual API 312, or implement a switch to flip between 750ohm and 1.5K. I am probably over thinking it, but that's how I roll.

My math may be off, but here is how it should work.

Cinemag 75101 wired 1:5- 750ohm=18.75K load, 1.5K=37.5k load
EA2622 wired 1:7- 750ohm=36.75k load, 1.5K=73.5K load

Depending on how lazy I am, or how much it really matters, I will most likely just go for the 1.5k input impedance. One thing that is not clear to me though. In the article I read, it stated that if you under load the mic, you get some LF roll off, but it says nothing about the effects of over loading the mic if any.

Time to get those transformers mounted. ;D

 

[mikefatom]

Overloading a mic has much less noticeable effects, especially with condenser mics (buffer circuitry inside the mic).  This usually yields a flatter bass response with dynamic mics and from experience, it can make them sound "brighter" as well.  Whatever that means in my world...  :  This is all subjective...

 

[Rybow]

Overloading a mic has much less noticeable effects, especially with condenser mics (buffer circuitry inside the mic).  This usually yields a flatter bass response with dynamic mics and from experience, it can make them sound "brighter" as well.  Whatever that means in my world...  :  This is all subjective...

Thanks Mike. I am not too sure what I plan to do just yet, but I am leaning towards a switch. Seems that its better to overload them a bit then underload them.

One last question while I await the last few pieces. I have a 15VA 2 x 18V toroid already, but will 15VA be enough for 2 channels?

 

[mitsos]

RE, loading.. another option would be to stick a pot (with an R in series) in place of the load R, so you can vary the load on the transformer. JLM uses this type of variable impedance pot, I've never done it but it seems like a more complete solution than having 2 choices on a switch. If you're set on the switch, stick a pot in temporarily to test out which load R's you prefer soundwise. Otherwise it's a crap shoot.

Not counting phantom or LEDs, you have about 416mA per rail.  That's plenty, unless your opamp is abnormally powerhungry (2520s, melcors, 990s are not). Now if you like to light your equipment like a christmas tree, it might be a different story

hope that helps!

 

[bobschwenkler]

An 18V secondary might be cutting it too close with the regulation it'll be feeding. I'm not so studied up on regulators. Anyone else know better?

 

[mitsos]

I assumed he'd be running them at +/-16V or +/-18V ... Either one should be OK in most situations, assuming mains fluctuation of no more than 10% less than nominal.

As I understand it, you generally want the same AC voltage as the required DC output of the PSU.  Assuming no main fluctuations, fullwave rectification of 18VAC will give 18 x 1.414 - 0.7V diode drop =  24.752 VDC... A regulator needs to see at least 3VDC above the DC output, assuming he's running at +/- 16 or 18 VDC, that leaves 8.752 or 6.752 to be "wasted" as heat. 

The only problem I could see is if your mains power fluctuates far below what they claim. Assuming mains power goes down 10%, that would leave  (18*.9)*1.414-0.7= 22.2068 VDC pre-regulators, still enough for +/-16V but coming close to the limit (limit would be 21V) for +/-18.

That's my story and I'm sticking to it! (until someone comes in and helps me put my foot in my mouth).

 

[bobschwenkler]

One thing I still don't know, are transformer voltages spec'd as peak to peak, RMS, or some other unit of measure?

 

[mikefatom]

One thing I still don't know, are transformer voltages spec'd as peak to peak, RMS, or some other unit of measure?

Power TX voltages are spec'd in RMS.

 

[mikefatom]

I assumed he'd be running them at +/-16V or +/-18V ... Either one should be OK in most situations, assuming mains fluctuation of no more than 10% less than nominal.

As I understand it, you generally want the same AC voltage as the required DC output of the PSU.  Assuming no main fluctuations, fullwave rectification of 18VAC will give 18 x 1.414 - 0.7V diode drop =  24.752 VDC... A regulator needs to see at least 3VDC above the DC output, assuming he's running at +/- 16 or 18 VDC, that leaves 8.752 or 6.752 to be "wasted" as heat. 

The only problem I could see is if your mains power fluctuates far below what they claim. Assuming mains power goes down 10%, that would leave  (18*.9)*1.414-0.7= 22.2068 VDC pre-regulators, still enough for +/-16V but coming close to the limit (limit would be 21V) for +/-18.

That's my story and I'm sticking to it! (until someone comes in and helps me put my foot in my mouth).

The minimum voltage difference between Vin and Vout for guaranteed proper regulation of a linear regulator is called "drop out" voltage.  There are very low drop-out (LDO) voltage linear regulators out there, including LDO 317/337 variants.  If I remember correctly, a regular 317 (e.g. LM317 from national semi) will have a drop out voltage of about 2V.  I find that even a dual 15VAC power transformer works well for powering up a +/-16v circuit such as a 312 pre.  It helps greatly with reducing heat generated from the regs as well.

 

[Rybow]

I will be running the 312's off of +/- 16V coming out of Mike's 4 rail PSU, so it looks like I'll have a little bit of extra heat to contend with if the mains behave. Interesting stuff so far. You guys are just full of great info!

I'll take a look into the impedance pot and maybe testing out some different options. I have basically looked at all the mics I own plus ones that I want, and picked 2 values that should basically cover them. An impedance pot would be the most flexible, however, I can see myself spending a lot of time messing around with it to get different sounds and not recording them. Thats why I thought the switch would be the best. Kind of one position for dynamics, one positions for ribbons kind of thing.

 

[scottlangendyk]

So I finally got around to wiring up both the 312's on my desk. I had originally had just one wired up, and I was using Canare StarQuad shielded cable (the only mic cable I had around at the time) to wire the XLR connectors to the card. This time around I decided to remove the shield from the mic cable and split up the 4 cores into two twisted pairs. I also swapped out the wiring on the pots to something similar in gauge to the Canare cores.

So I've wired everything up, and it seems to be working fine, but given the fact that I was no longer using shielded cable I decided to check out the noise floor. I put a 100ohm resistor (had a bunch on my desk), across pins 2 and 3 of the input, and cranked the gain. I compared the noise to the noise on my interface pres, and the actual noise seems a bit quieter, but there's a slight audible hum. I should also probably note to make sure I'm hearing the noise, I'm adding about 24dB of digital gain within Logic, while monitoring at a moderately loud level.

I'm just wondering what particularly could be causing this, or if I should even be concerned. I never did this test before hand while using the shielded cable, so it's hard to rule that out. I've read that it's important to use shielded cable for wiring up the gain pot particularly? Correct me if I'm wrong but, wouldn't the shield only be effective if it's connected to ground?

 

[Rybow]

Hey Scott. I don't know what could be causing your hum issue. I wired mine up today without shielded wire, and I get zero hum. Even with my toroid 3 inches from my output wires. All I can think of is if you don't have a ground connected somewhere, or maybe a ground loop.

As I said above, I finished wiring mine up today! Yippee! However, one channel has no output right now, but I am pretty sure I know what it is. If I can't figure it out tomorrow, I'll ask here. Too tired to troubleshoot.

Channel one (not working) has the EA 2622, EA 2503, and gar2520. Channel 2 has the cinemag 75101APC, CMOQ 2L, and gar1731. Both channels have output trims and an impedance switch that goes to 600 ohms or 1.5k. Haven't played with it yet, but I will in a couple of days. Power supply is one of Mike Fatoms 4 rail's (very nice). Case is from Dan Deurloo in the white market (also very nice). I also have output meters on each channel for 3 reasons. 1. Looks!! 2. Building a 2 channel, I had some extra space up front, and 3. Metering. All I have to do now is calibrate the meters and figure out whats up with channel 1. Then apply my graphics to the front. Oh, and clean up the wires. Anyways, here is a couple of pics.





Sound. So far, AWESOME! Tons of gain. Tried it with an SM 7B on my voice, gain on half and level was good. SM 57 on my amp. Ridiculous. Ree-freakin-diculous. Thats all I'll say for now.

Quick question. Wiring the EA2622 in series you just connect pin 2 and 3 right? Just making sure. Its most likely my opamp. It was the first DOA I have ever built, and I wasn't too confident. Plus, I think I heard something pop when I turned it off the first time.

 

[jsteiger]

...Quick question. Wiring the EA2622 in series you just connect pin 2 and 3 right?...

For typical 312 type circuit you will want the 2622's primaries wired in parallel. So, 1-->2 and then 3-->4.

Best, Jeff

 

[Rybow]

Thanks Jeff! Parallel. That's what I meant. I keep getting the 2 mixed up.

By the way, those Clarostat attenuators work insanely well. Although, I am a bit surprised by the amount of headroom in this circuit. Takes a lot of gain to get some grit, but the grit I get, I like. Anyways, worth every penny IMO.

 

[scottlangendyk]

Hey Scott. I don't know what could be causing your hum issue. I wired mine up today without shielded wire, and I get zero hum. Even with my toroid 3 inches from my output wires. All I can think of is if you don't have a ground connected somewhere, or maybe a ground loop.

As I said above, I finished wiring mine up today! Yippee! However, one channel has no output right now, but I am pretty sure I know what it is. If I can't figure it out tomorrow, I'll ask here. Too tired to troubleshoot.

Channel one (not working) has the EA 2622, EA 2503, and gar2520. Channel 2 has the cinemag 75101APC, CMOQ 2L, and gar1731. Both channels have output trims and an impedance switch that goes to 600 ohms or 1.5k. Haven't played with it yet, but I will in a couple of days. Power supply is one of Mike Fatoms 4 rail's (very nice). Case is from Dan Deurloo in the white market (also very nice). I also have output meters on each channel for 3 reasons. 1. Looks!! 2. Building a 2 channel, I had some extra space up front, and 3. Metering. All I have to do now is calibrate the meters and figure out whats up with channel 1. Then apply my graphics to the front. Oh, and clean up the wires. Anyways, here is a couple of pics.





Sound. So far, AWESOME! Tons of gain. Tried it with an SM 7B on my voice, gain on half and level was good. SM 57 on my amp. Ridiculous. Ree-freakin-diculous. Thats all I'll say for now.

Quick question. Wiring the EA2622 in series you just connect pin 2 and 3 right? Just making sure. Its most likely my opamp. It was the first DOA I have ever built, and I wasn't too confident. Plus, I think I heard something pop when I turned it off the first time.

Hum is only audible with gain maxed out, and adding an additional 50dbFS or so of gain in Logic. Never in a million years would I require that much gain and expect a better S/N ration. The only reason I bring up the issue is because I performed the same test with the pres on my interface, and there was the same amount of noise (expected) but no hum.

It's not a big deal, and with usual levels, I can only see this being an issue when I'm stacking a lot of tracks in a mix (though probably can't be much worse than stacking tracks from a cheap preamp, or stacking noisy guitar tracks). Just thought I'd post here and see if I was missing anything obvious. Probably doesn't hurt to shield my wiring anyways when I eventually get around to racking up the units.

I've been meaning to pick up an SM7B for tracking vocals, glad to hear you're really digging it through the 312.

 

[Rybow]

Yeah, I can see stacking being a possible issue, but with the amount of gain you are adding, that hum is pretty far down in the noise floor. You could just shield your input and output by connecting the shield to pin 1 or your XLR's chassis ground. That's as much as I know about using shielded wire. On most gain pots, pin 3 (I think) is connected to ground, so if you want to shield your gain pot, you could connect your shields to that. Check out your schematic to see how things are connected.

I was really surprised by the SM 7B through the 312. Really great sound. Compared to my apogee ensemble pres, there is a lot more mid range definition, and I used less gain. I still have to try taking the foam windscreen off the sm 7b. Apparently that makes a big difference, but it opens the mic up to plosives and windblasts. I avoided getting one for a long time just because everyone has one, and I was looking for something different. It is a good all rounder though. I have had good results on male vox, female vox, electric and acoustic guitars, and kick drum. Certainly usable.

 

[Rybow]

Sorry to double post, but I have figured out some things.

Channel one- Swapped the opamps, and 2520 from channel 1 worked fine. Great even! 1731 from channel 2 made no difference. This channel makes zero sound. No hum, no noise. So I think I have a short. Probably on the input transformer, or I wired the pins wrong. There is voltage going to the opamp pins. I'll look at that tomorrow. Fixed I must have done lots of drugs when connecting the primaries in parallel. Instead of connecting the starts together, and the ends together, I connected each windings start to its own end. Doh! I can be a real doofus.

Channel two- Phantom power. Nothing but hum. Measured 47.9 volts on both the hot and cold pins of the XLR. I wired pin 1 straight to the board, but maybe I have to wire it to the star ground instead? Fixed! I ran each XLR jacks pin 1 to the star ground. Its pretty messy right now as I ran each individual pin to star ground. Couple of Q's about cleaning this up as I see a bunch of possibilities. Let me know if any are big no no's.
1. I have connected each XLR jacks tab (the thing over pin 1) together, and then run a line to star ground. Can I just connect each jacks pin 1 to its tab?
2. Bob's boards- I have wired chassis jacks to the board instead of using PCB mount ones. On Bob's boards that are solder pads marked 1, 2, 3 which I had originally wired to. Under these is a solder pad marked G which does not have continuity to pin 1. Will this G pad give me a better ground?

Anyways, both channels a dead quiet now. Phantom power works perfectly. Loving the sound of these pre amps. Anyway, I'd type more, but I wanna play with my new toy! Thanks Bob, Jeff, Mike, and Mitsos for all your help.

Edit: Just wanted to talk about the impedance switch quickly. I got the chance to put it into use. I had a 57 to my channel 1, and an e609 to my channel 2. 57 on 600ohms sounds great. Not too dull, good definition in the highs, nice bass. On 1.5K the highs become much more pronounced. e609 was similar, except on 600 ohms it was dull as mud, and on 1.5K it really shined. I did a test recording using a distorted electric guitar with the 57 on 600 ohm, e609 on 1.5k. Sound was great! Both mics complimented each other nicely to create a real thick sound with lotsa definition. Anyway, impedance switch is highly recommended. Gives you the choice of more highs or more lows.

 

[soundcave]

My current work  ;D: