£500 pre's - SPL GoldMike v FMR RNP - Opinions.

  • Thread starter Thread starter Mr Funk
  • Start date Start date
Clever work-arounds aren't necessarily a bad thing... That is, unless it adds unnessary crap that would add crap in the "sound" or prevent a desired lack-thereof (which ever is applicable), but that doesn't seem to be the case here.

Of course, they're not necessarily all that "appropriate" or cost effective in many cases either... Yes, there's plenty of times where a different component would be desired.

How it applies to Earthworks preamps, I have no idea.
 
I think there's a slight misunderstanding here. The SSM2017 is not an opamp but what is called an "instrument amp". A chip like the SSM2015/2017 or the Burr Brown INA 103/163/217 (the latter being a drop in replacement for the SSM2017) is kind of a ready made preamp, there's really not much to be added. If you take a look at the datasheets (Burr Brown available from www.ti.com) you'll see that you really need only a very few additional parts, no elaborate design necessary. The crucial points perhaps are the input caps (BB suggests bipolar electrolytics but most preamps use regular electrolytics because they're much cheaper) and the output stage which can either use an additional opamp to eleminate the DC offset or again capacitators, again usually regular electrolytics. Or you could get rid of the electrolytics in the signal path by using quality input and/or output transformers. Jensen sells transformers specifically to enhance the performance of SSM2017 and 2015 chips. So in theory there are a few options to chose from, but the less expensive preamps use designs with electrolytics in the signal path. Perhaps the Earthworks uses something better, I don't know, but if they do, that might explain its cleaner sound.

A cheap way to enhance the performance of such designs is to solder a a small (0.1 uF) quality film capacitator in parallel with the electrolytics. Scott Dorsey keeps recommending this on RAP. I've tried this on my Behinger Ultragain 2000 (the old one without a fake tube stage), and it did in fact make it sound more transparent, quite noticably so. It's also a good idea to renew the electrolytics while you're at it. Especially if it's an older unit. Electrolytics tend to degenerate more rapidly if they don't get some voltage. In those preamps only the input capacitators get some decent voltage when blocking phantom power. The output caps are mostly in there for safety reasons and blocking what little DC offset is produced by the preamp. Thus they're prone to decay. Another easy mod is to exchange tantalum capacitators, should there be any on the PCB, for quality film types. Tantalums really cloud the sound. In a chip based design you probably won't find any tantalums, but in a less expensive discrete preamp like the Ultragain 2000 you might. Behringer, for some reason, always used tantalum 0.1 uF capacitators in their older through the hole preamps and processors, even though other caps were film types. After my various recapping experiments my ancient Ultragain 2000 sounds really good and clean, much more transparent than before. I'm sure you can enhance a chip based design that way, too.

BTW: The much rumored 1510 chip is from THAT, not THE. But I wouldn't expect too much by just replacing the 2017. The 2017 is not such a bad chip. The Burr Brown replacement INA217 has slightly better specs, the THAT chip is not available yet, so nobody knows. Some designs use really cheap opamps after the input stage. Those can compromise the performance more than the 2017 itself. There's good reading on RAP about stuff like that, in particular a post by Monte McGuire about modding a Symetrix 202 (you can find stuff like that via Google groups search). But my guess is that recapping and adding those small bypass capacitators in most cases will be the best bang for the buck, it's also the easiest to do. Removing a chip from a PCB is really not that easy, especially if you want to keep both the chip and the PCB intact.

Or just skip the whole modding part and buy a really good discrete preamp. I wouldn't go for another chip based preamp like the SPL, if you want something radically different and better than what you have. Personally, I really don't believe in hybrid designs. Either tube or transistor is fine, but having both in the same unit is like dining with your best friend and a supermodel at the same time.

BTW pt. 2: The last time I checked the INA103 was still available. The INA163 seems to be a slightly simpler version of the 103 at about half its price, but it's only available in SMD versions. The INA103 is through-the-hole.
 
Hmm, pretty long reply in retrospect. I hope this doesn't sound like I claim to be an expert. I'm not. I just like to read datasheets and RAP posts. Solder a bit.
 
Say, that's some great info.

Thanks, Rossi! :D

THE . . . THIS, THAT, THE OTHER . . . it's all the same. :D


Speaking of Monte, here's just one quote I pulled off RAP regarding the 2017. It's not hard to find a lot more just like it. And Scott has basically echoed the exact same sentiments (hatred) towards that chip in a few emails we've exchanged :

Monte: the 2017 has a rising THD vs. frequency curve that starts to take off around 3KHz, and that's a bad sign IMHO. It usually correlates with an overbright, slightly brittle sound.

My data sheet for the 2015 doesn't show a distortion curve directly, but by comparing their 1KHz and 10KHz THD numbers, I concluded that there is far less of a HF THD rise on the 2015 than the 2017. So, while the low frequency distortion of the 2015 is slightly higher (but that might not really be true), the shape of the THD curve for the 2015 is superior; it implies less 'nasty' forms of distortion than that of the 2017. So, in one way, the 2015 specs better, if you count the nature of the THD rather than the amount.

I also feel the 2015 sounds better; I've compared a modified dbx 760x (which uses a 2017) to a modified Symetrix 202 (which uses the 2015) and I definitely preferred the 2015. The modifications resulted in very similar support circuitry for both mike amps: the same model of coupling caps were used in both units, the same types of op amps were used in the output stage, and the power supply bypassing was similar in both circuits.



And judging from what my own ears tell me . . . I can't help but believe there's some truth to it. I've owned several mic pres through the years that utilized the ssm2017 . . . and have grown to dislike just about all of them for the same reasons Monte states . . . Cymbals just don't sound right -- the high end overall just sounds kind of funny. -- which might imply some sort of HF distortion going on.

The pres that I've liked all along, on the other hand, tend to use the same chip -- the INA163 or it's predecessor, the 103. And I had no idea until recently that that was the case. And the main reason is the sound I get from cymbals, and smoother esses from vocal mics.

Coincidence?
 
The distortion figures on the Burr Browns are indeed better than those in the SSM 2017 datasheet. But it's probably not just the figures themselves but the kind of distortion they produce. Some people say the Burr Brown opamps sound "euphonic", hence a kind of better sounding exciter-like distortion. That may be true for the instrument amps as well.

Yesterday, after my post I downloaded a product description for the Grace preamp. The link was posted in the DMP-3 thread. What I read in that description confirmed much of what I said. Just read the section about passive components which stresses the importance of the capacitators.

One thing I find odd is that everybody seems to use quite large electrolytics as input caps, usually about 47-100 uF. Such a value in high quality MKP film capacitators would cost quite a lot. The question however is if you really need that big a capacitor (actually 2 per channel). According to the formula C=1/2*pi*R*f you'd need about 5.3 uF for a 3K impedance input. The resulting lowcut frequency would be 10 Hz. For 20 Hz, which would still be enough ( the slope is only 6db/octave) half that value (2.66 uF) would be enough. High quality film capacitators at those values would be affordable. The higher the input impedance the smaller the values needed.

Since Grace apparently uses film types as input caps, I'd like to know how big theirs are. I'm sure they wouldn't use more than is really needed, as they are very expensive. With electrolytics the size hardly matters, a 100 uF costs about the same as 47 uF. If there's no other reason why a bigger cap is needed you could substitute those dull electrolytics by a much smaller MKP capacitor according to the formula above. I'm quite sure it's mostly the electrolytics in the signal path that mark the difference between DMP3 and Grace.
 
Rossi said:
I'm quite sure it's mostly the electrolytics in the signal path that mark the difference between DMP3 and Grace.

That says it all right there... Many people don't understand that caps can make as big of a difference as they do!

Electrolytics don't belong in the signal path... And forget about tantulams all together!

With that said, it's amazing that when when you open-up the RNC, you find tantulams all over the place! How is it that it sounds as good as it does? I can only imagine how it would sound without them!
 
Rossi said:
I'm quite sure it's mostly the electrolytics in the signal path that mark the difference between DMP3 and Grace.

That says it all right there... Many people don't understand that caps can make as big of a difference as they do!

Electrolytics don't belong in the signal path... And forget about tantulams all together!

With that said, it's amazing that when when you open-up the RNC, you find tantulams all over the place! How is it that it sounds as good as it does? I can only imagine how it would sound without them!
 
I don't like tantalums either, but even some Neve pres use them. Apparently there are ways to get around their shortcomings. They're either not in in the signal path or they are uses with a DC offset. The main problem with tantalums seems to be that they don't work properly around the zero level. That's also true for (polarized) electrolytics. From what I've heard the RNP does sound very clean, so I guess good engineering can do a lot. Still it would be cool to hear a statement about why tantalums were used. I suspect it was mainly their size. Yet another argument for a full blown 19" version, perhaps with a (variable) lowcut which I find strangely missing. I love variable lowcuts.
 
Is the Aphex 207 an improvement on the DMP2? I know that the RNP, Grace etc would be much better upgrades, but it is possible to get a used 207 for about $100 and I'm wondering if the 207 is $100 bettere than the DMP2?
 
man if your getting the new aphex 207 for $100 please let me in on your source cause I've been hearing great things and the cheapest I've seen used is about $250-$275
 
Back
Top