Driving Force for GRADOs

[jelosno]

What is the perfect driving force for Grados like the RS-1?
What makes a Grado rally swing?

To chose THE BEST MATCHING OPT for a tube amp to drive Grados there should be knowledge of what current and volatage the Grado likes best.

Any knowledge out there?

is this data available somewhere - haven't found it on the Grado site.
can it be calculated from the RS-1 data available.

Stefan

 

[Garbz]

There is no such thing as what it likes best. Only what it does with what you supply it for one and what it needs from the other.

Changes in voltage directly effect the volume. The driver being a reactive load to form a voltage over it, it must draw some current. Providing the required current for a given voltage can be supplied there's no more the amp can do to make it Grado friendly. From then on it's just how well the amp itself works.

Many people falsely attribute a huge abundance of overated current for a good sounding amp, when in reality this current is often not used even on fast loud transients.

On the headwize forum one member posted a lovely mathematical derrivation of all sorts based on the grado specs. What voltages are needed for certain volumes, etc.

 

[digitalmind]

Many people will agree that the Mapletree Ear+ HD is the best amp on the market for a RS-1. It's available as prebuilt or in kit form on http://hollowstate.netfirms.com/, and it costs $450 + $50 for the HD upgrade.

 

[jelosno]

Quote:

Originally Posted by digitalmind Many people will agree that the Mapletree Ear+ HD is the best amp on the market for a RS-1. It's available as prebuilt or in kit form on http://hollowstate.netfirms.com/, and it costs $450 + $50 for the HD upgrade.

YUP, I do have the EAR+HD

it uses a Hammond 119DA OPT. The EAR+ also works fine with some other headphones.

I would like to replace the 119DA with something that narrows the output down to only work with Grados and that even better.

The EAR+ IS excellent with the RS-1. The 119DA is a standard part that contributes to the very good price of the EAR+. If I could swap the 119DA with it's 600:8 (or 75:1 or turns ratio of 8.66:1) for a OPT with a differnt ratio - maybe 1000:8 (custom made Sowter?) or so I could get better for Grados.

To calculate the 'best' OPT ratio I would need the mA and mW environment the RS-1 likes best.

Stefan

 

[digitalmind]

I could be wrong here, but.

A really good amp for Grado's is the RA-1. A low gain Grado RA-1 amp uses a gain of 4.8 (non inverting opamp with 465k and 122k resistors). Say you listen at 1/3rd of volume to get the best results, with an average of 1 volt line input you'll be running roughly 50mA, or 80mW, through the RS-1. (U = IR, 1*(4.8/3) = I*32, I = 0.05).

I'm not entirely sure what the gain of the Ear+ is, otherwise that would be a much better starting point since you have it on hand (as do I, great amp!).
I think that that would be a good starting point. Are you planning on designing an amp from the ground up?

 

[jelosno]

digitalmind,

I really like the idea of the RA-1 data.

Haven't thought of that one. I have the RA-1 as well. I could put a 32 Ohm load on it and measure the rest.

I will not build an amp from ground up. But I am planning to do some modifications to the EAR+ over the course of the next months.

Since I plan to use some rather large parts like 50uF PIO caps before the OPT I will build a new case when I do have time. Most probably a metal case with a stainless steel front.

Thx for the input

Stefan

 

[dsavitsk]

Quote:

Originally Posted by Garbz From then on it's just how well the amp itself works.

This just ignores too much to be useful at all. The headphones do not constitute a seperate system after the amp, but are an integral component that influence how the amp behaves. This is especially true with tubes and output transformers, but is true for any amp.

Quote:

Originally Posted by digitalmind I could be wrong here, but.

I am going to go out on a limb and say that a single 4556 opamp with some low quality input capacitors in the signal path does not consitiute the end all and be all of Grado amplification. It is a dreadful little amp.

Quote:

Originally Posted by jelosno I will not build an amp from ground up. But I am planning to do some modifications to the EAR+ over the course of the next months.

I think that the place to start with modding the EAR+ is with the PS. Look around for posts by voodoochile for what he did there. The basica are, if you are going to stick with solid state rectification, add a snubber cap and a high value resistor in parallel to each diode, and upgrade the diodes to high speed ones. Otherwise use a tube. Also, try to remove the electrolytics from the PS -- Solens or ASC caps are very good replacements -- and add a choke or two.

The second thing to do is likely to swap out the output trassformers. I don't remember if the EAR is a single feed or a para feed, but there are some decent James (single feed) options for cheap-ish, or if you want to spend a little more, have some EP's or some Magnequests made. There are lots of options that are better than the Hammonds, and very few that are worse.

There is no great secret to driving grados with tubes -- use a tube that can put out enough power, and use a high quality OPT with a secondary that matches the Grado's impedance. I have found that paralleling a 30R or so resistor with the phones and using a 16 ohm trabsformer tap is a good way to go as you will liklely lose bass response if the transformers don't have a purpose made 32 ohm tap, though it does depend on the design.

 

[jelosno]

Quote:

It is a dreadful little amp.

Naaa. Dreadful is something else. It's actually quite OK. For a mobile battery powered one. For the price though - it is dreadful.

Quote:

I think that the place to start with modding the EAR+ is with the PS. Look around for posts by voodoochile for what he did there.

Yup, I do have close looks there! Also followed voodoochile's posts closely

but he went the tube way.
I will also build a stainless steel thingy but completely different from his design which is just way cool and extremely well executed.

Quote:

The basica are, if you are going to stick with solid state rectification, add a snubber cap and a high value resistor in parallel to each diode, and upgrade the diodes to high speed ones.

What's a 'snubber cap'? The HD version does have HEXFREDS.

Quote:

Otherwise use a tube.

Really that much better? I would like to stick to SS rectification.

Quote:

Also, try to remove the electrolytics from the PS -- Solens or ASC caps are very good replacements -- and add a choke or two.

I have ordered 100uF Solen FastCaps with the kit right from the start. Those are a reason alone to build a new chassis.
As for the chokes this is some kind of a tricky thing. Not to integrate them but to bring the voltage down to the required levels after the rectifier. One solution would be to use resistors in line with the chokes but then the whole idea of using a choke would be gone. I will experiment with the power transformer a bit. Maybe parallel = 115V instead of 230V as Lloyd Peppard the designer of the EAR+ suggested. Will see...

Quote:

don't remember if the EAR is a single feed or a para feed

Para Feed as far as I can recall it. There is a document describing it on www.mapletreeaudio/earplusdesign.pdf but I can't open it at the moment. It is under the headphone section in the start page.

Quote:

but there are some decent James (single feed) options for cheap-ish

Not that I know of. If I would stick with the 600:8 of the EAR+ the James OPTs would be $ 300+.

 

[digitalmind]

Quote:

Originally Posted by dsavitsk I am going to go out on a limb and say that a single 4556 opamp with some low quality input capacitors in the signal path does not consitiute the end all and be all of Grado amplification. It is a dreadful little amp.

It's a very, very basic design but it has been noted that Grado's sound great out of it. Simplicity can be good in some cases.

 

[dsavitsk]

Quote:

Originally Posted by jelosno What's a 'snubber cap'? The HD version does have HEXFREDS.

Basically it is a cap (and often a resistor) in parallel with each of the diodes that prevent the switching noise inherent to solid state diodes. With hexfreds it is less important, but will still likely do some good. For the caps, try something like a 10nF (0.01uF) high voltage (1.5KV) ceramic cap and for resistors, something high like 330K/2W.

Quote:

Really that much better? I would like to stick to SS rectification.

SS should be fine.

Quote:

As for the chokes this is some kind of a tricky thing. Not to integrate them but to bring the voltage down to the required levels after the rectifier.

Chokes do have some DC resistance, so they may bring down the voltage already. But, you could always just add an extra RC filter to the PS to drop some volts, and to add that much more filtering.

Quote:

Para Feed as far as I can recall it.

The link didn't work for me, but for parafeed you probably don't want the James transformers as they are air gapped. Look at some of the magnequest offerings. Mike can probably custom wind whatever you need.


Quote:

the James OPTs would be $ 300+.

That's not really so bad in OPT world, though there are some James trannies in the $150 range that are fine. http://euphoniaaudio.netfirms.com/ea...6bfd9db6759561 Not sure why you'd be looking at the high watt versions for headphones?

 

[dsavitsk]

Quote:

Originally Posted by digitalmind It's a very, very basic design but it has been noted that Grado's sound great out of it.

I will respectfully disagree, and leave it at that.

 

[jelosno]

Quote:

Originally Posted by dsavitsk The link didn't work for me,

Found it. Here we go: http://hollowstate.netfirms.com/earplusdesign.pdf

 

[cetoole]

Quote:

Originally Posted by dsavitsk I will respectfully disagree, and leave it at that.

On which part of the statement, that a monolythic device with ~17 active devices and about a dozen passives per channel is simple, or that it sounds good?

 

[dsavitsk]

Quote:

Originally Posted by jelosno Found it. Here we go: http://hollowstate.netfirms.com/earplusdesign.pdf

Interesting. There are a couple of options here, some minor, some major. A minor one would be to increase the transformer ratio (with a different transformer obviously) which would allow you to reduce the size of the coupling cap allowing somethig of higher quality. However, finding how to tame the resonant frequency is not easy, and might involve some trial and error. This would also reduce gain which may or may not be a good thing.

A slightly more major change would be to load the plates with a CCS. This will increase gain slightly and make the tubes operate in a more linear way.

If it were me, I'd do away with the cathode follower all together (at the expense that it would not be a line level preamp anymore), replace the 12ax7 with another 12b4, get a 5K:32 parafeed transformer, and attach this to the plates -- with a CCS load too. Obviously it is a different amp at that point, but at least it could still use the same chassis and most of the parts. Also, it is about unity gain in this config.

Quote:

Originally Posted by cetoole On which part of the statement, that a monolythic device with ~17 active devices and about a dozen passives per channel is simple, or that it sounds good?

I was only addressing the second, but the first is worth saying too.

 

[tomb]

Quote:

Originally Posted by dsavitsk I was only addressing the second, but the first is worth saying too.

With you on that, and could be indicative of a general theme. As you have stated - plainly - not a good direction to go around here.

To get off that, dsavitsk, my experiment with 470uF Nichicon ES's without a bypass box cap on C7 for the Millett was a dismal failure. To top it all, the box cap is needed for bass!

Besides obvious harshness without the box cap, the bass disappeared, too! Soldering the box cap back in fixed it.