[Veto]

Quote:

Originally Posted by Sugano-san .... Anyone's advice to use a 627 without compensation (I am aware it has a "built-in" one) instead of a 637 with the proper compensation should therefore be ignored. I tried it. The 627 does not work on a stand-alone basis in the RKV. I would be a waste of time.

Just checking but, the 627 with the cap works fine?

Veto

 

[ppl]

If I was Doing This Mod I would use anAD744 with pin (5) as opposed to Pin (6) as the opamps output. Pin (6) is the normal output of the opamp and pin(5) is the compensation Pin. Leave4 pin (6) Not Connected and use pim (5) as the opamp output Driving the Tube.

This is a far beter soultion that trying to use an OPA637 for previously stated reasons however any opamp operating from its internal output stage needs current sourcing to remove crossover distortion with the comp pin you compleatly Bypass the AD744's internal output stage and drive the Tubefrom the AD744's SE Class A Voltage gain stage and the ultra high input
Impedance of the tube's makes this confoguation marvolus. IMHO

 

[jcx]

Quote:

Originally Posted by Sugano-san You may not have read my longer post on page 1 of this thread. In that post I described that neither the 637 nor the 627 work in the RKV without such a compensation. And I should know because I tried all 4 variants (AP and BP of both). As you may have gathered from reading parts of this thread, the 637 works fine with the compensation. The value of that compensation was determined by the designer and manufacturer of the RKV, who knows its "topology" better than any of us. Anyone's advice to use a 627 without compensation (I am aware it has a "built-in" one) instead of a 637 with the proper compensation should therefore be ignored. I tried it. The 627 does not work on a stand-alone basis in the RKV. I would be a waste of time. If you are interested in details of how the RKV's circuit works, you may want to search for 00940's summary description, which he posted some months ago. You will find that the gain is way higher than 5.

You have seriously mistaken my point, I agree that the unity gain stable OP627 should work only when compensated with the additional 68 pF – and even if it establishes stability you may introducing interaction with source Z or volume pot setting with this type of compensation

The cap 68 pF Cap and input Z ( 1 K resistor + whatever is to the left) forms an integrating feedback that slows down substitute op amps much faster than the 3-4MHz LF351 to make the outer loop stable around the inverting gain tube output stage when using the faster op amp
//
edit: I need to sim this, after some thought I don't believe this is the correct view of the outer loop compensation if the goal is to replicate the gain/phase slope of the 351 with a faster op amp - the high frequency stability argument is unaffected
//

but the added 68 pF also determines a local high frequency gain around the faster op amp - the audio frequency closed loop gain of 40+ has little to do with the stability of this high frequency inner feedback loop

2 loops with 2 different requirements

Now that I can see the 100pF RF bypass to gnd at the op amp inverting input I can calculate the local “noise gain” seen by the fast op amp at high frequencies to be ~ 2.5 ~= 1+ 100/68 which is fine for the OP627 which is specified to be stable down to unity gain but I still can’t see why you expect the same compensation to work with the OPA637

What part of: “not recommended for gain less than 5” do you not understand about the OPA637? – maybe you can’t see that 2.5 < 5 ?

you have to check the stability on a per part basis when violating vendors’ recommended operating conditions – Burr Brown’s specs are very conservative and the mod may work with the OPA637 but it cannot be recommended to people who don’t have the knowledge or means to determine if the OPA637 is oscillating in the 10+ MHz region and are receiving different batches of op amps manufactured at different times from the ones you assure us “works” for you

USE the 68pF comensation C with UNITY GAIN STABLE op amps faster than the LF351 ~ 3-4 MHz gain bandwidth

DO NOT use the compenstion as shown for op amps with a minimum gain requirement >2.5, like the OPA637

rkv refs:

http://v3.espacenet.com/origdoc?DB=E...&QPN=DE3200517

http://www6.head-fi.org/forums/showt...8&page=1&pp=20

 

[sassysound]

Quote:

Originally Posted by dmb367 Is there a better place to look for a used one then here? After all the nonsense, I'm dying to get one... thanks alot Dan

Man, all of us who went to the meet is paying for it now....

haha so who's next?

BTW I think we need a new smiley.....one that has a smiley eating our wallets. It will be head-fi's most used smiley .... :eatwallet:

 

[Sugano-san]

Quote:

Originally Posted by Veto Just checking but, the 627 with the cap works fine?

No idea, haven't tried it. When I get a chance I will ask Helmut if he is aware of any reasons not to do it.

 

[Jon L]

Quote:

Originally Posted by ppl If I was Doing This Mod I would use anAD744 with pin (5) as opposed to Pin (6) as the opamps output. Pin (6) is the normal output of the opamp and pin(5) is the compensation Pin. Leave4 pin (6) Not Connected and use pim (5) as the opamp output Driving the Tube. This is a far beter soultion that trying to use an OPA637 for previously stated reasons however any opamp operating from its internal output stage needs current sourcing to remove crossover distortion with the comp pin you compleatly Bypass the AD744's internal output stage and drive the Tubefrom the AD744's SE Class A Voltage gain stage and the ultra high input Impedance of the tube's makes this confoguation marvolus. IMHO

Ah, I was wondering the same thing about AD744. This is known in Xin's SuperMacro3 circuits as "output bypass" mod. AD744 output bypass is my favorite opamp in SM3, more than 627 or 637. On the same token, AD829 can also be used in output bypass mode, which should make an interesting option for RKV mod..

 

[Sugano-san]

Quote:

Originally Posted by ppl If I was Doing This Mod I would use anAD744 with pin (5) as opposed to Pin (6) as the opamps output. Pin (6) is the normal output of the opamp and pin(5) is the compensation Pin. Leave4 pin (6) Not Connected and use pim (5) as the opamp output Driving the Tube. This is a far beter soultion that trying to use an OPA637 for previously stated reasons however any opamp operating from its internal output stage needs current sourcing to remove crossover distortion with the comp pin you compleatly Bypass the AD744's internal output stage and drive the Tubefrom the AD744's SE Class A Voltage gain stage and the ultra high input Impedance of the tube's makes this confoguation marvolus. IMHO

Phil,

Thanks very much for your input. The RKV mod as described by Dan & I is hopefully just the first step of an interesting journey exploring all promising ways for maxxxing out this amp. And it goes without saying that constructive input from amp designers is highly appreciated.

The RKV has responded so very well to the 637 mod that the first objective has been reached. This first objective was to demonstrate that the RKV has a huge potential for excellent sound, which is not sufficiently exploited in its stock configuration. Other options may be even more promising than the OPA627/637 as described above.

The second objective would therefore be to start off a discussion and identify those other options.

The approach you described sounds like such an option, but unfortunately I do not have any AD744s at home, so I cannot try it for the time being. I just had a quick look at the Analog Devices website. Which variant of the AD744 is the one to get? In PDIP size they only offer samples of the following types: JN, JNZ and KNZ. Which one of these would you recommend?

Would other AD opamps also work, e.g., the AD825ARZ, the AD829JNZ or the AD847JNZ?

Would your approach also work with opamps from another manufacturer, e.g., ON Semiconductor, specifically, the SA5534N or AN or the SE5534N or AN? I've got some of those at home.

Thanks in advance.

 

[Veto]

Did this 'mod' today.
WOW the amp sounds really different, it used to be a bit tubby but now it's tight. GREAT bass!
Really works well with HD650.
Don't really like it with the K1000 though, but I will check some more (comparing is not all that quickly done). I think the new highs don't mix well with the already bright K1000.

Thanks for the mod, if you have this amp and a HD650 you really should check it out. For the money it's a no brainer.

Veto

 

[Blorton]

Veto - glad to hear you like it. Did you go with the 627 or the 637?

Dan

 

[Blorton]

Sugano - would you be so kind as to post pics and information on the cap upgrades that were done on yours? Transformer too, please.

Might as well make this an official RKV mods thread.

Thanks,
Dan

 

[Veto]

Quote:

Originally Posted by Blorton Veto - glad to hear you like it. Did you go with the 627 or the 637? Dan

637, I didn't want to take the risk of it not working (I'm no electro wizard, barely know how to use a soldering iron

)

The SA5000 behaves real strange with the modded RKV, the bass is great but the rest nah.
I don't have the impendanzer (?) so I guess that one won't work whatever i try.

Veto

 

[Blorton]

Quote:

Originally Posted by Veto 637, I didn't want to take the risk of it not working (I'm no electro wizard, barely know how to use a soldering iron ) The SA5000 behaves real strange with the modded RKV, the bass is great but the rest nah. I don't have the impendanzer (?) so I guess that one won't work whatever i try. Veto

I've had the same problem with my SA5k's. I've had to give up using them with this amp for the time being until I can figure out a solution.

If my Impedanzer is any indication, you aren't missing anything. It doesn't help.

It may well be that what ppl and jcx are saying is reallu highlighted by the SA5k and/or the Impedanzer. Guess I'll try their suggestions and see how that goes. My SA5k's are down for probably at least two weeks, so it'll be awhile before I can report on that.

Dan

 

[Veto]

Well I tried some more with the K1000's and they definately don't respond well to the new opamp.
It seems that the lower mids are somewhat leaner with the mod, the bass is lower and tighter (you can hear that with the hd650 and sa5000) but I guess out of the reach for the K1000 to pick it up. Also the (lower)mids are somewhat fleshed out and that doesn't benefit the akg's and I guess the sony's don't like that either.
The hd650 can handle a somewhat leaner midrange a lot better maybe it remedies the bump it has in the lower regions.

Again with the mod the 650 really (really!) sounds excellent.

Veto

 

[Sugano-san]

Quote:

Originally Posted by Blorton Sugano - would you be so kind as to post pics and information on the cap upgrades that were done on yours? Transformer too, please.

Have a look at the first photo below, the silver RKV/Impedanzer combination, and note in particular the transformer in the back. You can see the copper color of the wire, because it’s not a shielded transformer. For all I know this transformer works very well in the stock RKV setup, which comprises an Alps volume pot (shielded).

However, in my RKV that pot was replaced with a stepped attenuator (which is not shielded), and that thing picked up a little background buzz, a bit like an antenna. I therefore had the stock tramsformer replaced with a mu metal-shielded transformer like the one you can see on the second photo below. The buzz is gone now (it was never prominent, but audible).

By the way, that RKV/Impedanzer combination is mine, the only silver one currently in existence. Some of the chrome plated parts (knob, stripes, feet) have been replaced with gold plated ones, and it looks even better now.

Regarding the 2 pairs of caps (the big blue ones), their stock capacitance is 220µF, but I had them replaced with 380µF types (plus the new ones are suitable for a higher voltage than the stock ones). In particular, the new caps are of a much higher quality which means that the internal resistance (or whatever the technical terms is) is only 20-25% of that of the blue ones.

I was really lucky to score three pairs of these extremely high quality French "elkos" that were part of a custom made batch for use in measuring instruments. 380µF is not a standard value, afaik.

You can see the new transformer, the stock caps and also the socketed Opamps very clearly on this photo (Note that this is not my RKV's board):

 

[Sugano-san]

Quote:

Originally Posted by Veto Again with the mod the 650 really (really!) sounds excellent.

I couldn't agree more. Welcome to the club, Veto!