AD797 Phono Stage Build and Help Desk Thread

[HalfApt]

I'll give the S/N ratio a rough calculation based on a quick read of the 1977 Leach article....

Geez, I wasn't expecting that, but if you're having fun ...!

1977 article? The first I saw was Feb. '78 Audio magazine, is there an earlier one?

He also issued an "update" in Oct. '81, w/major changes to resistor values, then in the '99 paper reverted back to mostly the same as the original values! Summarized here:



I built mine in '88, so it's based on the '81 parts revision.

 

[wyn palmer]

Geez, I wasn't expecting that, but if you're having fun ...!

1977 article? The first I saw was Feb. '78 Audio magazine, is there an earlier one?

He also issued an "update" in Oct. '81, w/major changes to resistor values, then in the '99 paper reverted back to mostly the same as the original values! Summarized here:

View attachment 1279024

I built mine in '88, so it's based on the '81 parts revision.

http://leachlegacy.ece.gatech.edu/papers/wbpreamp/feb77article.pdf

 

[HalfApt]

http://leachlegacy.ece.gatech.edu/papers/wbpreamp/feb77article.pdf

Aha! We're talking about different things.

I'm using this one (Audio '78 & updates):

http://www.NullTime.com/hobby/audio/Leach MC pre-preamp.pdf

And his 1999 paper:

http://leachlegacy.ece.gatech.edu/headamp/

 

[wyn palmer]

Thank you, much appreciated. Sorry for my confusion.
The AD797 has a input noise voltage of 0.9nv/rtHz that is equivalent to the Johnson noise of c. 50 ohms. The input current noise is 2pA/rtHz and the non-inverting/inverting resistances to ground are 11 ohms combined for the 34dB design, so that noise component is negligible.
A quick look- and I hope I'm not wrong- suggests that ideal re generated input voltage noise for the input pair of the Leach MC amp is about 3dB higher than this (100 ohms equiv due to the two emitters being in parallel), and that's before any noise generated by rb (the base spreading resistance) of the transistors- which probably adds a dB or 2
So, I would expect that the AD797 design would have a 3-5dB input referred noise improvement over the Leach amp.
The gain is 8dB higher for the AD797MC also. So, I would expect that if you were happy with the Leach amp and you did not request an overly high output from the 797 amp that would overload something, that you would be more than happy with the AD797 design.

 

[mkane]

Looks great on paper, sounds great in action.

 

[HalfApt]

Thanks Wyn!

Your analysis gives me confidence that the AD797 will be fine. And now that you have "learned" me in the finer points of design, I can always reduce the gain a bit to clean it up if needed.

w/LTSpice, I have been able to play with the Leach circuit and see different ways to increase its gain, but don't know which is appropriate.

I'm off to build the AD797! Will let you all know how it goes (but don't hold your breath, got many other things to do as well).

 

[tabarddn51]

Hi Guy's. just thought I would let you know I took the plunge & ordered 2 x AD797SQ from F/Bay, just to try them out, at $3-50 each with free delivery if they were crap then at that price it does not matter.

https://www.ebay.com/itm/1pcs-AD797...e=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

Well today I received them & promptly pooped them into my spare AD797 phono amp. Swapped over the unit from my main AD797 & sat back to listen. Well the 1st thing I noticed was a large reduction in "hiss noise" that I had experienced from this unit. Also even not broken in I sensed I was listening to more detail & the sound stage was sightly more forward.
So anyone wanting to try this phono but are put off by the high cost of the genuine AD797, I certainly can recommend these units, they are definitely staying my set up

Cheers Derek

 

[wyn palmer]

Hi Guy's. just thought I would let you know I took the plunge & ordered 2 x AD797SQ from F/Bay, just to try them out, at $3-50 each with free delivery if they were crap then at that price it does not matter.

https://www.ebay.com/itm/1pcs-AD797SQ-883B-8Pin-LOW-NOISE-Mono-Single-OP-AMP-Linear-IC-797ANZ-OPA604AP/153001716591?ssPageName=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

Well today I received them & promptly pooped them into my spare AD797 phono amp. Swapped over the unit from my main AD797 & sat back to listen. Well the 1st thing I noticed was a large reduction in "hiss noise" that I had experienced from this unit. Also even not broken in I sensed I was listening to more detail & the sound stage was sightly more forward.
So anyone wanting to try this phono but are put off by the high cost of the genuine AD797, I certainly can recommend these units, they are definitely staying my set up

Cheers Derek

Well, you can easily find out if they are genuine by doing two or three measurements- the quiescent current consumption of one of the opamps (c.8ma) the gain bandwidth product (110MHz- you need an oscillator/scope) and the offset.
The AD797 has a very low input v offset at 50uv nominally, a bias current of 250nA, and also a current offset of about 100nA.
I would recommend that you try and do the first two measurements if you can before recommending a source. Rip offs are common and the audible differences are not as profound as you might like to believe. The cheap amps I bought failed all three tests but sounded OK, just not as good as the AD797s.
Wyn

 

[wyn palmer]

Thanks Wyn!

Your analysis gives me confidence that the AD797 will be fine. And now that you have "learned" me in the finer points of design, I can always reduce the gain a bit to clean it up if needed.

w/LTSpice, I have been able to play with the Leach circuit and see different ways to increase its gain, but don't know which is appropriate.

I'm off to build the AD797! Will let you all know how it goes (but don't hold your breath, got many other things to do as well).

I ran the ADI generated model AD797 sims with the A weighting network, then ran the sims for the Leach preamp using the most recent .asc file. The frequency response was measured over 20Hz-12kHz- so I effectively added a "brick wall filter" at 12kHz. The AD797 gave a weighted S/N ratio with a 0.07mv 1kHz input of 56dB (5.5uvrms noise at the output), the Leach was 5.4uvrms noise scaled to x50 from x20 by modifying the load R to 10k with the 47k RIAA stage load in place and the cartridge 4 ohms source added with a 75 ohms replacing the 1k input R- about 56dB also.
So, the simulated result is essentially identical, with the Leach having a couple of tenths of a dB lower gain and slightly lower noise than the AD797.
The original transistors result in a performance that is, as he states, about 4-5dB worse.
So, if you want to stick with the Leach, change R1 to 75 and R4 to 10k and the ac response at least will be equivalent to the AD797 design-and the current consumption will be much less.
Incidentally this comparison is not entirely apples to apples.
The noise for the AD797 model is the worst case data sheet limit, whereas the noise for the transistors is typical. There could be a few dBs (3 or so) difference- to the benefit of the AD797 design- in practice, which is more or less what the back of the envelope calculation suggested.

 

[BigElCat]

No, you have to have an SOIC package to work with the adapter, whether it is an A or B device.
A part number for example would be the AD797ARZ or the AD797BRZ, both of which can be bought from Mouser.

I posted a link to a "BrownDog" adaptor, two AD797s on an 8 pin DIP.

Looks like plug-n-play.

You're saying I'd need a SOIC also ? What's a SOIC ?

 

[sachu888]

This preamp needs 4 mono opamps. You will need single opamp on the browndog adapter, and that adapter is different than you posted. You will need to use these adapters https://www.ebay.com/itm/2PCS-SOP8-...170255?hash=item33e3528d4f:g:uEgAAOxylhxRFxkk
In simple words you linked mono to stereo adapters whereas mono to mono adapters needed.

Regards
Sachin

 

[wyn palmer]

Looks like plug-n-play.

You're saying I'd need a SOIC also ? What's a SOIC ?[/QUOTE]

I just meant that you need to have the chips in SO (Small Outline) type packages.
I'm still confused as to your needs. Sachin seems to feel that you are asking for the adaptation for a single opamp SO IC to a DIP, whereas I was under the impression that your need is for the adaptation of two single opamp SO packages to a dual.

 

[absolon]

This preamp needs 4 mono opamps. You will need single opamp on the browndog adapter, and that adapter is different than you posted. You will need to use these adapters https://www.ebay.com/itm/2PCS-SOP8-...170255?hash=item33e3528d4f:g:uEgAAOxylhxRFxkk
In simple words you linked mono to stereo adapters whereas mono to mono adapters needed.

Regards
Sachin

He's piggybacking on this thread to ask about using the 797 chips in a DAC, not in the phono preamp. Presumably, duals are needed in his application.

 

[BigElCat]

He's piggybacking on this thread to ask about using the 797 chips in a DAC, not in the phono preamp. Presumably, duals are needed in his application.

That's correct. In my first post, I included a link to a commercial product.

https://www.cimarrontechnology.com/browndog-020302-single-to-dual-op-amp-adapter

It has two 797 chips on an 8 pin dip adaptor. It looks like it would be plug-n-play for my DAC. I could use either 797A or 797B.

I guess either one would work, but I'm not sure.

WYN had given me the impression, when he first joined AK, that the 797 would be an excellent choice for digital sound.

What say ye? A or B ?

 

[wyn palmer]

That's correct. In my first post, I included a link to a commercial product.

https://www.cimarrontechnology.com/browndog-020302-single-to-dual-op-amp-adapter

It has two 797 chips on an 8 pin dip adaptor. It looks like it would be plug-n-play for my DAC. I could use either 797A or 797B.

I guess either one would work, but I'm not sure.

WYN had given me the impression, when he first joined, that the 797 would be an excellent choice for digital sound.

What say ye?

A or B probably doesn't matter. That's a matter of specs that I would consider to be largely irrelevant for all but a few audio applications- parameters like lower offset, better controlled/higher transimpedance, for example. A or B doesn't matter for the AD797MCpre, for example.
What matters is the gain configuration of the amps and what they are used for, and that depends on the application details.
Do you know how the amps are used- for example as an inverting or non-inverting gain stage of what gain- or perhaps as an i-v converter?
Alternatively, you can just try them out.
Wyn

 

[BigElCat]

Do you know how the amps are used- for example as an inverting or non-inverting gain stage of what gain- or perhaps as an i-v converter?

I'm more of a 'stick a weiner in the warp drive, and call it a day' sort of guy.


I need three of these things, about $90. Is it worth the risk ?

 

[wyn palmer]

I'm more of a 'stick a weiner in the warp drive, and call it a day' sort of guy.

In that case - just go for it!
Better your weiner than mine ;-).

 

[tabarddn51]

Hi Wyn, Thanks for the comment on my AD797SQ purchase. I frankly don't not have the technical electronic's knowledge to check them out. I am quite certain that as they are from China, they are copies, but what surprised me is that they without a doubt they sounded better. Even more strange was that I was assured by the original builder that he used genuine parts!!.
Also what surprised me was the reduction of noise from the unit, very strange.
I recommended them because I know that some builders, retired like me, may not have the funds to buy genuine AD797's & when someone is contemplating building the kit, cost is for many a major consideration. Without a doubt if I had had to by genuine 797's I would not have built the kit, not willing to take the risk. (my kit building hit rate is about 2 duds to every 1 success), No matter how many good reviews the product has had people are often a little reluctant to shell out hard earned money, especially if you have to live on a tight budget.
I think they should be considered like the Russian output caps. In theory they should not sound great, but they do & so do the SQ op amps.

Cheers Derek

 

[wyn palmer]

Hi Wyn, Thanks for the comment on my AD797SQ purchase. I frankly don't not have the technical electronic's knowledge to check them out. I am quite certain that as they are from China, they are copies, but what surprised me is that they without a doubt they sounded better. Even more strange was that I was assured by the original builder that he used genuine parts!!.
Also what surprised me was the reduction of noise from the unit, very strange.
I recommended them because I know that some builders, retired like me, may not have the funds to buy genuine AD797's & when someone is contemplating building the kit, cost is for many a major consideration. Without a doubt if I had had to by genuine 797's I would not have built the kit, not willing to take the risk. (my kit building hit rate is about 2 duds to every 1 success), No matter how many good reviews the product has had people are often a little reluctant to shell out hard earned money, especially if you have to live on a tight budget.
I think they should be considered like the Russian output caps. In theory they should not sound great, but they do & so do the SQ op amps.

Cheers Derek

Thanks Derek,
however, if, for example, what you have as a replacement for a real AD797 is actually a 1970s era 741 equivalent- a 1MHz gain bandwidth product opamp with decent noise, high distortion- which seems to be what I was shipped from China - then is the fact that you like it really relevant? In addition if you recommend something that is fake, then you are enabling someone to commit a crime, and that doesn't seem right does it?
In this case there's really nothing special in the design of the circuit around the opamps, and a 741 would result in a variety of degradations in the response that you might enjoy but is hardly representative of what a 797 could do.
Can you measure, without an input signal applied, the voltages across R17 to R24- or at the very least R17,18, 21 and 22- that will almost certainly diagnose whether the opamps are 797s or not. You do have a dc meter don't you?
If you are correct and the input devices are 797s then great.
Heck, if they are not 797s I might consider sending you a few real ones as I have quite a few.

 

[HalfApt]

I ran the ADI generated model AD797 sims with the A weighting network, then ran the sims for the Leach preamp using the most recent .asc file....

Thanks Wyn!

If you're talking about the .asc file from the Leach web page, that's for the Common-Emitter circuit, not the Common-Base that I built (he didn't provide .asc files for the 2 Common-Base circuits). So your reference to "change R1 to 75 and R4 to 10k" may not be appropriate to my build.

I don't have the means to take real measurements, so I used the "Mark-I ear" method. With everything turned on and warmed up, needle lifted, turned the volume to maximum. Had to get my ears to within 6-8 inches of the speakers to just begin to hear hiss (and a little hum). An old analog RS SPL meter wouldn't even register 60dB w/mic 1" from either tweeter or woofer panels.

All-in-all, I'd say thats very quiet!

Will still do the AD797 though, and see how they stack up against one another.