I'm updating a phono section I built rather a long time ago. It's the two op-amp design, with a non-inverting 1st stage having a gain of 3.15X, followed by the inverting RIAA stage. I think it's from a Jung book. Both amps were originally 5534s, and the design used 100uF electrolytics after each stage because of the large offsets. I've seen it published with LT1028 for the front end, and that's how I'm using it now, with no coupling caps needed. I'm also using the AD845 for the RIAA section. Just curious, what would everybody recommend for op-amps in this thing, or ditch it altogether? I do have two OPA627s available. Only caveat is that I drive a fairly low impedance stepped attenuator- 7kohms.
depends on whether you make a single step amplification or a double step.
I achieved very good results with a AD797 in a one step amplification(and RIAA correction).
In a 2 step amplifier one can use the LT1028, LT1115 or AD797 as the first amplification either with or without active RIAA correction. If one uses passive correction between the two amplification step I prefer the AD797.
The second step amplification can with a good "normal" opamp which has to be able to drive de 7K attenuator. A Opa637 will suffice.
the offset is very diffcult to get rid of ecause of the high amplificationfactor. I always use the Blackgate NX Hi-Q, 47 uF, 6,3 Volts as cap. Its small, not too expensive and works like a charm.
A DC servo will give a sound degradation.
Rob
I achieved very good results with a AD797 in a one step amplification(and RIAA correction).
In a 2 step amplifier one can use the LT1028, LT1115 or AD797 as the first amplification either with or without active RIAA correction. If one uses passive correction between the two amplification step I prefer the AD797.
The second step amplification can with a good "normal" opamp which has to be able to drive de 7K attenuator. A Opa637 will suffice.
the offset is very diffcult to get rid of ecause of the high amplificationfactor. I always use the Blackgate NX Hi-Q, 47 uF, 6,3 Volts as cap. Its small, not too expensive and works like a charm.
A DC servo will give a sound degradation.
Rob
depends on whether you make a single step amplification or a double step.
I achieved very good results with a AD797 in a one step amplification(and RIAA correction).
In a 2 step amplifier one can use the LT1028, LT1115 or AD797 as the first amplification either with or without active RIAA correction. If one uses passive correction between the two amplification step I prefer the AD797.
The second step amplification can with a good "normal" opamp which has to be able to drive de 7K attenuator. A Opa637 will suffice.
the offset is very diffcult to get rid of ecause of the high amplificationfactor. I always use the Blackgate NX Hi-Q, 47 uF, 6,3 Volts as cap. Its small, not too expensive and works like a charm.
A DC servo will give a sound degradation.
Rob
I achieved very good results with a AD797 in a one step amplification(and RIAA correction).
In a 2 step amplifier one can use the LT1028, LT1115 or AD797 as the first amplification either with or without active RIAA correction. If one uses passive correction between the two amplification step I prefer the AD797.
The second step amplification can with a good "normal" opamp which has to be able to drive de 7K attenuator. A Opa637 will suffice.
the offset is very diffcult to get rid of ecause of the high amplificationfactor. I always use the Blackgate NX Hi-Q, 47 uF, 6,3 Volts as cap. Its small, not too expensive and works like a charm.
A DC servo will give a sound degradation.
Rob
The LM4562 will easily drive the stepped attenuator.
I have used the LT1115, LT1028 and NE5534 and as well, the lowly LM833. I haven't used the AD797 for a phono preamp -- just for noise measurements and in the super-reg.
By the way, we had a great time driving upstate -- took the Ontario Parkway across to Lockport, then over to PA and OH.
I have used the LT1115, LT1028 and NE5534 and as well, the lowly LM833. I haven't used the AD797 for a phono preamp -- just for noise measurements and in the super-reg.
By the way, we had a great time driving upstate -- took the Ontario Parkway across to Lockport, then over to PA and OH.
MM or MC?
IIRC.... the AD797 has back to back diodes which limits the input of .6v max... which might play havok if used in the second stage.
Also, for lowest johnson noise, it's best suited with low impedance input... so it's probably not well suited for the first stage if using it with MM cartridges.
I would love to see something really beautiful using the LT chips mentioned earlier.
IIRC.... the AD797 has back to back diodes which limits the input of .6v max... which might play havok if used in the second stage.
Also, for lowest johnson noise, it's best suited with low impedance input... so it's probably not well suited for the first stage if using it with MM cartridges.
I would love to see something really beautiful using the LT chips mentioned earlier.
Thanks all! This is a mm preamp, and right now it sounds very good with the LT1028 front end (flat) and AD845 in the active RIAA stage. My guess is it could be better with the National parts in the RIAA stage (I'm ordering some tonight), and I might as well try the OPA627 since I have 'em. I used to use that part in atomic force microscope designs, and it has some really good noise and offset properties that were equaled by nothing else with the same bandwidth, but I'm not sure that translates to the ultimate audio opamp- those new National parts look really impressive, especially for the money. BTW, the reason I use a low value stepped volume control is I don't use a buffer afterwards, and usually only go through about ten inches of cable to a digital interface.
Using the Linkwitz inverse RIAA network, the preamp is within a few tenths of a dB of the right curve. I found the Hagerman network with the additional HF time constant, and it's also within a few tenths of that, though I could make it better by fiddling the values a bit. No idea if it's worth it, since the differences are up at 20kHz, and I can't hear much above 15! edit- cart is an Ortofon OM30.
Using the Linkwitz inverse RIAA network, the preamp is within a few tenths of a dB of the right curve. I found the Hagerman network with the additional HF time constant, and it's also within a few tenths of that, though I could make it better by fiddling the values a bit. No idea if it's worth it, since the differences are up at 20kHz, and I can't hear much above 15! edit- cart is an Ortofon OM30.
Hi Conrad,
you mentioned the LT1028 in other threads too.
Perhaps you have some advice for me.
I have experienced random oscillations at 1MHz and can't get rid of them. Maybe they are no oscillations at all, but some sort of noise pickup.
Due to the random character I can't trace it down and any circuit changes which seems effective, don't work some times later anymore.
Is there a general issue with LT1028? I have heard collegues reporting problems and the "mystery" oscillations showed up in different LT1028 circuits and even in the high gain, wide bandwith instrumentation amplifier from the datasheet I tried to build.
regards
you mentioned the LT1028 in other threads too.
Perhaps you have some advice for me.
I have experienced random oscillations at 1MHz and can't get rid of them. Maybe they are no oscillations at all, but some sort of noise pickup.
Due to the random character I can't trace it down and any circuit changes which seems effective, don't work some times later anymore.
Is there a general issue with LT1028? I have heard collegues reporting problems and the "mystery" oscillations showed up in different LT1028 circuits and even in the high gain, wide bandwith instrumentation amplifier from the datasheet I tried to build.
regards
IMO, the LT1028 isn't any more susceptible to oscillations than anything else, but it is a bit unusual. One of the ways it achieves its low noise is via high current through the front end. It's also sort of pointless to use high value (high noise) resistors with it. Thus, I bypass heavily, with 47uF Oscon caps right up close to the part. I'm also using 475 ohms feedback and 221 ohms to ground. Some opamps give oscillation trouble if you put an RF or loading cap right from the non-inverting input, to ground. I've never understood this, but a few hundred ohms series resistance on the ni input to isolate the cap will help. Naturally you've fueled my general paranoia, so now I have to go back and check the circuit more carefully!
Put in my Digikey order yesterday, so I should have half a tube of LME49710 parts, plus a mess of polypropylene caps, arriving tomorrow.
Last night I compared the AD825 and OPA627 in the RIAA stage. It was hot, so I had a lot of background noise from fans, but have to agree with other posts about the AD825 sounding aggressive, and the OPA627 having a smoother wider soundstage. I don't understand why they should sound much different at all, and offer the caveat that I need to check out the 627 on the bench. It's always risky to pop a new opamp in a socket and reach conclusions without checking the performance. There may also be some issues with offset affecting stages further down the line. I also have a vestigial comp cap from the 5534s, and I need to be sure it isn't affecting the offset pins of any other amp.
Put in my Digikey order yesterday, so I should have half a tube of LME49710 parts, plus a mess of polypropylene caps, arriving tomorrow.
Last night I compared the AD825 and OPA627 in the RIAA stage. It was hot, so I had a lot of background noise from fans, but have to agree with other posts about the AD825 sounding aggressive, and the OPA627 having a smoother wider soundstage. I don't understand why they should sound much different at all, and offer the caveat that I need to check out the 627 on the bench. It's always risky to pop a new opamp in a socket and reach conclusions without checking the performance. There may also be some issues with offset affecting stages further down the line. I also have a vestigial comp cap from the 5534s, and I need to be sure it isn't affecting the offset pins of any other amp.
Speaking of oscillations -- I have noted that with the AD797 as you null the offset the opamp will start to sing -- in the case of a design I was working on yesterday it was modulating a 1kHz sine wave at 6.6 MHz --- so I tamed it with 100R in series with 56pF parallel to Rf. (Remember to read the datasheet to the end.)
Mike, no experience with the OPA27, but have used lots of the old OP27 and OP37. These seem to be modernized souped up versions of the old parts. The old ones were really optimized for DC performance, with decent BW on the 37, but I don't know how good they were for audio. My guess is the new ones will be better all around, but the only way to know for sure is get a couple and test 'em!
Even though my parts aren't here yet, I couldn't keep my hands off the preamp and did some more comparisons. I don't think the LT1028 is the best choice for the first stage. Yes, it has spectacular noise performance, but a mm phono amp isn't that challenging noise-wise. OTOH, because of the eq on the incoming signal, I think the AD845 is a better choice- it's strong suite is quick settling, and it seems to sound cleaner. I did have to select them for low offset. The OPA627 is the winner so far in the RIAA half of the thing, but I have really high hopes for the new National parts.
Right now, I think the preamp is ever so slightly harsh. I still have a couple silver micas in the eq section, and DA on silver micas isn't great. OTOH, I can't really account for any of this on the test bench. Differential comparisons between the channels, using different opamps, shows some differences in recovery time after an edge. These are in a frequency range that might matter, not up in the tens of kHz, but they're subtle, as are the sonic differences, if they exist at all. The builder is probably the worst person to evaluate the circuit!
Even though my parts aren't here yet, I couldn't keep my hands off the preamp and did some more comparisons. I don't think the LT1028 is the best choice for the first stage. Yes, it has spectacular noise performance, but a mm phono amp isn't that challenging noise-wise. OTOH, because of the eq on the incoming signal, I think the AD845 is a better choice- it's strong suite is quick settling, and it seems to sound cleaner. I did have to select them for low offset. The OPA627 is the winner so far in the RIAA half of the thing, but I have really high hopes for the new National parts.
Right now, I think the preamp is ever so slightly harsh. I still have a couple silver micas in the eq section, and DA on silver micas isn't great. OTOH, I can't really account for any of this on the test bench. Differential comparisons between the channels, using different opamps, shows some differences in recovery time after an edge. These are in a frequency range that might matter, not up in the tens of kHz, but they're subtle, as are the sonic differences, if they exist at all. The builder is probably the worst person to evaluate the circuit!
Conrad Hoffman said:
Try these guys for polystyrene caps -- http://www.surplussales.com/Capacitors/Poly-Unelco.html
Conrad Hoffman said:
I have a 1/2 dozen of them so I decided to build a board. The specs looked impressive as far as mating with my MC cartridge. Actually I was also debating trying some of the new National parts as well. I'm using the passive/active design from their application note for the LM4562. Decided to try it as a clean slate option to my age old discrete design, which hasn't fared well as my amp and preamp have evolved. I'm assuming the LME49710 would be a drop in swap in the same circuit. Keep us posted as to what they sound like.
I hesitated jumping into the National chip because of it's high-end pretentions... Normally this is a code phrase for "choke the life out of the sound with ultra low distortion" , thought I'd try the OPA27s first since they were already at hand.
It might make sense to use the OPA27's in the first stage and the Nationals in the second. The spec sheet on them indicate they are best, noisewise, with a low impedance source. Any thoughts?
Regards, Mike.
sorry but the LT1028 won't be the noise winner with mm cart, the ~5K coil resistance turns the heavily biased bipolar Q input noise current into much higher noise than you would have with a decent fet input op amp
http://www.diyaudio.com/forums/showthread.php?s=&postid=1193893&highlight=#post1193893
http://www.diyaudio.com/forums/showthread.php?s=&postid=1193893&highlight=#post1193893
jcx, I thought about the same thing, but thought mm source Z was lower. What I don't understand is how to evaluate it in terms of the noise formula. My Ortofon is only about 700 ohms DC resistance, but that's not (I think) the output impedance for an AC signal, and I'm sure it varies over the audio range. I could measure it, but fear damaging the cartridge. How much voltage do you thing those itty bitty wires could stand? If I could put a couple tenths of a volt across it, I could sweep the audio range and plot a curve. Or is the Z looking in the same as looking out? Though they don't say, I assume the LME49710 is a well biased bipolar, like the LT1028, based on the input current numbers, but aimed at a little higher impedance than the 1028. I think the OPA627 is down in the low pA, and little will beat it for higher Z applications. National is pretty quiet about the process used for the 49710 series, but this weekend will tell the tale soundwise. I also want to try the passive eq shown in their data sheet. BTW, nobody talks about the AD845, but it has a good combination of specs- what do you think of it?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.