i don't know if it's a problem but i have just checked the 20 boards again for voltages.
board1 (right)
IC pins 4&7 = 22v
In +/- = 0v
out +/- = -83mv
board2 (left)
IC pins 4&7 = 29.1v
In +/- = 0v
out +/- = -101mv
board1 (right)
IC pins 4&7 = 22v
In +/- = 0v
out +/- = -83mv
board2 (left)
IC pins 4&7 = 29.1v
In +/- = 0v
out +/- = -101mv
Your offsets are a little higher than I typically see, but not dangerously so.
The AD797* has max input offset voltage of 80 microvolts, so the offset should be a fair bit lower than what you measure. I am a little worried about high frequency oscillations. The AD797 datasheet is very clear about the importance of ceramic bypass caps next to the package to keep this op amp stable, advice you have blatantly disregarded.
Yes, one symptom of high frequency instability is higher than expected DC offsets.
You can check the offset voltage at pin 6, rather than the Sapphire output. If this is also 100 mV or so, its the op amp and not the diamond buffer that is generating the offset, and for an AD797 it means the op amp is out of spec and needs attention.
*iirc you substituted this one for the standard OPA134.
The AD797* has max input offset voltage of 80 microvolts, so the offset should be a fair bit lower than what you measure. I am a little worried about high frequency oscillations. The AD797 datasheet is very clear about the importance of ceramic bypass caps next to the package to keep this op amp stable, advice you have blatantly disregarded.
Yes, one symptom of high frequency instability is higher than expected DC offsets.
You can check the offset voltage at pin 6, rather than the Sapphire output. If this is also 100 mV or so, its the op amp and not the diamond buffer that is generating the offset, and for an AD797 it means the op amp is out of spec and needs attention.
*iirc you substituted this one for the standard OPA134.
Last edited:
You mean without the op amp? That won't give you a realistic value, you'll need to have the op amp in the circuit, the diamond buffer input is essentially disconnected and floating otherwise.
ok i popped the opa134 opamps in and i got very wide differences so i juggled them about and i now have
board 1 out = +13.1mv
board 2 out = +12.9mv
i was getting values from -21.9mv to +13.1mv so some of my opamps must be dodgy.
board 1 out = +13.1mv
board 2 out = +12.9mv
i was getting values from -21.9mv to +13.1mv so some of my opamps must be dodgy.
ok i done as recommended and installed the bypass caps on the opamps.
there is better/cleaner separation of individual instruments and notes with the bypass caps on the opamp sockets but vocals have taken on a slight spittle sound. this my be due to them bedding in.
i'm getting there and i do appreciate your patience Richard 🙂
there is better/cleaner separation of individual instruments and notes with the bypass caps on the opamp sockets but vocals have taken on a slight spittle sound. this my be due to them bedding in.
i'm getting there and i do appreciate your patience Richard 🙂
The OPA134 doesn't need bypass caps, I was talking about the AD797, a completely difference beast. Sorry, earlier I mistakenly though you were using that op amp.
You should give the Zobel filter a try.
You should give the Zobel filter a try.
Richard, i dont know if this is useful to you but i done some opamp swapping last night and here are the results of the output offset.
TI OPA134 range from -20/+20mv
TI OPA132 range from -20/+20mv
TI OPA551 i only have 2 chips but both were +0.1mv
Sig NE5534 range from +400/+600mv
AD 797 i only have 2 chips +80/+100
TI OPA134 range from -20/+20mv
TI OPA132 range from -20/+20mv
TI OPA551 i only have 2 chips but both were +0.1mv
Sig NE5534 range from +400/+600mv
AD 797 i only have 2 chips +80/+100
Hi! I'm Repente from Spain 🙂 (Auriculares HIFI)
Sorry in advance for my low level of the languaje.
I am going to make a Sapphire; I have allready buy the kit and the rest of the parts needed.
I will post my progress and my questions 😉
Greetings!!
Sorry in advance for my low level of the languaje.
I am going to make a Sapphire; I have allready buy the kit and the rest of the parts needed.
I will post my progress and my questions 😉
Greetings!!
nice to see another builder of the Sapphire.
small mistake in values. OPA551 should have read OPA177 and values are -0.1&-0.2mv
lets see if i have this right, the closer you get the output offset of the opamp in circuit to 0mv the lesser the high frequency oscillation?
so if you have a number of opamps to choose from you want to pick the ones that have a value of plus or minus closest to 0mv?
No, its not like that at all. Also, I realised that the AD797 is naturally going to have a large DC offset when used in the Sapphire circuit. In short - forget I ever said anything about oscillation. (Though really, when using the AD797, you should have the ceramic bypass caps.)
The AD797 has super-low voltage noise, and to get that the input currents are really large. If you stick it in a circuit with a large imbalance of the DC input impedances (the sapphire puts 100k on IN+, and just 1k on IN-) relatively large output offset will be generated. Same story with the NE5534. Maybe the NE5534A (the higher tolerance part) might give better results, I don't remember getting such high offsets when I tried them. FET input opamps will tend to give much better DC offset performance, since the input currents are several order of magnitude smaller.
There is no reason to cherry pick op amps for low offset, but at the same time the AD797 and NE5534 of yours are a bit too high to recommend.
The AD797 has super-low voltage noise, and to get that the input currents are really large. If you stick it in a circuit with a large imbalance of the DC input impedances (the sapphire puts 100k on IN+, and just 1k on IN-) relatively large output offset will be generated. Same story with the NE5534. Maybe the NE5534A (the higher tolerance part) might give better results, I don't remember getting such high offsets when I tried them. FET input opamps will tend to give much better DC offset performance, since the input currents are several order of magnitude smaller.
There is no reason to cherry pick op amps for low offset, but at the same time the AD797 and NE5534 of yours are a bit too high to recommend.
Here it's the official (in spanish) Shappire post :-D
El hilo del "Shapphire amp"
I have allready finish the board work and now I'm waiting por the Toroidys and the rest of the components.
I'm afraid with the "tune-up" part. I'm beginner in diy proyect and the terms of "bias" or "offset" makes me crazy jejej.
I will connect everything, plug my cheat Koss KSC75 and pray... 🙂
El hilo del "Shapphire amp"
I have allready finish the board work and now I'm waiting por the Toroidys and the rest of the components.
I'm afraid with the "tune-up" part. I'm beginner in diy proyect and the terms of "bias" or "offset" makes me crazy jejej.
I will connect everything, plug my cheat Koss KSC75 and pray... 🙂
So Sapphire becomes "Shapphire" in Spanish. Who knew?
Lucky you: The Sapphire has no "bias" or "offset" adjustments. I've already taken care of all that at the design stage. The kit/BOM is what it is, everything is preset and the bias is fixed at about 28 mA. The offset is "luck of the draw" but small enough with the OPA134 to be irrelevant. The only thing you can adjust if you want is the gain, by changing the value of R4.
Lucky you: The Sapphire has no "bias" or "offset" adjustments. I've already taken care of all that at the design stage. The kit/BOM is what it is, everything is preset and the bias is fixed at about 28 mA. The offset is "luck of the draw" but small enough with the OPA134 to be irrelevant. The only thing you can adjust if you want is the gain, by changing the value of R4.
Argg, sorry! I'm very bad writing in english 🙂
Sapphire it's sapphire here and everywhere. "Zafiro" would be in spanish but "Sapphire" looks more cool ;-)
Then, there's no bias tune up? Reading older posts and your manual y thought it was.
Zobel filter have no adjustement too? (I was solder everything opcional, zobel filter and all capacitators).
Here it some photos; many thanks and greetings!!
Sapphire it's sapphire here and everywhere. "Zafiro" would be in spanish but "Sapphire" looks more cool ;-)
Then, there's no bias tune up? Reading older posts and your manual y thought it was.
Zobel filter have no adjustement too? (I was solder everything opcional, zobel filter and all capacitators).
Here it some photos; many thanks and greetings!!
Thanks for the photos!
Originally I looked at many different bias settings, before settling on this one. So, no, there are no adjustments needed now.
BTW the heatsinks can be used the other way up so the resisters nearby have more clearance.
Originally I looked at many different bias settings, before settling on this one. So, no, there are no adjustments needed now.
BTW the heatsinks can be used the other way up so the resisters nearby have more clearance.
Last edited: