Hey guys covids got me eager for soldering stuff and drawing stuff on eagle.
One idea ive been wanting to try is nested opamp with discrete input & output in dip8 package. Looking around it seems like theres alrdy some like that coming out of china.
I bet they perform well, given the complexity of the architecture and especially if the mystery opamp is tamed for unity gain, allowing it to be versatile.
My idea for now revolves around 5534. Ltp input going into pin 1 and 8. Then output from pin 5 to a diamond buffer like ha5033 or buf634.
Is this a good idea?
One idea ive been wanting to try is nested opamp with discrete input & output in dip8 package. Looking around it seems like theres alrdy some like that coming out of china.
I bet they perform well, given the complexity of the architecture and especially if the mystery opamp is tamed for unity gain, allowing it to be versatile.
My idea for now revolves around 5534. Ltp input going into pin 1 and 8. Then output from pin 5 to a diamond buffer like ha5033 or buf634.
Is this a good idea?
i got my idea off the ground and it sounds really good. (please excuse the botched pcb job because hopefully with your feedback the next ones gonna be much better.)
The key of the topology is squeezing the most out of the venerable 5534 because in a way to me this is the most true to the source ic there is, from most mixing gear containing the chip and therefore its coloration.
What i did to it is basically use linuxguru's output on it and a resistor fed ltp on the front with their collectors going into pin 1 and 8 of the ic thereby replacing its internal input stage. Pin 2 and 3 r tied to the neg input and an additional 100k ohm resistor added between pin 1 and 8 like they did in 6010 preamp to make the 5534 unity gain stable. Resistor tail was chose over ccs for the ltp because from what i know it gives better harmonic profile and more 'mellow' sound. So external ltp, class a biased 5534, then cascoded output. The resulting sound is very reminscent of a triode- clean and smooth. No doubt the best opamp ive used in my whammy headphone amp.
Now id like to redesign the pcb so that i can use it everywhere in my future gears and i rlly wanna use to92 packages and dip8 opamp because i just cant with smd. And besides i plan on overbiasing the output and the ic and their larger casings will fuction as heatsinks.
i plan on drawing it as a multifloor design, and tho ive seen a few of em with soic8 chips in the mix, never with a dip8 chip, and am wondering how much more the risk of oscillation will be with the bigger package.
edit: here's the output circuit https://www.diyaudio.com/forums/analog-line-level/193768-ann-lf01-discrete-hybrid-opamp-module-6.html#post4420882
edit2: my pcb
The key of the topology is squeezing the most out of the venerable 5534 because in a way to me this is the most true to the source ic there is, from most mixing gear containing the chip and therefore its coloration.
What i did to it is basically use linuxguru's output on it and a resistor fed ltp on the front with their collectors going into pin 1 and 8 of the ic thereby replacing its internal input stage. Pin 2 and 3 r tied to the neg input and an additional 100k ohm resistor added between pin 1 and 8 like they did in 6010 preamp to make the 5534 unity gain stable. Resistor tail was chose over ccs for the ltp because from what i know it gives better harmonic profile and more 'mellow' sound. So external ltp, class a biased 5534, then cascoded output. The resulting sound is very reminscent of a triode- clean and smooth. No doubt the best opamp ive used in my whammy headphone amp.
Now id like to redesign the pcb so that i can use it everywhere in my future gears and i rlly wanna use to92 packages and dip8 opamp because i just cant with smd. And besides i plan on overbiasing the output and the ic and their larger casings will fuction as heatsinks.
i plan on drawing it as a multifloor design, and tho ive seen a few of em with soic8 chips in the mix, never with a dip8 chip, and am wondering how much more the risk of oscillation will be with the bigger package.
edit: here's the output circuit https://www.diyaudio.com/forums/analog-line-level/193768-ann-lf01-discrete-hybrid-opamp-module-6.html#post4420882
edit2: my pcb
Last edited:
Guys i did a second spin of this topology and i must say its the best opamp ive ever heard in my whammy. The distortion profile, and the snr makes it totally sound like a vacuum tube. If i bias it to where the bjts are toasty the clarity of the sound becomes unbelievable. Ofcourse its also dirt cheap, stable and probably more robust than a full discrete op since the 5534's internal safeguards are all still active. Ive made this thru hole version and also smd version thats half the size. Ill post gerbers upon request. Layout is prob not optimal.
Made holes on the top board so the led shined thru. Yeh im not an engineer, more like a lego builder ��
Made holes on the top board so the led shined thru. Yeh im not an engineer, more like a lego builder ��
Last edited:
Use the 2520 footprint.
Why not just adopt the 2520 footprint?
There are opamp and hybrid kits in that format from places like capi-gear, which has
both a 5534 and LM49710 both with discrete output stages. They also have their
version of the JE-918 discrete opamp.
There is the JE-990 kit from Hairball audio, for a discrete opamp much better than the 5534.
JLM audio has several kits as well.
The format provides a standard form to make your own - 1.1" square, 5 pin.
Just stay away from the opamps the replicate the API sound from the 1970s, unless
you want that mid-range forward sound.
Why not just adopt the 2520 footprint?
There are opamp and hybrid kits in that format from places like capi-gear, which has
both a 5534 and LM49710 both with discrete output stages. They also have their
version of the JE-918 discrete opamp.
There is the JE-990 kit from Hairball audio, for a discrete opamp much better than the 5534.
JLM audio has several kits as well.
The format provides a standard form to make your own - 1.1" square, 5 pin.
Just stay away from the opamps the replicate the API sound from the 1970s, unless
you want that mid-range forward sound.
The diyAudio Store sells PCBs for a 25W Class-A power amp designed by Nelson Pass; it's called "M2x". The PCB features removable, interchangeable daughter cards which implement the input stage of the power amp. One of those interchangeable cards, called Norwood, is a composite amplifier. It consists of an AD744 (externally compensated, JFET-input opamp) followed by an HA5002 video buffer. Schematic below. People who have built and listened to the amplifier, appear to find the sound of Norwood very pleasing. In plenty of cases, more pleasing than any of the other daughter card circuit designs.
_
_
Attachments
Last edited:
If u look at my whammy board i actually incorporated that pro audio socket. But in retrospect its more of a novelty factor than else cuz of how hard it is to source those opamps in korea. And besides, class a biased ne5534 is the bees knees for me
I will draw out the schematic when i find the time.
In the next iteration id like to incorporate jung style feedback loop directly on the board. Need to read more upon how thats gonna affect the versatility of the opamp.
Last edited:
Yes Mark, id like to try that combo in the near future coupled with the input upgradability of the ne5534. A nice combo would be the buffer, smd signetics 5534 and a super pair bjt in soic package. The footprint would end up being only as big as the buffer is, or as small.
I wonder how 5002 will behave if i take the output from pin 5 of the 5534 which i believe has 2v offset?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.