Hello all,
I've got an excellent BJT DOA in Sam Groner's SGA-SOA-v2 Simple Discrete Amplifier on an improved PCB.
I've also got an improved version of the MOX FET DOA featured here on DiyAudio: http://www.diyaudio.com/forums/analog-line-level/27651-mox-active-crossover.html.
The MOX DOA can use a legacy Toshiba 2SK389 part, the NEC uPA68H part or the new Linear Systems LSK389 part. All three work especially well in this design.
Both DOAs are packaged in the popular API 2520 plug format as used by the Hardy 990 DOA and many others.
I've been experimenting with DOA blocks for a while. I have a nice two channel preamplifier with a phono input section with balanced as well as RCA outputs that is great for vinyl systems. I also have Sam's excellent headphone amplifier running as well as other goodies in the works.
Anyone interested or is what I am describing too 'old school'?
I've got an excellent BJT DOA in Sam Groner's SGA-SOA-v2 Simple Discrete Amplifier on an improved PCB.
I've also got an improved version of the MOX FET DOA featured here on DiyAudio: http://www.diyaudio.com/forums/analog-line-level/27651-mox-active-crossover.html.
The MOX DOA can use a legacy Toshiba 2SK389 part, the NEC uPA68H part or the new Linear Systems LSK389 part. All three work especially well in this design.
Both DOAs are packaged in the popular API 2520 plug format as used by the Hardy 990 DOA and many others.
I've been experimenting with DOA blocks for a while. I have a nice two channel preamplifier with a phono input section with balanced as well as RCA outputs that is great for vinyl systems. I also have Sam's excellent headphone amplifier running as well as other goodies in the works.
Anyone interested or is what I am describing too 'old school'?
Last edited:
You're probably not going to find too many other people with API 2520 style DOAs here, or at least I don't think we've got a lot of pro audio people (even if Samuel Groner actually is a member if memory serves). Doesn't mean they aren't interesting, of course. One cool thing about the API 2520 pinout is that it features a dedicated ground connection. That certainly gives you an extra degree of liberty.
Samuel Groner is a member here on DiyAudio. His SGA_SOA-v2 DOA is a very clean and dynamic gain block on par with the best.
Do you have any of the un populated SGA-SOA-v2 Simple Discrete Amplifier PCB for sale or know where I can get them?
Thanks
John C.
Some of the studio-gear builders favor old-school modular opamps. A thread on such a forum may bring lots of orders.
> the popular API 2520 plug format .... this 'old school' modular stuff.
Before API, this was a Burr-Brown (or Philbrick??) opamp module footprint.
> Location: Bakersfield, CA
Last time I was in Bakersfield, an opamp was a yellow casting with two 12AX7 on top and an octal on the bottom.
Before API, this was a Burr-Brown (or Philbrick??) opamp module footprint.
> Location: Bakersfield, CA
Last time I was in Bakersfield, an opamp was a yellow casting with two 12AX7 on top and an octal on the bottom.
Last edited:
FWIW - I've built quite a few DOAs from the various kits and PWBs that have been offered.
My Sapphire 3 headphone amp is running a pair of WhistleRock BLU-18s. My DOA SeNNator runs Capi-gear's JE-918 PWBs using 2N4403s and BC337s, and they feed a pair of Hardy 990c DOAs.
In my stash are Sam Groner SGA-SOA-2, Pier Paolo Abbate APP992, Hairball Audio/MNATS 990, JLM Audio Hybrid, JLM Audio JLMHV25, Cappi-gear 2520BC, Capi-gear BB2521, FiveFish DOA-17, WhistleRock ML-918, and various SOIC-8/Dip8-to-DOA Adapters.
My Sapphire 3 headphone amp is running a pair of WhistleRock BLU-18s. My DOA SeNNator runs Capi-gear's JE-918 PWBs using 2N4403s and BC337s, and they feed a pair of Hardy 990c DOAs.
In my stash are Sam Groner SGA-SOA-2, Pier Paolo Abbate APP992, Hairball Audio/MNATS 990, JLM Audio Hybrid, JLM Audio JLMHV25, Cappi-gear 2520BC, Capi-gear BB2521, FiveFish DOA-17, WhistleRock ML-918, and various SOIC-8/Dip8-to-DOA Adapters.
And what are your thoughts as to the sound quality of DOAs versus typical audio op amps? Is the added expense worth it??
There is a perceived difference in the noise floor. DOAs seem to have more darkness. I wrote up several listening impressions in the Sapphire thread when trying out several different DOAs and IC opamps in the primary gain position. Posts 589, 590 - page 53, post 597 - page 60, post 603 - page 61, post 623 - page 63, and post 665 - page 67.
Richard (RJM) wrote up an impression of swapping a Sonic Imagery Labs 995FET in post 973 through post 979 on page 98.
In general, the closest IC to a DOA is the AD797, in that specific use. Some caveats to note - the rail voltage was +-10.5V at the opamp pins. It made for a disadvantage for the DOAs, especially the JE-918.
When I built the SeNNator, the sonic signature of the shelving caps masked perceived difference in the opamps. But, given it was specifically designed to feed Sennheiser HD650s, the DOAs stay in Class A the vast majority of the time due to their 15ma to 17ma idle current on the output stage.
My next project will be a simplified version of Doug Self's preamp and I'll probably mix ICs for the input stage and use DOAs for the gain/ouput stage.
I like Capi-gear's JE-918 PWB. It's cheap and allows me to select the transistors. While laid out for BC550/560, quieter transistors like selected 2N4403, BC327, or KSA992 can be used for the PNP And BC337 or KSC1845 for the NPNs.
So is the expense worth it? Perhaps. More so in headphone amps, perhaps in near-field monitoring situations. And anytime one needs to cleanly drive 150 ohms or 600 ohms.
They are worse in cases where DC coupling is involved - I haven't seen one that has bias compensation. So either a cap or DC offset adjustment or DC servo is needed.
The other downside is they idle at 25 ma. It's good that the regulators have a steady load - ie: LM317/LM337 should have at least 45ma loading. But four of them draw 100ma from a power supply during idle. If one believes that they will use max power, that's 260ma per DOA.
I wouldn't want to build an entire DOA Doug Self preamp like the IC version you've done. But I believe they have their place in such a design.
For instance, in Doug's preamp, A3a, the tone-control gain stage, and A6b - eliminating A7a in the volume control stage would be places for DOAs.
Richard (RJM) wrote up an impression of swapping a Sonic Imagery Labs 995FET in post 973 through post 979 on page 98.
In general, the closest IC to a DOA is the AD797, in that specific use. Some caveats to note - the rail voltage was +-10.5V at the opamp pins. It made for a disadvantage for the DOAs, especially the JE-918.
When I built the SeNNator, the sonic signature of the shelving caps masked perceived difference in the opamps. But, given it was specifically designed to feed Sennheiser HD650s, the DOAs stay in Class A the vast majority of the time due to their 15ma to 17ma idle current on the output stage.
My next project will be a simplified version of Doug Self's preamp and I'll probably mix ICs for the input stage and use DOAs for the gain/ouput stage.
I like Capi-gear's JE-918 PWB. It's cheap and allows me to select the transistors. While laid out for BC550/560, quieter transistors like selected 2N4403, BC327, or KSA992 can be used for the PNP And BC337 or KSC1845 for the NPNs.
So is the expense worth it? Perhaps. More so in headphone amps, perhaps in near-field monitoring situations. And anytime one needs to cleanly drive 150 ohms or 600 ohms.
They are worse in cases where DC coupling is involved - I haven't seen one that has bias compensation. So either a cap or DC offset adjustment or DC servo is needed.
The other downside is they idle at 25 ma. It's good that the regulators have a steady load - ie: LM317/LM337 should have at least 45ma loading. But four of them draw 100ma from a power supply during idle. If one believes that they will use max power, that's 260ma per DOA.
I wouldn't want to build an entire DOA Doug Self preamp like the IC version you've done. But I believe they have their place in such a design.
For instance, in Doug's preamp, A3a, the tone-control gain stage, and A6b - eliminating A7a in the volume control stage would be places for DOAs.
Interesting that you mention this!
I have been working on the same thing pretty much as you describe your project. In my case I crafted a custom input section and (like you suggested) went out of my way to kill off modern op amps. And I added a DOA phono section.
The only modern op amps left are in the tone control section as shown below and those can be switched out of the audio path.
I am using the the Samuel Groner headphone amp as the output section.
And while I am at it I decided to spin up an improved version of a series regulated power supply. This one is adjustable from +/-9 volts to +/-24 volts.
This is my current project on my dev bench. I could report on it's progress should there be an interest in me doing that.
For anyone interested ...
Here is a PIC of the Sam Groner v2 DOA
And here is a PIC of his excellent headphone amplifier that uses 2 of his DOAs
Here is a PIC of the Sam Groner v2 DOA
And here is a PIC of his excellent headphone amplifier that uses 2 of his DOAs
Ah Nice!
FWIW - I don't drop a DIP8 footprint because adapters are easy to come by. See Capi-gears - IC DOA PCB or Collective Cases' DIP8-to-2520 PCB.
And I have a ton of not-produced anymore WhistleRock adapters. I do see the issue - it adds $2.50 to $5.00 to each DOA socket. However, I argue if people wanted to use ICs, then there is the original kit. 🙂
Also, my crude work in documented in the blogs with pictures of the DOA SeNNator headphone amp.
I got tired of drilling holes and buying transformers and such, so I just re-designed Collective Case's DC-filter and regulator PWB, switched to newer regulators and a resistor change could lower the voltage from +-15V, plus there's the 48V for those discrete circuits and microphones...
FWIW - I don't drop a DIP8 footprint because adapters are easy to come by. See Capi-gears - IC DOA PCB or Collective Cases' DIP8-to-2520 PCB.
And I have a ton of not-produced anymore WhistleRock adapters. I do see the issue - it adds $2.50 to $5.00 to each DOA socket. However, I argue if people wanted to use ICs, then there is the original kit. 🙂
Also, my crude work in documented in the blogs with pictures of the DOA SeNNator headphone amp.
I got tired of drilling holes and buying transformers and such, so I just re-designed Collective Case's DC-filter and regulator PWB, switched to newer regulators and a resistor change could lower the voltage from +-15V, plus there's the 48V for those discrete circuits and microphones...
BTW - would that input stage and phono preamp fit the back panel of a Hammond 1455T2202 enclosure? (6.6" wide by 2.19" tall)
I ask as I could do the line stage on a 160mm x 50 mm PWB with two channels and add the phono/input and my DC-filter/Reg PWB and avoid too much chassis work - I suck at chassis work, even worse than SY!
I ask as I could do the line stage on a 160mm x 50 mm PWB with two channels and add the phono/input and my DC-filter/Reg PWB and avoid too much chassis work - I suck at chassis work, even worse than SY!
The Phono IO board as shown measures 190mm by 66mm and is designed to fit in a 1U chassis.
- Status
- Not open for further replies.