I'm thinking to prepare another RIAA pre-amp for MM cart with option for flat operation with ~+30dB gain. This pre-amp is based on example found on LT1115 data sheet + Andy C's RIAA stage using LT1028. I made some changes and simulations with the combined phono -schematic to find suitable changes needed to get original schematic work for flat operation as well. Ended to these changes:
RIAA EQ
RIAA stage itself is based on the one found from here - http://andyc.diy-audio-engineering.o...o_preamp.xhtml (C/R values can be found from this site). I did change the C4 to 470uF for this stage I'm working with.
Flat pre-amp (By simulation, Gain = ~+30.0dB ((1Hz) 20Hz - 20kHz (>100kHz)) , Phase < +-0.5deg ((1Hz) 20Hz - 20kHz)):
R1 = 54.5kohm
Disable R4, R7, R8 and C1, C2
Am I going in the woods with these "thoughts"? Is the LT1028 good selection (least it's not the cheapest one).
RIAA EQ
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
RIAA stage itself is based on the one found from here - http://andyc.diy-audio-engineering.o...o_preamp.xhtml (C/R values can be found from this site). I did change the C4 to 470uF for this stage I'm working with.
Flat pre-amp (By simulation, Gain = ~+30.0dB ((1Hz) 20Hz - 20kHz (>100kHz)) , Phase < +-0.5deg ((1Hz) 20Hz - 20kHz)):
R1 = 54.5kohm
Disable R4, R7, R8 and C1, C2
Am I going in the woods with these "thoughts"? Is the LT1028 good selection (least it's not the cheapest one).
Last edited:
jiitteepee fancy meeting you here, PM me and I will send you some AD797's for your project.
My suggestion would be to augment the LT1028 with a unity gain buffer (LT1010 might be a suitable choice). For a couple of reasons - first to remove any thermal interactions between the output drive requirements and the input stage. Second to preserve dynamics - the load seen by the LT1028's OPS isn't particularly benign - 0.1uF and 10R in series. Having a separate buffer allows separate decoupling of the supplies and means the LT1028 can work with much cleaner rails.
<edit> On re-reading your post I see you're intending 'flat' operation, meaning no EQ at all? In which case the OPS is going to see a much easier load with those largish fb caps deleted. But then again the output when working 'flat' will be tilted up to HF so dynamics are still a major issue vis-a-vis the supply noise.
<edit> On re-reading your post I see you're intending 'flat' operation, meaning no EQ at all? In which case the OPS is going to see a much easier load with those largish fb caps deleted. But then again the output when working 'flat' will be tilted up to HF so dynamics are still a major issue vis-a-vis the supply noise.
Last edited:
Hi Scott!
Thanks for the kind offering but, I'm sticking with this 1028 because of I already have handful of them to use, it's model is bundled with the LTSpice IV so I can make simulations before adding features (this project is more like an experiment / learning happening for me ... ) and it's not very commonly used op-amp in pre-amp design (least I have not seen many).
Thanks for the info. I added the schematics into my 1st post. Is the LT1010 placed OK there?
About the power.
I was going to use 12V battery as power source. I need to use a dual output converter to get +-12V. Is the Meanwell DCW03-12 good enough as used on phono preamp project here?
I was going to use 12V battery as power source. I need to use a dual output converter to get +-12V. Is the Meanwell DCW03-12 good enough as used on phono preamp project here?
Do you mean I should just send copy of the LT1010 output back to its input or to LT1028 input through some resistor (10k?) or something more complex (by documentation, LT1010 is an unity gain buffer and has less pins to use compared to say LT1028).
This topic has a long history and there are numerous DIY audio threads on using LT opamps in RIAA preamps.
See here. LT1115 - Ultra-Low Noise, Low Distortion, Audio Op Amp - Linear Technology
See here. LT1115 - Ultra-Low Noise, Low Distortion, Audio Op Amp - Linear Technology
still doesn't make low noise bjt Inoise any better with mm coil's high Z
~1 nV heavy bias bjt inputs LT1028/AD797 are really only OK with low Z source less than few 100 Ohms
Scott's AD743/745 would still be the winner - if AD still wanted to make them (seems they do but under protest Product Status:Not Recommended for New Designs)
~1 nV heavy bias bjt inputs LT1028/AD797 are really only OK with low Z source less than few 100 Ohms
Scott's AD743/745 would still be the winner - if AD still wanted to make them (seems they do but under protest Product Status:Not Recommended for New Designs)
OK, I did some changes to the schematic (se post #1). By simulation, it looks quite good and accurate in theory (+-0.0035dB) but, ... of course, component tolerances and wrong load capacitance kills that accuracy.
Anyway, I have couple questions:
1. Is variable capacitor an option for to control the load capacitance?
2. Does computer power (ATX) as a voltage source for LT1028/LT1010 need some extra filtering?
3. Is there any advantage in building the final R/C values by using multiple components instead of lowest count (tolerances repeals?)?
Anyway, I have couple questions:
1. Is variable capacitor an option for to control the load capacitance?
2. Does computer power (ATX) as a voltage source for LT1028/LT1010 need some extra filtering?
3. Is there any advantage in building the final R/C values by using multiple components instead of lowest count (tolerances repeals?)?
Last edited:
I have no knowledge of variable caps so to jump over that to your 2nd question. I'd definitely avoid an ATX power supply for this application. Its certainly possible to filter the output enough to remove the differential mode HF hash but the common-mode is going to be a serious problem. There's also the question of how much LF noise is going to be present as ATX power supplies are designed for computers not audio and tend to be highly cost-engineered.
While on the subject of power supplies - you've lost much of the advantage of using the buffer in connecting the power supplies together. The front end opamp should get some extra filtering (LC) so its not exposed to the nasty switching noise produced by the buffer.
While on the subject of power supplies - you've lost much of the advantage of using the buffer in connecting the power supplies together. The front end opamp should get some extra filtering (LC) so its not exposed to the nasty switching noise produced by the buffer.
Hmm... does it make any difference if the pre-amp is going to be used only with PC setup (-->pre-amp --> USB sound card --> PC)?
Any suggestions ... I did add the JFET system as suggested for LT1115 based design (link given in rayfutrell's post) but does that help anything in this?
If everything runs from the same PSU then yes, it makes a difference. You won't need to worry about common-mode noise, only differential mode.
You mean suggestions for actual values and part numbers for inductors? I would only be able to do that if I was designing this for real with knowledge of the desired performance relative to the PSRR spec. In general to get the best SQ its necessary to keep HF hash off the rails as much as possible - LC filters are the best way to do this. You could start with an online LC filter designer to design a filter which gets HF rejection to match the opamp's PSRR curve.
The JFET current source is a good idea to reduce HF hash generated by the opamp itself, the LC filtering is to keep the buffer's hash out.
LOL ... both questions got solved (I'm still learning the LTSpice software ... as well as electronics). Spikes was caused by PULSE -command for another small circuit which I had left active while running the .tran simulation. Level was just too high because of V was set to 0.1V.
This is just what I thought.
I made new transient analysis and now the result looks much better AFAIK.
As I'm in learning mode, can you or someone else point me some examples showing the usage of LC filters (schematics, LTSpice project files).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.