The 'setting' resistor is in effect the feedback leg to ground on the regulator. It's not, as I have read, recommended to filter it if you want the regulator to behave as intended. And these are not that quiet... Try listening to the output from them..
Sent from my x600 using Tapatalk
Last edited:
Member
Joined 2009
Paid Member
Well Nico, seems that your schematic is world famous DH-101 !
http://www.hafler.com/pdf/archive/DH-101_preamp_man.pdf
Seems to be very well regarded (some complaints about the original in terms of quality of switch gear, use of carbon resistors and RIAA not as popular - but overall favourable)
There's likely a wealth of information about this pre-amp on the internet (no shortage of old ones for sale) too.
Edit: the successor: http://www.hafler.com/pdf/archive/DH-110_preamp_man.pdf
http://www.hafler.com/pdf/archive/DH-101_preamp_man.pdf
Seems to be very well regarded (some complaints about the original in terms of quality of switch gear, use of carbon resistors and RIAA not as popular - but overall favourable)
There's likely a wealth of information about this pre-amp on the internet (no shortage of old ones for sale) too.
Edit: the successor: http://www.hafler.com/pdf/archive/DH-110_preamp_man.pdf
Last edited:
It's funny you say that. I tried most of the ideas around and few were better and the ones that were seemed to have stability problems. The gain setting resistors are in the data sheet. How it seems to work is all the 78 series regulators seem to have identical noise multiplied by the voltage. Because the LM317 is the 1.25V version if you like it can have noise advantages if the filtering is used ( 10 uF I seem to remember is said to be a good compromise ).
This was the most imformative read I found in the link ( LM317 is second page ). As they people building this are not the usual DIY Audio scientist types I feel LM317 is not a bad device and was used by Naim. One thread on regulators at DIY ran into the 1000's and never got to where I would have wanted to be if I had spent a day with the problem. Please read this link if nothing else. From what I understand LM317 is a 1970's op amp of about 20nV /root Hz with 1.25V band gap reference and pass transistor. Doubtless a 1 nV device would be better. Considering it's modest spec it's not bad. I've seen a LED and op amp suggested as a DIY idea.
Simple Voltage Regulators Part 1: Noise - [English]
In 1972 I was told the zener with pass transistor was history at college and should forget them. Reading the simple voltage regulator link I am again reminded how very good it can be. I remember in the DIY Audio regulator I asked what happened at 25 MHz. Oh that's the job of the output capacitor. Exactly as cheap regulators I thought to myself.
Member
Joined 2009
Paid Member
Since we decided to use the approach of an upstream regulator and downstream regulator we got lots of options.
The upstream will be at the power-inlet. It doesn't have to be fancy but should be robust. Perhaps nice if it's simple enough to avoid needing a pcb, just bolt it onto the chassis side and be done with it.
The downstream regulators are crucial to the audio circuit performance. We'll put separate dedicated downstream regulators at each circuit, or on each pcb 'module'. For low current circuits this could even be shunt regulation without creating a lot of heat sinking needs. People who write about such things on this forum say they always prefer the sound of their circuits when used with shunt over series regulators.
I'm looking for a good quality input-selector switch, any recommendations ? or just use toggle switches ?
The upstream will be at the power-inlet. It doesn't have to be fancy but should be robust. Perhaps nice if it's simple enough to avoid needing a pcb, just bolt it onto the chassis side and be done with it.
The downstream regulators are crucial to the audio circuit performance. We'll put separate dedicated downstream regulators at each circuit, or on each pcb 'module'. For low current circuits this could even be shunt regulation without creating a lot of heat sinking needs. People who write about such things on this forum say they always prefer the sound of their circuits when used with shunt over series regulators.
I'm looking for a good quality input-selector switch, any recommendations ? or just use toggle switches ?
Greetings! To switch low-level analog signals, you can use the Takamisawa relay. You can control the switching of the relay in any convenient way.
Member
Joined 2009
Paid Member
With respect to conventional mechanical switches, I want to note that: they tend to deteriorate contact with time. The DACT volume control used by me, if I do not use it for some time, starts to make "clicks and rustles" during switching. Therefore, the whole issue of choice is the material and quality of contacts.
I think you haven't stated your sonic aims. DNM 3D? Naim 552? Somewhere in between? Or just a basic Hafler/NAD in an inferior case?
What sonic traits are important to you?
Do you have access to a decent factory built reference?
Is the sonic part important at all or just the functionality?
DC coupled? Balanced inputs/outputs?
Member
Joined 2009
Paid Member
I’ve not had the pleasure of hearing those preamps. I once owned a Bryston integrated and it had a preamp that I used once or twice for testing amps (now sold. It was utterly transparent to my ears. I used it simply as a convenient volume control for some DIY amps.I think you haven't stated your sonic aims. DNM 3D? Naim 552? Somewhere in between? Or just a basic Hafler/NAD in an inferior case?
1) no pops, clicks, thuds or other parasitic sounds during operation of the controls
2) no discernible hum or hiss from the listening chair (except for what one expects with Vinyl sometimes)
3) transparent enough to allow me to use it to compare other items in the chain (e.g. power amps and speakers)
4) special case: when used with my Naim clone, maintains the ‘house sound’ e.g. NAC 62
5) special case: when used with poor source/recordings can dial in some optional ‘masking sweetness’
6) special case: when used with ‘aggressive’ sounding speakers/room can help to pull back on the top-end
7) historical interest: if there’s a ‘sound’ associated with a well-regarded proven design then maybe include it as a switchable option
no.
I can live with a passive pre-amp from a functionality standpoint, that’s a minimum requirement. But I’m spending the effort on something more complex to explore the sonic part.
I have no allergy to caps or transformers in the signal chain but will happily use dc-coupling when it's achievable without heroic effort.
No experience with balanced signal chain and no plans to go balanced with other components for home/DIY use. This kind of kills the DNM approach doesn’t it.
Much clearer now
If it were my project i would stick to nice opamps. Like perhaps an AD797. With a very good power supply: Salas shunt, maybe choke input for the raw dc, nice caps. Imho this beats a surprising amount of heroic discrete effort into a pulp, both subjectively and objectively.
Nothing wrong with your idea to combine a passive and active options. As other posters mentioned, very few switches are both good sounding and durable and those tend to be extremely expensive. If going for relays there is also the option of using bi-stable types.
If it were my project i would stick to nice opamps. Like perhaps an AD797. With a very good power supply: Salas shunt, maybe choke input for the raw dc, nice caps. Imho this beats a surprising amount of heroic discrete effort into a pulp
, both subjectively and objectively. Nothing wrong with your idea to combine a passive and active options. As other posters mentioned, very few switches are both good sounding and durable and those tend to be extremely expensive. If going for relays there is also the option of using bi-stable types.
Member
Joined 2009
Paid Member
I'm not allergic to opamps either. However, for item 4) on my list above, discrete maybe required. Nothing precludes both options existing on the pcb. I don't like to spoil the fun, the chase, but insisting only one design can reign at the end. Some of the discrete circuits are very simple and don't take up much space. I can include more than one design on a pcb and populate whichever I like or maybe all of them ! Sounds like that would be a mess doesn't it
If I include an Opamp buffer (the most neutral option is probably a buffer, I have little need for gain most times) there is a design I liked from Dr. Bora. I asked him to explain it to me once but he declined, said it was proprietary. Unfortunately, he passed away since. But his design is here in post #94 and #95: http://www.diyaudio.com/forums/solid-state/65175-bora-nice-guy-serbia-yugoslavia-3.html#post1183839
If I include an Opamp buffer (the most neutral option is probably a buffer, I have little need for gain most times) there is a design I liked from Dr. Bora. I asked him to explain it to me once but he declined, said it was proprietary. Unfortunately, he passed away since. But his design is here in post #94 and #95: http://www.diyaudio.com/forums/solid-state/65175-bora-nice-guy-serbia-yugoslavia-3.html#post1183839
Last edited:
Member
Joined 2009
Paid Member
If there is some consensus and wisdom that the proper approach is small relays I think this can work too. Is this the general wisdom ?
I can use one relay per input. For the option to use the box without power, the passive pre-amp, I can use a normally closed relay on one of the inputs.
I still use the dirt cheap Lorlin switches. Use it as a reference before spending tons of money to make it worse. Yes the switch will age, they last perhaps 5 years and are easy to replace.
I use 3 pole 4 way make before brake. I common the ground as a DIN plug needs with it's one switch section. I have one set of RCA . My Phono stage is self powered and DIN. As said before I have an expansion box on one input where the lower grade sources go it.
Lorlin do a 12 position swich. If I had good advice on relays I might use the Lorlin switch to switch 12 sets of relays, 12 way version of the switch. I would still switch the grounds. I might have to make a delay circuit to do make before brake.
On a project using a Nait style case we had to use a switch mode PSU. I made a probe to measure the RF and other fields. To aid switch placement at the back of the case I bored out some plumbing fittings used for pneumatic equipment ( 5 mm to 6 mm ). I used a drill as the reamer was missing. Then used stock 6mm aluminium bar ( all from RS UK ). Very cheap and gave the Lorlin switch an expensive feel.
I use 3 pole 4 way make before brake. I common the ground as a DIN plug needs with it's one switch section. I have one set of RCA . My Phono stage is self powered and DIN. As said before I have an expansion box on one input where the lower grade sources go it.
Lorlin do a 12 position swich. If I had good advice on relays I might use the Lorlin switch to switch 12 sets of relays, 12 way version of the switch. I would still switch the grounds. I might have to make a delay circuit to do make before brake.
On a project using a Nait style case we had to use a switch mode PSU. I made a probe to measure the RF and other fields. To aid switch placement at the back of the case I bored out some plumbing fittings used for pneumatic equipment ( 5 mm to 6 mm ). I used a drill as the reamer was missing. Then used stock 6mm aluminium bar ( all from RS UK ). Very cheap and gave the Lorlin switch an expensive feel.
Аh, forgot the requirement for Naim house sound. This means unipolar supply and the nastiest coupling caps money can buy. I kinda like the old Naim power amps but not so keen on the preamps.
The only justification for using solid state devices in an audiophile preamp that i see is the potential elimination of coupling caps. If this potential cannot be realised, the entire exercise appears pointless.
Seeing numerous coupling caps in old designs by Naim, Pass, etc makes me really scratch my head.
Btw, a conceptually simple and very modular design can be appropriated from the better quality Marantz preamps, per example the sc-11s1. Based mostly on a universal diamond buffer type module, these use voltage to current conversion when voltage gain is required and the same module for buffering only. I think current version is HDAM3, with circuits and even ready made modules available on the net.
Hi,
(Haven't posted here for years - still been reading though).
I made a preamp based on a Naim circuit a while ago. Chronicled here...
http://www.diyaudio.com/forums/solid-state/208965-spare-bits-preamp.html
Like the Naim, it has (still using it) 317/337 regulators.
Not sure it contributes to a particular Naim-esque sound. I regard it as a reasonably implemented buffer.
As I said, it is still used to this day with no issues. I think it's fairly transparent.
If I was to revisit this part of my system, having played with regulators in my DAC, I'd have monolithic pre regulators (like the 3x7) followed by discrete/opamp based regulators - similar to previous suggestions. In my DAC, I'm currently using regulators very similar to a Sulzer, but with green LEDs as the voltage reference (like a flea). I'd probably use something similar in a preamp. I've not used anything more complicated (i.e. Jung) as I'm not sure my kit is good enough to warrant it.
I'd also investigate a relay based R2R potentiometer controlled by a microcontroller and a
optoelectronic rotary encoder or IR - but that's just me.
I really enjoyed implementing the discrete logic for the input selection, but I wouldn't recommend doing it. Logic chips and/or micros can do it for you.
Phil
(Haven't posted here for years - still been reading though).
I made a preamp based on a Naim circuit a while ago. Chronicled here...
http://www.diyaudio.com/forums/solid-state/208965-spare-bits-preamp.html
Like the Naim, it has (still using it) 317/337 regulators.
Not sure it contributes to a particular Naim-esque sound. I regard it as a reasonably implemented buffer.
As I said, it is still used to this day with no issues. I think it's fairly transparent.
If I was to revisit this part of my system, having played with regulators in my DAC, I'd have monolithic pre regulators (like the 3x7) followed by discrete/opamp based regulators - similar to previous suggestions. In my DAC, I'm currently using regulators very similar to a Sulzer, but with green LEDs as the voltage reference (like a flea). I'd probably use something similar in a preamp. I've not used anything more complicated (i.e. Jung) as I'm not sure my kit is good enough to warrant it.
I'd also investigate a relay based R2R potentiometer controlled by a microcontroller and a
optoelectronic rotary encoder or IR - but that's just me.
I really enjoyed implementing the discrete logic for the input selection, but I wouldn't recommend doing it. Logic chips and/or micros can do it for you.
Phil
I hope everyone reads the regulator noise link at No23 as I think we can better Naim using the same parts as them. Or use one zener and emitter follower and perhaps a CCS to the zener, output impedance shouldn't be a problem. I honestly feel shunt regulators are a fad. The zener is a fine shunt device and is often ignored. Mostly where shunt and series regulators fail is at > 1MHz. Local decoupling is the answer with all and either. Most op amps hate RF. JFET's perhaps less so.
I think a descrete op amp well worth a try. Little mini NAP built exactly as you want it. BC337-40 and BC327-40 come to mind as they have everything and even low rbb' to give very low noise ( 0.6nV/root Hz I believe ). I think +/- 20 V would be very OK. That op amp used as often as you like in the design. As my brother said Naim seemed to have Naim DNA circuits they used again and again. This would be in the spirit of that. Run as a single supply if you like.
Also, to use a +/- PSU must be better. The Naim preamps were not very special, very OK for 1978. I am sure anyone with good ears can do better and still have the Naim qualities. I should be possible to have a faster yet more open version. Yes, faster than Naim. Polystyrene caps help that.
I think a descrete op amp well worth a try. Little mini NAP built exactly as you want it. BC337-40 and BC327-40 come to mind as they have everything and even low rbb' to give very low noise ( 0.6nV/root Hz I believe ). I think +/- 20 V would be very OK. That op amp used as often as you like in the design. As my brother said Naim seemed to have Naim DNA circuits they used again and again. This would be in the spirit of that. Run as a single supply if you like.
Also, to use a +/- PSU must be better. The Naim preamps were not very special, very OK for 1978. I am sure anyone with good ears can do better and still have the Naim qualities. I should be possible to have a faster yet more open version. Yes, faster than Naim. Polystyrene caps help that.
Member
Joined 2009
Paid Member
No reason to doubt your honesty: we all hear differently. As for myself, the annoying bit is i had a collection of built jung/sulzer varieties of most types known to man, mostly the ALW... They all went into the trash after the first encounter with Salas.
True, there is a price to pay in both extra voltage and current, which unfortunately translates into heat.
Obviously, not trying to suggest shunt regs for a project with no high end pretensions.
Member
Joined 2009
Paid Member
Not sure this comment you make is constructive.
Please do have a look at my TGM1i thread, where you can see my first attempt at a preamp. It uses a shunt regulator (it was also blessed by Salas). It's one of the options. In my previous attempt it was a fixed voltage reg. I also like the floating shunt approach for circuits that don't need perfect controls on supply voltage; instead we fix the shunt regulator open circuit current draw.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.