 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
120.16.242.34
In a recent thread Dave Slagle suggested I post my schematic of an LCR Phono stage I recently modelled. Jeff Davison was looking to build a Phono stage utilising a 600 ohm LCR RIAA, so for expediency I modified the Sim of my own Ref Phono and after several iterations this worked best.
Although I haven't actually built this LCR version I'm confident it will work as Spice predicts. Why? I started Simming many of my past designs recently for the heck of it and it's been incredibly accurate ... I mean exact, according all of the bench measurements I had recorded in my notes!
I also expect it to be extremely quiet. It should sound rich, very detailed and airy. Anyway it should be a good starting point for anyone interested in this approach. Comments welcome.
Note 1 - The reverse RIAA in the front end (in reverse color) is used only to plot the real world frequency response in the Sim. My thanks to Jim Hagerman, it was borrowed from his site.
Note 2 - The model for the 600 ohm LCR (also in reverse color) was borrowed from Tango but other equivalents such as S&B / Silk should work with similar results.
Note 3 – The RC network across the LCR is used to compensate for falling top end response, a trait of the LCR networks it seems. My choice, but those who do not agree may choose to omit this.
Note 4 – The Grid Choke is tuned with the coupling cap to restore bottom end frequency response arising from not terminating the LCR with 600 ohms. This technique significantly improves gain over correct termination and does not appear to cause other problems. LF response can be tailored with different tuning.
Note 5 – The Phono stage has sufficient gain to directly couple a low OP MC cartridge, though some may opt to use a SUT (I’m not one of them because IME all but the very best are BW limited to some degree).
The Line stage is optional and I’ve chosen Lithium battery bias in the Grid circuits for something different. It works very well. I’ve chosen Schottky diodes to bias both Phono stages. I haven’t actually tried this on the 2nd stage but anything else seems compromised on the 6C45P, including a single (standard) diode equivalent … a result of higher dynamic impedance I suspect.
Obviously tube selection and many other parts of the circuit could easily be changed to suit individual taste but I chose the 6C45P because I use it in my own RC RIAA version of this Phono and it’s dead quiet! I personally use Cca Siemens tubes in the second stage but the sky is the limit here.
For simplicity the CCS could be substituted with a 12 – 15K R with a loss of approx of 2db. Alternatively, the purist may opt to throw in another 6C45P like I do. For those with an aversion to the CF stage try Allen Wright’s suped up version.
Finally, I recommend a good Shunt Reg PSU … but only after some decent L filtering.
Feedback would be appreciated from anyone out there who chooses to build this.
Schematic Only ... hope this is readable!
Naz
Follow Ups:
Hi.
It may look immaculate on paper, but would it sound good musically is another story.
Let us know if it'll really sound good with so many active stages involved!
c-J
I get your drift and agree that the fewer the stages the better. My current Phono is similar but with RC RIAA and sounds VERY good, even compared to many well respected High end commercial systems that I've had in my system from time to time.
I currently don't use the final stage in my set up and still have enough gain to run a 0.6mV MC with full system OP. But that's one of the points, this is an all tube Phono that can run low OP MCs without a SUT ... and it's dead quiet.
Naz
Naz, just a thought... I'll be doing this basically verbatim with 6C45p's up front, but a thought occured to me...
Could the 6C45p be replace by a triode strapped E810F/7788 or a triode strapped C3g ? Both have pretty good characteristics in triode.
Jeff
Sure,
I used the 6C45P because I know how well it works and how quiet it is however there are any number of tubes you could try with minimal variation. I'd be interested in your impressions if you get around to comparing a few.
Naz
A triode-strapped E810F/7788 is very similar to the WE437A/EC8020/6C45.
The C3g makes a very fine triode. But it has about twice the internal resistance of those other tubes.
I would use a 10k LCR. 600 ohms is murder. With 10k either of these tubes can be used without having to be used as cathode followers.
A D3A is suppose to be similar to the E810F/7788, so a D3A triode strapped would probably work too. Also, the Russian 6J52P is suppose to be similar to D3A as well, making another choices possible.
Edits: 03/12/09
Now this is funny from the spec sheet..
I saw that too. I had to re-read it a few times.
It makes me wonder about the vintage of the data sheet and why it is in English at all. Surely the US was not importing tubes from Russia in the days of tubed TV's and US - Soviet relations were not wonderful after that. My only guess is that the translation happened after somebody found a gazillion of these tubes and needed to market them to the West. My only problem with that theory is that the sheet looks "original"
dave
Peter,
Jeff's advice from a manufacturer was that the 600 ohm LCRs work better (as lower impedance tends to) so that was a pre-requisite. I agree it would have made life a lot easier to use 10K. I'd like to hear from anyone who has tried both.
Naz
Naz, I've heard from someone lurking this thread, who has worked with the S&B, the Tango and the Silk, that the Silk does not behave the same and it's internals are different enough that it is not a "drop in" for those LCR's and would not work well with a circuit designed for them and would need specific tuning to do so.I'm waiting to hear back from him to see if I can pass on the specifics.
Kinda vauge I know, but I am a bit worried at this point to use the Silks in this circuit.Jeff
Edits: 03/10/09
Jeff,
It's always a problem when we don't know exactly what's inside the LCR can. I have no doubt that the Tango would work as predicted, save for the still unverified effect of the small amount of DC current flowing. Although I would not expect a problem here, the simple remedy would be to add a large cap in series with the ground leg of the LCR to isolate the DC. While I hate large caps and would avoid them at all costs, if you had to use one then I suspect it would sound better in the ground leg.
As I mentioned earlier you still have the option of either terminating the LCR with a grid leak of 600 ohms and eliminating the grid choke or adding a 560 ohm resistor in series with its input or both in which case any 600 ohm LCR should work. You would sacrifice some gain you have approx 75db to play with including the third stage.
I went with the grid choke primarily to avoid the high value coupling cap that would otherwise be required with a 600R grid leak.
I know you have some doubts about the direction, but any of the modules can be made to work, and I would not expect major tuning difficulties ... it would just help to know their make-up. The critical question that needs to be answered is whether the LCR of choice will handle DC current and if so what the limit is. With this info we can establish the basic configuration. I wouldn't be too concerned about the tuning as this can be resolved on the bench if we cannot obtain a detailed circuit.
Naz
a 560 in series withthe input could be a workble solution... a nice Tantalum or Vishay maybe? The cap to ground could be a workable "work-around" but DC at the levels mentioned elsewhere wouldn't be a issue i.e. +100v max, and the output cap in front of the grid choke works a treat to block it.JD
Edits: 03/10/09
Naz, the 600 ohm grid leak is a big no-no according to Kittipol ans he said it would..
"If you connect 600 ohm, then total output impedance of LCR become 300 ohm (paralled effect) and will ruin LCR RIAA frequency."
We already know how much DC the Silk can handle. Brian at DIYHiFi is good to work with as he "caught" my order before it went out the door when I emailed him my trepidations.
BTW.... sending you a P.M. that may perhaps shed some light on where my trepidations come from
Jeff
I asked Brian at DIYHiFi to comment on the suitability of using the Silk LCR with a 600R termination and he said it won't work that way.
Regards,
John
> > "If you connect 600 ohm, then total output impedance of LCR become 300 ohm (paralled effect) and will ruin LCR RIAA frequency." < <
My understanding is that it merely reduces the input impedance to 300 ohms but since we are driving with approx 20 ohms this should not affect the RIAA. Having said that, I don't favour this option because it forces you to use a high value coupling cap, so let's cross that one off the list.
Naz
Naz,
working wit PSUD designing a raw B+ to determin transformer capacity needed... what currents are the 6DJ8 and 6H30P run at?
Thanks
Jeff
Just under 10mA and a tad over 15mA respectively. In straight grounded cathode circuits I like to run them hotter but they lose nothing at lower current (possibly even lower than this) in this configuration.
Cheers,
Naz
Hi Naz,It looks like you used mu follower type ccts for stages 2 and 3 yet you don't use the mu out bit. Why is that?
cheers,
Stephen
Edits: 03/09/09
Good question Stephen. Simply because I prefer the sound of the “quasi” CCS. Technically the Mu Follower is better but somehow it just doesn’t sound as natural IME. The extremes are certainly there and one might argue, better but the mids are more inviting and have better depth. The same goes for SRPP and perhaps even moreso, but this is just my opinion.
Cheers,
Naz
Well you seem to have a similar taste to me in that regard :-)
Have you tried choke loading?
cheers,
Stephen
Yup, I don't like standard grounded cathode stages with choke loading because every one I've tried, even expensive ones, seem to be compromised at the extreme top end.
I use them in Mu (CCS) type circuits where I need high linear voltage swing eg to drive my 845s and this works extremely well but I don't bother when small signals are involved as I've never been able to hear a difference.
Regards,
Naz
Naz..
I think we're starting to get somewhere. I forwarded the schematic to Kittipol at Sacthailand and "just" got his reply 11:35pm est..
Here's his reply verbatim:
Hello,
The circuit should work, although I think 0.15uF is too small.
There will be almost zero DC current flow through the LCR since you use 0.15uF to block the DC from input.
Please be careful to not use more than +100V to the input of LCR.
Regards,
Kittipol Saengoumnuayporn
SACThailand
Cool,
The 0.15uF appears small but is an important part of the tuned circuit with the 300H grid choke. This is the effect of variations from 0.1uF - 0.22uF.
You can hardly see it but you get the idea. The yellow is 0.1uF, the blue is 0.15uF and the red is 0.22uF.
Cheers,
Naz
ideas where to source the 300H choke?
JD
I bought some really nice units through Ebay a few years back, I'll try to find the vendor. This came up in a quick search:
Naz
I saw that too, and the price is nice.... but are they of good enough quality? I wonder if Dave could do some? I have a pair of low nickle Magnequest grid chokes in my stash but the H is probably too large as the exact value is not specified in their data.JD
Edits: 03/09/09 03/09/09
Here's the link.
Naz
nt
here's accsess to much larger views I posted here:
Here's the Spics screen:
http://gallery.audioasylum.com/cgi/wi.mpl?u=45887&f=Phon_Pre_LCR_6-3-09.JPG&w=1680&h=1050
Here's a large image of the schemo:
bigger schemo
Thanks Jeff,
Although I'm a contributing member it doesn't seem to have been acknowledged and the BW limits otherwise imposed are a bit tough IMHO.
Hopefully we'll get some interesting feedback.
Naz
In order for it to block DC it needs to be the middle design.
hey-Hey!!!,
It seems the first two are the same, just the second has values installed. What sort of coupling is needed( allowed? ) for the second stage grid?
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
There is an additional resistor in the first unit. I find it interesting that leaving out a resistor actually improves response. I believe the origin of the linked sketch was Hiraga and at the time of publication the only game in town was the Tango EQ-600P and various historical units (pultec)
The type of coupling allowed has nothing to do with the version used but the overall circuit design and the requirements on parts selection. Use 630V caps in #1 or #2 and you could direct couple up to that voltage. The diagram of #3 can also be direct coupled but I doubt the tango would be happy in that situation.
dave
Thanks for your comments. I used the Tango model because it was the only commercially available unit I could get info on. I assumed most would want to buy the LCR rather than attempt to piece it together.
I have to admit the only real question in my mind was whether these things are OK with DC current. Silk specifies a max of 100V DC which implies that current is not a prob but not knowing what’s inside, they could be similar to the DC blocked circuits you are suggesting. Do you think that a few mA is too much for Tango circuit types?
As for the HF compensating components, I assumed that you can't get to the individual nodes for a simpler solution but as I said, this is optional anyway. I like wide BW but it doesn't seem to be everyone's cup of tea.
I had trouble logging onto your site to post the ASC file. Could be just my password but will post when I sort this out.
Naz
Kittipol never specified what's inside, but maybe we can get a glimpse from a few of the replies I recieved from him when I sent him some ideas......Here's a few messages from SacThailand I recieved from Kittipol from different question I asked..... just to keep it interesting..........:
Hello,
You should use SILK LCR600R version. Direct coupling is possible but the voltage limit is +100V to the LCR input.
This voltage will also appear at LCR output so you must not connect any shunt resistor nor grid resistor.
Output capacitor value depends very much on experiment. If you use SRPP output stage, the capacitor should be of large value (more than 1uF)
In case you have any question, pls email me.
Regards,
Kittipol Saengoumnuayporn
SACThailand
Hello,
Do you know why people call it "LCR 600 ohm"?
Because the LCR has approx 600 ohm input impedance, and approx 600 ohm output impedance.
In my web circuit, I use AC coupling to input of LCR, so in order to make second tube work, we need grid resistor. Value should be anything from 0.2~1.0M ohm. But not 600 ohm as people believe.
If you connect 600 ohm, then total output impedance of LCR become 300 ohm (paralled effect) and will ruin LCR RIAA frequency.
In your intend circuit with DC couple, all voltage from LCR input will appear at LCR output, suppose you use +100V, then this +100V will go accross the LCR to your 1M grid resistor. Then 0.1mA will flow through LCR and will ruin the supermalloy core inside.
If you decide DC couple, and want -3V bias for second tube, then you raise the cathode of second stage to +103V and not connect any grid resistor.
Regards,
Kittipol Saengoumnuayporn
SACThailand
Hello,
I have tried Loesch circuit long time ago as input to drive LCR, it was not perform well. You know why? Because the drive circuit has output impedance much larger than 600 ohm.
I recommend you stick with your previous CCS input since it has low output impedance.
After LCR, the output impedance is 600 ohm. So you will need much larger than 0.1uF in order to prevent low frequency roll-off. My Grid choke can be used instead of grid resistor but you'll need very very special mumetal shield since my grid choke is an open form type.
The best thing is to buy cheap resistors/capacitors and you MUST experiment by changing value, circuit. There is no short cut when designing your own circuit. Once you get the circuit, then start to change to better component.
DIY Audio is about experiment and learning. It trains your brain, skills and patient.
Regards,
Kittipol Saengoumnuayporn
SACThailand
I could use some clarification concerning Kittipol's second missive...
My understanding concerning 600ohm LCR modules in general is that either the input or load needs to be 600ohms. But, Kittipol seems to suggest that the Silk 600ohm LCR only works with 600ohm input impedence. Am I reading that wrong, or is terminating the LCR with a 600R off the table when working with the Silk modules?
Regards,
John
seems that Kittipol implies that the module is already terminated internally with 600R as I asked him about using a 600R grid leak in the first tube of the output stage.... he said it would be parralleling and the net would be 300R and something NOT to do.
JD
Jeff,
I mentioned this before. There can be no internal termination of 600 ohms since that would draw excessive current through the inductors when direct coupled. The warning to not exceed 100V simply means that the caps used have a 100V rating.
I have also mentioned (and documented via spice) that the input impedance of the LCR is ~600R independent of the load placed on the secondary. Leave the secondary open and it is 600R. Short the secondary to ground and the source still sees 600R.
The situation where the 300R number comes into play is when you attempt to load a pentode with a 600R resistor in order to define the source Z feeding the network. In this situation, the 600R plate load is in parallel with the 600 LCR network (which is 600R independent of secondary termination) which nets a 300R load on the pentode.
What I find clever about NAZ's design is that he dispenses with the 600R termination altogether which nets him an extra 6dB of overall gain (6dB less loss in the network) at the cost of a drop of in LF response. In order to get back the LF loss, he terminates the network with a CL resonant circuit and tunes it to get back the bass response. If you look at the frequency response plots you will see him flat to around 23hz and then -4dB at 20hz with a steep slope. I don't have a problem with this since the LF behavior is essentially from a few stacked filters rather than a single complex one and I'm a firm believer that keeping extreme LF behavior out of a phono pre is actually a good thing.
dave
I'm flip flopping whether or not to build Naz's circuit with the Silks as there is no way to determin what's actually happening in side their module. I have it on good authority that it is not a "drop-in" exact replacement for the S&B or the Tango, and that a lot at "circuit tuning" must be done around the Silk to work in a circuit designed for the S&B and Tango.Naz was suggesting that maybe we / I should work with you as you did with Andrew in your forums postings? The Silk is starting to scare me because of the lack of information about it and the secretivness of Kittipol on his proprietary design.
I was also thinkin of possibly using this (expensive) solution as it IS a drop in replacement:...
JD
Edits: 03/10/09
I find this page very confusing. They mention the Tango and show the same picture from Hiraga, yet show a schematic that is different using what is equivalent to the middle method of the hiraga diagram which will allow direct coupling up to the rated cap value at a small penalty in overall accuracy. In the same schematic they designate the two inductors as the "EQ-2L" as if they are a stand alone module so I am unclear what you would be purchasing.
below is the actual Tango EQ600P drawing from the spec sheet.
dave
Tango did do a module with just the inductors...
me too :(
Although I sent them an email to se if it would be a 1 to 1 replacement for the Tango in any given circuit.
JD
Jeff, I responded to Lars who provided some useful input at Dave's site. We may be getting somewhere.
Naz
Naz, I got a reply from Acoustic-Dimensions in the Netherlands who also make an LCR. They say theirs is an exact replica also exact in that theirs, and the Tango, have no tolerence to DC at all..... which unlike the Silk that allows up to +100vJD
Edits: 03/11/09
Thanks Jeff, though I think I'm still not clear. If it's terminated by 600R, then why all the hub-bub about about the 600ohm input impedence? Input stage could be whatever you want as long as it's comfortable driving 600ohms, if that's the case.
Regards,
John
Jeff
I have a Silk Phono LCR right now I am playing
with for Paul AKA xaudiomanx.
He has a seperate PS chassis that makes it very quiet.
It has a 6SL7 GC directly coupled to a 6SL7 CF.
Both are K&K Cascoded DN2540 CCS loaded.
The 6SL7 cathode follower has an output impedance
very close to the 600R recommended value.
One 6SL7 before the LCR and one after the LCR.
It has a .22uFd output cap coupling with a 470KR grid leak.
I am now having to add another stage for more gain.
It seems to sound fine but again I need more gain
to really give it a proper review.
I will have an Edcor 15K:15K as a 1:4 autotransformer
then another 6SL7 CF on the ouput.
I will compare it to my Octal Cornet when 1 am finished.
DanL
I removed the login requirement so you should be able to post away.
dave
COOL!
Thanks Naz! A CCS on the input is an interesting idea, even tho yo loose a bit of gain, you'd only have to tame 1 tube for ocillation. The 6C45p has 4 pins onthe cathode and an extra pin on the grid... in this tube they can be antenae for RF. On the WE437A I used in my SET amp I put stopper resistors on its two cathode pins ( between the bias battery and the tube) for a little bit of resisance which came to 2.5R. Doing this "trick" on the 6C45P may have an effect on the biasing with the Schotkies?
JD
Shouldn't be a problem Jeff. In fact you could even reduce the Schottky number to 2 which would also reduce the 6C45P plate voltage to about 73V but I wanted to keep the bias well away from 0V because grid current starts before you get to positive territory.
Naz
a few quick thoughts.
You will have DC current through the LCR with the design you use. Spice will tell you how much. Depending on the inductor design it may or may not be a problem.
I suspect the network to bring the high end back up will not be needed. The inductors typically have enough capacitance to take care of this naturally and in some cases the capacitance is so large it becomes problematic. If you want to buy back the top end you can insert around a 60-65 ohm resistor as shown in the pic below.
Interesting with the tuned CL to buy back the bass... I'll have to look at that one.
dave
Naz has come up with a nice cicuit here. He told be that voltage through the LCR would be +82vdc. The Silk LCT RIAA is tolerant to +100vdc, so it looks like (at least) that module would be workable in this design.I'm going to wath the evolution of this thread to see if any further mods are done with further insight and/or suggestions for improvement or problem fixes.
Starting to collect parts and I've already started build a couple of Allen Wrights "Super Reg" clones. I'll build this as a complete dual mono after the power transformer dial rectification and dual regulation and separete VCSS filament supplies for the input stage and output stage.
JD
Edits: 03/08/09
With 82V, the first inductor will have just over 2ma of current through it and the second about 1/2ma. In order to make the LCR like the Silk, R27 and R31 (14K and 68K) would need to be left out. Then no current would flow through the inductors. If the LCR's are built from scratch, the inductors could be designed to deal with this and a proper voltage rating selected for the caps.
dave
Dave, Naz was using a Tango in his model. I do not know the internals of the Silk as Kittipol is a little secretive. It may or may not be as similar on the inside.JD
Edits: 03/08/09
in order for it to block DC and keep curent out of the inductors it needs to be the middle design.
Here are the two sims:
And the plotted response:
Cool,
if you want you can host the .asc at the link beow and then link to it here. It would be soo much quicker to have the file than t build it.
dave
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
