LCR Phono using 6C45P

the road not taken.
Andrew
Posts: 201
Joined: Wed Apr 26, 2006 8:06 am

Post by Andrew »

Assuming the chokes can take the DC down via the caps to ground, couldn't you load the first stage with 7k3, then into the LCR and finally direct DC couple into the next stage with a 1 Meg doing both grid leak and termination of the LCR?
I thought of that some time back, but remember the 7K3 load resistor || 7K3 LCR means a 3600 R load on the 6C45... Now before you go asking for a 14K6 LCR why not this....
Hi Dave,

Could you post the .asc for this one?

thanks,

-- Andrew
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Now before you go asking for a 14K6 LCR why not this....
When I see circuits like this, I wonder about the large decoupling cap on the 2nd stage cathode and think, isn't this just another coupling cap? And a worse one at that? I mean, it looks kinda DC coupled but isn't really because of that cap. But then, I haven't tried it to see what it sounds like.

I think somewhere in this forum there was a thread on LCR drivers and one of the options discussed was a pentode with a transformer strapped across the load resistor. I'm a bit partial to a pentode first stage. It offers good gain, low noise and something I need which is low input capacitance.

These high gm options usually have huge miller capacitance and if a low O/P MC is used through an input Tx, the input capacitance seen by the cart can be enormous and move a previous benign HF resonant frequency right into the audio band. It was a post from Jonathan Carr that pointed me to this which prompted me to try wiring my input stage as a pentode from its triode strapped state.

So with transformer coupling, we could get rid of cap coupling, get some step down and therefore good drive to the LCR. Thoughts?

cheers,
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hi Stephen,

If you check my solution there are no caps!

To add a big series resistor of several kohms makes you loose it all. Must say I think even adding 600 ohms as I have proposed earlier is less good :oops: .

About stepdowns before RIAA I think it is the wrong way to go to this worsens S/N.
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Hi Lars,

I appreciate the "no caps" in your design but for me, it's got SS stuff in there and for my taste, it would be a non stater on those grounds. It's my prejudice that I've yet to be cured of :wink:

Some CSS designs have come close but when compared to a choke load, they so far have lost out for my taste. I can appreciate what they do though. I can't imagine a two stage buffer being "not there" but maybe it's fine. I have no idea. I don't do regulation or CSS loads anywhere in phono stages :)

Why does a step down worsen S/N? In my current config, I had a 1:1 tx changed to a 2:1 I think and it didn't make one bit of difference in reality but then I don't measure anything. Do you mean it worsens the 2nd stage S/N?

And the series resistor you mention .... I should have said I don't think that's the go if tx coupling. I would just couple the secondary to the LCR and terminate in the characteristic Z whatever that is for the LCR network. It's what I hope to do one day as I have it all ready to go when I can get Dave to wind some coils.

cheers,
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

hey andrew,

.asc is below.

Stephen,

I see your point on the cap. you could always use a 150V VR tube :-) I'm sure spice would love in but have to wonder about noise. At some point you need the cap to be there. A few years steve Bench made an NoR NoC amp and while it worked, even he admitted that a few well placed C's made it better.

For the pentode, I see your point, but haven't made it that far. Obviously starting with a tube like the 7788 is a wise choice since you can try and listen to both.

dave
Attachments
7K3 DC LCR Phono.asc
Normal disclaimer. I use Steve Bench's DD tube models, simply delete and replace the triodes with your model (.inc file or whatever)
(6.46 KiB) Downloaded 582 times
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

With MC pu and transformer there are no problems whatsoever to get rid of the caps 8) .
Attachments
10k DC LCR LS.asc
(5.72 KiB) Downloaded 508 times
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

When the cct is DC coupled with split supply as Lars has shown, the "coupling cap" is now in the supply and is a large one isn't it? But then I guess we then end up in the debate of AC loops and what is a coupling cap and what isn't. And then we end up in a debate as to which cap is more or less audible which I guess leads us down the debate over parafeed topologies. So many cccts, so little time :)

Another thing I wonder about is how effective hi Z LCR configs are? IME, low Z RC EQ sounds better than hi Z. I'd love to know why.

With a pentode and split supply, couldn't we DC couple a low Z LCR?

cheers,
Andrew
Posts: 201
Joined: Wed Apr 26, 2006 8:06 am

Post by Andrew »

In the end I went for a pentode driving the LCR, I agree with Stephen, for me as a personal choice, this seemed the best of all compromises.

As to DC coupling or not? I guess you either place a small the cap inline with the signal after the LCR or a larger cap in the cathode, choose your poison, as they say.

I guess I'm just asking, can you chase your tail too much trying to get rid of caps and as a result add compromise elsewhere? Just a thought....

cheers,

-- Andrew
Andrew
Posts: 201
Joined: Wed Apr 26, 2006 8:06 am

Post by Andrew »

hey andrew,

.asc is below.

dave
Thanks Dave, intuitively I would have assumed there would be a bass hump caused by the higher load into the LCR but there isn't, and yet the Z looking back into R10 is about 10k and a bit. I need to think about this some more.

I like the idea of OPT's been thinking about trying this myself, something like a 6C45, 417A or triode connected D3A or E180F.

cheers,

-- Andrew
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Stephen,
In a good PSU there is no cap wheter you go series or shunt :wink:!
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

I've tried this in LTSIPCE and would welcome feedback. As I said earlier, I thought it was as idea round here in another thread but I can't find it.

I use something very similar but parafeed and really want to try this one day but money is a problem at present.

I've ignored the LCR bit at present as it is a load that is constant. A 2:1 stepdown provides the best compromise for overall gain of the stage. 1:1 has less gain while 4:1 has no significant gain advantage. Stick any LCR on the end. The higher Z, the more gain but the rest I think stays the same. No coupling caps.

I initially tried this parafeed into an RC EQ going from direct into an RC EQ with small coupling cap to stage 2. For a reason I cannot work out, it sounds much better. It seems to remove the signature of the power supply cap among other things which being big, has a signiificant sound.

cheers,
Attachments
D3a_pentodetx.asc
(2.35 KiB) Downloaded 509 times
Last edited by izzy wizzy on Sun Mar 15, 2009 10:25 pm, edited 1 time in total.
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hi Stephen,

Adjusted your circuit using available data for a commercial 2,25:1 transformer. Haven´t simmed yet.

About the transformer-discussion you could check the 416 thread. With 416(mu=250-300) it would probably work as you need to go 4:1 or 8:1 to make the best of it.

More in depth info from Dave I suppose.....
Attachments
D3a pentodetxLS.asc
(2.07 KiB) Downloaded 482 times
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

hey guys,

I really like the pentode transformer idea....

here are a few pics...

I'll add discussion later. I find it very interesting that the gain of the pentode with increased load value outpaces the gain lost by the turns ratio.

I'll also vote for a 1:1 into a larger valued LCR. (given the needed inductance for the step down and my general distaste for low ratio transformers)

dave
Attachments
The inductance of the secondary is stepped to give the turns ratios on the plots
The inductance of the secondary is stepped to give the turns ratios on the plots
Picture 27.png (31.5 KiB) Viewed 32036 times
My "preferred" version using a 1:1 and a higher value LCR.  (note the primary inductance did not change but the bass did)
My "preferred" version using a 1:1 and a higher value LCR. (note the primary inductance did not change but the bass did)
Picture 29.png (33.37 KiB) Viewed 32038 times
Last edited by dave slagle on Sun Mar 15, 2009 2:16 pm, edited 1 time in total.
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Dave,

My understanding is an inductive load and a pentode don't work practically. Why do you think it a good idea for this app?

Lars,

I stuck to 2:1 as increasing it to 4:1 or higher didn't gain anything worthwhile in terms of overall stage gain. going 4:1 loses overall gain while 2:1 gives higher overall gain than 1:1. I think I got it wrong and should have used 100H/25H for a 2:1.

I wonder why you added the 1000u? In my sim, it doesn't do anything gain wise or maybe I've missed something. I don't use it in my current pre either.

I'll look at the other thread but for my needs, a pentode has much lower input capacitance.

cheers,
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey Stephen.

I like the transformer since it dispenses with the cap and the 1:1 can be made very high bandwidth.

Maybe my pentode model is messed up, but In my sims above, I had a distinct gain in overall response when going from 2:1 to 4:1 and given that the voltage loss should be 1/2, the load now becomes 4X which by simple approximations should net you a 2X gain. Going to 10:1 still gave me gain and I do eventually get to the point of diminishing returns and loss of amplitude by the time I reach 25:1 (this says nothing about distortion mind you)

I double checked the curves of the D3A in spice and they seem to be reasonably accurate


dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

The only way you can make it work is by triodestrapping.

Checked and the bypassed cathode resistor didn´t help the lousy lowend behaviour enough. Also gain got excessive and gave the stage a poor overload reserve.
Brgds
Lars
Andrew
Posts: 201
Joined: Wed Apr 26, 2006 8:06 am

Post by Andrew »

Question, is a TX loading different to a choke loading? How will the TX reflecting whatever it sees on its 'secondary' different for a triode to a pentode. I'm not questioning that a choke loaded pentode is out, a constant current source loading a constant current source is the problem. However, a transformer loaded is going to reflect what's on secondary. I was sure Dave had a sim of this working? Looking at this another way, don't we use tx-coupled pentodes on output stages. OK, I don't but amps like this do exist.

-- Andrew
Last edited by Andrew on Sun Mar 15, 2009 8:44 pm, edited 1 time in total.
Andrew
Posts: 201
Joined: Wed Apr 26, 2006 8:06 am

Post by Andrew »

Dave, I can't download Stephen's sim, just get a 'file not found' error.:?:

-- Andrew
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Andrew,

I re-posted the attachment.

cheers,
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I think things are becoming unclear so I'll take things inline for clarity.
Andrew wrote:Question, is a TX loading different to a choke loading?
no, the same rules apply.
How will the TX reflecting whatever it sees on its 'secondary' different for a triode to a pentode.
Simply put a pentode given its high source impedance forces you to size the primary inductance to match the reflected load. Swapping that out for a triode allows much less inductance to be used to get the same LF response which in turn allows you to design for much better HF response. Assume a 5K reflected load and a triode with a 1K Rp and a pentode with a 100K Rp. In order to be -1dB @ 10hz for the triode you would need 26hy's of inductance. For the same reflected load and a pentode you would need 151hy's

Look at the plots below where I increase the stepdown ratio of the transformer. The loss in bass is due to the need for more inductance as the reflected load goes up. If you triode wire the D3A you will not see the bass roll off with increased stepdown since the Low Rp of the source will dominate.

Image

I'll ask you guys a question about what I just said. Have any of you ever seen the argument made that a transformer for a pentode should be designed differently than one for a triode? I have been saying this to deaf ears for some time now and I'm trying to figure out why nobody seems to "get it" Who Knows i may be the one who doesn't "get it" but in the event that this concept makes it into some commercial spiel, please "have my back" so to speak. (sorry i'm in a ranting mood... now back to your regular program) :-)
I'm not questioning that a choke loaded pentode is out, a constant current source loading a constant current source is the problem.
an active device (ccs) and a pentode (high resistance) in series are not a problem. The CCS will easily dominate. A choke (or transformer primary) if unloaded attempts to be an ideal CCS so the two are similar. In order for your pentode to behave as desired you need a fixed defined load which neither a choke or CCS can provide. Put a resistor in parallel with the choke or the CCS and you suddenly have the load the pentode needs to see.
However, a transformer loaded is going to reflect what's on secondary. I was sure Dave had a sim of this working? Looking at this another way, don't we use tx-coupled pentodes on output stages. OK, I don't but amps like this do exist.
EXACTLY... If transformers and pentodes did not mix, how could we ever have a PP pentode amp?

dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Put a resistor in parallel with the choke or the CCS and you suddenly have the load the pentode needs to see.
Isn´t this what Stephen had in his initial schematic ?

The problem is that when adding the primary parallell resistor, gain gets as low as when triodestrapping. And still the pentode gets worse lowend performance.....

Also distortion gets unaccepatable with pentode.
Attachments
D3a pentodetjfr.asc
(4.05 KiB) Downloaded 463 times
Brgds
Lars
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

This is why you shouldn´t buy a readybuilt potted LCR-filter:

http://ansaht.altervista.org/phpBB2/vie ... c&start=80
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Lars,

The schemo you compared to the triode isn't mine. It's missing the power supply decoupling cap. Also I use 4k7 on the anode of the pentode at 250V and maybe even 10k is possible at 300V.

With that arrangement, the pentode produces more gain. Why do you say the pentode has higher distortion? Does the sim day this? I haven't used that part of it much yet. But in theory, at low signal levels, a pentode has lower distortion than a triode.

cheers,
Attachments
D3a_pentodetjfrmoda.asc
(4.35 KiB) Downloaded 447 times
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hello Stephen,

LTSpice doesn´t care much about PSU decoupling caps. Also the values of the components where altered to get the same current through both. That doesn´t change the function either.
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

Lars,

Yes but look where I stuck the PSU decoupling cap :wink:

Try it with and without and look at the gain difference - 3x more with it as I've shown.

This stage originally started as parafeed with batthery bias in the grid. All I did was move the ground connection and replaced the battery with a bais resistor.

I see what you're saying re the current. I've just moved things around in the pentode's favour a bit.

cheers,
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Stephen,
With a decopupling cap and a series resistor or with a lower voltage directly to the transformer will not change the sims at all.

I can´t see how a transformer coupled pentode could be useful driving a LCR :cry: !
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

The transformer coupled LCR can dispense with the need for a coupling cap and conceptually a stepdown could be used to inccrease the overall gain of the circuit.

dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

dave slagle wrote:The transformer coupled LCR can dispense with the need for a coupling cap and conceptually a stepdown could be used to inccrease the overall gain of the circuit.

dave

As the pentodes Zout is ca Gm*Ra things get a little more problematic as the LCR wants to be driven from lowZ.

I think Stephens idea with a primary termination is in the right direction but I can not imagine how to make this work IRL.....
Brgds
Lars
izzy wizzy
Posts: 64
Joined: Sun Jun 26, 2005 8:38 pm
Location: Folkestone, UK

Post by izzy wizzy »

I've sim'd 3 circuits with a pentode, transformer coupled, driving various LCRs. The LCRs are various circuits I've found around the place and are in no way optimised as I don't understand them fully yet.

They are 600R, 1.3k and 7.3k. The gain difference between 600R and 1.3k is as expected but the 7.3k version doesn't have that much more gain.

I would choose the 2nd stage gain for what is required by the system its being used in. Currently I use 5687 but 6C45, triode wired pentodes etc could be used for more gain.

cheers,
Attachments
D3a_pentodetx_lcr_600_a.asc
Pentode driven, tx coupled 600R LCR
(6.09 KiB) Downloaded 530 times
D3a_pentodetx_lcr_1.3k_a.asc
Pentode driven, tx coupled 1.3k LCR
(5.88 KiB) Downloaded 557 times
7K3_DC__LCR_Phonotx.asc
Pentode driven, tx coupled 7.3k LCR
(7.43 KiB) Downloaded 496 times
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hey Stephen,
Didn´t see your 150u :oops: ! Will take that in consideration when I do my next sim.
Brgds
Lars
Post Reply