New LCR Phono : Pentode Aikido, Slagle 1.5k, C3g

[IslandPink]

Time to start another very slow thread, I think !
Ever since I had to photography my current phono for the jFET phono thread .....
http://www.audio-talk.co.uk/phpBB2/view ... 57&start=0
....I have been embarrassed, and knew something would have to be done .
So, there is a new circuit planned, and quite a few parts already acquired, including a new top plate and box for the LCR chokes.
First, here's the circuit :

[izzy wizzy]

Is there a reason for the series resistor over a termination resistor on the second stage grid?

You'll lose quite a bit of gain in the first stage with the unbypassed cathode. Did you decide on that over a battery or some kind of diode biasing?

Just curious. Look forward to seeing how it turns out.

cheers,

Stephen

[IslandPink]

Is there a reason for the series resistor over a termination resistor on the second stage grid?

You'll lose quite a bit of gain in the first stage with the unbypassed cathode. Did you decide on that over a battery or some kind of diode biasing?

Stephen

Hi Stephen
(i) Are you talking about the LCR and it's need for 1.5k somewhere ?
If so there are two options, one which uses (initial) series resistance but keeps a small coupling cap afterwards ...and the other which uses a big coupling cap and no series resistance in front, but the 1.5k to ground after the RIAA . I don't know which is the best . The 22uF for the 2nd option will have to be a Black Gate VK to have a chance !

(ii) I was counting on not losing very much gain with the un-bypassed cathode resistor there, do you have a calc for the lost gain ? - I'm hoping to get over x100 in the first stage . I can use 1000uF/25V BG if necessary which have proved themselves in the current one.

Thanks for any advice

[IslandPink]

Andrew -
Just spotted something worrying on the sketch I have of the full LCR detail from you and Dave .
The caps to ground appear to have parallel resistors across them : 35k across the 2uf and 170k across the 0.05uF . This means I can only run the chokes and caps at zero DC potential , which means the option of the series (1.3k) resistor and small coupling cap after the chokes is now out of the question .
I'm surprised I hadn't noticed this, but it seems different to all the other LCR circuits I've seen where the caps are not by-passed .
I'm wondering how much difference it makes to omit the resistors ( & slightly re-design if necessary ) , as I definitely prefer to have caps biased-up when they are taking very small signals .

Edit : I notice Andrew, fairly early-on in your discussions over the 1.5k option, you have those resistors in place :
http://www.intactaudio.com/forum/viewto ... sc&start=0

Now I can understand the 170k might be to put in the extra 50kHz time-constant, but not sure about the 35k . I am happy to omit the 50kHz thing, have been down that road on RC networks and can take it or leave it !

[Andrew]

It's OK the filter draws a few milliamps but its fine, Dave designed the chokes to have some DC across them, says it helps close the gap, or something.

I run mine at 150v-ish above ground and just have a small cap after them, to no adverse effects. Of course, you can miss out the resistors to ground if you want and, yes, one of them is the 50kHz turnover point.

Otherwise, you can put caps to ground after the resistors but you need quite big ones to avoid changing the filter's response.

If you check the threads over at Dave's place you'll see he confirms the chokes are OK, you just need to budget for the current.

Andrew

[IslandPink]

It's OK the filter draws a few milliamps but its fine, Dave designed the chokes to have some DC across them, says it helps close the gap, or something.
I run mine at 150v-ish above ground and just have a small cap after them, to no adverse effects. Of course, you can miss out the resistors to ground if you want and, yes, one of them is the 50kHz turnover point.

OK thanks. Are you sure omitting the 35k across the first cap doesn't affect the RIAA ? How quickly could you check that -do you still have the Spice file ?

Otherwise, you can put caps to ground after the resistors but you need quite big ones to avoid changing the filter's response.

Yes, I suppose so, but then you don't get the benefit of the DC through the chokes, so not really worth doing ?

I might try it with one of the resistors actually , or a bigger one to reduce the current , I suppose, as you say it's just a little extra drain on the Aikido stage & power supply at the end of the day . I think 4mA might be enough to unbalance the Aikido, I'd go down to 1-2mA I reckon .

[IslandPink]

'Phono upgrade kit' shown below ( a selection of bits anyway ) :

[Andrew]

OK thanks. Are you sure omitting the 35k across the first cap doesn't affect the RIAA ? How quickly could you check that -do you still have the Spice file ?

Otherwise, you can put caps to ground after the resistors but you need quite big ones to avoid changing the filter's response.

I can check for you...

[izzy wizzy]

Hi Stephen
(i) Are you talking about the LCR and it's need for 1.5k somewhere ?
If so there are two options, one which uses (initial) series resistance but keeps a small coupling cap afterwards ...and the other which uses a big coupling cap and no series resistance in front, but the 1.5k to ground after the RIAA . I don't know which is the best . The 22uF for the 2nd option will have to be a Black Gate VK to have a chance !

(ii) I was counting on not losing very much gain with the un-bypassed cathode resistor there, do you have a calc for the lost gain ? - I'm hoping to get over x100 in the first stage . I can use 1000uF/25V BG if necessary which have proved themselves in the current one.

Thanks for any advice

i) I don't know which is best either. I'spose I'd go for some tx coupling someplace but who knows. LCR with valves involves some kind of compromise somewhere. Was just interested to hear you reasoning.

ii) Did a very quick model using a D3a coz I don't have an easy to access E83F model. I'm guessing the ratios would be about similar between the two configurations. Bypassed the gain is about 50dB, unbypassed about 35dB.

cheers,

Stephen

[IslandPink]

That's more of a difference than I expected, Stephen .
I will have to leave a bit of space for caps in that case.
I was always under the impression pentode gain was less affected by bypassing the cathode resistor, since the internal resistance of the pentode is much higher than triode ?
I know I use the C3m or C3g drivers in the power amp without caps .

[Paul Barker]

Sorry strike all the following I did it for the c3g. Now I see it was the e83f Mark is using in first stage I have to recalculate. See the ammendment in the next post. I leave this one alone for history.

Using your circuit values Mark (I averaged cathode resistor to 240 ohm), and assuming internal resistance of 300k and assuming transconductance of 14.
I have also assumed a grid resistance of the following stage of 1 megohm, which is what is usually used when direct coupling. I cannot verify if that assumption is acurate.

I get amplification (A) of 45 unbypassed and of 194 bypassed.

Your effective plate resistance is 15k parallelled by 300k parallelled by 1meg.
Your effective cathode resistance = 1/gm (71 in your case) plus cathode resistor if unbypassed. So your cathode resistance is either 71 or 311.

Your stage gain is largely affected by this because the effect the cathode resistor has is inverse.

To explain it more fully the gain equation is

(1/(1/15k+1/300k+1/1meg))/either 71 or 311

the bypassed gain is very impressive on paper.

[Paul Barker]

OK the E83F has IR of 500k and S of 9.

The new result is A of 40 or 129.

[IslandPink]

Thanks Paul, that confirms what Stephen was saying then .
Good to see the figures in detail . I must get the equation into a spreadsheet to use again .

I do need the gain of eg. 125 - 130 in the first stage to allow use of the C3g at the end. I'm hoping it will sound nice when backed-up by the oomph of the 6N6P aikido to drive the LCR . I know Mike Carrick is a big fan of the E83F in terms of tone .

I'll use the BG ( NX ) 1000u/25's again then . Let's hope no harm comes to them - I feel we are really stuffed nowadays as we can't get any new ones - there's nothing I'm aware of that has replaced them .

I must say however I'm still a bit surprised that I get loads of gain in my powers amps when running C3m or C3g un-bypassed driver ..

[izzy wizzy]

I must say however I'm still a bit surprised that I get loads of gain in my powers amps when running C3m or C3g un-bypassed driver ..

But I bet you're not looking for 1st stage phono type gain in yer power amps. Signal swing might be a bit of an issue otherwise.

You might like to try a battery in the 1st stage as well - an AAA alkaline could work with a bit of an adjustment on the anode resistor. It's my preference over a cathode cap or LED. Some like LEDs though, even schottkys.

cheers,

Stephen

[IslandPink]

But I bet you're not looking for 1st stage phono type gain in yer power amps. Signal swing might be a bit of an issue otherwise.

No, but I'm recalling the fact that I built a three-stage ECC40-45-300B ( with by-passed cathode caps ) power amp option to compare, and the gain was only slightly more than the C3m-300B with 11k load on the C3m's and no bypass caps . I can also recall that James D told me there was very little loss of gain by not using the caps for a pentode in this case. I'll see if I can find anything in old e-mails from James at that time .

You might like to try a battery in the 1st stage as well - an AAA alkaline could work with a bit of an adjustment on the anode resistor. It's my preference over a cathode cap or LED. Some like LEDs though, even schottkys.
Stephen

Thanks for the advice Stephen, but I'm with Paul on this one I think . Anything other than high quality by-pass caps or grid battery bias sounds poor in the treble to me - and just less realistic . I could try grid battery bias on st. 1 of course...