• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Suggestions for parafeed OPT - Any guidance out there?

I have a small head amp that I have been testing, which currently uses a relatively large 30VA toroid as OPT.

The approximate reflected Z is about 15k, using 300R load to the secondary.

Primary volts swing is about 30V RMS, using the toroid (240:30V) - primary L= 6H and lossy capacitance is a whopping 0.5uF!
Turns ratio comes out to something like 6.7:1

So far I have found Edcore XSM 10k:150 (oversize but tolerant of the voltage requirements)

Hammond 107V, which looks tiny...DIP mount???
But also has good primary inductance, 25k:600R, and appears to be suitable for an input level of 30V RMS.

Thus either OPT above, I am either using a lower Z tap for a higher load R, or vice versa.

Any suggestions, hints or advice, to help me select one of these OPT, or perhaps another more suitable, is gratefully received.

Thanks
 
Thanks, Sowter was a name I had forgotten.

With a ratio of 6:1 it's close to my 6.7:1 toroid. Interesting, this may offer more flexibility to me than say, Edcore WSM/XSM.
Thanks again :)
The only downside, is the primary reflected impedance is on the low side, 15k would probably work better, at least from a THD perspective, which is why I had looked at the Edcore 10k:150R, which would then come out a little too high with 300R load.

The output tube I am using is (not strictly) an output tube, 1J24B, a subminiature russian tube.

Triode wired it has a mu of about 8, and is biased at about 8mA, pretty much the maximum.

I dont know, and have not tried to calculate the Ra, so the reflected impedance is merely what iteratively has worked best, with a random toroid - I found 12k to 15k worked best.

This amp is to be used with Sennheiser HD650, since nothing I have seems to provide enough volts!

Even so, with the toroid, I can squeeze about 4V, 60mW from 7Hz to 27kHz -3dB points, which should be loud enough (HD650 are reputedly about 100dB/mW, or 105dB @ 1V)
 
Last edited:
Apologies for the double post....

PRR,

You may be correct, I recalled 0.5uF, but it may well have been 50nF, I was using a auto ranging handheld LCR meter I'm not all that familiar with. Will double check.

There is significant degradation in the THD with increasing frequency though, which I suppose is prompting my looking for a 'proper' OPT. It seems to load well up to a point, and my THD measures at 1kHz and 100Hz have been encouraging. It's just the 10kHz :D

I should make a spreadsheet to plot the transformer Impedance,
but "Reflected Z" I quote is based on the turns ratio, assuming I have calculated correctly.


What does surprise me is that I need such a small blocking capacitor, and have a good LF response.
 
Last edited:
Using a Agilent U173 LCR meter, incorrectly....

Hopefully now a little more sense...

Primary = 95R
Secondary = 3R
Primary Inductance (secondary open circuit) = 15mH
Primary inductance (secondary short circuit) = 20H
Secondary inductance (primary open) = 1.3H
Secondary inductance (primary shorted) = 300uH
Capacitance between primary and secondary = 270pF
All measured at 100Hz, except the DCR of each winding.

Hopefully enough numbers to characterise the transformer, or am I missing something else?

that is, assuming a simple LCR meter can be trusted for this type of thing
 
mondogemerator,

Your post #9 shows:
Primary Inductance (secondary open circuit) = 15mH
Primary inductance (secondary short circuit) = 20H

That is Backwards:
The higher primary inductance is when the secondary is open.
The lower value primary inductance is called leakage inductance, and is tested with the secondary shorted.
 
Nothing about that seems right.

The tube should have a mu closer to 20 from what I am reading online. If that is true, I would expect a mu of about 12-16 in real world applications.

Could you please post your schematic.

Hilariously, nothing about your observations based on reading online "truth" are factually accurate.

In pentode, then yes mu may be about 20.

Wow This forum has gone down hill...

Ask a question, receive berating criticism based on "what I read online", without considering what you read online was tosh.

I have discovered this myself, taking forum pointers,
only to realise I'm not as stupid as I credit myself, and I was ok the right track after all.


Cheers PRR, indeed my Z isnt 15k, but the Z ratio is.
However, I am not aware of an easy way to measure a transformer, without a £££s network analyser - so essentially whether my Z or Z ratio is 15k or not, is largely moot, much like the quality of responses like these.
Considering, one could have offered up some algebraic posing as a means of help - after my own investigation, it seems almost pointless without aforementioned £££s network analyser - the quality of measurements to define the various parameters being hard to achieve for the layman, and be representative of reality. Particularly so, when such information about "suitable" transformers is just as difficult to find.

MOD please close this thread -
 

Tjj226

Member
2012-12-02 2:49 am
Hilariously, nothing about your observations based on reading online "truth" are factually accurate.

In pentode, then yes mu may be about 20.

Wow This forum has gone down hill...

Ask a question, receive berating criticism based on "what I read online", without considering what you read online was tosh.

I have discovered this myself, taking forum pointers,
only to realise I'm not as stupid as I credit myself, and I was ok the right track after all.


Cheers PRR, indeed my Z isnt 15k, but the Z ratio is.
However, I am not aware of an easy way to measure a transformer, without a £££s network analyser - so essentially whether my Z or Z ratio is 15k or not, is largely moot, much like the quality of responses like these.
Considering, one could have offered up some algebraic posing as a means of help - after my own investigation, it seems almost pointless without aforementioned £££s network analyser - the quality of measurements to define the various parameters being hard to achieve for the layman, and be representative of reality. Particularly so, when such information about "suitable" transformers is just as difficult to find.

MOD please close this thread -

Woaaaaaahhh chiiiiiiil dude chill.

I was looking for tube data to help figure out what sort of primary impedance we actually need. I happened to run into a few posts where people were testing this tube in triode mode and they were claiming a mu of 20. They even posted the tube curves.

1Ж24Б, Tube 1Ж24Б; Rohre 1Ж24Б ID19748, Vacuum Pentode

Scroll down (almost to the bottom) and you will see what I am talking about.

While I understand you have experimentally figured out what transformers kind of sort of work, I was hoping to track down a more firm set of parameters.

Thats ALL I was doing. I wasn't trying to berate you. I was just trying to see if we could work from a more solid foundation so that we could HELP you.

I would hate to tell you to go buy XYZ transformer just to have it be the wrong thing.

That was my only intent. I had no intent of berating you or belittling you. I was JUST trying to be as accurate as possible so that I didn't run the risk of recommending the wrong thing.

I am genuinely sorry if I offended you.
 
Please accept my apologies.

I'm a little trigger happy today!

As PRR (I believe) pointed put in my other thread, I can approximate triode curves- I just havent bothered.

I should really go and do that groundwork, but I'm lazy, and I had fiddled until something worked in a reasonable fashion.

I figure, judging by nothing more than THD, that I really need an OPT which gives a impedance load of 10-15k across the AF band, and not just to 1kHz.
 
Last edited:

Tjj226

Member
2012-12-02 2:49 am
Please accept my apologies.

I'm a little trigger happy today!

As PRR (I believe) pointed put in my other thread, I can approximate triode curves- I just havent bothered.

I should really go and do that groundwork, but I'm lazy, and I had fiddled until something worked in a reasonable fashion.

I figure, judging by nothing more than THD, that I really need an OPT which gives a impedance load of 10-15k across the AF band, and not just to 1kHz.

NP. We all have trigger happy days. :)

My thing about trying to determine the "correct" mu and impedance ratio is 2 fold.

1: The output transformer is obviously stepping down the voltage from the tube. If you only have 1 tube, then the amplification of the tube should ideally exceed the step down ratio of the transformer. Or at the very least the ratio should be such that you are at least left with 1V rms at the output.

2: Lets say that by some miracle your tube can actually reach a gain factor of 20. It means we can increase the primary impedance by quite a bit which opens up some extra options.

For instance the Lundhal LL1930 might be a decent option IF you had the extra gain. It has a step down of 11.6 to 1. So if you absolutely for sure have a gain of 8, it may just be to large of a drain on your output.
 

6A3sUMMER

Member
2016-06-07 6:50 am
Regarding the use of u (Mu) and Gain:

The u (Mu) of a triode at a quiescent operating point is a constant (a fixed plate voltage and a fixed bias).

The Gain of a triode that has a bypassed cathode is u x (RL/(rp + RL)).
Of course RL includes the plate resistor in parallel with the grid resistor of the next stage.

A 6SN7 at a quiescent operating point may have a u of 20.
But if the plate resistance is 7700 Ohms, and the plate load resistor in parallel with the next stage grid resistor combination is 23,100 Ohms, the gain will be 15.

Please do not refer to triode u and Gain as the same thing (unless RL is infinite).

The only spec for the u of a pentode is the gain of G2 referred to G1.
The gain of the pentode plate referred to G1 is Gm x RL.
 
I didn't use mu in place of gain... pretty sure of that.

I've said "near mu" when CCS loaded giving 8V/V or the like.

Of course mu is not appropriate when discussing pentodes, bar the G1G2 mu.

Saying that, easy to get terms mixed when discussing triode wired pentodes!

Now...back to the load.

Sadly, as far as the Radiomuseum threads, there are a few, very difficult to spot, errors.

Much as the articles by Joe useful and full of information, some of the links within them are wrong - if I am constantly mixing up the use of 1J24B and 1J29B please forgive me - Mr Sousa also does the same on a few occasions!

Input stage drives 1M grid leak, as well as LNFB.
So if I had a real figure for mu, I could reverse calculate the gain to compare.

So the first stage load would be something like 1M||750k, I.e. RL||Rlnfb?
Using a CCS there is no Rp to use in calculation.

Time to hook up a conventional resistor anode load, and take a few measures of Vg1 against Ik in the region I am biasing, to give me a idea of ra (rp), and a better guesstimate of required transformer impedance.

If I recall though, I took a single beam resistance point which indicated ra of around 5k, for triode wired 1J29B, hence my (perhaps mistaken) belief that I require a 15k:300R OPT.
 
Last edited:

Deke609

Member
Paid Member
2018-11-05 11:37 pm
Toronto
Maybe the Lundahl 2774 Headphone OT

Thanks, Sowter was a name I had forgotten.

With a ratio of 6:1 it's close to my 6.7:1 toroid. Interesting, this may offer more flexibility to me than say, Edcore WSM/XSM.
Thanks again :)
The only downside, is the primary reflected impedance is on the low side, 15k would probably work better [snip]

The Lundahl 2774 has a 6.8:1 configuration option that would give you close to 14K reflected on the primary with secondary loaded by 300 ohms: http://www.lundahltransformers.com/wp-content/uploads/datasheets/2774.pdf

[Edit: and you can get them made "PP-Z": zero gap for parafeed]

cheers, Derek
 
Last edited:

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
> primary L= 6H and lossy capacitance is a whopping 0.5uF!
> much like the quality of responses like these.


Your data is bogus. Do not attack when this is pointed-out.

> I should really go and do that groundwork, but I'm lazy

Your life, your choice.

> I am not aware of an easy way to measure a transformer, without a £££s network analyser

Signal generator. Resistor. Audio voltmeter.
 

Attachments

  • Mondo-Z-42.gif
    Mondo-Z-42.gif
    11.4 KB · Views: 64