• 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.

Best 300B SE OPT?

Jump to Latest
#481
ISCMMS said:
Data doesn’t spec a drive impedance so naturally the freq response will also not be shown.

However, it does spec Lpri for a range of frequencies and signal amplitudes so the LF can be calculated wrt drive impedance and measurements confirm the inductance spec is accurate.

Wrt the HF, 20kHz no problem but as always there are things that unloaded (tube grid to signal common) secondaries show as ‘not so pretty’ despite it ‘sounding preferable’ (to me) compared with a resistor load which will quieten the squirrels. So it will be application dependant.

There are also phase balance issues wrt each half of the secondary, yet somehow it sounds overall ‘more whole’ or ‘complete’ than the tube based inverter/splitters that I’ve tried.
Click to expand...
That’s a pitty. It is like buying a tube without having seen the curves. But if it works for you i am happy too.
 
#482
tubes4all said:
That’s a pitty. It is like buying a tube without having seen the curves. But if it works for you i am happy too.
Click to expand...
I am saying the freq response is not given on the data sheet because the drive impedance is not specified. In other words it will depend on your driver tube.

These were custom built for my application; using my in circuit driver Rp, Ip, voltage swing, voltage gain and output impedance requirements.
 
#483
ISCMMS said:
I am saying the freq response is not given on the data sheet because the drive impedance is not specified. In other words it will depend on your driver tube.

These were custom built for my application; using my in circuit driver Rp, Ip, voltage swing, voltage gain and output impedance requirements.
Click to expand...
Ok for custom it's different, but I think/hope you measured it. Btw it is quite normal that the drive impedance is not specified, but usually manufacturers have some suggestions what would be suitable and I think you asked for a specific application.
 
#484
andyjevans said:
I've previously been rather dismissive of Lundahl's LL1664 as a SE OPT for the 300b, preferring the LL1682. This is still the case - to me ears a higher primary impedance cleans up the midrange from some of the 300b's euphonic sound.

But I'm now using the LL1664 with a pair of 6C4C outputs and the result is excellent. The 6C4C is a Soviet copy of early US biplate 2a3s and I rate it very highly - usual Svetlana quality build. So my 300bs are going back on the shelf and I'm converting my amps to 6C4C. Good to know that the LL1664 is a good option for this - it's not overpriced.
Click to expand...
Hello Andy! As I understand it, you used LL1664 for two parallel 6C4C. I just can’t decide what is better to take 1664 or 1627 for PSE 6C4C? I do not want to overpay and also have an additional two kg of weight. What do you think about this?
 
#485
No. not parallel SE. Just one per side. If that's enough power for you then LL1664 is fine. If you are going to parallel two 6C4C then you would use something like a 2.5K OPT. There may be cheaper options than Lundahl which are also good. You might want to go back to the 300b if you want more power.

An alternative suggestion is to use a high gain output tube. This allows you to use a DHT driver tube. A good suggestion for an output tube is a EL12n. These are available cheap in Europe and sound pretty good. You could then use a simple DHT driver tube like a 2P29L. So you get the DHT but in the input rather than the output. Good cheap option.

https://www.conrad.com/p/el-12-n-va...r-of-pins-num-8-base-y8a-content-1-pcs-130538
 
#487
Andrey2211 said:
Hello Andy! As I understand it, you used LL1664 for two parallel 6C4C. I just can’t decide what is better to take 1664 or 1627 for PSE 6C4C? I do not want to overpay and also have an additional two kg of weight. What do you think about this?
Click to expand...
There is substantial difference between those two transformers. One could already guess from the size and weight that the small 1664 can only do 10W at expense of high DC resistance. The 1664 is 3K with total primary Rdc=148R and secondary Rdc =0.5R for 8R load; the LL1627 is 2.3K but RDC is only 56R and 0.2R respectively. Secondaries can be directly compared. The large difference will remain even if the higher impedance is taken into account. Being more economical it's also got 4x leakage inductance of the LL1627. 8 mH of the 1664 is a bit high for a 3K transformer with 17H primary inductance, in general. You can use it but do not expect "best" performance.

The 6C4C PSE I did it with the LL1627/90mA. 1W at 0.3% THD, 3W at 1% THD and max 10W at 5% THD Class A1( i.e. without going into positive grid field). 6C4C running at 360V/50mA each. That's 18W dissipation each but it is fine for these tubes. Running the 1627 at 100 mA DC instead of 90mA will reduce the headroom at 30Hz from 13W to 10W which is the max Pout anyway.....
If you want to stay within 15W plate dissipation you can decrease the plate voltage from 360V to 300V with 50 mA per tube. That should still turn into about 7W+ Class A1.
 
Last edited:
#489
The LL1627 works just fine for the 2A3/6C4C PSE. Other transformers can be used of course. The availability and cost also depend on where you are.
If 7W is you target then also halving the impedance from 2.5K to 1.25K for the classic 250V/60mA operative point will work fine. If I were in the US, for example, I would certainly consider the Edcor CXSE25 a much better choice than the 1664 and a good cheaper alternative to the 1627. These exist in both 1.25K and 2.5K impedance.
 
#490
I found that the smaller fixed primary LL166* OPTs sounded significantly better than the larger LL162* series with multiple windings. The big Lundahls are 2.5kg. The smaller ones are 1.4kg. Closest would be the 3K LL1664 with a 100mA gap. You could run a pair of 6C4C at 50mA each and take the a-k voltage up to 300V which is allowed in PP on the data.

You find a list of primary values here:

http://www.lundahl.se/tube-output/
 
#492
LL1682 is a little great OPT with excellent FR. The 1664 is not at the same level. The fixed winding scheme they use fits the 1682 better. Same for 1627 and 1623. The 1627 is good and the 1623 is poor for the price. Same reason. They are not designed one by one so some models are more optimal than others.
 
  • Like
Reactions: andyjevans
#494
45 said:
LL1682 is a little great OPT with excellent FR. The 1664 is not at the same level. The fixed winding scheme they use fits the 1682 better. Same for 1627 and 1623. The 1627 is good and the 1623 is poor for the price. Same reason. They are not designed one by one so some models are more optimal than others.
Click to expand...
I measured the LL1664 and it can do 8Hz-100kHz with a triode mode KT88 -3dB. Only very high dcR losses
 
#495
The ones I had stopped at 80 KHz but that could be just due to the different source impedance however the problem was that mines at least were ringing quite a lot, which the LL1682 does not with a variety of source impedances. I suspect that the LL1664 and LL1682 have the same identical primary and only the secondary is adapted for different ratio. In fact if you look at the gap options for DC current or PP they have the same inductance. The fact his that it's very unlikely one is going to use a 5.5K (actually 8.8K into 8R) transformer with just 17H inductance. So, limiting the DC current at 50 mA inductance will be at least 35H and that doubles the ratio between primary inductance and leakage with (positive) consequences on the FR behaviour.
 
Last edited:
#497
Hi,
Coming from page 14 or 15 of this thread, I Just sent an "order"/Request for Quote to Monolith Magnetics for a pair of Summit S-11s.. We shall see what we shall see about where that leads (these are the budget MM 300B 5K trafos) and I fear they lead to better places....for 300B?

If I go for these they will add to the rest of a growing bunch I have waiting to try...as follows:

Onetics, Level 3, 6K
Tamura 7003, 5K
Hashimoto H-30-3.5S
No Lundahl
Slagle 5k:8
Also have Hammond 1638 SEA, and James 6123, but I figure they are out of the running.

I'm not sure which circuit I will try... to date I have built two Loftin-Whites.: 24A/45/james6123, playing now. Also, EL84 (pentode)/2A3/FS-030. I am tempted to try something quite different...what should it be?

Ack! Best,

Robert
 
#498
Try a 2 stage all-DHT version. That's my choice. Shown here for 2a3 but you can modify it. I prefer the 2a3.

2a3 + 10Y SUT.png
 
#499
While we're speaking of this, I once met an opt guru & he said if it wasn't due to the varied impedance of commercial speakers,
to obtain the best sound from an opt is to have only 1 secondary winding instead of multiple taps. Can't recall his reasoning, perhaps
experts here could shed some light.

Cheers
 
Last edited:
#500
There are multiple taps and multiple secondaries. That's not quite the same thing. A transformer with multiple taps will use all the secondary turns only with the higher secondary impedance and will leave a good part of them unused for lower impedance. This will make it less efficient and might reduce the high frequency response. A (properly made) transformer with multiple secondaries instead uses all the turns at all times and so the issues of the tapped secondary do not exist. Performance is the same for all secondaries. Needless to say that having multiple secondaries makes the transformer more expensive. No free lunch. Exact 8R/4R or whatever is not possible because the impedance depends on the square ratio of turns but can get very close like 3.6R/8R or 4R/9R/16R etc...Not a big deal really,
 
You must log in or register to reply here.
Top Bottom