Excellent points Franz. I wonder why they have the mains transformer oriented in that manner. It doesn't follow common sense for mounting a mains transformer in a tube circuit.
Doesn't the EU have radius standards for consumer electronics?
EU standards?
For (consumer-)devices with more than 40 or 45 VDC inside?
Quite hard if not impossible to certify such a device.
The responsibility for a self imported device from outside EU is imho by the importer himself. He has to verify safety and security labels.
BTW: is the main trannie tested for 4kV isolation?
Franz
Last edited:
I don't think so, but hey, you would notice that rather fast 😛
About the Gold Lion's, they might be not as well constructed as we think.
Just noticed a remark on Tubedepot:
"Due to excessive current demands found in the Vox AC-30, we do not recommend these tubes for use in these amps.
As far as I know, nothing (of current production) beats the JJ EL84.
But I would love to test these:
TAD EL84-STR PREMIUM Matched, Qartett TAD TubeAmpDoctor
Don't know who is building those tubes for TAD, they are shorter an thicker (22mm!) than "normal" EL84's...
What I am hearing is that the problem the problem Rhing had which caused him to loose his Gold Lion was related to the capacitor overheating, not the pin1 wiring. I have been running mine every night for 2 months, and nothing, yet. I will redesign the case to reduce heating
OneyedK- Thanks for the tip on the JJ EL84, I just ordered 4.
JimS
OneyedK- Thanks for the tip on the JJ EL84, I just ordered 4.
JimS
I am not sure what happened where the one Gold Lion went out. All I know is that I replaced the bulging Panasonic FM cap which I had installed a couple days before. It was the same exact value (220uF/50V) as the stock Chemicon cap that was in there. I, then, put in the stock 6p15 power tubes, and have had no problems since.
My Gold Lions lasted about two months. In fact, they were the tubes in the amp when I took it to the San Francisco Vacuum Tube Users Group meeting in February where everyone liked it.
Please keep us all updated on the service life of your Gold Lions. If it keeps going, I'll chalk my failure up to a bad cap. I do see Russian 6P15P-EV tubes on eBay at reasonable prices. I wonder how these sound in comparison to the Chinese stock tubes.
Are you curious enough to want your Gold Lion failure Xray'd?
I already got some 6P15P-EV internal pics up for comparison.
You could probably get the same info with a hammer and pliers,
but takes a risk you deform something and draw an incorrect
conclusion.
Now I been using JJ-EL84's in my old Fisher X-100-3, with sockets
labeled 7189. And the JJs appear by default to bias correctly for
that application. 6P15P-EV also works, but seems very underbiased
in my amp. So I would expect JJ-EL84 to run seriously hot in the
Suppo without changing a cathode resistor or something like that...
6P15P-EV and EL84/7189 definitely don't bias the same.
I already got some 6P15P-EV internal pics up for comparison.
You could probably get the same info with a hammer and pliers,
but takes a risk you deform something and draw an incorrect
conclusion.
Now I been using JJ-EL84's in my old Fisher X-100-3, with sockets
labeled 7189. And the JJs appear by default to bias correctly for
that application. 6P15P-EV also works, but seems very underbiased
in my amp. So I would expect JJ-EL84 to run seriously hot in the
Suppo without changing a cathode resistor or something like that...
6P15P-EV and EL84/7189 definitely don't bias the same.
Last edited:
Forgot to mention that the "triode mod" I presented earlier is correctly biased for JJ EL84 (195 Ohm cathode resistor). 6P15P is indeed a bit underbiased after the mod, but still sounds very nice. Did this to protect the tube rollers...
About the ebay 6p15p offers, I saw them too, but I'm leaning towards the 6p14 offerings, 14Watt anode power instead of 12 for 6p15 and EL84... Guess this will ask for accurate re-bias.
my gold lions only lasted a month in my el84 amp, which doesnt have any of the pinout issues. I now do not trust them at all.....
One blew with a similar failure to yours, a big blue flash and a bang. The other flash... then purple lightening.
So what value resistor should we use to properly bias the EL-84s for standard mode (not triode modified)?
Thanks,
Bill
Thanks,
Bill
Update: VitaminQ caps
Hello Suppos owners,
I just replaced the 4 coupling caps in my EL84 amp two days ago. I used .22uf 400 volt VitaminQ oil caps. The amp sounded terrible at first, but after 5 hours or so it started to open up. After 15 hours of burn in, this little amp definitely sounds better than my Fisher 800 receiver. My amp now puts out a clean, clear sound. Bass is deeper (because of the .22uf value?) and the stereo imaging is wider and much better defined. This amp is less fatiguing and it commands your attention. So, if you have one of these amps forget about the tube swapping until you swap out those stock .1uf caps with better and higher value caps. The .22uf may be two high an increase, but I can not hear any distortion due to OPT saturation on the bass. I'm glad it's
Friday because now I can listen deep into the night without worrying about getting up for work.
Have a good weekend,
David
Hello Suppos owners,
I just replaced the 4 coupling caps in my EL84 amp two days ago. I used .22uf 400 volt VitaminQ oil caps. The amp sounded terrible at first, but after 5 hours or so it started to open up. After 15 hours of burn in, this little amp definitely sounds better than my Fisher 800 receiver. My amp now puts out a clean, clear sound. Bass is deeper (because of the .22uf value?) and the stereo imaging is wider and much better defined. This amp is less fatiguing and it commands your attention. So, if you have one of these amps forget about the tube swapping until you swap out those stock .1uf caps with better and higher value caps. The .22uf may be two high an increase, but I can not hear any distortion due to OPT saturation on the bass. I'm glad it's
Friday because now I can listen deep into the night without worrying about getting up for work.
Have a good weekend,
David
Replacing those coupling capacitors with better ones is a great idea.
Choosing larger capacitors is a bad idea...
Rg for 6P15P or EL84 can be 1MOhm with autobias (cathode resistor).
Peepz at Suppo chose 220kOhm, don't know why, I suppose they misread the datasheets.
Changing Rg from 220kOhm to 1MOhm will place a lighter load on the phase splitter and the cutoff frequency will be 4 times lower.
To keep the same cutoff frequency, one could use 0.022µF capacitors and 1MOhm resistors. That's what I will try first. 22nF capacitors can be found smaller and in truly great quality, without a crazy price tag. (I don't want to spend al lot of money on my mods)
Choosing larger capacitors is a bad idea...
Rg for 6P15P or EL84 can be 1MOhm with autobias (cathode resistor).
Peepz at Suppo chose 220kOhm, don't know why, I suppose they misread the datasheets.
Changing Rg from 220kOhm to 1MOhm will place a lighter load on the phase splitter and the cutoff frequency will be 4 times lower.
To keep the same cutoff frequency, one could use 0.022µF capacitors and 1MOhm resistors. That's what I will try first. 22nF capacitors can be found smaller and in truly great quality, without a crazy price tag. (I don't want to spend al lot of money on my mods)
Little confused. First of all, would you recommend swapping those resistors on an amp without the triode mod? And, are you recommending the 1MOhm resistor with the original .1uF or with some other value?
.22uf versus .022uf?
Hello Oneyedk,
Even though I don't understand most of what you wrote, thanks for responding. I know I've doubled the value of the coupling caps and have heard that the increase may overload the OPT and thus produce distortion. Besides the bad sound can it harm my tubes or other parts on the amp? Could it also cause other negative consequences like reduced tube life or cause safety issures? So far I am very happy with the sound I am getting with the .22uf VitaminQs. But I'm ready to make changes if it would increase its performance. I'm a little unclear about which resistor you are referring to and its value Is R9 the one you are talking about?. You mentioned a cutoff frequency. What would that be? You also proposed using a .022uf cap and 1 mohm resistor. Can you explain what it is and how to calculate it?. If I keep my .22uf what value should the resistor be? I'm weak on a lot of the electronic details.So, if you don't mind.
Thanks,
David
Hello Oneyedk,
Even though I don't understand most of what you wrote, thanks for responding. I know I've doubled the value of the coupling caps and have heard that the increase may overload the OPT and thus produce distortion. Besides the bad sound can it harm my tubes or other parts on the amp? Could it also cause other negative consequences like reduced tube life or cause safety issures? So far I am very happy with the sound I am getting with the .22uf VitaminQs. But I'm ready to make changes if it would increase its performance. I'm a little unclear about which resistor you are referring to and its value Is R9 the one you are talking about?. You mentioned a cutoff frequency. What would that be? You also proposed using a .022uf cap and 1 mohm resistor. Can you explain what it is and how to calculate it?. If I keep my .22uf what value should the resistor be? I'm weak on a lot of the electronic details.So, if you don't mind.
Thanks,
David
Last edited:
Sorry for the long post, but some explanations need more than two words😱
Cutoff frequency:
f=1/(2*PI*R*C)
So when we take C2 and R9 as an example we get:
f=1/(2*PI*220000*0.0000001)=7.234Hz
If you increase C2 to 0.22µF you get:
f=1/(2*PI*220000*0.00000022)=3.288Hz
This amp consists of 3 high pass filters (HPF).
1st is C2/R9 (for the other tubes: C3/R10, C14/R10 and C15/R19)
2nd is R11/C4 (for the other pair of tubes: R20/C16)
3rd is the opt
Normally, we would use the 3rd as a reference, if we knew the transfer characteristics of the opt, we would have that reference. In this case, we don't 🙁 (at least, not yet, I'll measure it when I find the time)
If we had that frequency, we could calculate the first filter to prevent lower frequencies to mess op the amp. (garbage in, garbage amplified, garbage out! What doesn't get in, cannot distort the amp)
I presume Suppo did that, that's why I suggested to leave the frequency of that filter to be untouched.
--> that's why I suggested the 22nF (0.022µF) with 1MOhm combination, in that case, you keep Suppo's frequency of choice (or calculation), but you present the phasesplitter with a lighter load (1M instead of 220k)
Now the second "filter" is a different beast, if we had fixed instead of autobias, there would be no filter. In one way, it's a filter, in another it should keep the cathode voltage stable at every possible frequency that is presented to the power tubes.
In practice, the capacitor there acts as a charge-discharge device depending on the dynamics presented.
(the compressor effect of an autobias PP amp)
I'd keep the frequency of that "filter" as low as possible, but the bigger the electrolytic capacitor, the more difficult it will be to find (and to pay) a really good one. (low ESR, low inductance)
As for the triode mod, what did it actually do?
The power stage is a series circuit with three components:
1) opt
2) tube (Ri)
3) Rk/Ck
These three things define how we will move the speaker.
We did nothing to (1), the transformer stays as it was.
What we did with (2) is to switch the initial penthode, with a rather high Ri, to triode, wich has about 10x lower Ri. So we can keep the transformer (and speaker) and it's actions better in check.
Since we had to adapt the bias for triode mode, we ended up with a 2x lower frequency for the third filter. (Rk - R11 and R20 - became twice as big)
If we wanted to stay a decade away from the cutoff frequency of the 1st filter, we could start to calculate:
Rk = 195
f = 7.234/10 = 0.7234Hz (a decade lower)
Ck = 1/(2*PI*f*R) = 1/(2*PI*0.7234*195) = 0.001128 F = 1128µF
So a 1000µF/25V of great quality would improve sound/dynamics of your triode modded amp.
I wouldn't try that on the penthode mode version of the amp, since you would need 2500µF to get the same frequency and the high Ri will destroy the benefits of it.
Cutoff frequency:
f=1/(2*PI*R*C)
So when we take C2 and R9 as an example we get:
f=1/(2*PI*220000*0.0000001)=7.234Hz
If you increase C2 to 0.22µF you get:
f=1/(2*PI*220000*0.00000022)=3.288Hz
This amp consists of 3 high pass filters (HPF).
1st is C2/R9 (for the other tubes: C3/R10, C14/R10 and C15/R19)
2nd is R11/C4 (for the other pair of tubes: R20/C16)
3rd is the opt
Normally, we would use the 3rd as a reference, if we knew the transfer characteristics of the opt, we would have that reference. In this case, we don't 🙁 (at least, not yet, I'll measure it when I find the time)
If we had that frequency, we could calculate the first filter to prevent lower frequencies to mess op the amp. (garbage in, garbage amplified, garbage out! What doesn't get in, cannot distort the amp)
I presume Suppo did that, that's why I suggested to leave the frequency of that filter to be untouched.
--> that's why I suggested the 22nF (0.022µF) with 1MOhm combination, in that case, you keep Suppo's frequency of choice (or calculation), but you present the phasesplitter with a lighter load (1M instead of 220k)
Now the second "filter" is a different beast, if we had fixed instead of autobias, there would be no filter. In one way, it's a filter, in another it should keep the cathode voltage stable at every possible frequency that is presented to the power tubes.
In practice, the capacitor there acts as a charge-discharge device depending on the dynamics presented.
(the compressor effect of an autobias PP amp)
I'd keep the frequency of that "filter" as low as possible, but the bigger the electrolytic capacitor, the more difficult it will be to find (and to pay) a really good one. (low ESR, low inductance)
As for the triode mod, what did it actually do?
The power stage is a series circuit with three components:
1) opt
2) tube (Ri)
3) Rk/Ck
These three things define how we will move the speaker.
We did nothing to (1), the transformer stays as it was.
What we did with (2) is to switch the initial penthode, with a rather high Ri, to triode, wich has about 10x lower Ri. So we can keep the transformer (and speaker) and it's actions better in check.
Since we had to adapt the bias for triode mode, we ended up with a 2x lower frequency for the third filter. (Rk - R11 and R20 - became twice as big)
If we wanted to stay a decade away from the cutoff frequency of the 1st filter, we could start to calculate:
Rk = 195
f = 7.234/10 = 0.7234Hz (a decade lower)
Ck = 1/(2*PI*f*R) = 1/(2*PI*0.7234*195) = 0.001128 F = 1128µF
So a 1000µF/25V of great quality would improve sound/dynamics of your triode modded amp.
I wouldn't try that on the penthode mode version of the amp, since you would need 2500µF to get the same frequency and the high Ri will destroy the benefits of it.
Last edited:
Well OneyedK seems to know what he's talking about so maybe I'll have to follow suit (since I really don't know what I'm talking about ! 🙂 ).
Just for clarity's sake, would you recommended changing that 220K Ohm resistor to a 1MOhm resistor in a stock amp with the stock .1uF caps (well, stock value, not stocks caps)? I'm trying to do my mods in steps so I can get a feel for what each brings to the table. Also trying to spend the least amount of money possible; I'm not cheap, but its the principle of the thing 🙂
Just for clarity's sake, would you recommended changing that 220K Ohm resistor to a 1MOhm resistor in a stock amp with the stock .1uF caps (well, stock value, not stocks caps)? I'm trying to do my mods in steps so I can get a feel for what each brings to the table. Also trying to spend the least amount of money possible; I'm not cheap, but its the principle of the thing 🙂
F=1/2piRC 3.28Hz versus 7.23
Great post Oneyedk. Thanks. Please indulge me further. You stated 1 meg would be a better load for the splitter section. So, what would the upper and lower limits be for this part to maintain the appropriate load, 900K to 2megs? Also in setting the cutoff, what would the optimum cutoff point be? Again would there be a range of values that can work like 5 HZ to 8 HZ? From my calculations with my present values of 220k and .22uf, my amp's cutoff point is 3.28 Hz. I have various values of VitaminQ caps from .001uf to .22uf I can use. I found this: Pronine Electronics Design - RC Circuit Calculator Again thanks for your help.
Regards,
David
Great post Oneyedk. Thanks. Please indulge me further. You stated 1 meg would be a better load for the splitter section. So, what would the upper and lower limits be for this part to maintain the appropriate load, 900K to 2megs? Also in setting the cutoff, what would the optimum cutoff point be? Again would there be a range of values that can work like 5 HZ to 8 HZ? From my calculations with my present values of 220k and .22uf, my amp's cutoff point is 3.28 Hz. I have various values of VitaminQ caps from .001uf to .22uf I can use. I found this: Pronine Electronics Design - RC Circuit Calculator Again thanks for your help.
Regards,
David
Rg1 for EL84 and 6P15P cannot be higher than 1MOhm.
This can be found in the datasheet of these tubes.
Lower is not a problem for the power tube, higher will make the amp unstable
and/or prone to oscillations.
As for the phase spliter, in this case a concertina (sometimes called cathodyne),
an old rule of thumb was to load it with at least seven times the anode (or cathode, same value) resistor. In the case of our Suppo, 47k x 7 = 330k.
So, to answer your question, any value between 330k and 1MOhm would be fine.
The optimum cutoff frequency would be about 10x higher than the frequency where the opt
starts to saturate or roll off. We don't know that frequency... yet...
The good news is that IF you only use a CD-player or DAT-player as a source for the suppo,
no frequency lower than 20Hz will enter the amp. Simply because lower frequencies are not
encoded on the disk/tape.
The optimum cutuff frequency would be lower or equal to 2Hz (one decade below the lowest
frequency presented by the source).
@bigjppop: if you only use that kind of digital sources, you can safely use the 0.1µF/1MOhm combination.
btw, bookmarked that calculator, great find!
This can be found in the datasheet of these tubes.
Lower is not a problem for the power tube, higher will make the amp unstable
and/or prone to oscillations.
As for the phase spliter, in this case a concertina (sometimes called cathodyne),
an old rule of thumb was to load it with at least seven times the anode (or cathode, same value) resistor. In the case of our Suppo, 47k x 7 = 330k.
So, to answer your question, any value between 330k and 1MOhm would be fine.
The optimum cutoff frequency would be about 10x higher than the frequency where the opt
starts to saturate or roll off. We don't know that frequency... yet...
The good news is that IF you only use a CD-player or DAT-player as a source for the suppo,
no frequency lower than 20Hz will enter the amp. Simply because lower frequencies are not
encoded on the disk/tape.
The optimum cutuff frequency would be lower or equal to 2Hz (one decade below the lowest
frequency presented by the source).
@bigjppop: if you only use that kind of digital sources, you can safely use the 0.1µF/1MOhm combination.
btw, bookmarked that calculator, great find!
I really appreciate you guys helping out newbies like me. Looking at the schematic, I'm assuming the resistors in question are R9, R10, R18, R19; is that correct? Also, the schematic shoes them as 220K on one side and 200K on the other; I'm assuming that's simply a mistake and they should all be 1MOhm? Yes, I'll only be using digital sources (probably a DAC being fed by my music server).
P.S. A recommended wattage spec for these?
P.S. A recommended wattage spec for these?
Those resistors draw almost no current, so 1/4W is fine.
There is still one catch I forgot to mention, if you lower the cutoff frequency below the frequency where the opt shifts phase (we still don't know that frequency), there is a small chance that the amp will start to "motorboat" (oscillate at a low frequency).
Not dangerous, you won't blow up your speakers/tubes/opt, but you will not be able te enjoy much music unless you up the cutoff frequency again.🙄
Another thing to watch out for is noise, 1MOhm could inject more noise than 220k...
There is still one catch I forgot to mention, if you lower the cutoff frequency below the frequency where the opt shifts phase (we still don't know that frequency), there is a small chance that the amp will start to "motorboat" (oscillate at a low frequency).
Not dangerous, you won't blow up your speakers/tubes/opt, but you will not be able te enjoy much music unless you up the cutoff frequency again.🙄
Another thing to watch out for is noise, 1MOhm could inject more noise than 220k...
- Status
- Not open for further replies.