I was looking at this interresting schematic but I have a problem with some units :
Is 6ГH just mean 6H ( 6 Henry ) ?
The C1 and Ck capacitors seem to be unpolarised, not electrolytics like usual ... 68000Mk and 2200Mk ...
Did somebody know what mean this mysterious "Mk" capacitor unit ???
PicoFarad, NanoFarad ???
Alain.
An externally hosted image should be here but it was not working when we last tested it.
Is 6ГH just mean 6H ( 6 Henry ) ?
The C1 and Ck capacitors seem to be unpolarised, not electrolytics like usual ... 68000Mk and 2200Mk ...
Did somebody know what mean this mysterious "Mk" capacitor unit ???
PicoFarad, NanoFarad ???
Alain.
I don't think so because the filter electrolytic capacitors unit are already in "uF" on the schematic but the bypass unpolarised capacitors have the "Mk" unit ... Why ???
Also, 68000uF and 2200uF seem to be very high values for cathode bypass capacitors and they can't be paper or film capacitor (unless they are big like a garbage can) ...
Alain,
Yes, its a H and uF. The schematic attached is rather dumb. 68000uF cathode bypass is a nonsense. BTW, voltage on this monstrous 68000uF is rather low (several V probably).
I would be cautious to take schematic with so obvious flaws.
If you want to use similar but more proven design, take this schematic (it was printed in Russian-language Radio Hobby magazine published in Ukraine).
"Áþäæåòíûé" Loftin-White
Author says you can use output tubes 6S19P, 6P3S-E, GU50, 6S33S, 6S41S if you have corresponding output transformer.
Good luck !
I would be cautious to take schematic with so obvious flaws.
If you want to use similar but more proven design, take this schematic (it was printed in Russian-language Radio Hobby magazine published in Ukraine).
"Áþäæåòíûé" Loftin-White
Author says you can use output tubes 6S19P, 6P3S-E, GU50, 6S33S, 6S41S if you have corresponding output transformer.
Good luck !
Attachments
Last edited:
I think the solution is in the text of your link and google translate ...
The russian sentence :
The google traduction to english :
So, GloBug and LinuksGuru are right ...
But like said LinuksGuru said : "The schematic attached is rather dumb. 68000uF cathode bypass is a nonsense" ...
I think the C1 capacitor can be 470uF and Ck 47uF ... Plenty enough ... It is always funny to see how some people use a bazooka to kill a mosquito ...
Thanks to all !
Alain.
If you want to build a true Loftin-White amp, you don't need either of those caps, or power supply caps of the values shown. If you're serious about building it, I'll draw up a schematic.
Sheldon
Last edited:
I just made a little Spice simulation of this circuit ...
Very nice results from 100Hz to 20Khz but distortion seem to raise fast under 100Hz !
I use more "normal" values for capacitors and a real transformer specifications, a "little" Edcor GXSE10-8-5K.
The Spice models I use :
Alain.
Very nice results from 100Hz to 20Khz but distortion seem to raise fast under 100Hz !
I use more "normal" values for capacitors and a real transformer specifications, a "little" Edcor GXSE10-8-5K.
An externally hosted image should be here but it was not working when we last tested it.
The Spice models I use :
I will try with a better transformer to get better results under 100Hz ... A good negative feedback will help too ...
Alain.
I do the same test with the "Deluxe" Edcor transformer CXSE25-8-5K
I have to change the 6S19P bypass capacitor for a 6,8uF because there was a big resonance peak at low frequency ... Using this transformer, the response and distortion is very good from 20Hz tp 20Khz and the square waves at 20Khz look pretty good too.
The transformer inductance have a lot to do with the low frequency quality, 5H like the "little" one is much too low, this one have a 50H inductance and a better coupling factor. For SE, bigger transformer is always better ...
I will now try to make the distortion lower at 20Hz with the "little" transformer of the other schematic using a global negative feedback ... Of course, there will be less gain !
Alain.
I have to change the 6S19P bypass capacitor for a 6,8uF because there was a big resonance peak at low frequency ... Using this transformer, the response and distortion is very good from 20Hz tp 20Khz and the square waves at 20Khz look pretty good too.
The transformer inductance have a lot to do with the low frequency quality, 5H like the "little" one is much too low, this one have a 50H inductance and a better coupling factor. For SE, bigger transformer is always better ...
An externally hosted image should be here but it was not working when we last tested it.
I will now try to make the distortion lower at 20Hz with the "little" transformer of the other schematic using a global negative feedback ... Of course, there will be less gain !
Alain.
Look what the problem is with the "little" Edcor GXSE10-8-5K transformer ...
At 160Hz, the waveshape look good.
At 80Hz, there is already a bump in the slope.
At 40Hz, it is getting very bad.
At 20Hz, it is just horrible ...
Alain.
At 160Hz, the waveshape look good.
At 80Hz, there is already a bump in the slope.
At 40Hz, it is getting very bad.
At 20Hz, it is just horrible ...
An externally hosted image should be here but it was not working when we last tested it.
Alain.
Look what the problem is with the "little" Edcor GXSE10-8-5K transformer ...
At 160Hz, the waveshape look good.
At 80Hz, there is already a bump in the slope.
At 40Hz, it is getting very bad.
At 20Hz, it is just horrible ...
An externally hosted image should be here but it was not working when we last tested it.
Alain.
Well specs only list the response as 40Hz to 18KHz...
EDCOR Electronics Corporation. GXSE Series Output Transformers
But for $26 I'd call it a pretty good deal.
For a really serious SE, I think it worth paying 83,50$ each for the best ones ... It is only 57,40$ more and delivery is the same price ... Most other brands HI-FI 25W SE transformer cost over 150$ each so it is still a very cheap price ...
It is better to invest in good transformers because they last for life, but it is a good idea to save on tubes because they have to be replace periodically ... You can get four NOS 6S19P on eBay from Russia for about 10$, this is where you can save a lot of money. Those very good russian tubes are more linear than 2A3 and 300B and as powerfull, think about that ...
The GXSE transformer specs said 40Hz to 18Khz response but they don't said what happen at 40Hz ... They said the transformer distortion and noise are under 0,1% but it is at 1Khz ... Easy ...
Also, I use a "Ideal transformer model" in my Spice simulator and the only reals variables beside the turns ratio are the primary inductance and coupling factors between windings ... I have to approximate this spec because Edcor don't mention it ...
That mean the real transformer is worst than that because the simulator don't calculate Eddy current and hysteresis loss and also DC magnetic saturation effect ... Like I write before, those distortions are caused by a too low primary inductance.
I try to reduce those low frequency distortion with feedback in the circuit but there is nothing to do with the cheap GXSE transformer ... But since the waveshape is nice over 80Hz, those "little" transformers can be used in a "bi-amp" system for medium and high frequency (with a hi-pass filter) plus a more powerfull amplifier to drive the subwoofer(s) !
Alain.
I made the AC load line of the original schematic with the 5K load and I found out there is some problems with this circuit ...
1- The 218,1V operation is over the 200V maximum.
2- The 11,36W power dissipation of the tube is over the 11W maximum.
3- The 5K load is not the best choice ...
Like you can see, a 2,5K load, 195V operation point at 54ma and a -70V bias is much better ...
1- The 195V operation point is under the 200V maximum.
2- The 10,53W power dissipation of the tube is under the 11W maximum.
3- The 2,5K load result in a higher output power.
4- The 2,5K load result in lower amplitude distortion.
5- The maximum power can be reach with a smaller signal because bias is lower ...
I will redo my Spice simulations with those new parameters next week.
That's the first step before assembling any kinds of electronic circuits, specially when the parts are expensives, Spice simulations cost nothing and the very friendly and powerfull SIMetrix Spice simulator demo is totally free and for unlimited time use ... Circuits evaluation and drawing can be done really fast with it, I believe this fantastic simulator is actually the best in the word, it can run the Koren 8 parameters models very fast, not many other simulators can do that ... Anybody who like to learn more about electronic should download this fantastic software right now !
I made many triodes Spice models myself using the Andrei Frolov software from the datasheet curves and generate back the curves with the SIMetrix simulator to compare both and believe me, they are almost perfects ... The curves in this post are made by this simulator ... Then "the loop is looped" like some people said.
Alain.
1- The 218,1V operation is over the 200V maximum.
2- The 11,36W power dissipation of the tube is over the 11W maximum.
3- The 5K load is not the best choice ...
An externally hosted image should be here but it was not working when we last tested it.
Like you can see, a 2,5K load, 195V operation point at 54ma and a -70V bias is much better ...
1- The 195V operation point is under the 200V maximum.
2- The 10,53W power dissipation of the tube is under the 11W maximum.
3- The 2,5K load result in a higher output power.
4- The 2,5K load result in lower amplitude distortion.
5- The maximum power can be reach with a smaller signal because bias is lower ...
I will redo my Spice simulations with those new parameters next week.
That's the first step before assembling any kinds of electronic circuits, specially when the parts are expensives, Spice simulations cost nothing and the very friendly and powerfull SIMetrix Spice simulator demo is totally free and for unlimited time use ... Circuits evaluation and drawing can be done really fast with it, I believe this fantastic simulator is actually the best in the word, it can run the Koren 8 parameters models very fast, not many other simulators can do that ... Anybody who like to learn more about electronic should download this fantastic software right now !
I made many triodes Spice models myself using the Andrei Frolov software from the datasheet curves and generate back the curves with the SIMetrix simulator to compare both and believe me, they are almost perfects ... The curves in this post are made by this simulator ... Then "the loop is looped" like some people said.
Alain.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.